—_ ~ os en tenemos ne an a de inte ne oe a agrarian emer: a ee z eee nanos X am nope ng = ne Ep Oe dee tte ot ee ee ay ae THE AMERICAN JOURNAL OF PHYSIOLOGY VOLUME L 4 ore BALTIMORE, MD. 1919-1920 ae i a uf Tee 7 eton ake oul \ Ay > Na ( Bir | | a igor ‘air ~ 7. |] . | i erg Ae), Wt i mf ne H Weis \ ey lis : Wi Ne TAN on : # qr | ie : | ... CONTENTS No. 1. OcrosEerR 1, 1919 ACIMOSISE DURING STARVATIONG “Hiazime AISGdG.. 50.22 .a....056.-cessueee eel. SruDIES ON THE Extract oF Lune. Kowsaku Kakinuma................ MEAT SLATNING HAND AGIDOSISS | Hazeme Asada... 54). .a.1s sss saree ee. SrupiIges IN SECONDARY TRAUMATIC SHock. IV. THE BLoop Vo.tumME CHANGES AND THE HWrrect oF Gum AcaActA ON THEIR DEVELOPMENT. Herbert S. Gasser, Joseph Erlanger and W. J. Meek..................... Tue CaTALASES OF THE BLoop DurinGe ANESTHESIA. Stanley P. Reimann GRIN ES TOA BY YD OOS Gyo.) GSE EAS OIE ORG Cy COR SEER EEE CONTRIBUTIONS TO THE PHYSIOLOGY OF THE StomacH. LI. THE Controu OF THE Pytorus. A. B. Luckhardt, H. T. Phillips and A. J. Carlson... PHYSIOLOGICAL STUDIES ON PLANARIA. II. OxyGen CoNSUMPTION IN RE- TAG HMO MM anaKcioiNini NTO, Jy alse OUeapo tas Coto o ake eamen ben aoe so. Errect or ANESTHESIA AND OPERATION ON CERTAIN Merasouites. Stanley P. Reimann and Fred: L. Hartman................ es Me Dey! Santee oc EXPERIMENTAL SURGICAL SHocK. V. THE TREATMENT OF THE CONDITION oF Low Bioop PRESSURE WHICH FOLLOWS EXPOSURE OF THE ABDOMINAL Vasari e- SRR OR GTC). do, 6.15 elerey al Cace aeReee ee Ey Sea ae PR Be kg Cae W@XPERIMENTAL STUDIES ON THE REGULATION OF Bopy TEMPERATURE. III. Tue Errect oF INCREASED INTRACRANIAL PRESSURE ON Bopy TEM- PERATURE. (Preliminary Communication.) Lillian M. Moore......... STUDIES IN SECONDARY TRAUMATIC SHOCK. V. RESTORATION OF THE PLASMA VOLUME AND OF THE ALKALI Reserve. Herbert S. Gasser and Joseph LDC LTE RE i Se. SoSetibae Ss 3 cic cee ae MOR Bee act EN oe oh ee OE STUDIES IN SECONDARY TRAUMATIC SHocK. VI. STATISTICAL STUDY OF THE TREATMENT OF MEASURED TRAUMA WITH SOLUTIONS OF Gum ACACIA AND CRYSTALLOIDS. Joseph Erlanger and Herbert S. Gasser................- STUDIES IN SECONDARY TRAUMATIC SHocK. VII. NoTE oN THE ACTION OF Hypertonic Gum AcAcIA AND GLUCOSE AFTER HeMOoRRHAGE. Joseph PREG CT GN El CLOCMa eG ESSET oc. 56 eo oe 38 bones cine 2 wae Be eget 2 ees Tue INFLUENCE OF OXYGEN ADMINISTRATION ON THE CONCENTRATION OF THE Buioop wxHicn ACCOMPANIES THE DEVELOPMENT OF Lune Epema. D. Wis LVI On emote KC NGISe WiOUCLion gear AG Soe SAD es ee noises Gace aee ooo nr Tue Errect or ADRENALIN, DESICCATED THYROID AND CERTAIN INORGANIC SATS LONECATALASHVERODUGREON. Wis Biss UnGe. 2 cmies ce. os «srs clauses a A Note ON THE QUESTION OF THE SECRETORY FUNCTION OF THE SYMPATHETIC INNERVATION TO THE THyROID Gusanp. C.A. Mills................-5. il 82 86 102 104 119 1V CONTENTS No. 2. NovemBer 1, 1919 Tur HyPpEerRGLYCEMIA-PROVOKING ABILITY oF ASPHYXIAL BLoop. K. Yama- a eee A RE EA es See ICL SPREE hi SS ibys toe =. 177 Urea EXcrRETION AFTER SUPRARENALECTOMY. George Bevier and A. E. SYD) an CURR Oe ee PNY ie OAR, nt te Oo cr eA Ae 3 tole. ¢ 6:5 < 19] PosTURE-SENSE ConpUCTION PATHS IN THE SPINAL Corp. A PRELIMINARY — Report. Hugene S. May and John A. Larson..... ey ioe EA 0 Sc 204 STUDIES ON THE REGULATION OF THE BLoop Diastase. B. Fujimoto...-... 208 THE CHANGES IN THE CONTENT OF HEMOGLOBIN AND ERYTHROCYTES OF THE Bioop in Man Durine SHort Exrosures To Low GOxycen. Harold W. Gregg, Brenton kh. Lutz andslidward | Schneiders. Aaa ee 216 CIRCULATORY RESPONSES TO Low OxyGEN Tensions. Brenton R. Lutz and HdwardiG Sehneiders. ai. ghee 1 Do) oo ee eee 228 XVIII. Conpuction IN THE SMALL INTESTINE. Walter C. Alvarez and PASER CT MS LOT WEQUNET a: AR eae an eke bine) eee Boe 252 No. 3. DECEMBER 1, 1919 RESPIRATORY VOLUMES OF MEN Dourine SHorT EXPosURES TO CONSTANT Low OxyGEN TENSIONS ATTAINED BY REBREATHING. Maz M. Ellis... 267 ALVEOLAR AIR AND RESPIRATORY VOLUME saT Low OxyGEN TENSIONS. Brenton Re Euiz and Hdward (C. Schneiders... ..: 0529) eeee eee 286 CoMPENSATORY REAcTIONS TO Low OxycEn. Harold W. Gregg, Brenton R. ute and -dwandic Schneuder. acta at ee 2 See ee 302 Tue REACTIONS OF THE CARDIAC AND RESPIRATORY CENTERS TO CHANGES IN OxyGEN TENSION. Brenton R. Lutz and Edward C. Schneider.......... 327 EXPERIMENTAL STUDIES OF THE URETER. THE Cause oF URETERAL Con- TRACTION Ss tity ls 1S CLG Beka hae eats Seneecus eink see ee 0. eee 342 Tue Errect or INCREASING THE INTRACRANIAL PRESSURE IN RABBITS. Thalhieyppen il WOU. Veo bs oo bso bes cen eee SfSTo Suextucyc © 2..euNeRes yey ee 352 On THE DISTRIBUTION OF THE NON-PROTEIN NITROGEN IN CASES OF ANAPHY- LAXIS AND PrErTonrE Potsonine. Kambe Hisanobu...............------ 357 Tue EFFECT OF QUININE ON THE NITROGEN CONTENT OF THE EGG ALBUMEN oF Rinec Doves. Ellinor H. Behre and Oscar Riddle...............---. 364 Tue Nose-LickiInc ReErLex AND Its InuIBITION. S. J. Meltzer and T. S. GI CTIS sce te th i SEE ere, | eos oslo eee 377 INFRA-RED RADIANT ENERGY AND THE Hye. MM. Luchiesh................. 383 STUDIES ON THE CONDITIONS oF ACTIVITY IN ENDOCRINE GLANDS. V. THE IsoLaTED HEART As AN INDICATOR OF ADRENAL SECRETION INDUCED BY Pain, ASPHYXIA AND EXcITEMENT. W. B.Cannon........- SRM USO c 399 Errect oF Work AND HEat oN THE HypROGEN ION CONCENTRATION OF THE SWEATS, Ge Aw MOLWerl acme ee de ree tee: oe er 433 EFrecrt oF PHysICAL TRAINING AND PRACTICE ON THE PULSE RATE AND Buioop PressurES DurineG Acrivity AnD Durina Rest, with A Note on CrertTAIN AcuTE INFECTIONS AND ON THE DistrRESS RESULTING FROM BxmRcise, /Perey Dawson. oh a fonat o. 61 eee ee 442 CONTENTS No. 4. January 1, 1920 CARDIO-VASCULAR REACTION IN THE VALSALVA EXPERIMENT AND IN LIFTING witn A Nore on Parturition. Percy M. Dawson and Paul C. Hodges Some Aspects or THE Neuro-MuscuLtar ReEsprraATORY MECHANISM IN Guin NIPAINSen ME CLETE) GeRGOOTIOSS de... deca eiwnce: are aut tie. a } ‘ ae New cee “RA - olzeyg us ‘ 1 lee La re ee je RR ‘ THE AMERICAN JOURNAL OF PHYSIOLOGY VOL. 50 OCTOBER 1, 1919 No. 1 ACIDOSIS DURING STARVATION HAZIME ASADA From the Institute for Forensic Medicine of the Tokyo Imperial University, Japan Received for publication May 15, 1919 INTRODUCTION The physiological effects of starvation have been investigated by numerous observers and many papers have been published on the subject. Castellino noted a decrease in the alkalescence of the blood in starving rabbits in 1893, and in the following year Tanszk observed the same phenomenon in Succi’s blood during a period of starvation. This result was confirmed by the careful studies of Benedict (1) in 1907. London also observed a slight decrease in blood alkalescence in starving rabbits. The appearance of acetone bodies in urine and expired air, which is characteristic of the late stages of starvation, has also been extensively studied. The output of carbon dioxide is said by various authors to show a steady reduction throughout periods of inanition. Thus Benedict reports that the carbon dioxide content of the alveolar air was subnormal on the second day of a fast. There- after it held constant until the fourteenth day, when a second drop occurred, after which it suffered no further reduction. In each case the fall in tension amounted to about 4 mm. Hg. Thus it may be inferred that a tendency toward acidosis developed in this individual on the second day and that this tendency toward acidosis was accen- tuated on the fourteenth day. We know that the hydrogen ion concentration of the blood rather than the carbon dioxide tension is the predominating factor in the control of respiration, and also that the acidity of the blood may be divided into two parts, one due to carbon dioxide and another one 1 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 2 : HAZIME ASADA due to other acids. As the total hydrogen ion concentration necessary to stimulate the respiratory center must always be the same, it is readily seen that if the other acids in the blood increase in amount, the tension of carbon dioxide will decrease. Since alveolar carbon dioxide tension represents closely the carbon dioxide tension in the arterial blood, it affords a good index of the acidity of the blood. Therefore this index may be much more satisfactory and important than the urinary tests for acidity. Van Slyke and Cullen (2) have recently adopted the plasma bicar- bonate as the ideal measure for the alkaline reserve of the blood; a criterion much better than the measurement of the alveolar carbon diox- ide tension for the purpose of determining “‘acidosis” in the modern broader sense. For this determination they published a new method and designed a suitable apparatus. TABLE 1 Influence of two extractions DASE) OF BAS Diet crstolereisleles ee iejetee Clorlefaieiate sete 3 7 ) 11 ANIMATSNUMBDR2 -setenes «0 csitntive whic peice 7 9 16 12 25 CO sinalice dartonhlinelaain eas 0.3642 | 0.3522 | 0.3369 | 0.1723 | 0.2204 ; : e mgm..| 0.7150 | 0.6911 | 0.6616 | 0.3377 | 0.4328 CO» of the same animal pre-{ec....| 0.5170 | 0.4028 | 0.4112 | 0.3960 | 0.3268 Ceding (daypn. locas sat on \mgm..| 1.015 | 0.7902 | 0.8071 | 0.7778 | 0.6419 TABLE 2 Influence of second extraction on moribund animals DAY SVOR MAGS. ¢ ctoyoc'ssay stole cj svetsichel sinntojaes siereraie eiotyevele nis eave ieeiaie'e arslarcie te mevsic 5 14 20 ANTMAL) NUMBER ssc atest erste ce Oe ae + Oe Oe eee Oe eb ee 11 28 17 CON itis ooh Mca so. 0.4486 | 0.4280 | 0.4468 C02 01h ek a: | 9h eas 0.8812 | 0.8406 | 0.8770 TABLE 3 Arterial plasma bicarbonate in short time after death DAVAGOF (WAST ie gas cies ear Dee ee eee 12 i4 15 18 ANIMAT, NUMBER so: 5,015 001s fle + ec ee OE eee 2 26 23 27 { Ge. stetcene prqany AJ@AIs -u9}Uut Weeks L Glen esta ouT[ByB ApWysyg | onojyoyduy Penest Pees prqany por DUTULIB ouyPes]V i=} Pi SUTUIG4S 19} jv Imoy D nog, pot ; OUTULIBO prqany pot prqiny ATIYSYS | Iwao poy OUTULIGS) ATIYSIIS prow plow ATIYSTIS poy Apysyg | oejoydury sdure.1o WON SNA) POEM surureys I94ye P2T]IM Ajo} VIpouUr] Rent SUTUTG4S Iayye samoy GI moqy prqany Bet ‘O40 “IOTOD dUI[BY[V AyVYSIS UOTPVOY Be TEST Vv Jappey{q UuLDUNIQ es saia 26 HAZIME ASADA Since the intraperitoneal administration of 0.19 ec. of 5 per cent lithium carmine per 100 grams body weight for six days gave incom- plete vital staining, the dose was increased to 0.358 ec. per 100 grams, and continued for five days more. This yielded a satisfactory stain but left the animal moribund in extreme acidosis and loss of body weight. 3. Vital staining of animals previously made acidotic. In order to call forth acidosis, I injected a slightly acid or a slightly alkaline solution into the ear vein or into the peritoneal cavity of a rabbit or a mouse. It was possible then to vitally stain the animal to a very high degree with a single injection of the carmine solution. a. Vital staining of rabbits after preinjection of alkali or acid. To this series of experiments, 8 rabbits were used, classifying them into 3 groups; to the first group a tenth or thirtieth normal solution of hydro- chloric acid was administered into the ear veins, to the second a tenth or thirtieth normal solution of sodium bicarbonate, and the third remained without any preinjection. As the blood was very liable to coagulate in the presence of acid, I could not easily inject a sufficient amount of the tenth normal solution. A concentration of a thirtieth normal solution was relatively easily administered. It was the same in the case of alkali. After these pre- injections 5 per cent lithium carmine was administered intravenously in intervals of from 30 minutes to 24 hours. The animals thus stained vitally were killed by means of bleeding from carotid at various times after the final injection. At autopsy nearly all the glandular organs were carmine red, but the adrenal was almost always yellowish. In bladder the urine was very interesting: In the urine of No. 7, which was killed about 2 hours after the preinjection of 45 ce. 70 HCl and an hour after the intravenous administration of 8 ec. lithium carmine, carmine granula and bladder epithelia with carmine red nuclei were found abundantly under the microscope. The urine was slightly turbid, slightly acid and of carmine red color. The urine of no. 8, which was killed also about 2 hours after the preinjection of 54 ce. a NaHCO; and an hour after the intravenous administration of 9 cc. lithium ecar- mine, was also carmine red, turbid and of alkaline reaction, and con- tained no carmine granula nor cells with stained nuclei. The urine of no. 4 which was killed about 24 hours after the final preinjection of alkali and carmine administration, was carmine red, turbid and slightly i) “IJ VITAL STAINING AND ACIDOSIS acid. The urine of no. 6, which died with cramps immediately after the carmine injection to the amount of 8 cc. and 30 minutes after the third preinjection of a thirtieth normal solution of hydrochloric acid amounting to 40 cc., was slightly turbid, slightly red and of amphoteric reaction. The urines of those which were killed about 40 hours after the second preinjection of alkali or acid and about 20 hours after the carmine injection, gave, in general, a reaction the opposite of that of the preinjected solution. The urines of those which were not pre- injected with acid or alkali before the carmine staining, remained slightly alkaline or amphoteric. Generally speaking, in a short time after preinjection of acid or alkali, the urine is of nearly normal reaction, but 24 to 40 hours after the preinjection its reaction becomes quite the opposite of the preinjected solution in spite of the alkaline reaction of the carmine which was always administered afterwards. TABLE 5 Distribution of carmine in body organs of the preinjected rabbits after vital staining RABBIT Teel ie Ie iret sie Noe. di peso core a5 Distribution of carmine in: | MENTE cco > eee Oe Te ORE fy +/+} —(/+]/+/+/+4)+ IRUIGINN Ga a els S Ol aackonate tlre sea —}|—] +4 —} —}] +! -] - Spleen: 3 koe nck Prete oe eireys at —/—}—|—]—-]—-!—-]- 1 BPS Ta ee amie PRS sibel fee on) en ae ae Oe ee _ -- _ TSC EAR eee emetic aac res 10 haa Bare - = — This fact is very interesting in connection with explanation of car- mine-acidosis, because the repeated injections of alkaline carmine solution may of itself cause an acidosis in the experimental animals. At autopsy there were no special changes aside from the carmine staining of the internal organs. But on histological examination of the tissues I found nearly always cloudy swelling, coagulation, vascu- larization and fatty degeneration of different degree in the liver, kid- ney, spleen, and some other organs, without any conspicuous differ- ence under the three groups. With regard to the distribution of carmine granula in various organs also, as shown in table 5, I cannot find any special difference under the three groups. In the kidney of no. 6 which died immediately after the carmine injection, no carmine granula were found, but in that of no. 7 and 8 which were killed in 1 or 1.5 hours after the vital staining, the HAZIME ASADA 28 ea a ‘ouo [wouoyodvasyur ‘d puv UoTpolUl SNOUBARAZUT SOYROIPUL “A UOTPBIADIGGB OUT, “| “N = = = ‘say Z a0 A {NOUN Lg = eer], ze ‘say ST go d OYA 83 yods SE | ++ | 04} U0 pwag | ¢o A JNO G aE aor ae sr th ‘SIY Z god |-say¢t | god | ‘OOHeN 2 | Te at anal ate 4 oF "SI Z god |‘suyygt | god | *OOH*N 2 | 08 “Ie ae ar ts a3 “SIU Z ¢o-d | ‘say gt god | *OOHBN St | 6% oe = ‘say G°T god 8 | §0°d | *OOHBN SE | 22 = OO LL = ‘sIy ¢°T ¢'o°d 08 | ,¢0°4 | *OOH*N | 92 = = "sy § To°A 0§ | ¢0'd | *OOH®N Ny | 6 = 7 "sat § To°A 08) |, S200 0S | 2OORNEs als oo = "SAU § 10's 08) $04 | *OOH®eN | @ ze = "Say § TOA oe) Sod | *OOHeNEx | T = Se ene = = 0% 104 106). 7050 IOH x | a = a = = 106 Oe 06 | God lone | 07, S 3 a = 3 0% To A 06 | god OSE a 7 oF at = + — “SIU Z P(g) eG PSIG Ti 1G" () Xl OH 2% ee = OVAL, = oF = "SIU Z god |sayeT | god OH. | ace = oovdL, = "SIU % god | skepy, | god [OH x | 61 = = “Sa GT ZO A 0e | god OH * | 2 9081], = ‘sIy G'] ZO 0g | god IOH 9 oOBLI, oOBLT, ‘SI § T'0°A AUS esp IOH p a = Sal § L074 0g | god IOH g 09 nal AoUpLyy JOAUT LOAUT UNIO} UOBSO TAY pool¢T =. a ; F Ss qunouwy pul . UYZBLe Iwoulg ua a eG? IVa NIVAS AVAWUENI = an -Idvyoud WO ANIL — —— —— WSoow GQNIWUVD JO NOLO GIMISIG NOLLOULNIAUd yoyyp Lo prov fo uoyoaluradd sajyfp aod fo Bururvjs 7p} 4 9 WIAVL VITAL STAINING AND ACIDOSIS 29 staining was conspicuous; abundant in bases of epithelia of the tubili contorti of no. 8 and in traces in lumina of urine-tubules of no. 7. b. Vital staining of mice after preinjection of acid or alkali. Since rabbits require a large amount of carmine I resorted to the use of mice in order to spare the stains. Their weight was 10 to 15 grams. The preinjections were carried out to the amount of 0.5 cc. for the most part into the peritoneal cavity, and the stain was generally injected into the tail vein to the amount of 0.1 or 0.2 ec.; the intravenous ad- ministration of 0.3 ec. of the stain often killed the animals at once, but the intraperitoneal injection of 0.5 cc. was not so harmful. In from 20 minutes to 3 hours after the vital staining the mice were killed by decapitation, and immediately smear preparations on cover- glasses were made from blood and liver on the one hand, and on the other hand pieces of mesentery were stretched out between two cover- glasses. These preparations were stained with hematoxylin only or combining eosin thereto, after their fixation in ether and alcohol. Also specimens of liver and kidney of most of the mice were imbedded in paraffin, cut and stained with hematoxylin and hematoxylin and eosin. Microscopically I found no special difference between the two series, i.e., the animals treated with alkali and acid., Cloudy swelliag, coagu- lation, pyenosis of the nucleus, karyorhexis or karyolysis were found in nearly all preparations. But as shown in table 6, the animals treated with alkali (nos. 29, 30, 31). took the stains conspicuously as compared with those treated throughout in the same manner except for the pre- injection of acid (nos. 32, 33), in which only a trace of the stain was found. In the other animals even the carmine distribution did not differ. Those which were vitally stained by intravenous injection of less than 2 cc. carmine within 30 minutes of the preinjection nearly always gave negative results. Those which were vitally stained by intraperitoneal injection of 0.5 ec. carmine an hour and a half after the preinjection, on the other hand, gave positive results. DISCUSSION From a perusal of the above facts it may be readily seen that vital staining with lithium carmine produces an acidosis, and that if acidosis be established as a preliminary to the injection of lithium carmine, the staining is much more conspicuous than in animals not so treated. The relationship between acidosis and vital staining is difficult to explain. The current conception of cloudy swelling (5) is helpful. In vitally stained tissue the association of the carmine granula and cell edema (cloudy swelling) is generally recognized. If the edema is 30 HAZIME ASADA developed by a rise of intracellular osmotic pressure due to an abnormal splitting of the cellular proteins in the absence of an adequate oxygen supply the carmine solution may be carried in with the production of edema. It is conceivable then that the carmine granula are precipi- tated in the acid medium, thus completing the histological picture. This will explain the fact that carmine granula are found only in cells with easily permeable walls and which are exposed to a slow blood stream, e.g., the reticulo-endothelial cells or histiocytes, while these granula are never found in those cells which are constantly exposed to alkaline body fluids or are bathed in a rapid blood stream where water and carmine diffuse in hardly more rapidly than the crystalloids diffuse out, e.g., the cells of the brain, bile ducts and greater blood vessels. We may account, similarly, for the ease in staining surviving cells as opposed to living cells. If coagulation developed as the result of in- creased hydrogen-ion concentration before the stain could enter, there would be no deposition of carmine granula. Further, even after the cell walls lose their semipermeable character, the nuclei, possessing a relatively high hydrogen-ion concentration and still permeable to the dye, will take the stain, which has already entered the cell, and make the nuclear figures distinct. SUMMARY Vital staining can not be accomplished by a single intravenous or intraperitoneal injection of lithium carmine in subtoxic doses in a healthy animal. If, however, an acidosis be first established, a single injection of the dye will give a satisfactory stain. In animals which have been vitally stained, an actual decrease in plasma bicarbonate occurs. Hence the conclusion is drawn that it is incorrect to predicate the existence of specific stain-taking substances or granula in the cells. Rather, vital staining with lithium carmine is due to the development of an acidosis which so alters the function of the body cells that the dye diffusing in is deposited in granula. This deposition corresponds to the precipitation from colloidal solution of the dye when the normally alkaline solution is made acid in vitro. BIBLIOGRAPHY (1) Kktyono: Die vitale Karminspeicherung, 1914, 7. (2) Ktyono anp Karsunuma: Kyoto Igaku Zassi, xv, 6. (3) VAN SLYKE AND CuLLen: Journ. Biol. Chem., xxx, 289. (4) Asapa: This Journal, 1919, 1, 1. (5) Wetus: Chemical pathology, 1918, 394. STUDIES IN SECONDARY TRAUMATIC SHOCK IV. THe Buoop VoLUME CHANGES AND THE EFFECT OF Gum ACACIA ON THEIR DEVELOPMENT HERBERT S. GASSER, JOSEPH ERLANGER anp W. J. MEEK From the Laboratories of Physiology of the Medical Schools of Washington University and the University of Wisconsin Received for publication June 19, 1919 In our earlier shock studies the hemodynamic findings all pointed directly to a decreased effective blood volume as the only constant factor tending toward failure of the circulation. Attention has also ~ been called by numerous other observers to the significance of the reduction in both the effective and absolute volumes in this condition (1). The present experiments were designed to determine the rdle that the absolute volume of the blood plays in the reduction of the effective volume in shock, and to evaluate the possible modes by which such a reduction might be brought about. Three possibilities suggest them- selves as to ways the blood volume may be reduced: 1, hemorrhage, external or into tissues; 2, concentration of the blood, i.e., by filtration of the plasma; and 3, stasis in a portion of the vascular bed. Four forms of experimental shock were studied, namely, those pro- duced by injections of massive doses of adrenalin; by clamping the vena cava above the liver for a period of three hours so that the general arterial pressure was 30 to 40 mm. of mercury; by clamping the ab- dominal aorta above the coeliac axis for a period of three hours so that the distal pressure was 30 mm. of mercury and by exposure and manipu- lation of the intestines. Dogs were used in all experiments. External hemorrhage was absolutely excluded except for the small amounts of blood necessary for the samples. The blood pressure was determined by a small bore manometer connected directly to the artery. It was seldom necessary to ‘‘wash out’’ during the course of an experi- ment, so that the loss of very few cubic centimeters of blood was involved. , Samples were taken directly from a large artery through a clean dry cannula. The red blood cells were counted as an index to the dl 32 H. S. GASSER, J. ERLANGER AND W. J. MEEK relative plasma loss and the results obtained, therefore, indicate the minimum loss of blood through filtration of plasma as any local reten- tion of the erythrocytes would decrease the count. In the later experi- ments the hemoglobin changes were determined by means of the Dubosque colorimeter, the blood being diluted 1/100 with 0.03 per cent HCl, and the normal hemoglobin dilution being considered as 100 per cent. In case blood is withdrawn from circulation either by hemorrhage into tissue or stasis in a portion of the capillary bed, such a reduction could be appreciated only by a direct determination of the blood volume. Accordingly such determinations were made by the method of Meek and Gasser (2). Determinations may be made by this method to an accu- racy within 5 per cent. Four cubic centimeters per kilo of 20 per cent acacia were injected intravenously, and ten minutes after the injection a sample of blood was taken for the distillations. All determinations were done in duplicate. If stasis occurs it might be of a degree varying from a local retardation of the stream to an absolute removal from active circulation. As ten minutes were allowed between injecting and sampling, the acacia would be diluted by all the blood that passes through the general circulation in a period of ten minutes. : In this series the final shock determinations were made after the mean arterial pressure had fallen to 50 mm. mercury. This was the conventional level chosen as a criterion early in our experience with shock before we came to appreciate that the attributes of shock usually develop long before this level is reached. ADRENALIN SHOCK The first studies were made on adrenalin shock. The normal vol- umes were determined and then the volumes in shock (table 1). Every case showed a reduction in blood volume; in two cases this reduction amounted to one-third of the total blood, in another case the blood volume was reduced by half. The blood counts on the other hand indicated a relatively small concentration, a finding that occasioned some surprise in view of the large concentrations obtained by Lamson (3) and by Bainbridge and Trevan (4). This discrepancy in results was traced directly to the acacia injected for the determination of the normal blood volume. One experiment was, therefore, performed in confirmation of the above authors. Four injections of adrenalin were made as indicated in figure 1a; with each of the first three there occurred BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK 33 concentrations of the blood amounting to 29, 32 and 45 per cent of the normal respectively. In the intervals of lower blood pressure be- tween the first three injections the red cell count returned to a con- dition in the neighborhood of normal. Each succeeding injection produced a concentration greater than the preceding although the adrenalin was given in smaller doses. Each injection must therefore have left a residuum of damage which became such after the third injection that the plasma did not return to the vessels as before. It NIN 2 + 10 7 2 — 2 3 4 Fig. 1. (a) Lower: without acacia; (b) Upper: after the injection of 4 cc. per kilo of 20 per cent acacia. Light line: blood pressure in mm. Hg., indicated on the axis of ordinates. Heavy line: R. B. C. count in millions, indicated on the axis of ordinates. Axis of abscissas: time of day. Adrenalin injected at times indicated. Figures give numbers of cubic centimeters of 1/1000 injected. is probable that only in this latter condition was the damage to the tissues sufficient to be spoken of as true shock. A similar experiment was now performed in which 4 ec. per kilo of 20 per cent acacia were administered before adrenalin in comparable amounts was injected. The results (fig. 1b) show that the acacia not only has a marked effect on the final concentration attained but that it has an even more striking effect on the ebb and flow of plasma which THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 H. 8. GASSER, J. ERLANGER AND. W. J. MEEK 34 pword Ay yooyg ‘sopdures ul poAoutat Poorq *99 09 {}UNOD 4say 104 jB sanoy %¢ uorjoofur ulpeusIpy SuLoBly YOoys [Borda y, “qAvoY AB[NGa1 YIM [Pp oInssolg qaeBoy oy} JO osussvul Aq soyn -UIW QT penuryu0d uOoTye[NO “MQ “UOTPeUTUTIOJOpP oN -[OA Yooys oF uoTyoolur 104 -J@ soynurut Qe poddo4ys 4avoyyT UIOA dlloJUNSOUL GB O}UL poyooful SUM Ul[VuoIpe syuoUTIedxo G Sutpeooons oy} puw sity} uy ULOA [RLOUOJ OU} oyur uoryoofuy =“ uoOsTavdur09 10} Surpoosdons oy} 0} uor -uvduio9 & st yuowt1odxo sty, SYUVNAY Ores? | OG7OH | BG aeGice | Chen ley an) O00rTZ 2 el 000'9G5 On| ace | Sz'4| 2°00 | 2°6 Iz8'¢ (919 | 0009179 | O00‘9G8'¢ | AT J; Blowow TIA a 0O'SS | GP \9G'6 ) ) 0) 000.008‘ QOO'SOS'S | III 6'e Iceo | Fs | O000'089'F ooo'ozs'F | IT SLT | 2°01 | O'S | O00'9S6'L | 000'960'2 | I J *xBvul (0) SF ; Z6L | 89°L \x$°6 | 0°68 | F'E% | FO | 0007088 ooo'zez‘9 | I VINBOV NOY} “VW as) n A 3 a - 3 o (338 |828 | see osee) § @ esh | See | See ieee al ¢ 4 a5 y a5 a B SI 6 |e 's SI 6 4 MOOHS TVINUON 9 5 #9 S ae ae ae Be FI “INDOO'D a ‘UH | INNO ‘0 “a “U4 ro) a ie) 5 8 5 fo) e [P. (282 | Eel oeal & B | g&| e&| S| 2os| « Uuypuaspy fo Uotalur snouarmuur fig paonposd yoaoys’ T WIaVvi 35 BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK ‘SSOP JO SOLIOS BUC] GV JO SOUINIOA [VULLOU oY} JO VFVAVAV of} ST YOU ‘qys1om Apoq oy} Jo yuo aod 7°6 oq 0} potUNsse SI 41 poUlWIo}Jop JOU ST 9UIN[OA [VUILOU OY} OLOY A, “poyjow viovor ayy Aq SUIN[OA POOTY OY} JO SUOTFVUIULIOJOP JOoIIp oY} UOAIS oiv SUTUN]OD 9a1q} YSV[ OY} UT “pooyrq oy} Jo ournjfoa oy} Jo yuo aod 09 soynyTysuo. Vuise/d oy} yey UoOIYZdUINSsv oY} UO poyvUIT}SE SI oSvaIOOp vUISL[d oY, “yUNOD [[o9 poorq pot oy Aq pouryy -48 |vUILOU Jo vse UedI0d oY} JO [BIOIdID9I1 OY} SV PpoINSY st oUM[OA poorq Ur osvo1Iop OY} So[qv} JUoNDosqns puv Sty} UT y ‘Sy cur GG OMSSXIG ‘“Ul[RUTIpe Jo uorpooful Yyanoy az0qyB soyn } Blow’ YLAA “A -UIUT €Z SUOT}VUIUIOJap YoY | Y°2 Cao |8GeG | 8eG 4) (Glue AO ve 000 FZ9'E 00096 S | ITA JIvOY IBlNBIL1T Gable Gi males 000'926°¢ | oOO'zss’S | IA 36 H. S. GASSER, J. ERLANGER AND W. J. MEEK result from the pressure changes, the concentrations being in this case 4, 8 and 11 per cent of the normal for the successive injections. In experiment III of the acacia series a decrease in volume of 53 per cent by transudation alone would mean the loss of practically all the plasma. Obviously some other factors are involved in so large a depletion of the blood volume and of these, absolute stasis in some part of the vascular bed must be the most important. While the red cell count does not indicate any loss of plasma at all, this absence may be only apparent as the red cell count determines the minimum of concentration and the determination is misleading in proportion to. the number of corpuscles that are jammed in the capillaries. In ex- periment IV, which was very satisfactory, a large difference between the absolute loss and the loss by concentration again occurs. That the difference is not merely apparent but may be due mainly to stasis, may be inferred after comparison with experiment VII in which, with a very small decrease in blood volume, such a degree of shock developed that the pressure was 22 mm. of mercury. Here undoubtedly the defective circulation must be attributed to stasis, which however falls short of being absolute. In all but two of the experiments included in the table the adrenalin was injected into one of the small mesenteric veins. For this purpose care was taken to expose only one small loop of intestine. In these experiments curious localized constrictions of all branches of the mesen- teric veins appeared which were permanent in position and gave the vein a sausage-like appearance. This phenomenon is probably one of the factors accounting for the loss of blood to the circulation by stasis. SHOCK FROM CLAMPING THE AORTA It was evident from the experience with adrenalin that in the other experimental procedures for producing shock the volume changes should be determined both without and with a previous injection of acacia. When no acacia had been injected clamping the aorta for three hours (table 2) produced a quite constant amount of con- centration. As indicated by the red blood cell count the loss of plasma amounted to an average of 37.0 per cent of its original volume. On the contrary when acacia had been previously injected the plasma losses were 0 per cent and 8.8 per cent in two cases. Exactly similar results were obtained in the experiments in which the hemoglobin content was used as an index of the concentration (table 2 B., expers. suory “BUIIO}Op “CH wWory PV MEOrG. | LNG > A: rt suor} . * -BUTUOJOp “GY worl chz [ore Ions AI 0'F% | 99°24 | O'OT Ill BIOvoe UM “AT poyoolur sem viovoe a10joq UMBIP Pood ON | OTE 19 $8 |}€¢ |9¢ | 0002662 | 000'39¢2 ~) IT i pesowos durep> uayar & *oys punojoid ur soq 8'0F | G9'°¢S [FG'6 | O 0 0 000'FOI'9 | O00'OFS'9 | T J DM o suoly -& -suruuejep “qy wory vO | OFT | TI A : = B}L0B UOJ POAOUL 4 -91 SBM CuIe[D oY} 910} & -oq uoyn} svar qunod ; be eo “Bsoonur [eury FH --So}UT OFUT oTvYyALOWOHY | G°ST 62 1°98 | 0% | €°8Z | 000'S9Z‘6 | O00'S09'Z | AT a ourysoqul A ul oyVpnxo osRyI10 fe) t I ul ia iS -Woyoonul SNOUIUL Teak ae 2 -njoA “YBop ovIpavo | Mc area a q JO qJuNODDB UO poUTRy, | -qo you oUINjoa pool g6E | 4°82 | O'1S | O00'9EL'6 | OOO'FeHL | TIT || TS YOOYSs UL ION | a ‘puo 4B JusuToAO1d > -Ull jo “susIS poMoyg | PEI v8 ee | F'6L | LFS | OOO'FOK'G | OOO'FZI'Z | IT Q G6 | 4°8 | 8'0E | 000'SFO'S | OOO‘OST'D | I : UTE UG let er nec: fq 4 wd wd < od coal co |$se|Sse/ fee lzees| & Siaice| | JBge lua Hob | Seek |ocke| & SyUVWaa chan] 5 458 |48u)/ e's [Bato] 4 yoous __‘IVINUON Bo 8 PU Oa eu 26 lewd I “IND09 ‘0a “WU | \ENN00 ‘0a “u : ep (Gee lmee| sei eos &¢ By H2Z9g e: ioe] Bg tee seal ean oe Olga | ee Sel eaic cee iG Bil aia ere alg ee ee ee OS ee eee ee ye Oe } o) npowd yooyus’ 38 H. S. GASSER, J. ERLANGER AND W. J. MEEK IV and V). Here again, as in shock produced by injection of adrenalin, the large differences between the actual decrease in volume and the apparent decrease due to loss of plasma, amounting respectively in two experiments to 40.8 per cent and 20.7 per cent, point to absolute stasis as the principal factor in decreasing the effective volume in these animals which had received acacia. In figure 2 are plotted the red cell counts obtained in two experiments performed for the purpose of comparing transudation under as similar conditions as possible except for the preliminary acacia injection in the one case. The concentration in the case without acacia was more than four times that in the control. UNCLAMPED (UNCLAMPED MIL. Hr. 10:09 14:09 12:00 100 Fig. 2. R. B. C. count in shock produced by partial occlusion of the aorta. In the experiment indicated by the lower line 4 cc. per kilo of 20 per cent acacia were injected at the point designated. SHOCK FROM MANIPULATION OF THE INTESTINES In both of the two control cases (table 3), exposure of the intestines produced large concentrations of the blood. In these experiments the transudation of plasma could be observed directly on the peritoneal surface of the intestines. Beads of plasma would appear, grow larger, coalesce and start to run from the surface. In one experiment in which the drippings were collected, about 150 ce. were obtained. At the period of maximum transudation fluid was lost at the rate of 1 ce. per minute. This same phenomenon to a less degree can be seen also where shock is produced by other methods. If the abdomen is care- fully opened droplets of serum are often seen on the peritoneal surfaces, especially of the liver and spleen. 39 BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK ‘sojduies Jo SuIMvip snorAdad yNoyyIA poz olur vIOvde 107ze SUIUTe}GO sIN[OA UO poyenoye,) { ‘uoryeunxoiddy , rer | 0°92 | O'Ge | OOO'Ssr'Z | OOO'9ES'S | IIT loo. | t'6 | 2:02 | T'2E | 8°gt | 000‘0z6.S | O00‘;O0TE | II BIOvIv YUM “A 40'S 9°9 |c6's | TF | OO0'FOS‘'S | OOO'S9T'S | I . 5 . . . ‘ ¢ ¢ ¢ wor}esyueoU0D Aq Ajureul ssoT | 2°2S | S°L Ti | 2°72 | 0S | 000 7069 000 Z6I ¢ Tal BIO . . . . 5 ‘ ‘ ‘ E . S[OSSOA [BUI}JSOJUI UI SISVyS 9[qBqolg | OOF | 8G 9'cr | F'S2 | 28 | 000TES 000 99 9 if -B0B JNOUJM V ibe ealeace eae qaer see |e28|ses ogee] § sea | ahh | eee Zac] & n n ; ean | 2 ae Bilis a Bag 4 MOOHS TIVWHON SuUVWaY pe | emo 32 [bu 04 FI ‘ENQ00 ‘0a “U | “\ENNOD ‘9 ‘aU 2aq90 ig iS) Oy 4 S ro) to Rad mz Sw Z i eo ERAS >a! anal & Oy 4 S Dey HOY ia ° ° s PAS) oF af po BOF 5 sauysajur ay, fo aimsodxa fig paonpoud yooys € HIaAVL 40 H. S. GASSER, J. ERLANGER AND W. J. MEEK In this form of shock the difference between the experiments with and without acacia is not so marked. This may be due in the greater part to the fact that the damage produced can not be well controlled. The rapidity of the onset of shock in strong healthy animals varies with the severity of the manipulation. In one dog in the acacia series the concentration was maximal, in the other two the concentration amounted to less than one-half of that occurring in the controls. If we assume that the intestinal conditions are about the same in animals in which approximately the same time elapses between the exposure MM HG. 150 100 50 150 100 8 50 “NMIL 12 { 2 3 4 5 6 U & Fig. 3. Blood pressure and R. B. C. count in shock from manipulation of the intestines. Upper: without acacia. Lower: after the injection of 4 cc. per kilo of 20 per cent acacia. of the intestines and the intervention of shock, then comparison of two such cases (fig. 3) shows that there is a considerably longer interval before the plasma loss becomes appreciable if acacia has been injected previous to the intestinal manipulation. SHOCK FROM CLAMPING THE VENA CAVA Four cases are included in this series (table 4). In all a concentration of the blood took place, the degree being somewhat greater in the two animals unprotected by acacia. 41 BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK “UOTPIILUT BIOBVOV 10JJw OUINJOA UO poyzR[No[VyD | UOT} BIYUIIUOD Jo] [BUTS 0} spuodsai109 UWO0Ty -vUIUIOJOp daUIN[OA “(Asdoyny) o1svysto -uloy Ay[vioodsa suvs -IQ “WNUIxeyY = ( qunoo 4sB] 10qye imoy & poimsvour ouINnj[oA “SUIAOICUIL = [BUITUG “SyT “wu (jy oInssolg SMUVNAY 1&3 ANWOIOA NI @Svauo qoo ta -ad qa Lv Tao1vo N asSvauord ING) add G's S2S Tal Sohal) So Sle O00 Oge Z 000‘ 00F'9 06 |86 |8e | ee | Ez | O00'809'9 | 000°9SF'9 (@°6z) |(9' 21) |(€' T@)| (000‘z6S‘6) aw cor} ¢'9 | 89 | 000'SF'S | O00'SO6'L 78 0'ee | 8°61 | OFS | OOO‘OSF'S | O0O'9TS‘D Sesh eealeree| ae wd ba? 4a o4d90 see hele ea gen 4 MOOHS TIVWUON Bini || Sine 34 |b 34 5 ‘INQO9 ‘0°a@ "Mt | “LNA ‘Oa “A Go8/8o8 Sree eom ell ane ef| gf| ER| 4ee| & Il I Il BIOvI’ YIM “A 10) -Bov JOY “V paws vuaa ay) Burdwnjo liq poonpoud yooysy FUTAV SE 42 H. S. GASSER, J. ERLANGER AND W. J. MEEK In the second experiment in series A, the concentration of the blood was progressive until the clamp was removed. After this the red cell count decreased. At the end of the experiment the blood volume as determined directly was 23.7 per cent less than normal. If the circu- lation had been recovering as the cell count might indicate the pressure would not have progressively fallen and the final volume would not have been so low. A more probable explanation is that some of the corpuscles were jammed in the capillaries and therefore caused a de- crease in the count, a view which is supported by the hemorrhagic condition of the organs at autopsy. In another series of nine animals in which the cava was clamped for 2 to 23 hours so that the general blood pressure was 40 mm. of mercury, the concentrations at the end of clamping as indicated by hemoglobin determinations averaged 117.6 per cent of the normal, the individual variations ranging from 110.4 to 131.8 per cent. DISCUSSION In every case studied shock was associated with a concentration of the blood and loss of volume. This was not a phenomenon of severe shock alone, but started soon after the procedure was begun, the pur- pose of which was the production of shock. The concentrations did not always vary parallel to the loss in absolute volume. In the slightly larger portion of the total number of experiments the total loss of volume can in the main be explained by the amount of plasma loss. Examination of the tables shows that in a few cases the shock volumes as determined directly are not quite small enough to coincide with the volume calculated from the red cell count. These differences are without significance as they are obviously errors in procedure and technique and the reasons for their occurrence must vary with the individual cases. The following possibilities may be mentioned: 1, In one case the direct determination was made one- half hour after the red cell count, in an animal that was recovering. 2, When the volume of the blood in shock is determined without a previous determination of the normal volume, the percentage decrease is calculated on the assumption that the normal volume was 9.7 per cent of the body weight, although in some cases it might have been actually higher than the average for normal dogs. 3, It is probable that if we had followed the hemoglobin content instead of the cell count, many of the differences would have disappeared as it has been our experience that the variations in volume calculated from the BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK 43 erythrocyte count were almost uniformly greater than those calculated from simultaneous hemoglobin determinations. Scott, Herrmann and Snell (5) had a similar experience in their studies of the concentration of the blood after muscular exercise. These authors give a discussion of this point with a citation of the literature. In the other group, cases can be found in which the total loss of blood as determined by the acacia method is much greater than that indi- cated by the red cell count. The greatest variations between the total loss as actually determined and that deduced from the concentration were seen in the adrenalin cases in which the injections were made into a mesenteric vein. In this condition the venous constrictions mentioned above appear as a special factor. In three cases where the total losses were 53, 32 and 33 per cent of the total blood respectively, the losses by plasma filtration were 0, 6 and 13 per cent. There can be no doubt that this special condition can only be incidental and does not even account for the differences found in this particular type of shock. As has been already mentioned, in two cases after clamping the aorta there were differences of 40 per cent and 20 per cent of the blood volume between the total loss and the loss by transudation. Similar cases also occur in shock from clamping the cava or manipu- lating the intestine, although the differences seen are not so great. Before discussing the manner in which this blood is lost, it is necessary to review the post-mortem findings. In over one hundred animals examined it was found that no matter how the shock was produced the same general picture obtained. The most constant finding is an injection of the intestinal mucosa. In the milder cases the injection is confined to the duodenum; in more severe cases it extends along the whole of the small intestine giving it the appearance of deep red velvet. In the lower ileum the injection often again disappears although in severe cases it extends up to the ileo-cecal valve. In 60 per cent of the cases blood was found in the intestinal lumen; this was at times due to isolated punctate hemorrhages, more often a general sloughing of the tips of the villi took place. Microscopie sections show that except at injured tips of the villi the corpuscles are almost entirely within the vessels. The capillaries and small veins are greatly dilated and tightly packed with red blood cells. It is interesting to note that in animals that recover from this condition the injection has disappeared. How much this is due to sloughing and how much to clearing out of the capillaries we are unable to say. The spleen is usually swollen and hes dark raised areas which consist of hemorrhages into the pulp. This 44 H. S. GASSER, J. ERLANGER AND W. J. MEEK swelling and hemorrhage in some cases produces a spleen many times the normal size. The liver is occasionally found to be hemorrhagic on section. Where the loss of blood volume as directly determined is much greater than that determined by the red cell count, the difference can be at- tributed mainly to stasis. The red cell count is an index only to loss of plasma, as a result of which concentration of corpuscles takes place. As long as no other factor enters, the blood volume can be properly interpreted in this way. But as soon as the corpuscles reach a certain concentration, 60 per cent according to Trevan (6), they become con- tiguous, the internal friction rises rapidly and they have a tendency to jam especially when the arterial pressure is decreased as it is in these experiments. Insofar as the corpuscles are blocking the capillaries they are not included in the determinations of the red cells or of the hemoglobin content, and the latter fail as indices of blood out of circu- lation by the volume of the corpuscles and the volume of the plasma that would normally accompany them. In addition to corpuscular sedimentation the difference may be attributable to sequestration of both corpuscles and plasma and to blood lost into the intestinal lumen and to a much less extent to hemorrhages into the spleen pulp. In the group in which the total blood loss is to be accounted for by transudation, the congestion of the capillaries and the small veins is again seen post mortem; in these dilated and congested areas the blood must therefore be moving, however slowly. The blood volumes found in severe shock are often remarkably high. Typical shock was found to occur in animals whose absolute volumes were decreased by only 7 to 17 per cent of thenormal. For this there can be only one interpretation and that is an enlargement of the vas- cular bed and, therefore, a greatly reduced effective volume. Mann (7) in his experiments found that in shock from exposure of the intestines a portion of the blood becomes immoblized. On the assumption that the blood volume is 7.7 per cent of the body weight, he found that when all the blood that could be obtained was drawn from the femoral artery and heart, 24 per cent of the blood remained in the tissues of the normal dog while under similar conditions 61 per cent of the blood remained in the tissues of the dog in shock. We have often observed a similar phenomenon. From a dog in shock weighing 16 kilos, which would normally contain about 1400 ec. of blood, only 100 ce. could be obtained although artificial respiration and massage of the heart were resorted to. The decreased amount obtainable by bleeding is 4 Trceceiaeeges ee ee BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK 45 in part due to the decreased volume, but analysis of Mann’s data shows clearly that after allowance for the decreased volume of the blood in shock the actual amount of blood left in the body is more than in the normal dog. These findings are in line with the observa- tions that have been reported by surgeons that fatal consequences may supervene on the losses of apparently trivial amounts of blood. The experimental animals maintain their pressure in the face of de- creasing blood volume to what may be called a breaking point beyond which the pressure rapidly fails and death soon results. In spite of the fact that the amount of blood which may be obtained by bleeding is so small (i.e., that so little of it can be moved from the stagnant area by the remaining vis a tergo after hemorrhage is started), it is still possible for the shocked animal to gradually move the mass of blood in the stagnant area so that an injected substance, as acacia, is gradually mixed with the total circulating blood. The figures for the blood volumes represent all the blood that passes through the heart in a period of ten minutes, and as indicated in the preceding paragraph, may amount to a total not far below normal. While the evidence indicates that the deficiency of the circulation is mainly due to sluggishness of the flow in some parts of the vascular bed, the work of Gesell (8) shows that a decrease of 7 to 17 per cent of the blood volume is itself of significance. He found that a loss of blood by hemor- rhage of less than 10 per cent may elicit through constriction of central origin a decrease of flow through the submaxillary gland of more than 60 per cent. After simple hemorrhage the volume would rapidly be made up from the tissues and the flow restored. In these experiments, on the other hand, the volume-compensating mechanism is becoming ex- hausted and the decrease in volume, though small, is sustained. If the constriction would be sustained, then the reduction in flow might cause damage to the parts of the body compensating the areas primar- ily injured. Robertson and Bock (9) have found that after a large hemorrhage in wounded soldiers the blood volume even when aided by transfusion may not return to normal for 24 days. The organism is, therefore, able to maintain its existence after large losses of blood without complete restoration of the volume for several days. The locus in which the plasma is stagnant is certainly not the arteri- oles or portions of the vascular bed proximal to them. In the direct determinations of peripheral resistance to the inflow of salt solution (10) under a constant high pressure, it was found that while the periph- 46 H. S. GASSER, J. ERLANGER AND W. J. MEEK eral tone was decreased when the pressure falls below 50 mm. Hg, in the periods preceding this low level the resistance may be but very little decreased, normal or even above normal. The condition of the large veins in the abdominal cavity can be easily noted by direct obser- vation. Workers with experimental shock are agreed that in this condition the veins are collapsed and not engorged. There is left, then, only the capillaries and small veins. The full dilated capillaries and venules seen in the microscopic sections from the organs in shock prove their involvement beyond peradventure. It is interesting to note that Mall and Welch found that in the dog’s mesentery after occlusion of the superior mesenteric artery the smaller veins become distended with corpuscles even before the capillaries. An opening up of normally unused capillaries such as occurs in inflammation may also well be a factor in enlargement of the capillary bed, as has been suggested by Cannon (11). If one reviews the whole series of shock experiments and compares the concentrations obtained where shock was induced with and without an early injection of acacia, he finds that the average reduction of the blood volume by plasma loss was 6.8 per cent in the former case as compared with 20.3 per cent in the latter. It might be expected from such data that the animals that receive acacia would be more difficult to put into shock than their controls. There was, however, nothing observed in this series to contribute to this view. While on theoretical grounds it is to be expected that any benign method of interfering with transudation would be invaluable in the maintenance of normal blood volume, it should be fully appreciated that transudation is only one of the attendant circumstances in the development of the shock picture as a whole. That the transudation may be greater than indicated by the cell counts and may, therefore, alter the quantitative difference between the two series somewhat, should not be overweighted as the same phenomenon occurs in both series. To explain the conserving action of the acacia on the plasma volume a consideration is necessary of the factors that might determine its loss. The loss of plasma might be due to a decreased colloidal content, to an increase of capillary pressure or to a change in permeability in the vascular endothelium. The protein content of the plasma was studied in shock from clamp- ing the aorta by determining its specific gravity with the pycnometer and its power of refraction with the Pulfrich refractometer. The latter . method is theoretically more accurate as the salts have a relatively BLOOD VOLUME IN SECONDARY TRAUMATIC SHOCK 47 greater specific gravity than protein but a smaller refractive index. The findings obtained with one instrument in the main confirm those obtained with the other. They show the same sequence of events in each case (table 5). There is in two experiments a slight increase of the protein content up to the time that the clamp is removed; in the other two experiments the increase is absent. Water evidently filters more rapidly than protein at first but after unclamping when the pres- sure is raised in the splanchnic region this is no longer true, the increased TABLE 5 PER CENT OF : N IE EIN IN EXPERI |RBLOOD PRES-|HEMOGLOBIN| SPECIFIC PROTEIN IN ETER) I 105 100.0 1.0256 5.902 | Normal 58 112.4 1.0261 5.967 | After 3 hours clamping 40 113.3 1.0252 5.553 | Final 1a 105 100.0 1.0275 7.049 | Normal 58 111.4 10277, 7.049 | After 3 hours clamping 40 111.4 1.0272 6.768 | Final Ill 180 100.0 1.0284 6.970- | Normal 100 114.1 7.416 | After 3 hours clamping 43 121.6 6.719 | After second clamping 43 1225 1.0274 6.465 | Final ry* 100 100.0 7.675 | Normal Tah 102.1 7.675 | After 3 hours clamping 30 101.6 7.653 | Final * This animal received 4 cc. per kilo of 20 per cent sodium acacia before the aorta was clamped. permeability allows filtration of more protein and this increase is fol- lowed by a change in the opposite direction so that the protein content is finally 4 to 7 per cent less than at the start. It therefore follows that the plasma is lost mainly as a whole (see also Dale and Laidlaw (12) ), and that a change in the colloid content of the plasma is not an important factor in the filtration as the latter process occurs during the period in which the colloid content is increasing as well as in the later period when the plasma is again diluting. The explanation of the dilution in the final period probably lies in the fact that the organism is attempting to compensate the decreased 48 H. S. GASSER, J. ERLANGER AND W. J. MEEK volume. The decrease in the protein content of the plasma would accord with the assumption that the passage of fluid from tissue to blood is by osmosis and therefore consists mainly of water and salts (13). As the loss of plasma is progressive all parts of the vascular bed are not similarly affected; in some portions the normal reaction to decreased volume is possible synchronously with a condition allowing continuous filtration in other portions. The compensation of the loss of fluid by normal tissues was first noted by Cobbett and Roy (14). Our data give the additional information that some areas are still compensating even when the blood is concentrating, though with insufficient rapidity to keep pace with the filtration. The filtration of the plasma as a whole precludes the possibility that one finds suggested in the shock literature that the plasma is depleted by an increased affinity of the tissues for water. In many cases a rise of capillary pressure must be a factor. After adrenalin the observed (15) rise of portal pressure must determine a rise in the mean capillary pressure of the intestinal area. When the cava is clamped there is a slight rise of portal pressure while the venous pressure in the liver which is normally very low must be relatively greatly increased. When the aorta is clamped it is difficult to see where in the splanchnic area the capillary pressure can be increased unless it be in capillaries plugged at their distal ends. Aside from this possibility transudation in this condition must be due to a change in vascular permeability. This latter condition is seen most clearly when the intestines are manipulated. Here, to be sure, alterations in pres- sure may also occur from various mechanical possibilities but the im- portant changes are those which also occur in the inflammatory proc- ess, the analogy to which has been noted by a number of observers. Regardless of the cause of shock, it would be expected that the ex- pansion of the blood volume, which would result from the effort of the organism to compensate the increase of colloidal osmotic pressure of the blood produced by the injected acacia, would aid in maintaining the volume. To test the degree to which this might take place, data on the osmotic pressures developed by acacia and blood serum are necessary. Osmotic pressures of acacia and blood serum. As Bayliss has correctly pointed out the determinations should be made against Ringer’s so-_ lution. The two variables, acidity and CO, content of the solution, must be controlled. oun | 0.419 | 0.45 Regenerated parts cut off 23 days since first section; several hours | 0.26 in 2 hours 0.308 0.48 since second section 0.34 in 2 hours- * Whenever practicable, two separate determinations of the oxygen consump- tion were made on each occasion. 0.5 gram. The worms were then cut into small pieces and in several determina- tions taken during the first eight hours after section, the oxygen consumed ranged from 0.29 to 0.36 cc.—a distinct increase over the preceding figure. On the fol- lowing day the oxygen consumption was about the same —0.31 to 0.35 cc. On March 31, four days after section, the oxygen consumption had fallen to 0.27 to OXYGEN CONSUMPTION DURING REGENERATION é i 0.29 cc. On April 4, eight days after section, it was rising again, being 0.28 to 0.31; and on April 12, sixteen days after section, all of the pieces having undergone complete regeneration, the oxygen consumption was found to have increased greatly, having risen to 0.53 to 0.55 cc. The regenerated worms were now fed four times, on April 12, 13, 15 and 17, and on April 19 (the worms thus being in TABLE 3 Record of lot 8, Temperature 21°C. CALCULA- TION, AVERAGE AMOUNT OF CONDITION OF WORMS en ee IN) weieut | oxYGEN CONSUMED IN 2 HOURS BY 0.5 GRAM ce, grams ce. Decapitated, intact, one day since feed- | 0.37 in 14 hours | 0.952 0.26 ing 0.39 in 13 hours Cut into small pieces 1 to 8 hours after section 0.40 in 14 hours 0.46 in 14 hours 0.914 OFot 20 to 22 hours after section 0.48 in 14 hours 0.813 0.38 2 days after section 0.45 in 13 hours | 0.755 0.37 0.40 in 13 hours 7 days after section 0.25 in 13 hours | 0.598 0.27 0.24 in 14 hours 15 days after section 0.35 in 2 hours 0.435 0.39 0.33 in 2 hours Worms fed three times 0.48 in 2 hours 22 days after section; 1 day since feeding 0.46 in 2 hours 0.502 0.46 Regenerated parts cut off 23 days since first section; several hours since second section 0.33 in 2 hours 0.28 in 2 hours | 0.369 | 0.41 the same state regarding nutrition as the original lots), the oxygen consumption was found to be 0.56 to 0.62 cc., aslight increase. The regenerated tissue at both ends of the worms was now cut off and two days later, on April 21, the remaining portions, corresponding as nearly as possible to the original pieces, were tested. Their oxygen consumption was 0.53 to 0.63 cc. The temperature throughout the experiments was 22°C. + 0.5. we L. H. HYMAN Record of lots 4 to 6 (table 1). The worms in this lot were collected during the early winter of 1919, and were large and well-fed specimens. They were last fed on April 7, their heads cut off on April 9, and the decapitated worms tested on April 10. The oxygen consumption was 0.26 to 0.27 cc. They were then cut into small pieces and within the first few hours after section, their oxygen con- TABLE 4 Record of lot 9. Temperature 21°C. CALCULA- TION, AVERAGE OXYGEN CONSUMED IN TEST WEIGHT OXYGEN CONDITION OF WORMS CONSUMED IN 2 HOURS BY 0.5 GRAM [ cc. grams Cos Decapitated, intact, one day since feed- | 0.33 in 13 hours | 0.838 0.27 ing 0.35 in 1} hours Cut into small pieces + to several hours after section 0.38 in 14 hours 0.790 0.33 0.40 in 13 hours 21 to 23 hours after section 0.45 in 14 hours 0.714 0.42 2 days after section 0.39 in 14 hours 0.658 0.39 0.38 in 14 hours 7 days after section 0.21 in 13 hours | 0.530 0.28 0.24 in 14 hours 15 days after section 0.32 in 2 hours 0.375 0.41 0.30 in 2 hours Worms fed three times 22 days after section; 1 day since feeding 0.41 in 2 hours 0.41 in 2 hours | 0.448 | "O47 Regenerated parts cut off 23 days since first section; several hours since second section 0.36 in2 hours -| 0.338 0.55 0.39 in 2 hours sumption was found to have risen to 0.31 to 0.35 ce. Forty-eight to seventy-two hours after section, it was about the same, 0.32 to 0.36 ce. On April 17, seven days after section, it had fallen slightly to 0.30 to 0.32 ec. On April 23, thirteen days after section, it was rising again, being 0.32 to 0.38 cc.; and on May 2, twenty- two days after section, when the pieces had undergone complete regeneration, the AMOUNTOF _ OXYGEN CONSUMPTION DURING REGENERATION lo oxygen consumption was much higher than at the beginning of the experiment, being 0.41 to 0.45 ce. The regenerated portions were now removed and the oxy- gen consumption of the remaining pieces, corresponding to the original pieces, was 0.36 to 0.37 cc. The temperature throughout these experiments was also 22°C. = 0.5. TABLE 5 Record of lot 10. Temperature 18°C. CALCULA- TION, AVERAGE AMOUNT OF CONDITION OF WORMS Re on pt ata WEIGHT GonNuGutet IN 2 HOURS BY 0.5 GRAM : cc, grams ce. Decapitated, intact, four days since feed- | 0.23 in 2 hours 0.599 0.18 ing 0.21 in 2 hours Cut into small pieces 1} to 33 hours after section 0.29 in 2 hours 0.512 0.28 18 to 22 hours after section 0.29 in 2 hours 0.469 0.27 0.23 in 2 hours 2 days after section 0.23 in 2 hours 0.452 0.23 0.20 in 2 hours 4 days after section 0.18 in 2 hours 0.403 0.22 8 days after section 0.17 in 2 hours 0.332 0.25 0.17 in 2 hours 17 days after section 0.15 in 2 hours 0.181 0.42 0.16 in 2 hours Regenerated parts cut off 22 days since first section ; few hours since second section 0.05 in 2 hours 0.113 0.24 0.06 in 2 hours Record of lots 7 to 9 (tables 2 to 4). The stock from which these worms were taken was collected in January, 1919. They were last fed on April 23, their heads removed on the same day, and their rate of oxygen consumption tested on April 24. They were then cut into small pieces and their rate of oxygen consumption tested at various intervals after section, as in the preceding experiments. As be- fore, the oxygen consumption was found to be greater during the first few hours 74 L. H. HYMAN up through two days after section than it was in the same worms before section; it then fell, as shown by a measurement on May 1, seven days after section. From this time on, the oxygen consumption rose and on May 9, fifteen days after section, it was considerably higher than that of the original intact worms. The regenerated worms were then fed on May 10, 12 and 14, and the oxygen consump- TABLE 6 Record of lot 11. Temperature 18°C. CALCULA- TION, AVERAGE a < AMOUNT OF CONDITION OF WORMS ie are WEIGHT OXYGEN CONSUMED IN 2 HOURS BY 0.5 GRAM ce. - grams cc. Decapitated, intact, four days since feed- | 0.23 in 2 hours 0.560 0.21 ing 0.24 in 2 hours Cut into small pieces 1 to 3 hours after section 0.29 in 2 hours 0.540 0.26 18 to 22 hours after section 0.29 in 2 hours 0.25 in 2 hours 0.503 0.26 2 days after section 0.25 in 2 hours 0.476 0.25 0.23 in 2 hours 4 days after section 0.20 in 2 hours 0.430 0.23 8 days after section 0.20 in 2 hours 0.360 0.27 0.19 in 2 hours 17 days after section 0.17 in 2 hours 0.153 0.52 0.15 in 2 hours Regenerated parts cut off 0.06 in 2 hours 0.05 in 2 hours 22 days since first section; few hours since second section 0.085 0.32 tion tested on May 15, one day after feeding, as was the case with the original lots. As before, feeding resulted in a distinct increase in the rate of oxygen consumption. On May 16, the regenerated tissue was removed, and the oxygen consumption of the remaining pieces, corresponding to the original pieces, was again tested. The temperature throughout was 21°C. + 0.5. OXYGEN CONSUMPTION DURING REGENERATION 05 Record of lots 10 to 12 (tables 5 to?7). The members of lots 10 and 11 came from a stock which had been for some time in the laboratory (date of collection not known). They were large, well-fed individuals. The worms in lot 12 came from a general mixed stock containing material which had been used for other experi- mental purposes. The three lots were last fed on May 2, their heads were removed TABLE 7 Record of lot 12. Temperature 18°C. } CALCULA- TION, AVERAGE CONDITION OF WORMS waite ocredk if, W EIGHT Oa, % : | CONSUMED IN 2 HOURS BY 0.5 GRAM co. grams cen Decapitated, intact, four days since feed- | 0.23 in 2 hours 0.681 0.17 ing ; 0.24 in 2 hours Cut into small pieces 2 to 24 hours after section 0.30 in 2 hours 0.615 0.24 18 to 22 hours after section 0.31 in 2 hours 0.566 0.23 0.22 in 2 hours 2 days after section 0.28 in 2 hours 0.532 0.25 0.27 in 2 hours 4 days after section 0.22 in 2 hours 0.499 0.22 8 days after section 0.21 in 2 hours 0.422 0.24 0.21 in 2 hours 17 days after section 0.18 in 2 hours 0.263 0.36 0.20 in 2 hours Regenerated parts cut off 22 days since first section; 1 day since sec- | 0.11 in 2 hours 0.166 0.31 ond section 0.10 in 2 hours on May 5, and the oxygen consumption of the decapitated worms was tested on May 6. The worms were then cut into small pieces, and the oxygen consumption of the pieces tested at intervals after section as in the preceding experiments. The rate of oxygen consumption was again found to be greater after section than before, remaining high through two days after section, then falling, and finally rising as regeneration was completed. The regenerated parts were then removed, \ 76 L. H. HYMAN and the rate of oxygen consumption of the parts corresponding to the original pieces tested. The temperature throughout these three experiments was 18°C. + 0.5. A word is required regarding the degree of movement of the worms during these experiments, since movement increases the rate of oxygen consumption. The decapitated whole worms with which the experi- ments are begun are always perfectly quiet, as are also the pieces cut from them during the first ten or twelve hours after section. Twenty- four hours after section and from this time on through the greater part of the regeneration process there is some slight movement among the pieces but this is not sufficient to affect the measurements of oxygen consumption. The completely regenerated worms, however, usually move about considerably, and the figures obtained at the end of the regeneration process are probably slightly too high on this account, although the amount of movement was reduced in most cases by shad- ing the worms and placing them in the flasks in which they were to be tested some time before the test was carried out. In experiments 10 to 12 inclusive, movement was practically completely eliminated in all cases by testing the worms at a temperature of 18°C. Since the results in these three experiments do not differ from those of the other nine, it is reasonably certain that movement is never a significant factor in the general result. After the regenerated worms are fed, they remain perfectly quiet, and the pieces obtained by cutting off the regenerated regions are also motionless. CONCLUSIONS AND DISCUSSION The data upon twelve different lots of worms are in accord with each other, and justify the following conclusions: 1. The oxygen consumption per unit time per unit weight of a given lot of worms is greater after these worms have been cut into small pieces than it was while they were intact. Section therefore increases the rate of oxygen consumption. The same conclusion had already been reached in this laboratory by means of other methods. Thus Child (2) long ago observed that the susceptibility of newly-cut pieces of Planaria to cyanide and other toxic solutions is much greater than that of intact worms. This increase in susceptibility is greatest at the cut surfaces but involves the entire piece also, in the case of small pieces. In long pieces the increased susceptibility is observable only at the cut surfaces and adjacent regions. This demonstrates that the OXYGEN CONSUMPTION DURING REGENERATION 77 increase in susceptibility and rate of oxygen consumption following section are due primarily to the injury of cutting and spread from the cut surfaces with a decrement to the remainder of the pieces. The phenomenon is indeed only one example of the general physiological fact that injury is a form of stimulation, expressing itself in an increased rate of respiratory exchange, increased production of metabolic prod- ucts, and electrical negativity. The increase in susceptibility after section has been observed in this laboratory not only in Planaria but in a variety of the lower organisms. I found it to be true for several species of small fresh-water annelids (3) and incidentally in the course of numerous investigations upon the susceptibility of the lower forms, we have invariably observed that injured places are more susceptible than adjacent uninjured regions. Scott (4) noted that the oxygen consumption of the sea-anemone Sagartia is increased after section. Not only are the rate of oxygen consumption and the susceptibility to toxic solutions of pieces of Planaria increased after section but, as Child (2) has shown with the aid of the Tashiro biometer, the carbon dioxide production is likewise accelerated. Recently Miss Robbins has repeated and extended these observations using the phenolsul- phonephthalein method of Haas (5). Tashiro in his book A Chemical Sign of Life has shown that such an acceleration of carbon dioxide production as a consequence of injury is a practically universal biologi- cal phenomenon and he suggests that the occurrence of this change may be regarded as a proof that the material in question is living. 2. The increased rate of oxygen consumption following section was observed in these experiments to continue in most of the lots through forty-eight hours after section. Tests by the susceptibility method show that the susceptibility of the middle portions of the piece gradually falls after section and has returned to the condition found in intact worms in about twelve hours. However, the susceptibility of the cut and regenerating surfaces always remains higher than that of the middle part of the pieces. Hence the oxygen consumption, since it includes all parts of the pieces, must be greater than that of intact worms during this early part of the regeneration period. Nevertheless, it was ex- pected that the oxygen consumption twenty-four and forty-eight hours after section would be somewhat less than it was within the first few hours after section. In lots 2, 5 and 10 this was the case but not in the other nine lots. It is therefore evident that extraneous factors enter into the determinations during this period, and this is further 78 L. H. HYMAN indicated by the greater variability in the figures obtained during the first forty-eight hours after section than at later periods. It seems highly probable that the frequent weighings and handling during this period acted as sources of stimulation. This is further rendered prob- able by the greater uniformity in the measurements obtained with lots 10 to 12 which were tested at a lower temperature in order to eliminate some of these factors. 3. During the period from three or four days to a week after section the oxygen consumption is falling. Since the susceptibility method shows no such fall in metabolic rate during this time, the decrease must be regarded as due entirely to starvation. As shown in the first paper of this series (1), the rate of oxygen consumption decreases continuously during the first two weeks of starvation, this resulting from lack of activity of the digestive tract. In the case of regenerating pieces the total oxygen consumption does not decrease as a result of starvation to as great an extent as in non-regenerating worms, since this decrease is partially compensated by the increased metabolic rate of the regen- erating ends of the pieces. The total oxygen consumption, therefore, of the pieces during the period from three or four days to a week after section is the resultant of two conditions, an increase due to regen- eration and a decrease due to starvation. During this period the latter factor predominates. 4. From a week after section to the completion of the process of regeneration the rate of oxygen consumption is continually increasing. The oxygen consumption of the completely regenerated worms is very much greater than that of the intact worms from which the pieces were taken. The amount of this increase ranges in the twelve lots of worms from 50 to 150 per cent. Regeneration is thus a method of increasing the metabolic rate of Planaria. This conclusion had already been reached in this laboratory for Planaria and other forms through the use of the susceptibility method and further for Planaria through a study of carbon dioxide production during regeneration. Scott (4) found that the rate of oxygen ae of a sea-anemone is increased by regeneration. The objection may be raised that the increase in rate of oxygen consumption per unit weight during regeneration may be only apparent since the decrease in weight might be due to a loss of non-respiring materials. But if this were the case, the oxygen consumption should increase during the early stages of regeneration, when the weight is decreasing most rapidly. As a matter of fact however, both in the OXYGEN CONSUMPTION DURING REGENERATION 79 present series of experiments and in the experiments on starvation previously reported, the oxygen consumption is also decreasing during this time. In the later stages of both regeneration and starvation, the loss of weight occurs at about a uniform rate or even decreases slightly; yet during this time the rate of oxygen consumption is con- tinually rising. It is therefore reasonably certain that the observed acceleration of oxygen consumption in regeneration and starvation rep- resents a real increase in the basic metabolism of the cells of the organ- ism. The same conclusion was reached by Benedict (6) in his study of the metabolism of man during prolonged fasting. The present experiments therefore support Child’s conception of regeneration as a method of bringing about rejuvenescence—that is, restoring the organism to a metabolic condition comparable to that of young animals. 5. The regenerated worms of lots 1, 2, 3, 7, 8 and 9 were fed in order to bring them to a state of nutrition comparable to that of the original pieces, since, as already explained, starvation tends to lower the rate of oxygen consumption during the period in which the pieces are re- generating. In all six cases, the rate of oxygen consumption of the regenerated worms was increased by feeding. The tests were of course made at the same time interval after the last feeding as had been the case with the original pieces. It is naturally impossible to assert that after three or four feedings the regenerated pieces are in the same state of nutrition as the original pieces from which they came but at least they are as the result of such feeding more comparable to the latter. It is only after the effect of starvation has been eliminated that the real extent of the rise in oxygen consumption resulting from regeneration can be detected. 6. The rise in rate of oxygen consumption due to regeneration in- volves not only the newly regenerated portions, but also the old tissue of the piece. This was determined by removing and discarding the new heads and tails and testing the rate of oxygen consumption of the pieces remaining after this operation, such pieces corresponding as nearly as possible to the original pieces. Presumably these pieces are stimulated by section as were the original pieces although we cannot say whether they are stimulated to the same degree through the re- moval of the regenerated portions as are pieces cut from whole worms, as this point has not yet been subjected to experimental test. It is probable that they are less stimulated by section than are pieces cut from whole worms since in newly regenerated worms the old portions 80 L. H. HYMAN have not yet established connections with the newly formed ends. Probably the severance of morphological and physiological connections and conduction paths is one of the chief factors in the stimulation observable after cutting. Since, however, we have no definite information upon the degree of stimulation by section in pieces cut from regenerated worms, it is necessary to assume that it is the same as that in pieces cut from in- dividuals not recently regenerating. One must therefore compare these pieces comprising the old portions of the regenerated worms with the original pieces at approximately the same lengths of time after section. The necessary data for this comparison are given in the tables. It is evident that when the regenerated worms were fed, the oxygen consumption of the old portions cut from them is always considerably higher than that of the original pieces the same length of time after section and after feeding. When the regenerated worms were not fed, the oxygen consumption of the old tissue is less but in all cases, except in lot 10, it is still higher than that of the original pieces con- sidered the same length of time after section. This one exception may be due to variability in the degree of starvation. It is therefore certain that when restored to the same condition of nutrition, pieces from which regenerated tissue has grown out have a higher rate of oxygen consumption than the same pieces before such growth occurred. It is further evident from the data at hand that the rate of oxygen consumption of the old portions of the regenerated worms is less than that of the regenerated regions since in all cases where the worms were not fed, the oxygen consumption is reduced by cutting off the regener- ated tissue. Where the worms were fed, a comparison cannot be made since the rate of oxygen consumption after feeding depends upon the number of worms which feed and upon the amount of food which they ingest, and this in return depends upon the morphological condition of the digestive tract, a factor which is very variable in a mixed lot of regenerating worms such as those used in these experiments. SUMMARY 1. The oxygen consumption of a given lot of Planaria per unit weight is increased when they are cut into small pieces. This increase is due to the stimulation of injury. 2. This increase persists through about forty-eight hours after cut- ting but may fall slightly during this period. The persistence of the OXYGEN CONSUMPTION DURING REGENERATION 81 increase is associated with the activity of the cut surfaces, the original tissue in the pieces probably not being involved. 3. The oxygen consumption of the pieces then falls and remains at a low level for about one week. This fall is due entirely to starvation. The fact, however, that the decrease in rate of oxygen consumption during this period is not as great as in starving non-regenerating worms indicates that the oxygen consumption of the regenerating regions has remained high throughout. 4. The oxygen consumption then begins to rise and continues to rise as regeneration proceeds. The oxygen consumption of the com- pletely regenerated worms is 50 to 150 per cent greater than was that of the worms from which the pieces were taken. 5. When the regenerated worms are fed in order to eliminate the factor of starvation their oxygen consumption rises to a still higher figure. 6. The increased rate of oxygen consumption in regenerated worms is due not only to the high metabolic rate of the regenerated tissue but also in part to an acceleration of the rate of oxygen consumption of the old tissue comprising the original pieces. In other words, when a piece of Planaria undergoes regeneration, its metabolic rate is thereby accelerated; its rate is not, however, as high as that of the regenerated regions. 7. These experiments confirm the conclusion already reached by Child through other methods that the process of regeneration is a rejuvenating process, restoring the organism to a metabolic condition comparable to that of young organisms. BIBLIOGRAPHY (1) Hyman: This Journal, 1919, xlix, 377. (2) Cutty: Journ. Exper. Zodl., 1914, xvi, 413. (3) Hyman: Journ. Exper. Zo6l., 1916, xx, 99. (4) Scorr: Proc. Soc. Exper. Biol. and Med., 1916, xiii, 121. (5) Haas: Sci., 1916, xliv, 105. (6) Benepictr: Carnegie Inst. Washington Publ., no. 203, 1915. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 EFFECT OF ANESTHESIA AND OPERATION ON CERTAIN METABOLITES STANLEY P. REIMANN anp FRED. L. HARTMAN From the Department of Pathology of the Lankenau Hospital, Philadelphia Received for publication July 1, 1919 A previous study of anesthesia demonstrated certain changes in the acid-base equilibrium during that process (1). For practical needs in the matter, the question of whether to give carbon dioxide with the anesthetic or alkali before or after the anesthetic, was continually the basis of the work. A certain few surgical patients show uncompen- sated acidosis after anesthesia and operation. These studies were to show, if possible, changes in the nitrogen metabolism. TECHNIQUE Blood was collected before anesthesia was begun, and after the anesthetic was stopped, from the arm veins of unselected surgical patients admitted to the Lankenau Hospital, and operated in the clinic of Dr. John B. Deaver. Oxalate was used to prevent clotting. Blood-urea was determined by the urease method. Non-protein ni- trogen was determined on 10 ce. of blood; coagulation of the proteins was accomplished by trichloracetic acid, and nitrogen estimated on the filtrate by the Kjeldahl method. Urine in the first series was collected twenty-four hours before, and twenty-four hours after opera- tion; and in later experiments, it was collected at shorter intervals. The results were exactly the same. Urea was determined by the urease and titration method; ammonia by aeration and titration; phosphates by the uranium titration method; total acidity by titration with deci-normal sodium hydroxide, using phenolphthalein as indi- cator to the first faint pink. Our a priori arguments were as follows: If acids are increased dur- ing anesthesia, the titrable acidity of the urine should increase. The blood-urea should diminish as a result of increased demand for ammonia, The urine urea should possibly decrease and the urine 82 ME EFFECT OF SURGICAL PROCEDURES ON METABOLITES 83 ammonia increase. The figures shown in the table express the results of determinations on about ninety patients. The urinary acidity was increased very definitely in every case until in some instanees, almost half normal acid was excreted. The blood-urea and non-protein ni- trogen were increased in every case. In the urine the general ten- dency of the ammonia was to increase, whereas the excretion of urea was either increased or diminished, with apparently no relation be- tween the urea in the urine and in the blood. The question of retention or increased production immediately suggested itself in the case of the increase in the blood-urea and non-protein nitrogen. TABLE 1 BLOOD URINE Non-protein U : rea s ] : ae Lae teem per 100 cc. Ammonia Urea Titrable acidity o o 5 5 3 5 ” 2 oI 2 Before After Before After Before After faa) < fea} < mgm. |mgm.|mgm.\mgm.| gm. beet gm. Bes gm. bse gm eit ce pee ce. me 40 .4* |387.8) 35 | 43 |0.150/0.097/0.12110.060|2 .21)1 .43/1.93/0.97|106.0} 69 |302 |101 44.2 |49.2] 13 | 38 |0.232/0.082/0.016|0.055/3.96 39.8 (48.1) 12 | 15 |0.084)0.030\0.240/0.092)/4.11 39.4 |44.8) 25 | 45 |0.053,:0.070,0.094,0.049,0.66 — i=) eS) Or 23.0 |40.0} 13 | 20 |0.006/0.007|0.045'0.043/0.36/0.41/1.30/0.92)} 1.7] 2 |176 |126 33 .6* 21.5] 13 | 22 |0.073/0.048/0.039)/0.060}1 .62|1 .06|0.47|0.73] 79.0] 55 | 87 |134 42.8 |44.5) 24 | 34 (0.166/0.070/0.365)/0. 121/38 .27/1.39'5.17|1.72| 42.0) 18 |498 |166 44.2 |48.1| 22 | 27 |0.026,0.052/0.047|0.047/0.43|0.87|0.57/0.57) 29.0) 58 | 80 | 80 29.9 |34.4| 12 | 16 (0.022/0.019/0.041/0.045}1 .45|1.25/0.72/0.80) 19.0] 17 | 83 | 98 36.9 |43.6] 11 | 30 (0.020|0.081|/0.268/0.063/0.59)1. 74/7 .35'1.73| 14.0} 40 |408 |115 do|1. 1] 1.¢ 0. ol wm. CO ow rm oS) * These were the only cases in the series which decreased. MacNider’s excellent work has shown that during anesthesia, in a non-nephropathic animal, the response to diuretics is approximately normal (2). From the fact that the blood-urea was universally in- creased, and the fact that among these patients there were numbers whose kidneys were non-nephropathic, as far as it was possible to learn, we can say that if retention does play a part, it does not tell the whole story. Therefore, there must have been an increased metabolism in the usual way down to urea. There must, however, also have been some slight relative changes in the nitrogen metabolism; for the in- creases in the non-protein nitrogen do not entirely parallel the increases 84 S. P. REIMANN AND F. L: HARTMAN in blood-urea. As regards the urinary findings, the urea excretion bears no constant relationship to the variations in the blood-urea. This is in harmony with observations of urea excretion and blood-urea levels in other circumstances. We are not convinced that any definite and strict mathematical relationship has been shown as yet. That the ammonia is not increased in each case is perhaps not as surprising as first thought would suggest. A markedly increased elimination of phosphates accounts for some of the acidity of the urine (8). These no doubt play a part, as well as the sodium bicarbonate, in neutralizing the increased acids. In any given case, then, alkali is available in a number of forms, and in all probability, could a complete analysis be TABLE 2 Urinary elimination of phosphates (P2Os) BEFORE AFTER per cent per cent 0.89 1.2 0.304 eal 0.392 0.922 0.25 0.865 0.09 0.46 0.31 0.732 made of all the acids and alkalies, both in the blood and urine, before and after anesthesia, a mathematical proportion between them could be worked out. , The possibility has been kept in mind that the changes in metabolism may not be altogether along the normal route. Further investigations, particularly directed to a search for abnormal intermediate products, as well as, perhaps, the products of an entirely abnormal metabolism,— those due, for example, to direct destruction of cytoplasm or nuclear material by the anesthetics,—must be conducted to clear up some points. This was afield of our practical needs in the matter, and we have not carried them out. SUMMARY It has been demonstrated that small but very definite changes in metabolism take place during anesthesia. The reduction in bicar- bonate of the blood plasma which occurs in that case must be due to EFFECT OF SURGICAL PROCEDURES ON METABOLITES 85 these changes, and not to mere over-yentilation, as suggested by Y. Henderson (4). Therefore, giving carbon dioxide with the anesthetic is contraindicated. Alkali should be given in selected cases. BIBLIOGRAPHY (1) Remmann AND Bioom: Journ. Biol. Chem., 1918, xxxvi, 211. (2) MacNiprer: Journ. Exper. Med., xxviii, 501, 517. (3) Sawyer, BAUMANN AND Stevens: Journ. Biol. Chem., 1918, xxxiii, 109. (4) Henperson aNp Haccarp: Journ. Biol. Chem., 1918, xxxiii, Bees Byile EXPERIMENTAL SURGICAL SHOCK V. THe TREATMENT OF THE CoNDITION OF Low BLoop PRESSURE WHICH FOLLOWS EXPOSURE OF THE ABDOMINAL VISCERA! F. C. MANN From the Section of Experimental Surgery and Pathology, Mayo Clinic, Rochester, Minnesota Received for publication July 1, 1919 This work was undertaken for the purpose of investigating under standard experimental conditions all the more important methods of treating a condition which exhibits the clinical signs of surgical shock in order to determine the relative value of the methods. While numerous such investigations have been made, only a portion of the entire field has been covered by a single investigation. Since in the different series of experiments various methods were used to produce the condition called shock it is impossible to compare the therapeutic results. The method of producing the signs of shock was the same through- out in the present investigation and while the results may not be applied directly to all cases clinically diagnosed as surgical shock, the different methods of treating a condition which presents a common symptomatology can be accurately compared. Since it is obvious that the condition which the surgeon terms shock is due to a variety of causes, it is useless to attempt to find a specific therapeutic procedure, but as the symptoms are usually the same it is reasonable to suppose that some general therapeutic measures will be found. The method of producing shock has varied greatly with the different investigations. In my work exposure of the abdominal viscera has seemed to afford the nearest approach to the production of shock presenting all the clinical signs (17). Our routine method was as follows: The animal, a dog, which had been fasted for from twelve ' Presented before the American Physiological Society, April, 1919, Baltimore. 86 EXPERIMENTAL SURGICAL SHOCK 87 to eighteen hours, was etherized in a closed cabinet, incubated, and a constant surgical anesthesia maintained by means of a Connell appa- ratus. Carotid blood pressure was recorded by means of a mercury manometer; sometimes the membrane manometer was also employed. After a record of normal blood pressure was obtained, the abdomen was opened and the viscera exposed. The only trauma to which the exposed viscera were subjected was the occasional gentle sponging with dry gauze or changing them from one side of the body to the other. When blood pressure had decreased and remained rather sta- tionary at the desired level, which usually occurred about one to two hours after the exposure of the viscera, they were returned to the abdominal cavity and the wound was closed. After a length of time sufficient definitely to determine that blood pressure did not increase, procedures to improve the condition of the animal were instituted. The blood pressure was taken as a criterion of the animal’s condition because it affords the easiest method of comparison. All other clinical signs of shock were also noted. The maintenance of a constant anesthetic throughout the experi- ment removed the possibility of an error either in the interpretation of blood pressure records or in the general condition of the animal (19), (20). Anesthetic control experiments were carefully performed, the etherization being maintained at the same tension and for a length of time equal to that of the shock experiments. Practical conclusions can only be drawn from the results which apply to the condition in which the signs of shock were produced by exposure of the abdominal viscera, although it would seem that they should also be of value in a condition of a lowered and progressively decreasing blood pressure. The experiments were acute because it seemed that the therapeutic procedure would be put to a greater test if the animal was maintained under an anesthetic throughout the experiment and because the char- acter of the experiment would not warrant the complete withdrawal of the anesthetic. The fact should be emphasized that blood pressure which has been decreased and remains stationary below a certain level, and is allowed to remain there even for a very short time, is never restored and maintained by any known method of treatment. We have estimated the pressure below which no hope for restoration could be held as half the initial pressure maintained constant for one hour. In our experi- ments the methods of treatment, with very few exceptions, proved of no permanent value if the blood pressure had been decreased to less 88 F. C. MANN Fig. 1. Kymograph record showing the successful use of acacia. Record I, normal blood pressure 118. Record II, after the abdominal viscera had been exposed 134 hours; blood pressure 70. Record III, taken 30 minutes after the viscera were replaced; blood pressure 74; 160 ec. (20 ce. per kgm.) of a 6 per cent solution of acacia in 0.9 per cent sodium chlorid solution were injected (signal A-B); the injection was probably too rapid. The blood pressure increased to a maximum of 84. Records IV, V, VI, VII and VIII were taken at succeeding hours after the injection. The blood pressure was 84, 90, 95, 95 and 96 respec- tively. This is one of the few successful results following the use of acacia in the series of experiments. Fig. 2. Kymograph record showing the favorable action of acacia. Record J, normal blood pressure 105. Record IT, after 1 hour of exposure of the abdominal viscera; blood pressure 70. Record IIT, 15 minutes after replacing the viscera; blood pressure 68; 100 cc. (20 ee. per kgm.) of a 6 per cent acacia solution in 0.9 per cent sodium chlorid were injected slowly; blood pressure increased to a maximum of 82. Record IV, taken 30 minutes after injection; blood pressure 76. Record V, taken 1 hour after injection; blood pressure 80. Records VI, VII, VIII and IX were taken at successive hours after the injection with blood pressure of 92, 90, 98 and 85 respectively. These experiments seem to show that the result is as good as can be hoped for with acacia. Fig. 3. Kymograph record showing the results of the injection of an alkaline acacia solution. Record I, normal blood pressure 148. Record IT, after exposure of the abdominal viscera for 1 hour; blood pressure 88. Record III, after ex- posure of the abdominal viscera for 2 hours; blood pressure 75. Record IV, after replacing the viscera for 10 minutes and the injection of 146 cc. (20 ce. per kgm.) of a 7 per cent solution of acacia and 4 per cent solution sodium bicar- bonate; the blood pressure increased to a maximum of 118. The succeeding records were taken at one-hour intervals after injection. The decrease in blood. pressure occurring in the last records is characteristic of the action of an alkaline acacia solution, but.usually blood pressure is not maintained for so long a time. Fig. 4. Kymograph record showing a failure of acacia solution followed by a success with gelatine solution (Hogan’s). Record I, normal blood pressure 140. Record IT, 13 hours after exposure of the abdominal viscera; blood pressure 82. Record III, 30 minutes after the replacing of the viscera; blood pressure 94; 104 ce. (20 ec. per kgm.) of a 6 per cent solution of acacia in 0.9 per cent sodium chlorid were injected (signal A-B). The injection may possibly have been made too rapidly; at any rate there was little improvement in blood pressure. Record IV, taken 35 minutes after the injection of acacia; blood pressure 70; 104 ec. of gelatine solution (Hogan’s) were injected at the same rate the acacia had been injected; blood pressure increased to 112. Records V, VI, VII and VIII were taken at succeeding hours after the last injection. The blood pressure was respectively 100, 112, 118 and 110. The gelatine produced a much better result than acacia. -ERIMENTAL SURGICAL SHOCK {| Seca: renin naman 3 {mong RH yma nannnsnnnunn, aman A THE AMERICAN JOURNAL OF PHYSIOLOGY VOL. 5 no. J ’ 90 F. C. MANN Fig. 5. Kymograph record showing the action of the citrated blood after a failure of acacia and modified acacia solution. Record I, normal blood pressure 112. Record II, after exposure of the abdominal viscera for 1 hour; blood pres- sure 52. Record III, after the replacing of the viscera for 5 minutes; 70 ec. (20 ec. per kgm.) of a 6 per cent solution of acacia in 0.9 per cent sodium chlorid solution were injected (signal A-B). The blood pressure increased to a maximum of 80, but soon began to decrease and in 30 minutes it was 55. Record IV, taken 40 minutes after record IIT; blood pressure 50; 70 cc. of a modified acacia solution (6 per cent acacia, 10 per cent glucose, 1 per cent sodium carbonate, 1 per cent sodium sulphate were injected (signal C-D). The blood pressure increased to a maximum of 110, but soon began to decrease. Record V, taken 40 minutes after record IV; blood pressure 60. Record VI, taken 1 hour after the last injection; blood pressure 56; 70 ce. of citrated blood were injected (signal E-F). The blood pressure increased to a maximum of 85. Records VII, VIII, IX and X were taken at succeeding hours after the last injection. Note the beneficial action of blood. Fig. 6. Kymograph record showing the restoration and maintenance of blood pressure by the injection of blood after a failure of acacia solution. Record I, normal blood pressure 105. Record IT, 1 hour after exposure of the abdominal viscera; blood pressure 70. Record III, immediately after the replacing of the viscera; blood pressure 40; 70 ce. (20 cc. per kgm.) of a 6 per cent acacia solution in 0.9 per cent sodium chlorid solution were injected (signal A-B). The blood pressure increased to a maximum of 70, but within 30 minutes had decreased to 50. Record IV, taken 50 minutes after record IIT; blood pressure 48, 70 ce. citrated blood were injected (signal C-D). The blood pressure increased to a maximum of 86. Records V, VI and VII were taken at succeeding hours after the last injection, with blood pressures of 88, 90 and 88, respectively. Fig. 7. Kymograph record showing the effect of injection of dextrin followed by gelatine solution. Record I, normal blood pressure 115. Record II, 1 hour after the exposure of the abdominal viscera; blood pressure 65. Record III, 15 minutes after the replacing of the viscera; 150 ec. (20 cc. per kgm.) of a 20 per cent dextrin solution were injected. The blood pressure increased to a maximum of 120, but soon decreased to 60. Record IV, taken 1 hour after record III; blood pressure 58; 150 cc. gelatine solution (Hogan’s) were injected; blood pressure increased to a maximum of 82. The succeeding records were taken at one-hour intervals after injection. Note the slow recovery and the failure of- blood pressure. q 91 SHOCK SURGICAL EXPERIMENTAL GSI a a mc 6 is” ST << 1/7) ID ==} Lea bers o° UTM ag ered z en a SS it os 92 F. C. MANN Fig. 8. Kymograph record showing the results of the injection of dog serum. Record I, normal blood pressure 110. Record II, after exposure of the abdominal viscera for 1 hour; blood pressure 64. Record III, after the replacing of the viscera for 15 minutes; 110 ec. (20 ce. per kgm.) of dog serum were injected. The blood pressure was Increased to a maximum of 110. The succeeding records, IV to VIIT, were taken at one-hour intervals after injection. The serum re- stored blood pressure to normal and maintained it for 5 hours, until the experi- ment was interrupted. Fig. 9. Kymograph record showing the beneficial action of dog serum after a failure of normal salt solution. Record J, normal blood pressure of 112. Record IT, after exposure of the abdominal viscera for 1 hour; blood pressure75. Record III, after the viscera had been replaced for 10 minutes; blood pressure 65; 156 ec. (20 ce. per kgm.) of normal salt solution were injected (signal A-B). The blood pressure increased to a maximum of 85, but in 15 minutes had decreased to 72. Record IV, an equal amount of dog serum was injected (signal C-D). Blood pressure increased to a maximum of 120. Records V, VI, VII, VIII and IX, were taken at succeeding hours after the injection, with blood pressures respectively of 105, 110, 120, 115 and 105. Fig. 10. Kymograph record showing the beneficial results of citrated blood after a failure of acacia. Record I, normal blood pressure 130. Record II, after exposure of the abdominal viscera for 1 hour; blood pressure 80. Record III, after the replacing of the viscera for 15 minutes;blood pressure 70; 116 ec. (20 ce. per kgm.) of a 6 per cent acacia solution in 0.9 per cent sodium chlorid solution were injected slowly (signal A-B). ‘The blood pressure increased to a maximum of 90, but soon began to decrease. Record IV, taken 30 minutes after the injection was stopped. Blood pressure 60. Record V, taken 1 hour after the injection; blood pressure 70. Record VI, taken 13 hours after the injection; blood pressure 74; 116 cc. of citrated blood were injected slowly (signal C-D). The blood pres- sure increased to a maximum of 110 and then decreased slightly after the in- jection was stopped. Record VIJ, taken 30 minutes after the injection was stopped; blood pressure 118. Record VIIT, taken 13 hours after the injection; blood pressure 118. Record 1X, taken 23 hours after the injection; blood pres- sure 112. Record. X, taken 4 hours after the injection; oben pressure 98. The animal was used for another experiment. lone SHOCK SURGICAL EXPERIMENTAL OL “Si ~ fh pe ——--——- 4a = pee seg bay si? penunabeg aro LLM CTT SUED fery ee Pepwmogigs Phen 5 7 ¥ ae CL AN 4 LL VS @ a, em eee 94 F. C. MANN than one-half its initial value; this is true regardless of the means by which blood pressure is lowered, for example, by hemorrhage, exposure of abdominal viscera, and by obstruction of the venous return, such as partial occlusion of the vena cava. Conclusions should, therefore, not be drawn with regard to a therapeutic procedure when it is tried out in an experiment in which the blood pressure has been decreased below one-half its normal value and because of the variability of the different animals unless, of course, the animal recovers. We believe however, that it might be of some clinical value if a therapeutic pro- cedure could be found which completely or in greater part restores and maintains blood pressure for two hours, with the animal under a con- stant anesthesia and with a constant artificial temperature condition, after the blood pressure has been decreased from one-third to one- fourth its initial pressure by exposure of the abdominal viscera. These conditions were the standards we used in judging the value of the various methods of treatment. The treatment of shock may be described under four headings: a, general measures; b, special measures; c, the use of drugs; d, attempts to restore fluid volume. General measures. The most important general measure in the treatment of shock is the ancient practice of applying heat. The employment of heat is of value not only because shock is commonly associated with exposure to cold but also because the thermogenetic and thermo-regulatory mechanisms are impaired. It is probably not true that this impairment of the mechanism which keeps the body temperature constant is the primary cause of shock but the artificial maintenance of body temperature during the period of impairment produces beneficial results. It should be noted, however, that this deficiency in regulation applies to heat as well as to cold and that too much heat is harmful. In most instances the temperature of our animals was kept almost constant by the judicious employment of an electric heating pad. In some experiments the heat was only applied after low blood pressure had been produced and at the same time the other therapeutic measures were instituted. Except that the blood pressure decreased more slowly. when the heat was used from the beginning of the experiment no notable difference was observed in the results of the therapeutic procedure. In order to increase the circulation around the bulbar centers it is usually recommended that the head be placed in slight Trendelenburg _ EXPERIMENTAL SURGICAL SHOCK 95 position. Theoretically this should be of value, practically it may be; but in our experiments little effect could be noted. Special measures. The purpose of most of the many special measures which have been devised for treating shock is to increase blood pressure either by decreasing the vascular capacity or by aiding in the return of blood to the heart. Strapping the limb and increasing intra- abdominal pressure should be of benefit, inasmuch as such measures decrease the vascular capacity, but their value is difficult to demon- strate experimentally. Rebreathing has also been recommended; ac- cording to Porter, it increases the action of the respiratory pump and thus aids the return of blood to the heart by sucking it into the thorax. The rationale of rebreathing in treating shock from the chemical standpoint is an integral part of Henderson’s acapnial theory. The value and limitation of rebreathing in surgery and anesthesia were first carefully studied by Gatch (11). In previous studies on rebreath- ing in shock I have shown that the process is similar in the normal and in the shocked animal but that no measurable benefit results; this was confirmed by the present series of experiments. Drugs. Drugs are usually employed in the treatment of shock for one of two purposes; first, either as a general stimulant, with particular reference to their action on the circulation and on the central nervous system (strychnin, camphorated oil, alcohol, ete.); and second, as vasomotor constrictors (epinephrin, pituitrin, etc.). Following the theory that shock is due to excessive vasoconstriction, the nitrites have been recommended for the purpose of decreasing vasoconstriction. Morphin is also recommended, mainly for its depressing action on the central nervous system. Many investigations have been made with regard to the value of strychnin in the treatment of shock. Most investigators are agreed that the drug is of no value, although many surgeons, relying on their clinical experience, still use it in large doses (9). In experimental shock it is impossible to observe any effect of strychnin in doses smaller than those necessary to produce definite convulsive movements. It is questionable whether even these large doses produce a beneficial action; in our experiments it could not be said that any of the so-called stimu- lants were of value. The value of the use of vasoconstrictors in the treatment of shock is still an open question. In the first place, although the decreased blood pressure is of great importance in shock, it is not known whether or not its increase by means of vasomotor constrictors is in itself of much 96 F. C. MANN permanent benefit to the organism, and it would seem that they might be of distinct harm by decreasing the fluid supply to the tissues. In the second place, none of the vasoconstrictor drugs produce a very prolonged effect. Epinephrin is the most popular of these drugs to be employed in shock. It easily restores the decreased blood pressure and by continuous injection, the blood pressure can be maintained for a considerable length of time. As soon as the injections are stopped, however, the blood pressure sinks to its former level or usually lower. In our experiments pituitary extract produced a more prolonged action and seemed to be of somewhat greater benefit than epinephrin. Of course, repeated doses of the former cannot be employed as in the case of the latter drug. In general it may be said that experimentally the vasoconstrictor drugs produce little if any permanent benefit in the treatment of surgical shock although they might be employed clinically. The nitrites produce their characteristic depression of blood pressure when it has been decreased by exposure of the abdominal viscera, but certainly no beneficial result has been observed from their use. Neither is the effect of morphin marked, but since it has been shown that the drug changes the regulations of blood volume, it should be studied more fully (2). Attempts to restore fluid volume. It has been shown that a definite and marked loss of circulating fluid accompanies low blood pressure after the exposure of the abdominal viscera (18). This also seems to be true in other forms of experimental shock (12). In many clinical cases of surgical shock there is a loss of circulating fluid (4), (5), and it seems logical to treat the condition by an attempt to restore the lost fluid to the circulation. A large number of artificial fluids have been devised for this purpose. We have investigated the use of most of these solutions under the standardized conditions mentioned. We usually injected the fluid to be tested with a burette although in some instances a continuous injection machine was employed. The former method proved the most practical, although it was impossible accurately to control the rate of injection. The effect of the injection depends somewhat on the rate at which the solution enters the vein. In general it seemed that the best results were produced with a rate that was just a little less than the amount which produced cardiac disturbance. Better results were obtained when the temperature of the solution was below 37° rather than above. In our experiments the use of blood gave far better results than the use of any other substance except blood serum. If blood pressure is EXPERIMENTAL SURGICAL SHOCK 97 not decreased to less than one-half its initial value after exposure of the abdominal viscera, the intravenous injection of citrated blood in relatively large amounts, 20 cc. per kgm., will practically always restore and maintain it for many hours. Asa rule, equally good results have not been secured with any of the artificial solutions. Blood frequently restored blood pressure aiter other solutions had failed. Homologous blood serum will produce practically the same results (19). Blood or blood serum show in many ways their superiority over all artificial solutions. They do not raise blood pressure more than some of the other solutions and quite frequently the blood returns more slowly to normal than after the use of some artificial mediums, but, whereas in most instances in which artificial mediums are used blood pressure soon drops to the shock level or below, after the injection of blood or serum, the increase in pressure is usually maintained for many hours. In shock the injection of any solution brings about a return of sensibility requiring higher ether tensions. The degree of sensibility is more marked after the injection of blood or serum than after any one of the solutions. Physiologic sodium chlorid solution is usually employed to restore lost fluid volume; this is the least valuable of the artificial fluids used if the blood pressure has been lowered in the manner employed by us. Hypertonic saline solutions have been recommended and in some of our experiments they produced a definite beneficial action but the in- creased blood pressure was never long maintained. None of the saline solutions alone will maintain blood pressure for more than a very short time even when it has been reduced to but slightly below normal by exposure of the abdominal viscera. The saline solutions will usually pass out of the vascular system almost as fast as they are run in. The use of sodium carbonate and bicarbonate in hemorrhage and shock was experimentally investigated several years ago; their use clinically has been emphasized recently. ; Howell (16) seems to have been the first to study the effect of an alkaline salt in shock. He studied the effect of injection of sodium carbonate in a condition of shock produced by different methods. The beneficial results of such injections in the experimental conditions of low blood pressure which he had produced were due, he concluded, chiefly or entirely to a direct action on the heart. Dawson (6) in continuing Howell’s study, investigated the effect of the injection of sodium bicarbonate in a condition of low blood pressure produced by hemorrhage. He found that it produced better results than the THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 98 F. C. MANN sodium chlorid solution, and suggested that the bicarbonate solution be used in those cases of shock accompanied by hemorrhage. Seelig, Tierney and Rodenbaugh (23) obtained marked beneficial results by the injection of sodium carbonate in the condition of experimental shock, and concluded that the results were not due to the bulk of fluid in- jected, the hypertonicity or alkalinity of the fluid, or to the free carbon dioxid, but to the specific action of the salt on the heart muscle. Cannon (4), in his study of shock in the front line trenches, found that there is a definite decrease in the alkalinity of the blood in cases of shock. The injection of sodium bicarbonate relieved this and pro- duced very marked benefit. Patients on whom the surgeon refused to operate were tided over the critical period by the injection of either sodium carbonate or sodium bicarbonate which produced a rise in blood pressure and especially an increase in pulse pressure, thus making it possible to operate in a very short time. In our experiments more lasting benefit was secured by the injection of sodium carbonate or sodium bicarbonate than by normal salt solu- tion. Neither of the alkaline salts, however, completely restored blood pressure, nor was the increase long maintained. Glucose has been suggested and used in post-operative treatment by several clinicians (3). Erlanger and Woodyatt (8) investigated its action in experimental shock and found it to be of some benefit. In our experiments the injection of such solutions was of definite value although rarely was there a complete restoration of blood pressure, nor was the increase long maintained. Glucose when added to some of the other artificial solutions seemed to enhance their value. Hogan (15) first recommended gelatine as a medium to restore lost fluid volume. His formula was used in several of our experiments and gave good results in some. In general it is as satisfactory as any of the artificial mediums. Great care should be taken, however, in its preservation, because it deteriorates very readily and may produce untoward results. It was very difficult to modify the gelatine solu- tions by the addition of other substances, and no modification was found to be as safe or to give better results than Hogan’s original formula. We have used acacia and its various modifications as recommended by Bayliss (1). The addition of acacia to a transfusion solution certainly increases the power of that solution to restore and maintain blood pressure. The results following the use of acacia, however, were quite variable and sometimes disastrous. This variability of EXPERIMENTAL SURGICAL SHOCK 99 action seemed to depend on both the acacia and the condition of the animal. It is quite possible that the results of our use of acacia have not been so good as those which others report because our acacia was not the same (21). We obtained the best we could, however, and I am quite sure the average surgeon who wishes to use it clinically would ob- tain no better. The alkaline acacia solution, when properly made by the addition of sodium carbonate or sodium bicarbonate, usually produced a better result than acacia alone, but it is difficult to prepare an alka- line acacia solution and more difficult to sterilize it and, on the whole, it did not seem to be a safe solution to use. Good results were produced by the addition of glucose to acacia (7). The modified acacia solu- tion, which gave the best results in our experiments, consisted of 6 per cent acacia, 10 per cent glucose and 1 per cent sodium sulphate. Many other methods for restoring fluid volume besides those already mentioned were tried. The rapid injection of 35 per cent solution of cane sugar, as recommended by Guthrie (13), usually fully restored the blood pressure but it was not long maintained. In acute hemor- rhage, however, it produces good results. Various strengths of dextrine solutions were used; they restored the blood pressure more satis- factorily than the other artificial solutions but they failed to maintain it. A 1 per cent sodium sulphate solution produced fair results. It is interesting to note that distilled water gives better results in experi- mental shock with regard to blood pressure than do normal salt solu- tions. Crude preparations of hemoglobin from dog’s blood produced good results and seems to warrant future study. In summarizing the various methods employed to restore fluid volume it should be emphasized that a, in these experiments blood or blood serum produced by far the best results; b, the colloidal solutions were the best artificial solutions used; c, in general the gelatine solutions produced a more favorable action than the acacia solutions although some of the modifications of the acacia solutions produced as good or a better action than the gelatine; and d, care must be exercised in the use of gelatine and acacia because dangerous reactions may be pro- duced with either. SUMMARY All the more important methods of treating under standard experi- mental conditions a state that exhibits the clinical signs of surgical shock which is produced by the exposure of the abdominal viscera of a dog, under a constant ether anesthesia, until blood pressure de- 100 F. C. MANN creases to the desired level, were tested. The therapeutic measures were tested after the viscera had been replaced and after determining the curve of the blood pressure. The treatment of shock is described under four headings: 1. General measures. Heat, keeping the head down, ete. The value of the classical use of heat as well as the effect of cold in helping to produce the condition, was corroborated experimentally. 2. Special measures. Strapping the limbs, rebreathing, ete. Ex- perimentally, rebreathing was not found to be of importance. 3. The use of drugs. Stimulants, vasoconstrictors. None of the drugs usually employed in the treatment of shock were found to be very effective. 4. The restoration of fluid volume. The best results in the treat- ment of experimental shock were obtained by the injection of fluid media. The data of the experiments justify the conclusion that none of the artificial solutions give such good results as the use of blood. The so-called colloidal solutions and their various modifications give better results than normal salt solution, but their potency is certainly not equal to blood or blood serum and occasionally they might be harmful. BIBLIOGRAPHY (1) Bayuiss, W. M.: Methods of raising a low arterial pressure. Proc. Royal Soc., London, Series B., 1915-1917, lxxxix, 380.; Intravenous injection in wound shock, London, Longmans, 1918. 172 p. (2) Bogart, L. J., UNDERHILL F. P. AnD MENDEL, L. B.: The regulation of the blood volume after injections of saline solutions. This Journal, 1916, xli, 189-218. (8) Burnuam, A. C.: The administration of glucose solutions as a prophylactic against post-perative shock. Amer. Journ. Med. Sci., 1915, el, 431-436 (4) Cannon, W. B.: Shock and its control. This Journal, 1918, xlv, 544-545. (5) Dae, H. H. anp Larpuaw, P. P.: Surgical shock and some allied conditions. Brit. Med. Journ., 1917, i, 381-383. (6) Dawson, P. M.: The changes in the heart rate and blood ‘‘pressures’’ re- sulting from severe haemorrhage and subsequent infusion of sodium bicarbonate. Journ. Exper. Med., 1905, vii, 1-31. (7) Erutancer, J. AND GassEr, H. 8.: Hypertonic gum acacia and glucose in the treatment of secondary traumatic shock. Ann. Surg., 1919, lxix, 389-421. (8) Ertancrr, J. AnD Woopyatt, R. T.: Intravenous glucose injections in shock. Journ. Amer. Med. Assoc., 1917, lxix, 1410-1414. (9) Fox, J. F.: The prevention and treatment of surgical shock. Amer. Journ. Surg., 1918, xxxii, 143-146. EXPERIMENTAL SURGICAL SHOCK 101 (10) Gassrer, H. 8., Merk, W. J. AND ERLANGER, J.: The blood volume changes in shock and the modification of these by acacia. This Journal, 1918, xlv, 547-548. (11) Garcu, W. D.: Nitrous-oxid-oxygen anesthesia by the method of rebreath- ing. With special reference to surgical. shock. Journ. Amer. Med. Assoc., 1910, liv, 775-780; The use of rebreathing in the administration of anesthesia. Journ. Amer. Med. Assoc., 1911, lvii, 1593-1599. (12) GEsELL, R.: Observations on the volume flow of blood through the submax- illary gland. This Journal, 1918, xlv, 545-546. (13) Gururiz, C. C.: The interpretation of the manifestations of shock. Penn. Med. Journ., 1918, xxii, 123-126. (14) HenpERsoN, Y. AND Harvey, 8, C.: Acapnia and shock. VIII. The veno- pressor mechanism. This Journal, 1918, xlvi, 533-553. (15) Hoaan, J. J.: The intravenous use of colloidal (gelatin) solutions in shock. Journ. Amer. Med. Assoc., 1915, lxiv, 721-726. (16) Howey, W. H.: Observations upon the cause of shock, and the effect upon it of injections of solutions of sodium carbonate. Contrib. Med. Re- search (Vaughn), Ann Arbor, Mich., 1903, 51-62. (17) Mann, F C.: The peripheral origin of surgical shock. Johns Hopkins Hosp. Bull., 1914, xxv, 205-212. (18) Mann, F. C.: Shock and haemorrhage. An experimental study. Surg., Gynec. and Obst., 1915, xxi, 480-439. (19) Mann, F. C.: Further study of experimental surgical shock Journ. Amer. Med. Assoc., 1918, lxxi, 1184-1186. (20) Mann, F. C.: Studies on experimental surgical shock. This Journal, 1918, xlvii, 231-250. (21) Merk, W. J. anp Gasser, H. S.: The effects of injecting acacia. This Journal, 1918, xlv, 548-549. (22) Porter, W. T.: Respiratory suction an aid in surgical shock. Boston Med. and Surg. Journ., 1917, elxxvi, 699. (23) Sepuic, M. G., TrerNey, J. AND RoDENBAUGH, F.: An experimental study of sodium bicarbonate and other allied salts in shock. Amer. Journ. Med. Sci., 1913, exlvi, 195-204. EXPERIMENTAL STUDIES ON THE REGULATION OF BODY TEMPERATURE Ill. Tur Errect or INCREASED INTRACRANIAL PRESSURE ON Bopy TEMPERATURE (Preliminary Communication) LILLIAN M. MOORE From the Rudolph Spreckels Physiological Laboratory of the University of California Received for publication July 7, 1919 In a previous paper I (1) reported a series of “heat puncture’’ ex- periments on rabbits, the results of which indicated that in certain cases hyperthermia followed puncture of the brain, but since there was no correlation between the location of the lesion and the occur- rence of the hyperthermia, the rise in temperature did not depend upon injury to the corpus striatum or other special ‘‘cénter.” The existence of special “‘centers’”’ in the brain for the regulation of body temperature was, therefore, not confirmed. The hyperthermia which occurred in 22 per cent of the cases, how- ever, remained unexplained, as also did the fatal symptoms accom- panying a number of these and other cases of puncture; the symptoms noted in such cases suggested the possibility of a condition of increased intracranial pressure due perhaps to hemorrhage in the cranial vessels. This possibility was tested out by a series of experiments in which the intracranial pressure in rabbits was increased artificially. The symp- toms produced by greatly increasing the pressure, 250 mm. or more of water, were identical with those observed in fatal puncture cases; increase in rate of respiration, slowing of heart beat, vasoconstriction, pupilo-dilatation, rise in body temperature followed by a fall before death. A moderate increase (150 to 200 mm. of water) produced less marked effects, the most apparent being an increase in the rate of respiration and vasoconstriction with a subsequent rise in body tem- perature. Practically every case with a definite increase in the pressure in the cranial cavity showed a rise in temperature sufficiently high to 102 INTRACRANIAL PRESSURE AND BODY TEMPERATURE 103 be considered experimentally produced. The findings accord with the well-known symptoms in clinical cases of traumatic brain lesions. The explanation of this influence of increased intracranial pressure on body temperature and also the fatal symptoms of higher pressure would appear to be that the pressure acts as a stimulus to the principal bulbar centers causing an increased rate of respiration by stimulation of the respiratory center, decrease in heart rate by stimulation of the cardio-inhibitory center, vasoconstriction by stimulation of the vaso- motor center, and dilatation of the pupils by stimulation of the cervical sympathetic. In case of greatly increased and prolonged pressure these centers finally become paralyzed, their functioning ceases and death follows. Dixon and Halliburton (2) offer a similar explanation of the pressure symptoms observed by them in their experiments with dogs. On the other hand, with a moderate increase in intracranial pressure there is less evidence of bulbar stimulation, the most apparent excitation beng that of the vasomotor center. The vasoconstriction resulting from this stimulation causes more heat to be retained in the body and consequently a rise in the body temperature. In fatal cases the body temperature falls again just preceding death probably because of the paralysis of the vasomotor center. These results support the view that temperature regulation is de- pendent upon physico-chemical factors without the intervention of hypothetical ‘‘heat centers.” The rise in body temperature and other symptoms attending in- creased intracranial pressure correspond so closely to those of ‘heat puncture” in which there is generally sufficient brain lesion to cause an increase in pressure in the brain cavity, and those of clinical brain lesions, that it seems possible to apply the same explanation to each of these cases. I think the rise in temperature which is reported by advocates of the “heat center” theory as due to “heat puncture” can be explained in a like manner, as can also the rise obtained in 22 per cent of my punctures and the fatal symptoms in a number of these and other cases. The detailed results of this and further investigations will be published in a later paper. BIBLIOGRAPHY (1) Moore: This Journal, 1918, xlvi, 253. (2) Drxon anp Hatuisurton: Journ. Physiol., 1914, xlviii, 317. STUDIES IN SECONDARY TRAUMATIC SHOCK VY. RESTORATION OF THE PLASMA VOLUME AND OF THE ALKALI RESERVE HERBERT S. GASSER anp JOSEPH ERLANGER From the Physiological Laboratory of Washington University, St. Louis Received for publication July 7, 1919 The method of estimating the blood volume employed in the pre- ceding paper (1) brought to light the fact that the intravenous injec- tion of a concentrated solution of gum acacia tends to prevent the con- centration of the blood which otherwise practically invariably develops while shock is being induced. This observation suggested the present set of experiments which had as their first object a more detailed in- quiry into the mechanism of this action of gum acacia, and ways of facilitating it. SERIES I. OBSERVATIONS ON NORMAL ANIMALS Procedure. Three experiments were performed on each of four normal dogs. The animals were anesthetized with morphine and ether. The first experiment on each dog consisted in the administra- tion of a hypertonic crystalloid (18 per cent glucose, usually 5 ec. per kilo of body weight). After the animal had recovered from the effects of this injection, usually on the next day, a hypertonic colloid (5 to 6 cc. of a 30 or 25 per cent solution of gum acacia, respectively) was given. An interval of 3 days was then allowed to elapse. This was regarded as sufficiently long to permit of the disappearance of the gum acacia. Then in three of the instances the animal was given, first, the acacia and, immediately after, the glucose, both in the same concen- tration and amount as had been previously employed. To the fourth animal was given a 7 per cent solution of gum acacia. All injections were made as rapidly as possible. The hemoglobin percentage, deter- mined as described in the preceding paper (1), served to indicate the effects these injections had upon the blood volume. The injections were made into the femoral vein and the samples of blood were taken 104 PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK 105 from some large artery. The arterial pressure was not followed in this series of experiments. Results. When glucose is injected, the well-known fact is confirmed that the blood comes into osmotic equlibrium with the tissues within the first minute or two. The maximum theoretical dilution produced by the dose of glucose given may be calculated thus:—A 5.52 per cent solution of glucose is isotonic with blood ea (osmotic pressure of blood in atmospheres) X 18 _ : 52); 22.4 (osmotic pressure of 18 per cent glucose) the 0.9 gram of glucose contained in each 5 ce. of the solution, the dose per kilo of animal, would therefore suffice to make ——— 0.0559 05 59 OF 16:3' Ce: of an isotonic solution. If we take 92 cc. as the quantity of blood in TABLE 1 (1) GLUCOSE ACACIA ACACIA AND GLUCOSE =~ co SS (8) ~ I oS (5) (6) => w ~~ = ~ ww (2) 0 1 © ae ae Pon Be “Ee B=ige) we wE EXPERI- 5 a5 Bs a5 ag MENT i sie Se aS Bre ‘ o-4 So : oes So : >o NUMBER | Dose per kilo Ee 3 2 | Dose per kilo | 3 4 Foun 2 | Dose per kilo | & Q Ao |) RIBS a8o | 282 Zee Sog vss Sos 956 250 Sone Reg one QDRe = ne f= PONS EI] rey == rele 3} Quetes Foi B=| a fo) a fo) ) |per ceni\per cent per cent\per cent per cent oy 5 ce. 30% Q0 O7 ) 1 3 Ce. 18° 10.6 7.0 5) G@s 30% 31.2 19.0 ott. 18% 24.0 ae J 6 cc. 25% (orf a % Fo Lee |) “620*\) 6.625% |.31.2; | 1020 {8 ec. 18% bo o ic} Q — oo So 6 ce. 25% 35.0 - Q07 7 fon 9 57, 3 5 ec. 18% | 17.7 | 12.0 | 6 ce. 25% | 31.2 | 11.0 eee 4 mec: 189, | Ti2Z 7) AOL Ouleeo cowo0%| ol. 2) | 12.0.1 18 ce. 17% A) * Solution administered too slowly. each kilo of animal the addition of 16.3 ec. per kilo will make a blood volume of 108.3 ec.; the blood volume is increased 17.7 per cent. Of the water required to effect this increase 5 cc. are furnished by the injection; the remaining 11.3 cc. must be taken from the tissues. The results obtained are given in table 1, and two of the four sets of observations are plotted in figures 1 and 2. The table shows that in 106 HERBERT S. GASSER AND JOSEPH ERLANGER 105 0 Hr 1 6 Fig. 1. Experiment 2, table 1. Blood volume, in per cent of the original, plotted against the time. Zero hour marks the completion of the injection of 5 ec. of 18 per cent glucose (—x—x—) on the Ist day, 6 ce. of 25 per cent gum acacla (— @— e—) on the 2nd day, and 6 ce. of 25 per cent gum followed immediately by 5 ce. of 18 per cent glucose (— O— O—) on the 4th day. 130 120 110 ~ 100 0 Hr. ] ay, 3 4 Fig. 2. Experiment 3, table 1. Blood volume in per cent of the original plotted against the time. Zero hour marks the completion of the injection of 5 ce. of 18 per cent glucose (—x—x—) on the Ist day, 6 cc. of 25 per cent gum acacia (— @— @—) on the 2nd day, and 6 ce. of 25 per cent gum followed immediately by 5 ce. of 18 per cent glucose (— O— O—) some days later. PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK 107 each instance the observed maximum dilution after the injection of glucose (average = 8.7 per cent) falls far short of the theoretical maxi- mum (17.7 per cent). This difference, presumably, is merely the result of the rapid loss of glucose from the circulation; it may be disappearing while the injection is proceeding, and also during the short interval elapsing between the termination of the injection and the taking of the first sample of blood. The largest discrepancy is seen in experiment 2, and in this instance it was recorded that the injection proceeded too slowly. The maximum dilution was attained within 4 to 2 minutes. The blood then began to concentrate, rapidly at first, and then more slowly, and became normal within 5 to 45 minutes. The results following the injection of the acacia solution differed quite decidedly from those obtained with the glucose. We present first our method of calculating the anticipated dilution of the blood as based upon the osmotic pressure of the acacia solution. According to data presented in the preceding paper (1) we can consider the osmotic pres- sures of 7 per cent acacia and blood serum 22 mm. and 16.4 mm. of mercury, respectively. On this basis 5 ec. of 30 per cent acacia would have to expand to 28.7 cc. to have the same osmotic pressure as the serum colloids. This would entail a 31.2 per cent expansion of the blood volume (normal 92 ee. per kilo). The maximum dilutions observed experimentally (19, 10, 11 and 12 per cent) were in every instance much less than the theoretical 31.2 per cent. The dilution of 19 per cent was obtained in an experiment in which acacia was injected as soon as the dilution resulting from the glucose had disappeared; in the other experiments the acacia was in- jected one to three days after the sugar. This maximum dilution in every instance was attained very much more slowly (in 25 to 50 min- utes) than when the glucose was injected (in 4 to 2 minutes). Of greater interest, however, is the slow return of the blood concentration toward the normal. In the four experiments the return to normal required more than 3 hours, more than 6 hours, 24 hours and 3 hours respectively. It is obvious that the colloid attracts water very slowly compared with the crystalloid glucose; therefore, at no time during the period of observation does the serum regain its normal osmotic properties. In spite of the increased colloid content of the plasma the maximum vol- ume reached is not maintained. The plasma volume is reduced toward normal in spite of its increased colloidal osmotic pressure. This is 108 HERBERT S. GASSER AND JOSEPH ERLANGER most probably to be attributed to filtration resulting from the increase in capillary pressure that would arise from the plethora and the in- creased viscosity. As acacia can maintain for some time a volume greater than normal it is to be expected that it would be even more effective when injected into an animal whose blood volume had been previously depleted, the volume-compensating mechanism not being then called into play. This is an example of the well-known pharma- cological principle that it is easier to change a function toward normal than away from normal. It follows from these observations that neither the concentrated solution of gum acacia nor the concentrated solution of glucose is per- fectly satisfactory as a means of expanding the blood volume; the for- mer attracts water slowly, though it holds it in the circulation for rela- tively long periods of time, whereas the latter attracts water with great rapidity, but the increased blood volume disappears with cor- responding rapidity. To maintain an increased volume by the injec- tion of glucose, the injection must be continuous. It was conceived, therefore, that the shortcomings of each of these methods of expanding the blood volume might be corrected by com- bining the two; that if concentrated acacia were injected it would hold the water which a subsequent injection of concentrated glucose brought it. Experiment proved this to be true. When glucose is injected after acacia we might expect, if the acacia holds all the water that the sugar can bring in, that the maximum theoretical blood volume increases of 10.6, 17.7 and 17.7 per cent respectively in the first three experiments would be added to the blood volumes resulting from the acacia injection. If we add the former percentages to the percentage increases obtained experimentally in the acacia control series (columns 3 and 7, table 1) we have theoretically possible expansions of 29.6, 27.7 and 28.7 per cent, that is, volumes within the theoretical holding power of the injected acacia. The cor- responding expansions found were respectively 24, 18 and 35 per cent (column 9). Explanation of the variations from theory would demand a longer and more carefully controlled series. The data, at any rate, clearly support the original supposition, that glucose would bring in fluid which would be held by the acacia. The maximum dilution was reached very quickly; indeed, in two of the instances it was attained within 4 minutes or practically as quickly as when glucose was injected alone. In the other case the maximum was attained with the reading made 16 minutes after the injection, but 8 minutes had elapsed between this and the preceding reading. PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK 109 After attaining the maximum, the blood volume at once begins to decline. In the two instances in which the maximum was quickly attained (experiments 2 and 3) this decline was slowly progressive as long as it was followed. At the time of the last readings made, the blood volume has always been higher following the injection of the two solutions than in the case of the injection of acacia alone. This is prob- ably due in part to the greater initial increase in volume when the gum and the crystalloid are injected together, although it has been our experience in several instances that the increase in volume resulting from a rapid injection of very hypertonic colloid may be rapidly dis- posed of. The reason for this we have not analyzed. The slow return of the blood volume to normal is seen also after the injection of more dilute (7 per cent) acacia. In the third part of ex- periment 4, 18 ec. per kilo of such a solution were injected, a volume sufficient to increase the blood volume (normal 92 ce. per kilo) 19.6 per cent. The actual dilution observed was about 21 per cent directly after the injection. The blood volume began to decline at once, but so slowly that 5 hours later it was still 111 per cent of the original. The 7 per cent acacia was chosen in this experiment on the assumption that we were working with an isotonic colloid. According to our pres- ent calculations, however, even the 7 per cent would be appreciably hypertonic. SERIES II. OBSERVATIONS ON ANIMALS IN SHOCK 1. The effect of injection of solutions of colloids and crystalloids on the circulation The main conclusion to be drawn from the experiments reported in part 1 of this paper is that in norma! animals a hypertonic colloid will hold in the circulation not only the water that it itself slowly attracts, but also the water brought to it rapidly by a hypertonic crystalloid, so that the combined injection of the two results in a rapid and well- sustained expansion of the blood volume. The colloid employed in those experiments was gum acacia; the crystalloid, glucose. In view of the acidosis demonstrated to be present in the types of shock with which we were working it was deemed advisable in the tests carried out on animals in shock to employ as the crystalloid either Na,COs; or NaHCoO;, instead of glucose. It may be assumed that the principles involved, as regards changes in the blood volume, are not altered by this change in the crystalloid employed. 110 HERBERT S. GASSER AND JOSEPH ERLANGER Methods. The animals were first traumatized by holding the arterial pressure down to 40 mm. Hg. for a period of about 2 to 2} hours by partly occluding the inferior vena cava (2), (3). This damage in our experience in time usually brings on a shock-like failure of the circula- tion. The blood volume was followed, as previously, by estimating the hemoglobin. In addition, the pressure in the carotid artery was followed by a method that entailed practically no loss of blood or mixture of the anticoagulant salt with the blood in theanimal. In a few experiments the CO, capacity of the blood was followed by the Van Slyke method. In most of the experiments the solution injected was 25 per cent sodium acacia in 4 or 5 per cent NasCO; or NaHCOs, the amount, 4 or 5 ec. of the solution per kilo of body weight, though many other com- binations, including the first Bayliss solution (4), were used. As a 1.37 per cent solution of Na2CO; is isosmotic to blood, each cubic centi- meter of a 5 per cent solution injected would have to expand to 3.65 ec. before becoming isosmotice to plasma. Our 5 per cent solution of NaHCO; lowered the freezing point of water 1.879°C. One cubic centimeter of this would, therefore, expand to about 3.35 cc. Five per cent solutions of these substances are, therefore, comparable with respect to their osmotic pressures to the 18 per cent glucose in the pre- ceding experiments. Owing to the inadvisability of rapidly injecting concentrated acacia solutions into animals in shock, the injection in these experiments was prolonged usually to 20 or more minutes. Blood volume. The changes in the blood volume that occurred are collected in table 2. In every case the increase in volume after the injection of 25 per cent gum in combination with 4 or 5 per cent car- bonate is at least as great as that calculated to be necessary for all the crystalloid injected to become isosmotic. And in no instance 1s the increase in blood volume, determined shortly after the injection, any greater than the volume of water that the injected gum acacia is theo- retically capable of holding. But in experiments 22, 25 and 30 the blood volume continues to increase so that the ultimate dilution is greater than the theoretical holding power of the acacia, though within the limit of error of the methods, in the case of experiment 30. The explanation of this result is not obvious. There is the possibility that the improvement in the circulation permits the organism to partici- pate in increasing the blood volume. * A glance at the table will show that other solutions of gum acacia, such as simple 7 per cent gum acacia and 6 per. cent gum acacia in faa SHOCK PLASMA VOLUME AND ALKALINE RESERVE IN PeId |0cT-08| Z8T | OST | 08 (2°26 11 | Z& 18°96 |8't6 | 9°18 opty tod *09 gy ‘wrovow %y | py ped | SIL | 86 OIL V'OS |Z GOT |Z SIT |Z°SOT | 8°98 | OTPY tod 99 BT “OQOHYN %Z ul viovow %g | gy ped | 89 8g Ocl | €f |O'VIT | FSS} SL IS PIT |Z TIT | 2°S8 | O[Fy 40d 00 ZT “OOHYN %Z ul viowoe %g | 9z Ped | O1L | 06 GOL | Sct |OTIT| 828 | € |F'S2r |F'Z2T | 8°88 JoP{ aod 09 7 “OOPUN Mppy'g Ur viovov %y | egy ped | O01 | 2 GOL | #2 |F OOT FOOL | 0°08 | OTP] tod 09 F “OOHYN %p ur vrovov %ez | Tp Peary | GIT | OOT 9 Gla: aleOmtc 616 [6°16 | 6°SZ | OTP ted 00 Gg “OQOHVN %p ur vrowov %ez | ep ped | ZOL | G6 OCT CO SOs Gn alah eed 06 |0'F6 | 0°08 | OTF] tod ‘00 F “OQOHYN %p Ul viovov %cz | Fe ped | #8 0S OSI | OOL |9°ZIT | 9 TE} OOT |9°ZIT |O'6OT | 9'°S8 | OFT tod 00 F “OD*wN %G ut viovov %ez | og POAYT | OST | &6 Ocl | IGE |2°86 | OGL | ZS |Z’ FOT |Z FOT | 9°28 | OTF] 10d 00 F “OO**N %G ur viovov %ez | Hz PetGh Oz SP OFT | SL jo Le \éh' 1G | 82 |Z LZ1 JO'SIT [40 F8 | OTF tod “00 Gg “ODN Ye ul viovow %qz | z Ped | O& | SOL | GOL | OCE |e ScT | O'Sh | 29 |O'TET |8'SIT | 2°06 | OTP{ 49d *909 g “OD UN %e ut vrovow %az | zz 3 6 é 7DUL LOU JDULLOW | JDULLOU | JDULLOU ees oH Ree saynurue fo quad Jad |saynurvue fo fo fo quad sad quad sad\juao sad\quaoa sad Desaali ators 101 u01 iy y P -vofur |. osvoaro | -ovofur |. y Joye | -yvorq | perpuy qoyyu aUIN[OA] Ly 19}j8 OUIN]|OA i. Ajeyzerp | o10jogy Tale: soqn a : s Ou, OUILy, is ~-OWUIy : . “UIUL § | wanton = eT [<> UMM | woos INGWLVOLL 4 UOIPVAIOSGO 4SB'T WIN UIIX By 4 GUASsaud IVINALAV 5 is} NOILOALNI WOUT SHDNVHO AWOIOA aootd E) 6 WTAViL 112 HERBERT S. GASSER AND JOSEPH ERLANGER 2 per cent NaHCO; (4), when osmotic properties as well as volume of fluid injected are taken into account, are quite as efficacious in restoring and maintaining the blood volume as are the stronger solutions. While there is often some recession from the maximum dilution this is never large, and in five of seven animals in which the concentration was followed for over 1 hour after the injection, the volume at the last read- ing is as large as that obtaining within the first 3 minutes after the injection. When these experiments were performed we believed that 7 per cent acacia was isosmotic with the colloids of the blood and that its effect on the blood volume was merely additive. On the basis of data presented in the preceding paper (1), we now consider the colloidal osmotic pres- sure of serum to be 16.4 mm. of mercury and of 7 per cent acacia 22 mm. We have recently dialyzed a7 per cent solution of acacia in the same Locke’s solution used in making the above determinations, against serum in a Moore and Roaf osmometer, arrangement being made for intermittent renewal of the serum in the lower chamber. The pressure developed was 6.6 mm. of mercury. The pressure on the acacia was released after the first determination and the experiment then con- tinued with the salts in the acacia solution modified by the interchange of salts with the serum during the first determination. The pressure developed in the second determination was again 6.6 mm. ‘This pres- sure is to be compared with the difference of 5.6 mm. between the separate direct determinations. After the injection of 18 ce. per kilo of 7 per cent acacia the osmotic pressure of the plasma colloids is calculated to be 8.5 per cent above normal. After the injection of 5 ec. per kilo of a solution containing 18 per cent glucose and 25 per cent acacia the colloids would be 10.7 per cent hypertonic after enough fluid had been brought in from the tissues to render the crystalloid isotonic, assuming that all the glucose stayed in the vessels till diluted. It therefore follows that after the injection either of the 7 per cent acacia solution or of the very hyper- tonic gum and glucose solutions in the above quantities the blood plasma is left with a colloidal osmotic pressure somewhat above normal. In the solution containing 6 per cent acacia and 2 per cent NaHCO; both the colloid and crystalloid are slightly hypertonic, even after mak- ing allowance for the loss of carbonate in the precipitation of CaCOs. Table 2 shows that the blood dilution resulting from the injection of the 6 per cent gum acacia solution in 2 per cent NaHCO; was greater PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK 113 than could be accounted for simply by the addition of the volume of water actually injected. Blood pressure. Practically invariably the arterial pressure is raised by the injection of any of these solutions containing gum and crystal- loid (see table 2) even when the injection of the crystalloid alone is without effect (see a, fig. 6).. The effect of the injection upon the blood pressure, however, seems to be merely an incident in the course taken by the pressure as determined by the reaction of the animal to the clamping of the cava. If at the time the gum solution is injected the pressure is rising, the injection facilitates the rise somewhat; but if the pressure is falling, the solution while usually causing the pressure to rise for a time, fails to maintain the pressure. Sooner or later the pres- sure begins to fall again, and the animal invariably dies. This fall in pressure may occur even while the blood volume apparently is increasing. 2. Alkali reserve in experimental shock Before presenting the effect on the alkali reserve of the acacia-car- bonate mixtures, the effect of the shock-producing procedures which we have employed, upon the carbonate content of the plasma should be recorded. This study was made because at the time these experiments were performed considerable importance was being attached to the reduction in reserve alkali which had been found to obtain in clinical shock (5). In a number of animals we have followed with the Van Slyke ap- paratus the changes in the CO: capacity of the plasma of arterial blood drawn without loss of CO. This is regarded by Van Slyke and Cullen (6) as the ideal method of estimating the alkali reserve. Most of the determinations were made in duplicate. The data are collected in figure 3 in which the volume per cent of CO. is plotted against the mean arterial pressure. The first blood sample in each case was drawn after anesthetizing the animal with morphine followed by ether and after making all of the operations preparatory to starting the procedures by which shock-like failure of the circulation was to be induced. It is seen that these readings ranged between 50.5 and 26.1, and usually were below 40.0. : In order to avoid any complications that might arise as a result of the withdrawal of blood, it was necessary to reduce the number of readings of the CO. capacity to a minimum. As a rule, therefore, only one other reading was made and that at a time when we felt as- THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 114 HERBERT S. GASSER AND JOSEPH ERLANGER sured that a shock-like failure of the circulation had become established. The readings made at that time, with but one exception (no. 2), show a markedly reduced COz capacity. And, with but this one exception, the inclination of the line joining the initial with the shock reading in almost every instance is remarkably constant; these lines are all nearly parallel to each other. JGOMMISOHQI40 «130 120010090 80 70 60 50 40 30 Pee Fig. 3. Showing the relation of CO, capacity of arterial blood (ordinates) to arterial pressure (abscissae). The first reading in each case is made before start- ing to induce shock; the subsequent readings later by the time (in minutes) indicated in each case. Shock by intestinal exposure, — @—; shock by caval occlusion, — O—; shoek by aortic occlusion, — O—; shock by adrenalin injec- tion — A—. On the other hand the gradient of the curves or of the parts of the curves joining readings made before the fall in pressure in general had exceeded 50 per cent of the original value is not quite so steep. This possibly may mean that the tissue alterations that lead to the reduc- tion in alkali reserve do not progress rapidly until the condition that brings them about has reached a relatively advanced stage of develop- PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK 115 ment. This inference is borne out in the main by the curves of the relation obtaining between the CO, capacity and the volume flow of blood published by Gesell (7). His curves as a rule show that the flow of blood through the salivary gland may be below 50 per cent of the original for some time before the COz capacity begins to fall. It would seem, therefore, that the reduced COs capacity is the result and not the cause of the fall in arterial pressure. The gradient of the line joining the initial with the final readings as may be seen in figure 3 is largely independent of the time. It can also AP. 120 150 co, 60 0 100 40 0 Hrs. ! 2 3 4 5 Fig. 4. Experiment 22. Arterial pressure, — @— @—; blood volume, -- @-- @- -; per cent of normal, determined by the reciprocal of the Hb. content; CO, capacity, --@---@-- The figure shows the acceleration of the rise of pressure, that follows release of the cava, by the injection of 5 cc. per kilo of a solution consisting of 25 per cent sodium acacia in 5 per cent Na,CO;, and the subsequent maintenance of a good pressure. It shows, also, a tremendous in- crease in the blood volume resulting from the injection. The blood volume is still high at the close of the experiment, more than 2 hours after completing the injection. It also shows the rise in the CO» capacity. This animal died before twenty-three hours had elapsed. be seen that though the lines tend to parallel each other they may be on widely separated levels, so much so that the final readings in a case of advanced shock may be within, or almost within, the range of the nor- mal readings of other cases. While there is no clear direct relationship 116 HERBERT S. GASSER AND JOSEPH ERLANGER between the CO. capacity and the arterial pressure, it nevertheless is obvious that a reduction in the carbonate content of the plasma from arterial blood is a constant accompaniment of shock as we have induced it. The alkali reserve after injections of solutions containing acacia and carbonate. The carbonate content of the arterial blood was followed in two of the cases that received 5 per cent Na,CO; with the acacia. (0, 60 Vv. AP. 100 100 60 0 y 2 3 4 5 Fig. 5. Experiment 24. Four cubic centimeters per kilo of 25 per cent acacia in 5 per cent Na,CO; were injected. Essentially the same results were obtained in this case as in experiment 22. This animal recovered. 0 Hr 1 2 3 4 Ry Fig. 6. Experiment 34. In this case the pressure began to fall when some- thing less than 54 minutes had elapsed after removing the clamp from the cava. At a time when the pressure was stationary at 100 mm. of Hg. the animal (at a) was given 4 cc. per kilo of a 4 per cent solution of NaHCO;. The pressure fell somewhat while this was being given. A solution consisting of 25 per cent so- dium acacia in 4 per cent NaHCO, 4 cc. per kilo, was then given (at b). Under the influence of this injection the pressure rose and the blood volume increased, but neither was quite restored to normal. Almost immediately after terminat- ing the injection the blood pressure began to fall again despite any measureable decrease in blood volume. This animal died. PLASMA VOLUME AND ALKALINE RESERVE IN SHOCK ELF One received 5 cc. per kilo of body weight, the other 4 ec. In the first case (fig. 4) the volume of combined CO: was raised from the shock level of 19.8 per cent to 50.6 per cent; and in the second case (fig. 5) from 20.6 per cent to 48.6 per cent. The values attained in each case, therefore, were quite normal for dogs under ordinary conditions of experimentation. The ultimate results of the injection of a solution containing hypertonic acacia and carbonate. A general discussion of the effect of the injections on the blood pressure, the blood volume and the alkali reserve has been given. There remains to be considered the fate of the individual animals. In figures 4, 5 and 6 typical experiments are graphically presented. It is obvious from these and similar experiments (see table 2) that an intravenous injection which raises the blood pressure, and maintains the rise for the period of the experiment, which restores the blood volume and the alkali reserve in an animal that has been damaged to an extent that in time usually results in a shock-like failure of the circulation, may not prevent a fatal issue. One cannot help but conclude, therefore, that these alterations in the state of the animal do not constitute the primary cause of death from shock; in other words, they are merely symptoms of some more fundamental process, which determines the condition of the animal but to which the former may be contributary. SUMMARY 1. When glucose in 18 per cent solution is injected into the circula- tion of a normal animal the blood comes into osmotic equilibrium with the tissues within the first minute or two, the average maximum dilu- tion amounts to but half of the theoretical maximum and the blood regains its normal concentration within 5 to 45 minutes. 2. When gum acacia in a concentrated solution is injected the aver- age maximum dilution of 41.7 per cent of the theoretical maximum is attained within 25 to 50 minutes; the decline of the blood volume to normal requires 23 to 6 or more hours. 3. When the concentrated acacia is immediately followed by the glucose the maximum dilution is quickly attained and is much greater than that resulting from the injection of either of the two substances alone. The dilution is well maintained. 4, Comparable results are obtained in animals in shock when a strong solution of gum acacia is followed by a solution of NasCO; that is isosmotic to 18 per cent glucose. 118 HERBERT S. GASSER AND JOSEPH ERLANGER 5. Withsuch a combination of solutions given in appropriate amounts the blood volume, the blood pressure and the reserve alkali of animals in shock often can be brought to normal and held there for the usual duration of an experiment. Yet such animals, as well as shocked animals treated with other combinations of gum acacia and carbonate or bicarbonate, often died within 24 hours. BIBLIOGRAPHY (1) GassER AND ERLANGER: This Journal, 1919, 1, 31. (2) JANEWAY AND JACKSON: Proc. Soc. Exper. Biol. and Med., 1915, xii, 193. (3) ERLANGER AND GaAssER: This Journal, 1919, xlix, 151. (4) Bayuiss: Private Rept. of the English Committee on Shock, November 29, 1917. (5) Cannon: Journ. Amer. Med. Assoc., 1918, lxx, 531. (6) Van SLYKE AND CULLEN: Journ. Biol. Chem., 1917, xxx, 366. (7) GesELL: This Journal 1919, xlvii, 468. STUDIES IN SECONDARY TRAUMATIC SHOCK VI. SrTaTIsTICAL STuDY OF THE TREATMENT OF MEASURED TRAUMA WITH SOLUTIONS OF GumM ACACIA AND CRYSTALLOIDS JOSEPH ERLANGER anp HERBERT 8. GASSER! From the Physiological Laboratory of Washington University, St. Louis Received for publication July 7, 1919 INTRODUCTION With the growth of our personal experience with experimental shock we came more and more to feel that a solution consisting of hypertonic gum acacia and of a hypertonic crystalloid would prove helpful in counteracting the disturbances which seemed to be responsible for the failure of the circulation that constitutes the main feature of that state. The growth of this idea may be briefly outlined. Our first experi- ments had as their object a study of the hemodynamics of shock. They convinced us that a giving way of the vasomotor center or of the heart is to be regarded, speaking broadly, only as a relatively late conse- quence of a low arterial pressure, and not as its cause. In the early stages of shock as induced by exposure of the intestines (1), by partial occlusion of the aorta or of the vena cava (2), or by the administration of adrenalin (3), the only constant circulatory fault seemed to be a reduction in blood volume, both actual and effective (4). We found in all types of shock, among other less constant lesions, marked dila- tation of the capillaries and venules of the villi of the intestines. Since our methods of inducing shock all seemed to depend upon the effects of slowing the blood stream, since furthermore the same vas- cular changes can be seen to develop during partial occlusion of an artery (5), and since the shock that follows the administration of adrenalin seems to be referable to the constriction of the arterioles it produces by local action, we ventured to suggest (6) that in man the 1 With the assistance in some of the experiments of Paul C. Hodges, 119 120 JOSEPH ERLANGER AND HERBERT S. GASSER prime factor, also, might be a strong vasoconstriction compensating hemorrhage and wound weeping, and aggravated by exposure and pain. The changes thus produced in the capillaries and venules by the slowing of the circulation would account for the concentration of the blood found in experimental shock by attributing it to transu- dation of plasma and they account for the reduction in effective blood volume through stagnation in the dilated vessels. Although the disappearance of plasma from the circulating blood is of constant occurrence in shock experimentally produced, there is evidence that while plasma is disappearing, fluids are being added to the blood (4). The latter process may be regarded as an effort on the part of the organism to combat the diminution in blood volume. Evi- dence has been obtained indicating that the ability of the organism to thus make good the loss in blood volume decreases as shock develops. It was while we were studying the blood volume in shock that we happened upon the observation that the concentration of the blood that develops during shock induction does not occur, or at least is not so marked, when shock is induced after the administration of a dose of hypertonic gum acacia. This result we were able to show is due in part, at least, to the osmotic pressure the hypertonic gum exerts (7). It was found, however, that hypertonic gum solutions attract tissue fluids into the circulation very slowly. Hypertonic crystalloids in- jected intravenously, as is well known, attract water quickly. When the two, the hypertonic gum and the hypertonic crystalloid, are in- jected to all intents and purposes, simultaneously, the gum holds the water the crystalloid quickly brings into the circulation. Presumably, the blood volume is thus increased by a process that resembles the one the organism itself employs in combatting a reduction in blood volume. Inasmuch as the reserve alkali was found to be reduced in the types of shock we were studying (7), it was felt that sodium bicarbonate or sodium carbonate might be made to act at one and the same time as the hypertonic crystalloid and asa supply of base. We therefore studied the action on animals in shock of hypertonic gum acacia and hyper- tonic bicarbonate, as well as other combinations of gum and crystalloid and found (7) that while all equivalent combinations acted on the arterial pressure, the blood volume and the alkali reserve approxi- mately alike and about as had been anticipated, many of the animals, nevertheless, died. It thus became obvious that in order to ascertain whether these or other similar solutions that have been proposed or may be proposed TREATMENT OF TRAUMATIC SHOCK OA for the intravenous treatment of shock are of value, it would be neces- sary to study their action on a standard condition, one whose course, if left to itself, could be predicted with a reasonable degree of certainty and to use as the criterion of their efficacy not their effect on the blood pressure, or on the blood volume, or on the alkali reserve, but actual recovery from the state of shock. In the present state of uncertainty of our knowledge with regard to the nature of shock, no other criterion can be convincing. It was felt that the conditions for such a study would be supplied best by exposing an animal to a form of trauma which in time brings on a shock-like failure of the circulation, and to so grade that injury that it would just cause death if things were allowed to take their course. To state this conception in another way, the interpretation of therapeutic tests would be entirely free of ambiguity if they could be made on animals on which a minimal fatal amount of injury had been inflicted, and if recovery from the im- mediate effects of the damage were taken as the criterion of efficacy. Of the various shock-producing procedures with which we were familiar only two seemed likely to yield to such standardization, namely, partial temporary occlusion of the aorta or of the inferior vena cava. Exposure of the abdominal viscera has been so used by Mann (8), but we doubt if such damage can be uniformly applied. Other forms of mechanically produced tissue damage are very uncertain in their effects or cannot be inflicted aseptically (9); massive doses of adrenalin do not always cause death through shock; and it is still very questionable whether acapnia (10) or fat embolism (11) bring on shock properly so-called, either clinically or experimentally. THE STANDARD DAMAGE After a few trials we decided to use as the damage partial occlusion of the vena cava (12). The exact procedure finally adopted as the result of a number of preliminary trials was as follows: Under mor- phine and ether anesthesia, and under strict asepsis, a clamp, adjustable by means of a finely threaded screw, is placed on the inferior vena cava of the dog between the diaphragm and the liver, and the vein is so compressed as to attempt to hold the arterial pressure at 40 mm. Hg. for a period of 2 hours 15 minutes. The ether is discontinued after applying the clamp, and owing to the apathetic condition of the animal, it need not be administered again. As a rule the arterial pressure can thus be brought down to 40 mm. Hg. at once and held there. But sometimes the arterial pressure at 122 JOSEPH ERLANGER AND HERBERT S. GASSER first does not fall to 40. In such instances, however, the pressure as a rule tends to fall, and usually reaches 40 mm. Hg. in the course of a few minutes, when the clamp can be opened and adjusted so as to hold the pressure at 40. Occasionally, however, the pressure falls quite slowly and in a few instances it has failed to reach the level of 40 mm. Hg., even by the end of the 2 hour and 15 minute clamping period. We have in several of the latter instances determined the position of the clamp by post-mortem examination and invariably have found that it was occluding the cava. Nevertheless, recognizing the danger of including in the series cases in which the position of the clamp may have been faulty, we have once and for all excluded all instances in which the pressure by the end of the period had failed to fall to within a millimeter or two of 40. It may be added that the ptessure is just as apt to fail to fall when the initial arterial pressure is high as when it is low. Experience has indicated that the cases in which difficulty is ex- perienced in lowering the pressure are rather more apt to recover from the effects of the caval occlusion than cases in which the pressure can be brought down to 40 mm. Hg. at once. Indeed, in one or two in- stances the former, after removal of the clamp, have recovered without exhibiting any of the evidences of shock. The more favorable course of such cases seems capable of two explanations: either a, the high pressure protects the heart and medullary centers from the damage we have reason to believe (2) is done by the long-continued low pres- sure of the clamping period; or b, the high pressure of the clamping period is the result of an unusually good collateral circulation which protects the posterior parts of the body, as well as the anterior, from the effects of the occlusion. But whatever the explanation, the fact remains that failure of the pressure to fall seems to favor recovery. Some rule must, therefore, be made that will take this fact into account. We feel that recovery would be favored but little, if any, where the period during which the clamp is tight and the pressure above 40 mm. Hg. does not exceed 30 minutes. This arbitrary decision is applied in the statistical treatment of the experiments; it will be seen that the exclusion of such of these cases as recover does not materially alter the results. When the clamp is removed the arterial pressure rises with a bound, sinks again almost at once, usually to somewhere between 50 to 75 mm. Hg., and then mounts slowly. Experience has shown that the animal will die if now the pressure begins to fall consistently, even TREATMENT OF TRAUMATIC SHOCK dye though slowly, before 2 hours have elapsed.? This outcome we have not succeeded in preventing by any of the forms of treatment we have employed. We therefore exclude from the series those cases in which the pressure begins to fall consistently before 2 hours have elapsed, and treatment is not begun until this two hour period of observation has passed. One case (no. 215, table 7) has not been accepted although, strictly speaking, it is not excluded by this rule. The pressure in this case did not fall during, though it began tofall immediately after, the conclusion of the observation period and before treatment was begun. This has not occurred in any other case. Furthermore, the pressure at its highest did not get above 64 mm. Hg., a level which is 6 mm. below the lowest maximum observed in the whole series of 168 cases. If the animal does not die within 48 hours, timed from the beginning of the clamping period, as a consequence of the damage done through clamping the cava, very marked improvement almost invariably is apparent particularly as regards the state of apathy. In eight instances, however, animals that had shown this improve- ment were found dead unexpectedly, seven at the close, the eighth at the beginning, of the third day. Such animals, obviously, had re- covered from the immediate effects of the injury, but died of other causes. They are, therefore, included in the category of those that recovered from shock. Our reasons for thus treating these cases will be discussed later. A number of animals have become unavailable on account of certain accidents. Thus, animals that have died within 2 days and in which, at autopsy, extensive abdominal hemorrhage was found, must be discarded, for here the animal has had to contend not alone with the - 2 There has been but one exception to this rule, and that exception was not a clean-cut one. In the case of dog 208 (see table 3) after removing the clamp the pressure rose from the minimum of 55 mm. Hg. to a maximum of 97 mm. Hg. in the course of 27 minutes, but during the next 30 minutes it fell gradually to 80. Toward the close of the 2-hour observation period it began to rise, reaching 90 mm. Hg. 2 hours 10 minutes after removing the clamp. The animal’s pressure, followed 2 hours longer, continued to rise, eventually reaching 95 mm. Hg. This animal recovered. It can scarcely be said that in this case the pressure began to fall consistently before the close of the 2-hour period; or as a matter of fact it eventually began to rise. It should be added that this was also a cardiac case (see below) and in the effort to carry the animal through the clamping period of 2 hours 17 minutes, it twice was necessary to partially open the clamp, once for a period of 8 minutes and once for a period of 2 minutes. 124 JOSEPH ERLANGER AND HERBERT S. GASSER effects of the temporary anemia of the clamping period, but also with the effects of the hemorrhage. In most instances this hemorrhage was unavoidable and usually due to tearing of the liver through tension exerted by the clamp on an unusually broad suspensory ligament. Cases exhibiting slight abdominal hemorrhage at autopsy have been included in the series. The decision as to whether a case that is on the border line is to be included in, or excluded from, the series must of necessity be arbitrary. We have therefore presented the results obtained after excluding every case showing abdominal hemorrhage of any degree whatever. It will be seen that this way of viewing the data affects them quite distinctly. Abdominal hemorrhage, it might be added, has been no more frequent in treated than in untreated animals. This fact is mentioned in order to allay any suspicion that the hemorrhage might be a consequence of treatment. We add, also, that we have observed no evidence of a tendency to bleed other than that attributable to the rise in pressure produced by the injection. The control series, that is, the series of animals used for the purpose of determining the mortality rate from the standard damage, includes 23 “acceptable” cases. The acceptable group, to repeat, is made up by excluding cases in which the arterial pressure began to fall before the 2 hours had elapsed after removing the clamp, or in which at autopsy considerable abdominal hemorrhage was found. It is seen (table 1) that 52 per cent of the animals recovered. This method of doing damage, therefore, does not meet one of the conditions required of the ideal test; the damage inflicted does not just kill all the animals. We were unable, however, to come any closer to the ideal. When the clamping period was prolonged beyond 2 hours 15 minutes or when the - arterial pressure was held below 40 mm. Hg., so as to get every un- treated animal to die subsequently, treatment was absolutely without avail. The method adopted of preparing animals for treatment is not, therefore, as satisfactory as we had hoped to evolve, but though the method is not ideal, we felt convinced that by using a sufficiently large number of animals and by basing conclusions only on decided results, it would suffice our purposes. On the other-hand the method has the advantage of indicating, through variations in the death rate, deleterious action as well as beneficial action. Analysis of the control series, as well as of the pre-treatment stage of the several treated series, indicated the possibility of using in the interpretation of the results of treatment other data than those furnished by the number of deaths. Thus, it is seen that the span of - TREATMENT OF TRAUMATIC SHOCK 125 life in the fatal cases varies. It is obvious that if any particular form of treatment were injurious, its deleterious action might manifest itself not alone in an increase in the number of deaths but also in a shortening of the duration of the anti-mortal stage. As there were long periods during the night when the animals were not under ob- servation, and as most of the animals, especially those dying within TABLE 1 Summary of mortality statistics EXCLUDING wp ae AFTER EXCLUDING ete le eae CASES ay =f S 4 aavonadiGe 24 CALIDIUNG a ae ae t 3 : = HOURS HOURS CASES LIVED < $| a & Deaths 3 Deaths Deaths eo z | = within #2! | within within a S 48 hours = g 48 hours 48 hours =) — Reese eet © Peer Mei es ice SolGRY ecules ee 2 ail ies te || R 5 5 <|/e/8/°/;]8le/8/ Sl Sishle/s)/cle]s] ce Be coeoumen teal caeeiiosa oles ete lcs ls |e Sl yA et SE 74 |) fe Sr Re Sil 4 |e Sia ay feu 1 2 8 4 6 6 7 8 9 LOMSLIANE LA Sal Le |S N16! a7 Controlst:.4......... 40} 23}11 | 48} 4 | 19) 7 |86.9] 4 | 1 | 22/10 | 45} 18/11 (61.2 6 in 2 per cent gum- ; 16 56.3 bicarbonate.......| 27| 20] 9 | 45] 3 | 17] 6 35.3] 2 | 2 | 19) 8 | 42) or! 9 | or 17 53.0 25 and 5 per cent gum-bicarbonate..| 27| 16] 9 | 56] 2 | 14) 7 |50.0/ 8 | 6 | 15] 8 | 53] 13] 9 [69.3 25 and 18 per cent gum-glucose....... 43} 20) 9 | 45] 4 | 16) 5 |31.3] 1 | 0?) 18] 7 | 39) 19) 9 |47.4 25 in 18 per cent 18] 2 |11.0 gum-glucose...... 33| 21} 5 | 24] 4*) or] or] or | 3*/ O | 20) 4 | 20} 16] 5 [31.3 Peer ie o-9 * One of these cases had no abdominal hemorrhage, but the pleura contained a large amount of bloody exudate and the liver was the seat of a chronic process. 24 hours, succumbed during the night, we find it impossible in most instances to give the duration of life in hours. It is, however, possible to accurately divide the fatal cases into those that die within 24 hours and those that die within 48 hours, and this we have done. The arterial pressure also furnishes information of value in the analysis of the results obtained. Tables 2 and 3 show that in the control series the average initial arterial pressure is very much higher 126 JOSEPH ERLANGER AND HERBERT S. GASSER in the case of the animals that died (109.8 mm. Hg.) than in the case of those that lived (89.4 mm. Hg.). Excepting one group, this is true also of the groups of animals that were subsequently treated. This observation can mean but one thing, namely, that a high initial pres- sure prejudices the animal’s chances of recovery from the effects of holding its pressure down to 40 mm. Hg. for a fixed period. To what this may be due is a question we need not attempt to answer here. TABLE 2 Summary of average pressures | DIED RECOVERED OF SHOCK a PA > = ole to “ 2 ~ es me S 2 2 B wn a) & as A a = 2 3 3 o 2 a 5 2 oe 3 gees) 8s | 23! os) 8] Sees = : 3 2 a o w | S A 20 © Ez a Q, Zn g2& 28 o = Zn £8 £8 4d |a2/2| a | 25/28] 23 |e! s | 22) e8) ge = 2.O = es} a 2o ne £) -s no no ne = orc 5 - Oo oF oo = ba o= o8 oo q = =) te ao a> & A i= | Ton! qe me m | ae pe |ley2e a a Zila 84 a a 1 2 3 4 5 6 7 8 9 10 11 12 ‘ | S Controls.......| 99.6|102.9)11 |109.8/102.4|105.0/105.0 {12} 89.4/103 .4|109.2|109.2 6 in 2 per cent gum-bicar- bonate...... 100.9|101.7| 9 |107.9/101 .9|102.6]111.1 |11| 93.9/101.5/103 .7|113 .4 25 and 5 per cent gum-bi- carbonate....| 94.2/103.2| 9 | 98.8) 98.3] 99.8/113.0 | 7] 89.7|108.1/109.3/118.0 25 and 18 per cent gum- glucose...... 105.0)105.0} 9 |106.6) 97.5} 99.6/115.1 |11/103 .4)112 .6/115.1)131.2 25 in 18 per cent gum- glucose...... 103.1/103.1) 5 |102.0)100.2/103.0;118.0 |16/104.2)106 .0/106.8/117.6 Averages..... 100.6103.2) /105.0/100 1/101 .2}114.3* 96 . 1/106 .3/108 . 7/120 .0* * Controls not included. The important point is that on the average the animal with the high initial arterial pressure evidently is handicapped. On the other hand it is seen, and for this purpose again all of the groups are available, that by the end of the 2-hour period of observa- tion the arterial pressure of the cases that live, on the average, reaches a level considerably above the initial, whereas the average arterial pressure of the cases that die usually falls short of reaching the initial level. Presumably, therefore, speaking generally, a high post-decom- pression pressure favors recovery from the effects of clamping. G 601 P S01] F'68 | “AAV SOT F COT! 8601) “PAV ~ ay OF PAOG’ GL | O' OTT] 2 0 801) 082) SIT OF PAOQB /CT OF 9A0G” |FF | O'OLT] O OOIT) O82 | LOL | ‘sthoy FZ UTYyIIM psig Ta 98 FE | 068] 006] OTT OF PAOGE /JF | O' COT] ST 0°86 | 008 | 6IT SINOY FF UIYIIM po9iq ar OL 0 O'OLT) O86 | 90T OF PAOQK 61 0° OOT) 21 0:96 | 008] 2Z¢ “SNOW FG UIYYIM pod ar O€I 96 | O'SZI| O'OOT| FOT QO SEI] Sel | 0 STI] oss | 202 AP C6 0 0°96 | 0 2OT) ZIT % /, uedo re Afjaed =“ oF sanoy a O'GIT] FS | O'OOT) 0°06 GIL | @ podureja fovrpaep = 06 && | OF8 | O.OTT) ZO e Q OT; €@ | O'OOT] 0:06 | ZOT = SOL GI O° OOT; O'OTT) TIL iS OT uedo Ap,aed : A ‘LT sanoy Zz podureyo S ‘OVIPIVD “Of PAGE ,F | 0'S6 | G6 | O88 0°S6 | 806 = 66 GL 086 | O'SII| 2 = OF PAOG’ (ET | O'O0T! O 0 O01) 0°: $6 | OZI SINOY FZ ULYIIM pol ate GIL CT O'SOl) O'STT} S01 BH PTET fea ‘[[9M pouloes jeuITUE 3 ‘3yT “UU GOT fsanoy A 6G }8 oInssord poor | 0'OOL! O 0 001) 0°96 | Gs OF VAOGE |) = £6 Go | 0:06} 0: 0G} 86 S OF PAOG’ 98 | O'SZI] GE | O' Zell] O'ZOT| OOT OF VAOGE 78 ral oll LS |) OF LLL) OVOGE) “96 = Q'SOl] OG | O FOL] O'GOT] &6 = CEI Io | 0 OGI| O ScTi TOL Fa “ULUL $ Un UL "UU “ULUb “Ubu “Uh Ub "UU “Wut ial rg Lar a) Lal 102) wf ta] lar} tg Lo 7) BSS ee al a lay re | ¢ |3ob| 88] & | & @ |880) 22 | & B Bisur| Pepa \ aero | Waehe| aie 2 a | ad Bas 9 E 3 Eh olf ene Ea g 3 SyIvuLoy sie | So 2 =| SYAIVULIY ae Ea 5 8 B Bal g=| 2 | ae Badges eee eae sho ® . EA Hom a A surpwor qserq Bo Fs Sulpvod ysery 2, P 1p ®, "P ae ee a Ee OL ee | eae eee MOOHS JO GIUNAAOOTU avid en ee ee a ee $91U0S JOLJUO GASSER JOSEPH ERLANGER AND HERBERT S. 128 i a a as Se OF 2A0Q® ,9T | 086 | 0°86 OF PAOGE ,0E Asdoy -ne oN ‘ABp pag JO pus perdi ‘OF edge ,9¢ Asdoy -ne ON ‘Atp pig Jo pus pad OF PANG’ | FET Oy PA0GG TT OF PAOGS (FET SYIVWUII YT Ne ee eee eee ee) P Sit] 2 SOL 0 OTT) 0 OOT 0 OTT} 0'S8 0° 021} 0 O2T 0 O0T} 0°06 0 OFT} 0° O€T 0 O0T} 0°'S8 0 OdT| 0 9TT 0 OOT| 0'S8 0° OZT| 0 SOT 0 O&€T) 0° OZT UU “Ubu poe | Seay Be o 8 me | BE oO Bo Se | aie quoUlzVILy, G TOT] 6°86 | “PAV 8& If 9F qusulyeel} pus potsed woryeAses -qo uaaMjoq BUILT, MOOHS JO GAURAONTH 0'F6 | 0:02 0°26 | 0 SL 0°S8 | 0:08 0° 021) 0°08 0°28 | 0 €8 0° O&8T! 0 OOT O FS | 0 OOT 0’ 9LT) 0 OOL 0'SL | 0 SOL 0’ SOT} 0 OLT 0' O21) 0 O&T “UUs “WU ail E aoa | BF acd] 9 oO eg} 2 geil L&I T¥1 eal 2) 621 Jaquinu [BIIag 9°cOT + | 0901) 096] @ 0°96 | 008} S2r — | 0021} O'OTT) 9f | O'80T| OS8 | EFT Sano FZ UIYIM peqd | — | O'SO1] 0S6| EF | 0 F6 | 0°06} £9 == ORDER 0-008 0 OOT| O' TOT) 8g + | 0°02T| O'STT| €& O'STT] O'OTT] Eat 4 uedo Ayaed = dureyo “GT smoy Z pedureja ‘fovrp “IBD ‘OFAA0GR,ZT | — | 0001; 096) ZI | 096] 0 0cr Let OF PA0gB 24 | — | 0 OTT) O'OOT) O 0001) 0 O21) 6&T OF PA0Q® ,9 + | 0/96 | 0°86 I 0°€6 | 0081) 9ZT OF V9AOGB ,& “SAMO 7% SuIyIM potq | — | O'SET] O'OZT] EST | O'LTT| O'GET| FET “WU “UU Ulu “ULUL wu a 3° el pieces lara ae x ane Oo loz ge o pete 5 ge | 2 | gf [ees (ob) F | e ge 8 | Be | gee | ees) og B SHIBULIYT oF 5 ae BS g a FS 5 A a Nae Psa as 5 BHo a e cs = qyuourzywaly, oc ao adqdia TIIt opry wad +99 BT “OOHBN 729 wad g ur DrIvID WNB yuad sad wy) ¥ ATAVL 6 101] 6201) “eAV 129 TREATMENT OF TRAUMATIC SHOCK ‘MOTJVUIUTIOJOP "GFT LOJ poAOUIaA POOTq ouIoOg "BIOVIV YJIM pouIquiod *OQOHVN yueo aod g | “RIOBOR YYIM PoUrquiod "OOYRN Jueo aod F , SIT |€ 601 T 801/2Z°68 /2AV €IL |8 66 €°86 [8°86 | eAV /g uedo Ay}aed “ST sanoy Z podurvpo ‘foerpary “SInNOY FE UIYPIM pold | — 76 «10 92 8 10 SZ \0 SZ | PLZT SINOY FZ UIYJIM polqd | — G 0 92 6G |0 92 |0 SZ | F9T OF 9A0Q” /OL GIT |O' O01; 6 (0 O0T €P SINOY FZ UIYIIM poiq | — OOL |0'S8 € |0°S8 [0°06 |46F OF 9AOQ®B {09 ZIT |0' LOT) OZ \0' ZOT/0 09 | €9T EaNOU 7G, Ce Bera is PCL |O LIT} 9G |0 SITIO S6 | 69T OF PACK” ,6Z1I | SIL |O FIT) FE O 6010 06 | TZT SIMOY FZ ULYIIM polq, | — OZI |O'SIT} TS jO SITIO OOT} 6ST Asdoynv on ‘Aep pag jo pus pord | ZI |O'ZI1| ZE |O OTTO G6 | ZAT sanoy FZ UIYYIM pod, | + OOL |0 26 FL 10 16 |0 SOL|x IF T@L |O' CIT] O |O ZITO S6 | SOT sinoy FZ UIYYIM porq | + OZT |0 S6 6 |0'S6 |O OTT) OAT GZI |O'SIT} 8 |O ZIT/O'86 | Z8T = OST |O O€T] F 0 62T\O FIT} 9AT SsImoy FZ UL 601 |0°ZOT) O (0 ZOT|O OOT) 99T | -UIIM pold “OF 9A0QB ET | — 2G |O STL} FL JO TITIO Sat) SAT UU “ULUL Ulu ULUL ULUL ULUL UL UiUl “Ubu “UbUb fog Pe oe |e | Ase = | eo eats ae 85/3 B88|8e| = | = we (83 |e (888/85 | & | | Pa eres 2h ee eye a Freese | mess | Uae ea areata cine ete Sly Ee epee) Nay ie B 2 ee | Bi ccielieieeoall eae 3 i 4 eed MO Ml irs a 4 4 oa Opens Oley 3 SYIVUIOYT : Be? BP a =e : 5 SYAIVUIIYT ge 3 ge ae Bg : e. SEP ae |e ae | ee ae eee quowyrory, oS 5. 5 quourzved yf, oS ae aqdid MOOHS JO CAUATAONOAH ojry wad *390 G “OQOHVN 1429 wad ¢ pun o7ry sad *99 g ‘(VN) DiovID WN judd sad aay-fhyuon T $ WIAVL THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 130 JOSEPH ERLANGER AND HERBERT 8S. GASSER Cardiac cases. A certain number of animals die of, or are threatened — by, cardiac failure during the period of occlusion. We have not had an opportunity to study this condition carefully. It comes on as a rule during the first hour of the clamping period. Its imminence is indicated by an irregularity in the force and rhythm of the pulse as registered by the arterial manometer. If, when this is noted, the clamp on the cava is cautiously opened and the heart massaged through the thorax, the pulse in the course of a few minutes may become perfectly regular again. Once the heart has thus recovered, the arterial pressure often can again be held down by the clamp for the rest of the clamping period and the heart, as a rule, will not again become irregular. But quite as often opening the clamp and allowing the pressure to rise at the time cardiac failure becomes imminent does not help matters. Sometimes, indeed, it seems to do harm: the irregularity increases, the pressure falls and the heart stops. The further course of the cases showing these cardiac disturbances, but which, through temporary opening of the clamp, can be carried to the period of observation, seems to indicate, however, either that the damage that led to the cardiac disturbance was not altogether tran- sient, or that the effect upon the heart is a measure of the damage the occlusion of the cava has inflicted on the body generally. If we include in this analysis only those cases in which it was possible to carry out the clamping stage as usual, with the exception of the few minutes during which it was necessary to open the clamp in order to save the heart, and exclude the cases of marked abdominal hemorrhage, the whole series contains 22 cardiac cases. In 10 or 11 of these cases the arterial pressure after removing the clamp began to fall before the 2-hour period of observation had elapsed. In accordance with the rule, all of these died excepting the one of this character previously referred to. Twelve of the cases belong, therefore, to the series from which we deduce the results of applying measured damage. Six of these died within 48 hours despite the fact that their clamping period is usually somewhat curtailed by the period during which the heart is being given a chance to recover. ; That the cardiac cases.are somewhat handicapped is indicated also by the behavior of the arterial pressure during the observation period. While the average arterial pressure of all of the cases that died of clamp- ing the cava was 100.1 mm. Hg. at the close of the 2-hour period, of all that recovered, 106.3 mm. Hg., the average of the 6 cardiac cases that died was 85.2, of the 6 that recovered, 100.7. This last group TREATMENT OF TRAUMATIC SHOCK il includes one case (see table 6) whose arterial pressure rose to 135 mm. Hg., the highest pressure reached after 2 hours in the whole series of 219 animals. If we exclude this exceptional case, the average pres- sure of the cardiac cases that recovered becomes 93.8 mm. Hg. These cases, therefore, have to contend with a circulation that is inferior to that of the general run to a degree that is indicated by a lower arterial pressure amounting to from 12 to 15 mm. Hg. It seems only fair that these facts should be taken into account in considering the effects of treatment. With this end in view, we have included in the statistics the figures obtained after excluding the cardiac cases that died. It will be seen, however, that the general relation of the several groups to each other is not materially altered by this method of presenting the results. THEORY OF THE TREATMENTS The object we hadin mind when this investigation was begun was to determine the, relative efficacy of solutions containing gum acacia and a crystalloid in the prevention of death due to trauma. The number of solutions that might have been studied in such an investi- gation was almost limitless. Inasmuch as there were reasons for reaching an early decision, and inasmuch as it was realized that large numbers of animals would be required to satisfactorily test each solution, it was necessary to reduce the number tested to the smallest that might suffice to indicate the advantages and disadvantages of the different types of combinations. At the time, a solution containing bicarbonate as the crystalloid had been recommended for use at the front by the English Committee on Shock and Allied States (13). By starting the tests with this first Bayliss solution, therefore, we felt it would be possible at one and the same time to test the claims made for this solution and to determine the efficacy not alone of a solution practically isotonic as regards both colloid and crystalloid but also one containing an alkali which also was being recommended for use in the treatment of shock (14). The results obtained with such a solution could then be compared with the results obtained with a hypertonic solution of the same ingredients. For reasons that will become evident, solutions containing glucose instead of bicarbonate were also tested. 1. Gum acacia, 6 per cent, in NaHCOs;, 2 per cent. This solution was made up about as described in the private report of the Special GASSER JOSEPH ERLANGER AND HERBERT S. 132 /g uodo Apjavd dureyo “0% sanoy z podureyo (ORIpArwd — ESIFIMOFLLOT ‘sanoy %][G url porq |0'OST/0'80T| 6 [0 €OT\0 G2 | 88 (O' SZtl0 #2I| ZZ |O' FZTO' SS | SST ig uodo Apyavd durepo Oz OF PAOGL ,TT 0'0Z1/0'86 | 9 [0°86 0°86 | O08 sinoy Z podureyo fovrpreg | — (0°26 |0'8L | FZ |O'SL \0'84 | V8 0 SZIlO' SOT] F jO SOTO OOT! ZFT — |O'STT]O OOT] 4Z 0 S6 \0'S8 | €2 0° 91/0 SST} 8% |O O€T\0 OOT) YET OF PAOK*’ ,FET | — |O' SZLiO SIT] FL |O FETIO OOT| 9ST OF PAOGV {OT O'OST/O' OST) ZI [0 O&O POT) GZ + {0° 0ZT/0 OLT| Z |O SOTO OOT) 92 (0 LZET/0' OZ} ST |O STLO SOT) 69 sanoy $z ulIyyIM porq | + jO STO SOL OL jO SOTO OTT) 68 Asdoyne on ‘Aup pig jo puso poiq 0° GOTO 26 | 8% |0 06 |0 GOT) FS OF PAO’ ,Gz | + 10°06 |0'28 | Go |O' FS |O SIT} F9 4 uodo Apyavd dureyo “cy sanoy Zz podureyo LOBIpPIwI — | STFTUOATALO ‘Aep pig jo puo pod |0'SITi0 96 |G j0 €6 |0 Ocal) 22 — |O'STTJO OOT! OL |0'66 |0 OTT} 62 OF PAOGV {SE O'GSTIO' SIT] @ |O STO OZl! o2 OF PA0QV ,OL | — {0 O&ST|O ZOT| 9 0’ LOL/O Sal} 06 6, uodo ApPyred dureyo “QT sanoy Z% podurryo 0'OSTIO' ZET| Of |O'LZZ110 S2t| EST | 9¥IpIegQ ‘OF eAoqe ,¢ |] + [0 Z21|0' F6 | GT [0°06 JO DET! OST “Ubu “UU “Ulu “ULUL “ULUL “ULUb “ULL : UU “ULUL “ULUL sae] ae] 4 rg ay ™| D kg bg 4 ine] foal m! ey Meee AE epwr ial) esa be |@a|e2 Se | fe] 2 | & Am | me jago; ae) w | BE | ae | @o jago| ce) & = CePeculcieeibee los 8 S | ereenl Sieh oul) seed ~ 8 SyAvu9 yy BOWES og ie 2 5 syavuloy Be, Al ERA sos gs fs 5 Beles lee |e lee | ae Bee |S Bo) Seale ede heme See ehes so Ia oe Bre aleaee ete quourzywoLy, “3's B.S. Ey quUOUywaL J, oS RS MOOHS JO GAATUMTAOOUW agdia opuy wad 99g ‘asoongb quad wad gy pun opry sad 99 g (VQ) DLaDID wnNb ywad sod auf-hyuan J, 9 WIV TREATMENT OF TRAUMATIC SHOCK 133 Committee on Shock and Allied Conditions of November 9, 1917. Enough sodium bicarbonate to make a 2 per cent solution was added to a 6 per cent solution of gum acacia in water. The solution was heated and the precipitated CaCO; centrifuged off. The solution was then heated again, and if another precipitate formed, this again was removed, this process continuing until heating the mixture suf- ficiently long to sterilize it no longer gave rise to a precipitate. The only essential divergence from the method of preparation recommended by the English Committee consisted in not conducting the heating of the solution in a sealed vessel. Our method of prepa- ration probably converted most of the bicarbonate into carbonate. In this respect, however, the difference between the two methods is only relative. For even admitting that in the process of preparation as recommended by the English Committee the bicarbonate is dissociated only to the extent that occurs ordinarily at room temperature and in atmospheric air, only a certain proportion of it would be present as such (15), and the dissociation would tend to go further at the time the solution, heated for injection, is exposed to atmospheric air. This change in the bicarbonate can not be avoided by sterilizing the dry substance by dry heat, for unless this is done in the presence of CO2 under tension, the bicarbonate at the sterilizing temperature also would be converted into carbonate practically completely (15). As a matter of fact, unless solutions containing bicarbonate are kept exposed to CO, at a relatively high tension at the time they are being injected more or less carbonate will always be present. In order to understand the action of this solution on the blood volume and on the alkali reserve, it is necessary to know approximately the salt concentration remaining in solution after precipitating the lime and after boiling. The amount of calcium contained in gum acacia seems to vary within very wide limits. If we take the figure given by Bayliss, namely, 2.23 per cent (16), the minimum amount of NaHCO; left in the 2 per cent solution after adding it to the 6 per cent solution of gum acacia would make a 1.9 per cent solution. If boiling converts all of the bicarbonate into carbonate, the minimum salt con- centration would be 1.5 per cent NasCO3. The freezing point of the solution, if the conversion into carbonate were complete, would be about —0.6°C. or —0.77°C. for the unconverted bicarbonate; in other words the freezing point of the solution as injected would probably be somewhat lower than that of the blood plasma (—0.56°C.). When these experiments were done, we were of the opinion that this first 134 JOSEPH ERLANGER AND HERBERT S. GASSER Bayliss solution was isotonic with respect to the blood colloids. We now know that it is slightly hypertonic in this respect also. But the differences from isotonicity are probably not so large as to be of any particular significance (7). The dose of the solution used in these experiments, 12 ec. per kilo of body weight, is somewhat larger than that (500 cc.) recommended by the English Committee for use in man, which for a 60 kilo man is 8.3 ec. per kilo. The report does not specify the rate at which this dose is to be administered. We used the larger dose because 1t was about the amount needed to restore the blood of a shocked animal to its normal concentration (7), to say nothing of the amount needed to make good the stagnated blood. The larger dose was used also in order that the alkali would come closer to the amount needed to combat the average acidosis that develops in experimental shock. But even in the larger dosage the alkali, presumably, was not nearly sufficient for this purpose (7). The time taken to administer our dose averaged 29 minutes. 2. Gum acacia, 25 per cent, and sodium bicarbonate, 5 per cent. The solutions for this treatment were made up and administered in several ways. At first enough NaHCO; to make a 5 per cent solution was dis- solved in the 25 per cent gum, the solution heated and the precipitate centrifuged off. The gum, however, turned deep brown in color, especially during sterilization, and while we had no evidence that the colored gum was harmful, it was deemed best not to use it. The next method consisted in adding a very slight excess of the bicarbonate to the gum solution, applying heat and then centrifuging off the pre- cipitated CaCO 3. The solution was then neutralized carefully with HCl and sterilized in sealed centrifuge tubes. If, during sterilization, a new precipitate formed this was thrown down in the centrifuge. This solution of sodium acacia was decanted into a sterile burette as needed and the bicarbonate, dry sterilized, added in sufficient amount to make a 5 per cent solution of bicarbonate. In by far the largest number of experiments, 5 ec. per kilo of the sterile, neutralized sodium acacia, made as described above, were first given and immediately followed by a solution made by adding dry-sterilized bicarbonate to sterile water in sufficient amount to make a 5 per cent solution of NaHCO;. From time to time the proportions and the dosage were varied (see table 5) but never by enough to materially alter the con- ditions as here described. The administration of the acacia solution ~ TREATMENT OF TRAUMATIC SHOCK 135 occupied about a half-hour, of the carbonate solution about 28 minutes. The rate of injection of the solution, calculated as bicarbonate, was therefore 0.65 gram per kilo and hour. The rate recommended by the English Committee (17) was about 0.95 gram per kilo and hour in the form of a 4 per cent solution of bicarbonate made up, presumably as we have made our solution. The dosage we employed was based upon the view (16) that 7 per cent gum acacia was isotonic with respect to the colloids of the blood and that, therefore, the water the colloid theoretically could hold would be about the amount needed to restore the blood volume of shocked animals to normal. Newer estimations of the osmotic pres- sure of gum acacia (4), (7) indicate, however, that the figure used by us was somewhat too low. The bicarbonate was added in sufficient amount to draw into the circulation approximately the amount of water the gum on the basis of the available osmotic data was capable of holding. The amount of alkali administered was greater in the proportion of 25 to 21 than in the dose of the solution consisting of 6 per cent gum and 2 per cent bicarbonate. In neither case, though, was it present in amounts large enough to balance, by chemical means, at least, the acidosis usually present at the time the injection is begun (7). 3. Gum acacia, 25 per cent, and glucose, 18 per cent, were given in two ways. In one series the 25 per cent gum acacia was given first and was followed immediately by the 18 per cent glucose. The care- fully filtered gum solution was sterilized under pressure in stoppered centrifuge tubes and in case a turbidity developed the solution was cleared in the centrifuge. In order to avoid the change in the glucose solution that is produced by sterilization at high temperatures, as indicated by the yellow color that develops, the glucose solutions at first were sterilized by pasteurization. Subsequent experience showed, however, that the autoclaved glucose is not injurious; therefore, the latter method of sterilization was finally adopted. The dose of each of the solutions was 5 ec. per kilo of body weight. The injection of the gum again occupied about 30 minutes. The glucose injection averaged about 19 minutes. The latter, therefore, was given faster than the tolerant rate of 5 ce. per kilo and hour. The rationale of this dosage was the same as that underlying the acacia-carbonate dosage. The acacia was in such a concentration that through its osmotic tension it could easily hold the water which the 18 per cent solution of glucose would draw into the circulation. The effect of GASSER JOSEPH ERLANGER AND HERBERT §&. 136 eee 4 uodo Afyaed due foerpaeg OV PAOQ’ 16 stpruoytaoed fsAep SU pera: OF 940% 7 OF BA0Ces Ler St}UoOyLaod yas SAvp pag jo pues ped “07 e40q8 .€ 4 wedo AT,a0ed QOL ‘sanoy z podureyo OVIPAV) “Op AOL OF poo][q *99 OT ySorT OF BAO Sich Op XAOGV |G ,, uodo Ayperjved dureypo fowrpaeg SYAVULOYT 0 001/0°S8 0° S210 O11 0 061|0' S21 0 SZ1|0' SOT 0 SE1|0' 0&1 0° SZ1/0° SIT 0° S11/0 001 0°96 |0'16 0 OST|O' S21 0° Z21|0 O11 0’ SE1/0' 8&1 “UL “UU a) ac) Bo | So Sa lf Heh | Be @ | Bo & | & JUOUTZVOTT, 0 0 GS Oh, 8 0% 9% wu queur}ee1} pus poried Uor}eAIes -qo uaeMjeq BUILT, MOOHS AO GAUTAOOMIW potsed uoreAres -qo Jo pua ainsseig 0° OLT ainsseid [vI}IUy 0 OST 0° 061 O° LST 0 GO% 0 861 : 0) 681 oyepnxo Apoolq jo ][Ny wano[g ‘OSVOSIP ADATL OTUOLYD : O'GI¢) “SInoYy FEZ UYPIA pod Gor | GOT | ST |O' SOT] 06 | G0Z 0 161 Oc | LOL | 6 |O ZOT) 06 | 66T 0 606 SINOY FZ ULYPIM por | + | OST | GOL | SE |0'OOT] OOT | 261 Op OAOGR 0002] ,2 “Stnoy 8 Ul pold}] + | OZI | GOT | GE |O'90T] SIT | 10% ge] vedo ATeryaed dureyo O'LIZ] -OvVIpawy ‘Op oAoqe ,F\++ | G6 | G8 | SL |O'S8 | SIT | S6r ULuUL “ULUL “UU “Ubu Ubu E | Bt) Paley cevle ge lame E | Fl BE [sae | ae] 9 | & 8 SAVULOYT ae aul sic Soo| ses ® 5 z ge| e| "2 ldogle=| 2 | § B. sg2/ 8s] 3 B | quowyvory, oF ag aqgid unoy pun opry dad +99 ¢ ‘asoonpb yuao dod gy UL (VO) DDID WNB yuId Jad day-fhyuan Tf, 4 WTavb 137 TREATMENT OF TRAUMATIC SHOCK OVIPIVA RINOU 7G: UTS Bergh 59:8) 09) (22 Sieve Bode eile 9° LTT8 901 901 |% POT) PAV SIL | Or 6 OOT) COL | PAV 0 OOT/O'06 | 0 06 0 08 0 FIG 0'Oc1|0' 26 | 8T | c6 0 S8 (0 LIZ 0°96 0 G8 | SG | 08 (0-98 \0 SST quoUL -}bo1} SULINpP UAT TOU “OP 9AOGE OE (0 OTTO OOT) ZI | GOT |0'88 0 EST 0 OTTO S6 | &T | 16 (0:06 0 Iz 138 JOSEPH ERLANGER AND HERBERT S. GASSER this combination upon the blood volume of shocked animals has not been determined. It has, however, been shown in normal animals that the results obtained with it are in general accord with theory (7). 4. Gum acacia, 25 per cent, and glucose, 18 per cent, in combination. In the foregoing series the acacia and glucose were given in the suc- cession, in the dosage and at the rates described in order to make the conditions comparable with those obtaining in the series that received 25 per cent gum.and 5 per cent bicarbonate solutions. Thetwo latter substances could not conveniently be given in one solution because, as has been said, the gum acacia in strongly alkaline solution seemed to be considerably changed during sterilization. And it seemed advisable to give the aftercoming crystalloid as rapidly as was safe in order to bring rapidly into the circulation the water needed to dilute the gum acacia. There were reasons for suspecting, however, that the combined administration of acacia and crystalloid might have some advantages over their successive administration. For until this water was added, the circulation of blood would be slowed by the increase in its viscosity produced by the strong gum solution. By combining the glucose with the gum this necessity no longer existed and it then was possible to inject the glucose at a slower rate, one which in normal animals does not lead to glycosuria (18). In this series of experiments, therefore, the gum and the glucose were combined in one and the same solution. The method of making up this solution has been described in another place (19). It was injected at the rate of 5 ce. per kilo and hour; the glucose, therefore, entered the circulation at the subtolerant rate of 0.9 gram per kilo and hour. The effect of the injection of solutions 3 and 4 upon the alkali reserve has not been determined. According to Macleod and Hoover (20), alkaline glucose injections increase the amount of lactic acid in the blood. Acid glucose solutions, they showed, do not have this effect. The present solution, through the reaction of the acacia, is faintly acid: presumably, therefore, there would be no acid formed through the process described by Macleod and Hoover. STATISTICS Excluding, as has been explained, the cases in which the arterial pressure began to fall consistently before the close of the 2-hour period of observation, it is seen (tables 1, 2, 3) that of the 23 acceptable cases composing the control series, 11, or 48 per cent, died within 48 hours, TREATMENT OF TRAUMATIC SHOCK 139 4 of these, or 17.4 per cent of the group, within 24 hours. In the series treated with the combination of 6 per cent gum acacia and 2 per cent sodium bicarbonate (tables 1, 4) there are 20 acceptable cases of which 9, or 45 per cent, died within 48 hours, and 1 of these, or 5.0 per cent of all the acceptable cases, within 24 hours. In the series receiving in succession 25 per cent gum acacia and 5 per cent sodium bicarbonate (tables 1, 5) there are 16 acceptable cases; 9, or 56 per cent, died within 48 hours; and 8 of these, or 50 per cent of all in the group, died within 24 hours. It was so obvious that this treatment was doing harm that it was not regarded as necessary to accumulate as many cases as we have in the other groups. The group receiving 25 per cent gum acacia and 18 per cent glucose in succession (tables 1 and 6) comprises 20 acceptable cases, 9 of which, or 45 per cent, died within 48 hours, and of these 1, or 5.0 per cent of all the cases, within 24 hours. And, finally, the series receiving the gum and glucose in com- bination (tables 1 and 7) contains 21 acceptable cases: 5 of these, or 24 per cent, died within 48 hours, and 3 of these, or 14.8 per cent of the whole number, within 24 hours. The two most obvious results coming of this comparison of the several groups are a, the unfavorable showing of the series receiving the 25 per cent gum acacia and 5 per cent sodium bicarbonate in succession, as regards both total mortality and duration of life of the fatal cases; and b, the favorable showing of the series receiving 25 per cent gum in combination with 18 per cent glucose as regards total deaths. In the matter of one day deaths, the latter series is only slightly better than the control series and is not so favorable as the series that received the 6 per cent gum in 2 per cent bicarbonate or 25 per cent gum and 18 per cent glucose. The number of one day deaths in these three series is so small, however, that a difference of one case in either direction would practically have eliminated all differences. The high 24 hour mortality of the series receiving 25 per cent gum and 5 per cent bicarbonate cannot, however, be accounted for in this way. This treatment unquestionably does harm. As has been said, a high arterial pressure at the end of the observa- tion period seems to favor recovery. The average pressures of all of the groups at this time (table 2) are, however, so nearly alike that it is not necessary to make any allowance for such effect as this factor may have. From the cases that died and are otherwise acceptable have been excluded, as has been explained above, the cases in which considerable 140 JOSEPH ERLANGER AND HERBERT S. GASSER abdominal hemorrhage is found at autopsy. The necessity for ex- cluding such cases is obvious. On the other hand, as has been said, the decision as to whether the hemorrhage is marked or shght rests upon a judgment which often it is difficult to make. In order to take this into account we give in columns 6, 7, 8 and 9 of table 1 the statistics as obtained by excluding all of the fatal cases in which at autopsy even a small amount of bloody fluid was found in the abdo- men. It is seen that the relative number of hemorrhage cases in each group is practically the same, demonstrating again its accidental origin. It follows that when they are excluded the relative mor- tality rate is not markedly affected (column 9); the differences already presented (column 5) are merely somewhat emphasized. In columns 10 and 11 (table 1) it is seen that with the exception of two groups, practically all of the deaths occurring within 24 hours were complicated by slight hemorrhage. The exceptions are in the case of the groups receiving bicarbonate and seem to emphasize from still another aspect the deleterious effects of alkali. Set over against this is the fact that all of the one day deaths occurring in the series receiving 25 per cent gum acacia in 18 per cent glucose were com- plicated by the slight hemorrhage; there are left in this group altogether only 1 or 2 deaths if those with slight hemorrhage are excluded. It has been pointed out that, in general, cases with high initial arterial pressures are more apt to die than cases with low initial pres- sures. On this basis it is seen (table 2, columns 1 and 9) that the group with the best initial chance, namely, those treated with 25 per cent gum acacia and 5 per cent sodium bicarbonate, actually is the one that did the worst; while the group with the worst chances in this respect, namely, the one receiving 25 per cent gum acacia followed by 18 per cent glucose, did quite as well as any; and that the treat- ment that gave the best results, namely, the 25 per cent gum acacia in 18 per cent glucose, was tried out on a group whose initial chances were not particularly favorable. The results obtained by excluding, for reasons already given, the cases in which the pressure failed to fall to 40 mm. Hg. within 30 minutes after applying the clamp to the cava and lived are seen in columns 15, 16, 17 of table 1. These exclusions cause no important changes in the relative positions of the several groups. The treatment consisting of 25 per cent gum acacia and 5 per cent sodium bicarbonate still gives the worst result; that consisting of 25 per cent gum acacia in 18 per cent glucose the best. The group receiving 6 per cent gum TREATMENT OF TRAUMATIC SHOCK 141 acacia in 2 per cent sodium bicarbonate gains a little, that receiving 25 per cent gum acacia followed by 18 per cent glucose, somewhat more, on the control group. Finally, we take into consideration the handicap of the cases in which the heart during the clamping period threatened to fail, by calculating the results after excluding those of these cases that die. Columns 12, 13, 14 (table 1) show that this precaution also changes the results but little. Its main effect is to improve slightly the results of the administration of the solutions containing glucose. It has already been stated that the series of acceptable cases includes some 12 so-called heart cases, and that of these 6 recovered of shock. It is interesting to add here that of the 6 that recovered, 5 were in the series that were treated with glucose, namely, 2 in the group receiving acacia and glucose in succession, and 3 in the group receiving the solutions simultaneously. One was in the control group. While the number of cardiac cases is not sufficiently large to permit of any definite conclusion, it certainly is suggestive that with but one ex- ception, all of the cardiac cases that recovered of shock had received intravenously a substance, glucose, which is known to affect favorably the contraction of heart muscle. Thus far we have confined ourselves to the action of the solutions on the state that develops, to all intents and purposes, immediately after unclamping the vena cava, that is, on shock. It has already been stated that 8 of the animals that had shown marked improvement from this first state died rather unexpectedly, for the most part, toward the end of the third day. The possible significance of such deaths did not at first occur to us and as a consequence we have autopsy notes on only four of the cases. These four presented signs of peritonitis ranging from a fibrinous deposit on the peritoneum to a sanguino- purulent exudate. The fatal issue in these instances, therefore, can be attributed neither to shock nor to a late deleterious action of the injected solutions. The other four cases not autopsied were scattered among three of the groups. In view of the relatively small number of these cases—4 out of the 100 that were acceptable—and in view of the regular presence of peritonitis in the four cases that came to autopsy, and in view, furthermore, of the many other possible causes of death, even if peritonitis were excluded, it certainly seems unneces- sary to attribute these deaths to a delayed action of the injected solutions. It might be added that the incidence of peritonitis was not high when it is taken into account that the peritoneum was open for 142 JOSEPH ERLANGER AND HERBERT S. GASSER 2+ hours at a time when its resistance, owing to the slowing of the circulation, must have been quite low, and that there was the possi- bility of infection through the walls of the gastro-intestinal tract damaged by the long-lasting ischemia. DISCUSSION The foregoing analysis shows that no matter how the statistics are viewed, the relative efficacy of the solutions tried is — 25 per cent calcium acacia in 18 per cent glucose > 25 per cent gum acacia fol- lowed by 18 per cent glucose = or > 6 per cent sodium acacia in 2 per cent sodium bicarbonate = or > untreated controls > 25 per cent sodium acacia followed by 5 per cent sodium bicarbonate. In attempting to account for these results it should first be recalled that in shocked animals all of the solutions that have been tested are capable of restoring the blood volume and the blood pressure and of maintaining these to practically the same degree under similar cireum- stances. Considering the end in view it certainly seems justifiable to assume that these responses are desirable. The fact, therefore, that some of the solutions, 6 per cent gum acacia in2 per cent sodium bicar- bonate and 25 per cent gum acacia followed by 18 per cent glucose, for example, effect but little, if any, reduction in the death rate of traumatized animals, may mean that these solutions are not entirely innocuous and that the harm they do just balances the good that results from the improvement they effect in the circulation. There can, of course, be no doubt but that the treatment in which the 25 per cent sodium acacia is followed by 5 per cent sodium bicarbonate is harmful. Is it possible to ascertain from the limited data available wherein this harmfulness consists? Gum acacia in 6 per cent solution we believe is wholly innocuous. This is proved by our own experience. Even a 25 per cent solution of gum acacia in the dosage which we have used is quite innocuous to normal animals. When such a solution is given rapidly to dogs until the blood plasma becomes a 10 per cent solution of gum acacia no obvious symptoms develop (22). Czerny (23) found the maximal non-lethal dose for cats and rabbits to be 4.66 grams per kilo. This is almost four times the amount contained in the dose we give in shock and in the concentration (24 per cent) used by Czerny, would convert the plasma of these animals into an 8 per cent solution of gum. And Czerny did not even use aseptic precautions. The innocuousness of the hypertonic TREATMENT OF TRAUMATIC SHOCK 143 solution is illustrated also by the biological tests we have been making of the solution consisting of 25 per cent gum in 18 per cent glucose made up for the treatment of shock in man (19). Each batch of this solution is tested by injecting in the course of a half-hour 10.0 ec. per kilo into an animal previously bled to the extent of more than 3 per cent of its body weight. In none of the animals so treated has there been the slightest evidence of harm. There is no question, though, but that concentrated gum in the dosages we have employed may do harm if it is injected rapidly into an animal about to die of shock. Under these circumstances, in our experience, the heart occasionally has stopped as though it had fibril- ated. Presumably, this is the result of the slowing of the circulation that comes of the high viscosity of the injected gum solution. At any rate this harmful action cannot be attributed to the osmotic properties of the gum, for nothing of the kind is seen when the gum in equally hypertonic solution is injected more slowly, not even when it is in combination with hypertonic glucose. Nor can it be due to any chemical action of the gum, for inasmuch as the injected gum remains for some time in circulation, there is finally just as much of it in circu- lation when it is injected slowly as when it is injected rapidly; yet, under the former circumstances stoppage of the heart does not occur. To be sure, calcium acacia has been used almost exclusively in these tests; but as regards osmotic (4) and viscosity (16) factors, as these affect the blood, the sodium and calcium acacias are presumably practically alike. In the course of our experiments we have given large doses of both of these acacias and have not been able to satisfy our- selves that there is any difference in their action. Furthermore, if sodium acacia acts deleteriously through chemical means, the results obtained from the injection of the sodim acacia in combination with isotonic bicarbonate should have been almost as bad as the results of the injection of the 25 per cent solution followed by the hypertonic carbonate, for the amount of gum given with the former (0.72 gm. per kilo) is not very much smaller than the amount given with the latter (1.25 gm. per kilo). The difference in the results following the injection of these two solutions, though, is quite striking. The result of this analysis, as far as it has gone, is to indicate, therefore, that it is neither the sodium acacia per se that does the harm nor its hypertonicity, but rather the alkali that is injected with or after it, to a certain extent also the high viscosity of the more concentrated gum solution that lasts until it is diluted by its own osmotic action 144 JOSEPH ERLANGER AND HERBERT S. GASSER and by that of the crystalloid subsequently injected. It is not neces- sary to conclude at this juncture, however, that the bicarbonate is injurious by virtue of its alkalinity. It may be that salinecrystal- loids are undesirable in the treatment of shock, possibly because they accumulate and disturb the salt balance of the body. It was considerations of this character that led us, when it was found that the gum in combination with bicarbonate is without value in the treatment of shock, to try out glucose as the crystalloid, es- pecially in view of the indications obtained by one of us in collaboration with Woodyatt (24) that it ameliorates the symptoms of experimental shock better than do salines. The series of experiments in which the glucose replaced the bicar- bonate but comparable in every other respect save the nature of the gum, clearly demonstrated the superiority of the glucose as a means of drawing the water to the gum. Yet this succession of gum and glucose was without any marked effect upon the death rate due to the measured trauma. The next step was to eliminate the period of increased viscosity of the blood by injecting the gum and the crystal- loid simultaneously. It was by this method that the maximum saving of life was effected. It 1s conceivable that this beneficial action is due not alone to the osmotic action of the glucose but also to some specific action 1t may exert. With regard to the relative merits of salt and glucose as a means of bringing water into the circulation, it may first be pointed out that there is relatively little reason, other than osmotic, for ad- ministering salts in shock, for there is, so far as is known, no salt deficit in that state. Neither is there any lack of glucose (14). But glucose has the advantage over most other crystalloids that might be used for this purpose, that through oxidation and polymerization, the organism can quickly put it out of the way after it has accomplished the purpose of bringing water to the gum acacia. However this may be, there is abundant evidence in the literature indicating that glucose of itself acts beneficially in clinical shock and allied conditions. Wood- yatt, with Sansum and Wilder, obtained favorable results from sus- tained injection of glucose in hypertonic solution in two cases having features of shock (24). Hypertonic glucose has been found to be of benefit in pneumonia (25) and in a number of other conditions. There are a number of ways in which glucose might prove of functional value to the traumatized organism. The fact that in shock there is no deficiency of blood sugar does not necessarily mean that TREATMENT OF TRAUMATIC SHOCK 145 in this respect the administration of glucose would prove futile. The beneficial results said to come of feeding sugar to soldiers on the march (26), for instance, can not be attributed to any deficiency of glucose in the blood (27). Whatever the mechanism of the relief from fatigue may be, it seems likely that the same or a similar mechanism might work to relieve shock when glucose is administered. It is well known that glucose in excess of the normal blood content is oxidized in the tissues (28), and that it is used as a food when it is introduced into the system parenterally (29). Glucose also improves and sustains the beat of the perfused heart (80), (31), as well as the contractions of other types of muscle (32), (33). In this connection it is of interest to recall that of the six cardiac cases that recovered of shock, five were in the groups that received glucose. Over and above all of the considerations that have thus far been discussed, there was another which held out hopes that hypertonic solutions might prove of value in the treatment of shock. Eyster and Wilde (384) have found that certain erystalloids in hypertonic solution, glucose among others, cause an immediate increase in cardiac output and a vasodilatation. These actions apparently are independent. of the resultant hydremia; indeed, they seem to be specific. But what- ever may be the cause, it is obvious that these are exactly the responses best calculated to improve the circulation in shock. It will be noted that we have not tested the efficacy of blood in the treatment of shock. One reason for not doing so was the number of animals that would have been required, while another was the fact that our animals were not suffering from any loss of blood. When, as in pure shock, there is no actual loss of blood, but rather a con- centration of the blood and a crowding of the small veins and ecapil- laries with corpuscles, the introduction of more blood, unless it promptly betters the condition, would seem to be contra-indicated. At least this is the inference one is led to draw from the effects of the injection of blood into normal animals. Starling points out (35) that under these circumstances the blood fluids do not remain long in the vessels but pass into the lymphatics, leaving behind the corpuscles and a certain proportion of the proteins of the plasma. This concentration of the blood raises its viscosity and tends to embarrass the circulation; there is produced a state of affairs similar to the one we are trying to combat in shock. For this reason it is conceivable that blood plasma may have some advantages over whole blood in the treatment of non- hemorrhagic shock. Furthermore, the injection of blood even in suit- THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 1 146 JOSEPH ERLANGER AND HERBERT S. GASSER able cases is by no means entirely free of danger (36). It is, there- fore, of interest to recall in this connection that blood transfusion is not essential to recovery even from the severest acute hemorrhage, if only the blood bulk can be restored in other ways (37). The present investigation possibly indicates that a restoration of blood volume in which the tissues are made to participate through osmotic action is more efficacious than one effected through the injection of the full amount of fluid needed to make good the reduction. Results in man. After we had convinced ourselves through the animal experiments described in this paper that hypertonic gum and glucose are not alone innocuous if given slowly, but actually save a certain number of animals from death by trauma, we began the use of this solution in the treatment of shock in man. The results obtained in eleven cases have been recently described in another place (19). It need only be stated here that they fully confirm the conclusions reached in this paper. SUMMARY Of animals traumatized by holding the arterial pressure down to 40 mm. Hg. for 2 hours and 15 minutes by partially occluding the inferior vena cava— 48 per cent die within 48 hours. When treated with: a. 6 per cent gum in 2 per cent sodium bicarbonate 12 ce. per kilo of body weight 45 per cent die within 48 hours. b. 25 per cent gum followed by 5 per cent sodium bicarbonate, of each 5 ce. per kilo of body weight 56 per cent die within 48 hours. c. 25 per cent gum followed by 18 per cent glucose, of each 5 ee. per kilo of body weight 45 per cent die within 48 hours. d. 25 per cent gum in 18 per cent . glucose, 5 ec. per kilo of body weight and hour 24 per cent die within 48 hours. Not only is the death rate increased by treatment b, but death occurs earlier. These results are taken to indicate that bicarbonate and the high viscosity of a strong gum solution,are somewhat harmful, at least, in TREATMENT OF TRAUMATIC SHOCK 147 traumatized animals; that the harmfulness of the strong, viscid gum ean be avoided in part through the osmotic action of hypertonic glu- cose subsequently injected, but not by bicarbonate; and that when the hypertonic gum and the hypertonic glucose are given simultane- ously and slowly so as to avoid altogether the period during which the high viscosity of the gum is hampering the circulation, a maximum saving of life can be effected. The beneficial results presumably are due to the internal tranfusion (38) effected by the hypertonic solutions, to the maintenance of the increased blood volume through the colloidal and possibly other properties of the gum acacia, to the action of the hypertonic solution on the heart and blood vessels, and to the specific action of glucose on nutrition in general and on that of the heart muscle in particular. BIBLIOGRAPHY (1) ERLANGER, GESELL AND GAsseER: This Journal, 1919, xlix, 90. (2) ERLANGER AND GaAssER: This Journal, 1919, xlix, 151. (8) ERLANGER AND GaAssER: This Journal, 1919, xlix, 345. (4) GasspR, ERLANGER AND MEEK: This Journal, 1919, 1, 31. (5) Weuicu: Allbutt’s System of medicine, 1910, vi, 779. (6) ERLANGER, GESELL, GASSER AND ELuiotT: Journ. Amer. Med. Assoc., 1917, Ixix, 2089. (7) GassER AND ERLANGER: This Journal, 1, 104. (8) Mann: Journ. Amer. Med. Assoc., 1918, Ixxi, 1184. (9) WiaGeRS AND Epwarps: Report to Committee on Shock of the National Research Council, July 9, 1918. (10) AtMErpA AND ALMEIDA: Journ. Amer. Med. Assoc., 1918, lxxi, 1710. (11) McKissen: This Journal, 1919, xlviii, 331. (12) JANEWAY AND Jackson: Soc. Exper. Biol. Med., 1915, xii, 193. (13) Private Rept. of the English Committee on Shock, November 29, 1917. (14) Cannon: Rept. no. 2, Special Investigation Committee (English), reprinted in Journ. Amer. Med. Assoc., 1918, Ixx, 531. (15) Asrea: Handb. d. anorganisch. Chemie, 1908, ii, 1, 303. (16) Bayuiss: Proc. Roy. Soc., 1916, B Ixxxix, 380. (17) Cannon, Fraser AND CoweEL.: Rept. no. 2, Special Investigation Com- mittee (English), reprinted in Journ. Amer. Med. Assoc., 1918, Ixx, 618. (18) Woopyatt, SANSUM AND WiLpER: Journ. Amer. Med. Assoc., 1915, lxv, 2967. (19) ERLANGER AND Gasser: Ann. Surg., 1919, Ixviii, 389. (20) Mactreop anp Hoover: This Journal, 1917, xlii, 460. (21) Bayuiss: Intravenous injection in shock, London, 1918. (22) Meek anp Gasser: This Journal, 1918, xlv, 548. (23) Czprny: Arch. f. exper. Path. u. Pharm., 1894, xxxiv, 268. (24) ERLANGER AND WoopyatTtT: Journ. Amer. Med. Assoc., 1917, Ixix, 1410. 148 JOSEPH ERLANGER AND HERBERT S. GASSER (25) LircHFIELD: Journ. Amer. Med. Assoc., 1918, exxi, 503. (26) Taytor: Digestion and metabolism, 1912, 540. (27) Taytor: Digestion and metabolism, 1912, 256. (28) McGuiean: This Journal, 1908, xxi, 334. (29) Zuntz AND VON Merine: Arch. f. d. gesammt. Physiol., 1888, xxii, 173. (30) Locke: Journ. Physiol., 1907, xxxvi, 205; also: Zentrabl. f. Physiol., 1914, 78. (81) Evans: Journ. Physiol., 1914, xlvii, 407. (82) NEUKIRCH AND Rona: Arch. f. d. gesammt. Physiol., 1912, exlviii, 290. (33) BENEDICT AND CatTucartT: Publ. Nutrition Lab. Carnegie Inst., 1918, no. 187, 94. (84) EysterR AND WILDE: Journ. Pharm. Exper. Therap., 1910, 1, 391. (35) StartinG: The fluids of the body, 1909, 135. (36) PEMBERTON: Surg., Gynec., Obst., 1919, xxvii, 276. (37) Rous anp Witson: Journ. Amer. Med. Assoc., 1918, Ilxx, 219 (88) GEsELL: This Journal, 1919, xlvii, 468. a ‘ STUDIES IN SECONDARY TRAUMATIC SHOCK VII. Nove on THE AcTION OF HYPERTONIC Gum ACACIA AND GLUCOSE AFTER HEMORRHAGE! JOSEPH ERLANGER anp HERBERT 8S. GASSER From the Physiological Laboratory of Washington University, St. Louis Received for publication July 12, 1919 The injuries that lead to shock in man almost always are accom- panied by more or less hemorrhage. When the hemorrhage is so large as to be dangerous of itself the patient suffers not. alone from the im- mediate and direct consequences of the blood loss but according to pre- vailing opinion also runs the risk of passing in time into shock properly so-called, possibly, as a consequence of the action of the slowed blood stream upon the peripheral vessels (1). However this may be, it soon became obvious to us after we had acquired some experience with clinical shock that the surgeon not infrequently is called upon to treat cases with the symptoms of shock in which the extent of hemorrhage is unknown, or in which it is known to be dangerous but for which blood for transfusion is not immediately available. And the question arose in this connection, is it justifiable to employ the hypertonic gum-glucose solution in such instances? Experience gained through the use of the isotonic gum-saline solution in the British Army has led Bayliss to conclude (2) that this substitute for blood is especially useful in those cases in which shock is compli- cated with a certain amount of hemorrhage. We have not the slightest doubt but that this is true also of the hypertonic gum-glucose solution (3). But the question that concerns us here is not this, but rather, is a hemorrhage that is so extreme as to be apt to prove rapidly fatal as the result of the blood loss itself, a contraindication to the use of the hypertonic gum-glucose solution? 1 Reported to the Committee on Shock of the National Research Council, July, 1918. 149 150 JOSEPH ERLANGER AND HERBERT S. GASSER EXPERIMENTAL The method employed in seeking the answer to this question has been to select two animals as nearly alike as possible as regards weight, vigor and breed, to bleed them both, under ether, to about the same extent in proportion to body weight, and then to give to one of each of the couples 5 ee. of the 25 per cent gum acacia, 18 per cent glucose solution per kilo of body weight in the course of an hour. The usual aseptic precautions were observed. It will be noted in table 1 that the extent of the hemorrhage was not always exactly the same in the two animals of a couple. This came about because of differences in the reaction of the animals to the hemorrhage; the development of threat- ening symptoms, such as an exceedingly low arterial pressure or marked slowing of the heart rate, indicating that further hemorrhage would probably prove immediately fatal, sometimes forced us to desist in the withdrawal of blood in one or other of the two animals before the in- tended amount had been drawn. First hemorrhage. In the dog, according to Fredericq (4), a loss of blood amounting to 2.3 to 4.5 per cent of the body weight is dangerous, the loss of over 4.5 to 5.0 per cent, fatal. Our aim was to remove from the animal by a rapid hemorrhage an amount of blood that would just about prove fatal. As will be seen in table 1, the amount at first TABLE 1 Showing the amount of blood drawn in per cent of body weight GROUP 1 GROUP 2 GROUP 3 Dog 1 | Dog2 | Dog3 | Dog4 | Dogd | Dog6 Lore NCMTGD, ae make GOs Sia seas dae 4.41 | 3.96 | 4.45 | 4.64 | 4.80 | 4.80 pecond hemorriagese. tne) cette ee 3.16 | 3.19 | 3.14 | 3.18 | 2.67 | 2.60 ROGALMS A NERS ee EOS: W570 | 715 VTL SO TTT: eae ee removed ranged between 3.96 and 4.8 per cent of the body weight, yet not one of the animals succumbed, treated or untreated. In the case of group 1 (dogs 1 and 2, fig. 1, dots) more blood was removed from the test animal (dog 1, large dots) than from its control (dog 2, small dots); the fall in pressure in the case of the control was so profound that the hemorrhage had to be discontinued for fear of imme- diately killing the animal. This animal had the lower initial pressure. The pressure of the test animal in the course of 1 hour 30 minutes came back to within 10 mm. of its initial value; whereas the pressure of the untreated animal remained 37 mm. Hg. below its normal. GUM ACACIA AND GLUCOSE AFTER HEMORRHAGE 151 The initial conditions were reversed in the case of the second group (dogs 3 and 4, crosses). For the reasons mentioned above, it was not possible in this case to remove from the test animal (dog 3, large crosses) quite as much blood as from its control (dog 4, small crosses). The difference, though, was inconsiderable. The arterial pressure of the test’ animal was affected by the hemorrhage much more profoundly than that of the control. The arterial pressure of the treated animal eventually rose to 88 mm. Hg. (the initial was 87); of the untreated animal to 100 mm. Hg. (the initial was 95). ’ Haem. % body weight MM. Hg. i Dog No. O Hours | Fig. 1. Chart showing the effects of the first hemorrhage. Deg 1 (treated) @; dog 2 (control) e; dog 3 (treated) X; dog 4 (control) x; dog 5 (treated) Oo; dog 6 (control) o. The injection period is the time included in each case be- tween the parentheses. In the case of the third couple (dogs 5 and 6, circles) the hemorrhages were exactly equal in amount; 4.8 per cent of the body weight was removed. The arterial pressure of the test animal fell lower as the result of the hemorrhage than that of the control. The arterial pres- sure of the control animal rose to within 4 mm. of the initial level, that of the treated animal to within 30 mm., but the pressure of the treated animal at the end was a bit above that of the untreated animal. The only conclusion these observations justify is that the injection of the gum-glucose solution did not prejudice the chances of recovery from the effects of a severe hemorrhage. Whether any of the treated GASSER JOSEPH ERLANGER AND HERBERT S. 152 “QAO JXOU SOUO PoyoYoVqun oy} SB OWES OY} OLE SBUIpPvot oInssoid popoyovig, LL ty 198] 0 0 {éloot]| 28 0 196] 09 (G8) 8 he a a ‘* BUIpPvoL SL] OINSSOLg 88 | 62 | 98 62 cé |éloot] | [ss] | sr | 96 |[z9] |«loor] | "°°" es 5 eS omnssoid sous ty] G8 gL tS pvoqd C6 L00T 88 prog CG Z9 OOT elemelle/.e)cejie tshie).e @) ej aie, o. atelial ve) ele electors) ester oe *Aantut jo puo YIM Sutpuodsoii09 10 48 yussotg éLl g9 8G 6L OL cy 06 688 IV ike coh ak es oa a ie a NT Aanfur jo suru -uIs0q YIM Sutpuodsoii09 Io 4v oinssotg QS QO) Ng PL Q) Se SI 0g 82 ST (Ve Dlr rats ae ‘osvyIIOULOY jo pus 7e oInssatg 9% 9G Sv SIS VS 29) ee Ghare | Olona) OGeS ie i is es a: 1y810M Apoq yuoo sod oBvyAoUlo]] 96 SS 06 SOT 06 G6 18 o8 SOT OOT QUE is ge SS aes eee ae ainssoid [VryUy (21) (11) (OT) (8) (L) (9) (¢) (F) (8) (2) (1) [Ouju0o 4so0} | [O1qUO0D jor3u03 yso} |[orquoo| 4s03 | [OrgUOD] 4s0} | [OAJUOD] 4sS0} ‘g soq | ‘¢ soq | ‘9 s0q ‘p Boq | ‘eg 30q | ‘p doq | ‘g soq | ‘Z soq | ‘T soq | ‘Z 30q | ‘I 20q savy lowly os By ILOULOY, osBy Llouey oSVYILOULOY osey IOuley, osvYy LIOULoY puoova, 4SIT puooves 4ST puodog 4SIt iy € qn0u» Z an0up I qaaoun ¢ W1TAVL 153 AFTER HEMORRHAGE GUM ACACIA AND GLUCOSE ‘[ oINGY UL SB SUOT}VUSTSIG e ‘gsVYALIOULBY PUODIS OY! O $1J00]9 949 SUIMOUS JaIBye | l ‘) C ! e PO utaor e nem LY = °N 80g 3tH WW 154 JOSEPH ERLANGER AND HERBERT S. GASSER animals would have died had the solution not been administered can- not be determined. Second hemorrhage. Some two or three days subsequent to the experiments above described each couple was again bled, the amounts taken this time varying between 2.6 per cent and 3.2 per cent of the body weight (see table 1). The amount of blood removed was almost exactly alike in the two animals of each of the couples. The total of the two hemorrhages ranged between 7.15 and 7.77 per cent of the body weight. In the case of group 1 (dogs 1 and 2, fig. 2, dots) the immediate effect of the hemorrhage was to carry the arterial pressure down to the same level but the absolute drop was greater in the case of test animal. After treatment (dog 1, large dots) the pressure rose to 96 mm. Hg.; this animal recovered. The pressure of the untreated animal rose only to 45 mm. Hg., then fell and the animal died 1 hour 33 minutes after the hemorrhage. The animals of group 2 (dogs 3 and 4, crosses) were bled to the same extent (3.14 to 3.15 per cent of the body weight). Their pressures as a consequence did not fall very low, but they fell to about the same level. After the hemorrhage for about 36 minutes the pressure of the control animal practically maintained its level while that of the test animal slowly fell. Then the pressure of the control animal began to fall and this animal died 2 hours 20 minutes after the hemorrhage. At the same time under the influence of the injection the decline of the arterial pressure in the treated animal was converted into a rise; it mounted until, by the end of the injection period, it was 5 mm. Hg. above the initial level, but then it began to fall, though slowly, and the animal died 4 hours 30 minutes after the hemorrhage. Group 3 was bled to the extent of 2.67 per cent (treated) and 2.6 per cent (untreated) of the body weight. The initial pressure of the former was somewhat lower than that of the latter and during and subsequent to the hemorrhage it fell the lower. Both of these animals survived. DISCUSSION This analysis of the experimental data shows that the animal of each couple that received the solution did quite as well as, or better than, its control. In thus stating our conclusion we do not desire to give the impression that we are advocating the use of hypertonic gum-glucose solution in the treatment of pure and immediately dangerous hemor- GUM ACACIA AND GLUCOSE AFTER HEMORRHAGE 155 rhage, though we do believe that when blood is not instantly avail- able it is safe and perhaps advisable to give the gum-glucose solution pending the obtaining of blood. This opinion is based not alone on the results of the present experiments, but also upon our clinical expe- rience (see cases IV, IX and X in our paper on the treatment of shock in man (3) ). Penfield (5) has compared the effect of the following solutions, a, 0.9 per cent sodium chloride; b, 6.0 per cent gum acacia and 2.0 per cent sodium bicarbonate; c, 6.0 per cent gum acacia and 6.0 per cent glucose,—injected into animals after so bleeding them as to hold the arterial pressure down to 40 mm. Hg. for periods ranging between 60 and 89 minutes. The volume of solution injected equalled the amount of blood drawn and varied (as we calculate it) between 4.15 and 5.05 per cent of the body weight. If we except three of his eleven cases, the amount removed and injected averaged 4.44 per cent and did not exceed 4.65 per cent of the body weight. So far as can be determined there were no untreated controls in his series. One of the three of his animals that received the sodium -chloride solution died; the average hemorrhage amounted to 4.61 per cent of the body weight. The amount of blood withdrawn at the time of the first hemorrhage in our experiments averaged 4.51 per cent of the body weight and therefore was practically as large as in Penfield’s and not a single animal died whether treated or not. There is, therefore, no necessity for con- cluding that the isotonic solution of sodium chloride in Penfield’s hands accomplished any good. Three of the four animals into which Penfield injected his gum- glucose solution died; the average hemorrhage was 4.29 per cent of the body weight. At this place we desire to correct the statement made by Penfield that he used ‘‘gum-glucose solution as recommended by Erlanger.’ Ours is hypertonic gum-glucose, not isotonic. Having removed this possibility of a misunderstanding, we may say that we are surprised not that three of the four animals treated with the weak gum-glucose solution died, but rather that one of them lived. For Penfield replaced approximately half of the blood in the body with the same amount of a solution which within a very few minutes, through oxidation and polymerization of the glucose, came to consist practically of a solution of 6.0 per cent gum in pure water. The effect that this must have had upon the tissues and the salt balance of the organism, unquestionably accounts for the results Penfield obtained. 156 JOSEPH ERLANGER AND HERBERT S. GASSER The hypertonic gum-glucose solution, of course, is changed in the same way in the organism, but the dose we give, namely, 5 cc. per kilo an hour, is so small, and it is given so slowly, that such disturbance in salt balance as it may cause is negligible in comparison with that caused by the dose of 40 to 50 cc. per kilo employed by Penfield. CONCLUSIONS 1. The use of hypertonic gum-glucose solution is not contra-indicated in the treatment of shock even when it is complicated by dangerous hemorrhage. 2. The fact that the hypertonic gum-glucose solution does not preju- dice the recovery of animals from the effects of a hemorrhage that is apt to result fatally furnishes another proof of the innocuousness of this solution. BIBLIOGRAPHY (1) ERLANGER AND Gasser: This Journal, 1919, xlix, 151. (2) Bayuiss: Intravenous injection in shock, London, 1918. (3) ERLANGER AND GassER: Annals of Surgery, 1919, Ixvili, 389. (4) Freperica: Travaux du Laboratoire, 1886, 1, 133. (5) PENFIELD: This Journal, 1919, xlviii, 121. THE INFLUENCE OF OXYGEN ADMINISTRATION ON THE CONCENTRATION OF THE BLOOD WHICH ACCOMPANIES THE DEVELOPMENT OF LUNG EDEMA D. W. WILSON, Capt., and 8S. GOLDSCHMIDT, Capt., C. W. S., United States Army From the Physiological Laboratories, R. E. Experimental Station, Porton, England' Received for publication July 14, 1919 The enormous and rapid development of edema of the lungs which results from severe gassing of animals with the lung irritants used in warfare offers an unusual opportunity for studying the physiological effects accompanying this pathological condition. The rapidity with which the edema develops precludes the possibility of infection com- plicating the symptoms observed, and the condition which may develop after exposure to high concentrations of poisonous gas is so severe that the correlated symptoms can hardly be overlooked. Loss of water from the blood is one of the most characteristic phe- nomena accompanying the development of edema of the lungs in animals gassed with lung irritants. A concentration of the blood becomes evident at about the time when the edema of the lungs can be first demonstrated. Thereafter the loss of water from the blood and the increase in severity of the edema run roughly parallel. The conclusion was therefore made that the two are interrelated and that the pouring of water into the lungs is the cause of the concentration of the blood. Other considerations however make it necessary to proceed with caution before accepting this hypothesis. During the acute period after gassing there develops a deficiency of oxygen carried by the blood. Probably due to the poor aeration of the blood in the damaged lung the oxyen content of arterial and venous blood may drop to levels much below normal. The transport of oxygen to the tissues may be still further reduced by the decreased rate of blood flow with the probable result that the oxygenation of the tissues is seriously interfered with. 1 Published with the permission of the American and British military authorities. 157 158 D. W. WILSON AND S. GOLDSCHMIDT Physiologists have shown that muscle tissue imbibes water when supplied with insufficient oxygen. Based on this observation the hypothesis may be presented that the concentration of the blood is due, not primarily to the development of lung edema, but to the imbibition of water from the blood by the tissues which are not suf- ficiently oxygenated. To throw some light on the validity of this hypothesis the experiments reported below were carried out. Goats were gassed with lethal concentrations of chlorpicrin. As soon as possible after gassing, half of the animals were fitted with masks and given oxygen continuously in known quantities by means of a Haldane oxygen apparatus. The other animals were used as controls. The hemoglobin content was used as an index of the concentration of the blood. Hemoglobin determinations were made frequently using blood obtained by pricking an ear vein. Blood from the heart punc- tures was also used. The Haldane method was used for the hemoglobin determinations. When the concentration of the blood was sufficiently marked in the animals to which oxygen was being administered, heart punctures were made and the percentage saturation of the hemoglobin of the bloods from the right and left hearts was determined by means of Barcroft’s differential blood gas apparatus.’ Some difficulty was experienced in obtaining blood from the hearts of animals in which the lungs were large and edematous but the sample was considered satisfactory when it was obtained quickly and with little struggling on the part of the animal. The protocols are given below as well as curves showing the con- centration of the blood. In the curves, the percentage varia- tions from the normal are plotted to make all of the curves directly comparable. The maximum concentrations observed in the control animals varied from 30 per cent to 60 per cent above normal (average 43 per cent) while in the animals receiving oxygen the variation was from 28 per cent to 75 per cent above normal (average 48 per cent). It is apparent that, on the whole, the blood of animals which received oxygen con- * The Haldane oxygen apparatus furnishes oxygen to a mask fitted with valves for incoming and outgoing air. When the mask is worn by the animal, breathing is easy and the air in the mask may be enriched by varying amounts of oxygen. 3 We are indebted to Mr. Barcroft, Captain Dunn and Captain Peters for these data. a EFFECT OF OXYGEN ON BLOOD VOLUME IN LUNG EDEMA 159 TABLE 1 Chlorpicrin 1/8500 for 25 minutes (10.10 to 10.35 a.m.), August 19, 1918 GoaT 4537 GoaT 4567 9.35 a.m. Hb. 80 10.00 a.m. Hb. 60 10.35 a.m. Gassed 10.35 a.m. Gassed 11.00 a.m. Continuous oxygen by mask,| 11.00 a.m. Continuous oxygen by 1 liter per minute mask, 1 liter per minute 11.40 a.m. Hb. 79 (100 per cent) 11.50 a.m. Hb. 70 (117 per cent) 2.20 p.m. Hb. 85 (106 per cent) 2.40 p.m. Hb. 84 (140 per cent) 2.45 p.m. Continuous oxygen by} 2.45 p.m. Continous oxygen by mask, 3 liters per minute mask, 3 liters per minute 4.25 p.m. Hb. 105 (131 per cent) 3.45 p.m. Hb. 105 (175 per cent). 4.45 p.m. Heart puncture. Arterial Heart puncture. Venous blood 55 blood 93 per cent saturated. Venous per cent saturated. Animal died blood 45 per cent saturated on table. L:H 8.0 5.45 p.m. Hb. 110 (137 per cent) 5.59 p.m. Died. L:H 8.2 Goat 4406 (CONTROL) GOAT 4542 (CONTROL) 9.45 a.m. Hb. 42 9.50 a.m. Hb. 74 10.35 a.m. Gassed 10.35 a.m. Gassed 12.10 p.m. Hb. 48 (114 per cent) 12.20 p.m. Hb. 76 (103 per cent) 2.30 p.m. Hb. 60 (443 per cent) 2.50 p.m. Hb. 88 (119 per cent) 3.00 p.m. Died. L: H 6.0 5.30 p.m. Hb. 96 (130 per cent). Found dead next morning. L:H 8.3 Continuous oxyge aS) S cy c Ss S o/4 st ~~ S Se = < 100 a — N 7 2 J ¥ B 6 xt Hours after gassing Fig. 1. Showing the concentration of the blood after gassing. Solid line: animals receiving extra oxygen. Arrows indicate time of heart puncture. No. 4567, venous blood 55 per cent saturated. No. 4537, arterial blood 93 per cent saturated; venous blood 45 per cent saturated. Dotted line: control animals. 160 D. W. WILSON AND S. GOLDSCHMIDT TABLE 2 Chlorpicrin 1/8500 for 30 minutes (9.30 to 10.00 a.m.), August 21, 1918 GoaT 4526 Goat 4446 (CONTROL) 5.35 p.m. (8/19). Hb. 53 5.30 p.m. (8/19). Hb. 72 10.00 a.m. Gassed 10.00 a.m. Gassed 10.15 am. Continuous oxygen by | 12.15 p.m. Hb. 96 (133 per cent) mask, 13 liters per minute 1.45 p.m. Hb. 118 (164 per cent) 12.00 m. Hb. 67 (126 per cent) 2.00 p.m. Died. 12.15 p.m. Continuous oxygen by mask, 4 liters per minute 2.10 p.m. Hb. 70 (132 per cent) 3.20 p.m. Hb. 73 (138 per cent) 3.50 p.m. Hb. 76 (143 per cent). Heart puncture. Arterial blood 90 per cent saturated. Venous blood (obtained only after struggling) 20 per cent saturated 4.15 p.m. Oxygen stopped 5.30 p.m. Hb. 76 (143 per cent) 6.05 p.m. Died. L:H 7.7 es ae iid S S160 3 Fa g a, t = Hor S77) : KZ Ry 7 y 7 xg 7 26 a S §/20 zZ S NS Ss st § /00 S t= ce uf ¥ / Hours after gassing Fig. 2. Showing the concentration of the blood after gassing. Solid line: animals receiving extra oxygen. Arrow indicates time of heart puncture. 4526, arterial blood 90 per cent saturated. Dotted line: control animals. No. EFFECT OF OXYGEN ON BLOOD VOLUME IN LUNG EDEMA 161 TABLE 3 Chlorpicrin 1/8500 for 25 minutes (9.16 to 9.41 a.m.), August 23, 1918 GoaT 4577 GoaT 4631 5.50 p.m. (8/22). Hb. 60 5.45 p.m. (8/22). - Hb. 58 9.41 a.m. Gassed 941 am. Gassed 9.55 a.m. Continuous oxygen by] 9.55a.m. Continuous oxygen by mask, 3 liters per minute mask, 3 liters per minute 10.23 a.m. Hb. 70 (117 per cent) 10.10 a.m. Hb. 54 (93 per cent) 12.12 p.m. Hb. 81 (135 per cent) 12.00 m. Hb. 58 (100 per cent) 2.35 p.m. Hb.95 (158 percent). Heart | 2.50 p.m. Hb. 61 (105 per cent) puncture. Both samples obtained| 5.00 p.m. Hb. 67 (116 per cent) only after struggling. Arterial blood | 5.55 p.m. Hb. 69 (119 per cent) 58 per cent saturated. Venous blood} 6.20 p.m. Heart puncture. Arterial 4 per cent saturated 2.40 p.m. Died on table. L: H 7.4 Goat 4490 (CONTROL) blood 95 per cent saturated. Venous blood about 50 to 60 per cent 6.20 p.m. Hb. 74 (128 per Died'on table. L: H 5.7 cent). GoaTv 4457 (CONTROL) 9.41 a.m. 10.43 a.m. 12.30 p.m. Zaloppsny. 2.20 p.m. Gassed Hb. 85 Hb. 100 (118 per cent) Hb. 124 (146 per cent) Ibe WSLS). 7/ Died. mal=/00) Cle / ve x rcenta ~ Hemoglobin pe Hours after gassing Fig. 3. Showing the concentration of the blood after gassing. animals receiving extra oxygen. Arrows indicate time of heart puncture. 4577, both blood samples obtained only after struggling. 941 a.m. Gassed 10.54 a.m. Hb. 94 12.20 p.m. Hb. 96 (102 per cent) 3.15 p.m. Hb. 106 (113 per cent) 6.30 p.m. Hb. 124 (132 per cent). Found dead next morning. L: H7.1 Solid line: No. No. 4631, arterial blood 95 per cent saturated; venous blood 50 per cent saturated. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, No. 1 162 D. W. WILSON AND S. GOLDSCHMIDT centrated as rapidly and to as great an extent as that of the control animals. In order to demonstrate the efficiency of the oxygen administration, samples of blood were taken from both sides of the heart at suitable intervals and analyzed for oxygen. In most of the experiments the venous and arterial blood samples were obtained from the heart with- out difficulty and contained hemoglobin which was normally saturated with oxygen. With the increased concentration of the hemoglobin the oxygen content of the blood was even above normal. Occasionally the blood was obtained only after considerable struggling on the part of the animal so that the reduced oxygen content of such bloods was to be expected. These observations are reported here merely to make the experimental record complete, as obviously the low oxygen content of such bloods is without bearing on the present problem. These experiments demonstrate that, by breathing oxygen-rich atmospheres, oxygen may be absorbed through the damaged and edematous lungs in quantities sufficient to maintain a practically normal level of oxygen in the arterial blood. The high saturation of the hemoglobin of venous blood with oxygen would seem to prove that the blood flow is sufficiently rapid to normally oxygenate the tissues. Nevertheless, in spite of the normal oxygenation of the tissues in the animals receiving oxygen, the blood concentrated as rapidly and to as great an extent as in the control animals. The conclusion therefore seems justifiable that the lack of oxygen in the tissues and consequent imbibition of water is not an important factor in causing the concentration of the blood in animals developing edema after being gassed with lung irritants. An indication of the severity of the lung edema was obtained by comparing the weight of the lung to the weight of the heart at autopsy. The high lung to heart ratios obtained in practically all of the animals studied show that a severe grade of edema had already developed. The extent of the edema as indicated by this method was as great in the animals receiving oxygen as in the controls. - Although the data are necessarily few, it is apparent that the efficient oxygenation of the lung tissue in the animals receiving oxygen failed to diminish the tendency for the development of the edema of the lungs. With the enormous accumulation of fluid in the edematous lungs and the loss of water from the blood running roughly parallel, it is a tempting study to estimate even in a rough way the possible water EFFECT OF OXYGEN ON BLOOD VOLUME IN LUNG EDEMA 163 interchange. An attempt has been made with data which are more or less incomplete and with calculations involving gross errors but the relations are so striking that they are presented here (table 4). In the table are recorded data and calculations from animals in which the hemoglobin was not determined immediately before death but is estimated from the curve obtained from the various determinations. These estimated values are quite similar to average values obtained at death on other animals. TABLE 4 Comparison of calculated amounts of fluid lost from blood and extra fluid in the lungs in gassed animals M/e®|/@®;}@/o]@]@]|] ® | ® | a | ay | a | as re a § o ! ay a 3 Qa re) 5 a < zn | Sx S io ee : i 3 4 eon] 6 2 EE | RE tees e | ae | 8 Zae| Ze p eee es eh (Mae | ag | ae | PSH | Ses] “2 felis | 54.) 82.|.6° S23 /.986] 22 | $3 | 883/964] 23 6 | 2 | es | efi as] @ | es eoae] 2 | Be | obal| BES] ge o A E 5 2 FS) z eS | z =) & g kgm. | gram | gram gram |per cent| per cent| cc. cc, ce. ce. 3602 | 6/13] 31.8] 157 | 1472 377 | 163 | 200 | 1750 | 875 | 875 | 1095 3638 | 6/12| 29.5} 126 | 623 302 | 125] 140 | 1625 1160] 465] 321 3787 | 6/20) 16.8) 74 | 566 4098 | 7/ 4| 35.0} 195 | 1103 4406 | 8/19) 15.5) 85] 510 3920 | 7/ 4| 37.8) 200 | 1700 3600 | 6/12) 25.5) 130 | 849 4047 | 7/12) 33.6) 210 | 1556 178 | 128] 145} 925] 638 | 287] 388 114 | 140 | 1925 | 1375 | 550} 635 204} 143] 150} 850] 567 | 283] 306 480 | 180 * | 2080 | 1155 | 925 | 1220 312 | 154 * | 1400; 910] 490| 537 504 | 130 * | 1850 | 1420 | 430 | 1052 Se UNC TEN eee) Foner © is (or) 100) *Time of death not known. Calculations made using last hemoglobin de- terminations (column 8). Column 6 = column 5 ~ column 4. Column 7 = column 4 X 2.4. Columns 8 and 9. Values calculated using the normal as 100 per cent. Column 9 = extrapolation to time of death. Column 10 = column 3 X 0.055 X 1000 (Boycott and Damant, Journ. Physiol., 1907-8, xiv, 36). Column 11 = column 10 + column 9. Column 12 = column 10 — column 11. Column 13 = column 5 — column 7. Examining the last two columns of the table it is evident that in only one instance the amount of extra fluid in the lung is less than the calculated loss of fluid from the blood. In some instances the extra fluid in the lung is much greater than that lost by the blood. Little or no water is drunk by goats in this condition and the volume of urine 164 D. W. WILSON AND S. GOLDSCHMIDT excreted is small, so that the external factors do not confuse the picture. Even with the relatively large errors of calculation involved, the conclusion seems Justified that the loss of fluid by the blood can be accounted for by the excess of liquid in the edematous lung. The evidence suggests that the muscles, etc., do not imbibe water and cause the concentration of the blood. In fact it would appear that water may be drawn from some tissues to make up part of the volume of liquid in the lung. We are thus finally led back to our original point of view that the development of the edema of the lungs and the concentration of the blood are interrelated and are the im- portant factors in the pathological condition studied. With this fact established it is justifiable to conclude that the development of the edema of the lungs is the primary factor in the condition and that the development of the edema causes the concentration of the blood. SUMMARY The continuous administration of oxygen to goats gassed with chlorpicrin did not inhibit the concentration of the blood. The percentage saturation of the hemoglobin with oxygen was normal even after a considerable concentration of the blood had occurred. The concentration of the blood is not caused by the imbibition of water by the tissues as the result of oxygen want. The loss of water from the blood is therefore due to the development of the edema of the lungs. We wish to thank Mr. J. Barcroft and the staff of the Physiological Laboratories of the R. E. Experimental Station, Porton, England, for the many kindnesses extended to us throughout the course of our investigations there. THE EFFECT OF ADRENALIN, DESICCATED THYROID AND CERTAIN INORGANIC SALTS ON CATALASE PRODUCTION W. E. BURGE From the Physiological Laboratory, University of Illinois Received for publication July 19, 1919 Dahm and Steck (1) found that the ingestion of urea and of sodium chloride increased oxidation. This observation has been repeated and confirmed by Tangl (2) on curarized animals with their kidneys removed. Raeder (3) also found that saline injections especially if hypertonic increased the respiratory exchange. Loewy (4) observed that the in- gestion of small amounts of water (100 cc.) produced no change in oxygen consumption, while Speck (5) found that drinking large quan- tities (1250 ec.) did. According to Lusk (6) water, sodium chloride and urea have no effect on the respiratory exchange. Magnus-Levy (7) observed an increased carbon dioxide output in a man fed thyroid extract and an increased oxygen intake in cases of exophthalmie goiter. It is recognized that tetrahydro-6-naphthylamin increases oxidation 1n the body and decreases heat elimination. Grafe (8) has shown that the administration of ammonium carbonate, ammonium chloride and sodium carbonate increases oxidation. We (9) had found that. whatever increased oxidation in the body, the ingestion of food, for example, produced an increase in catalase by stimulating the alimentary glands, particularly the liver, to an in- creased output of this enzyme, and that whatever decreased oxidation, narcotics, for example, produced a decrease in catalase by direct de- struction and by decreasing its output from the liver. The object of the present investigation was to determine if adrenalin, tetrahydro-§- naphthylamin, desiccated thyroid, water, sodium chloride, ammonium chloride, sodium carbonate, ferric chloride, sodium citrate, am- monium carbonate, urea, triacetin and saccharin would or would not produce an increase in catalase. The amounts of the substances used will be given in the description of the individual experiments. The animals used were dogs and rabbits. After etherizing these animals 165 166 W. E. BURGE and opening the abdominal wall, each of the substances was introduced into the upper part of the intestine. The catalase in 0.5 ec. of blood was determined before, as well as at fixed intervals after the intro- duction of the materials. The determinations were made by adding 0.5 ec. of blood to hydrogen peroxide in a bottle at approximately 22°C. and the amount of gas liberated in 10 minutes was taken as a measure of the catalase content of the 0.5 ec. of blood. In figure 1 are shown the effects of the introduction into the intes- tines of rabbits of urea, sodium chloride, water, triacetin and glycerine. The amounts of the substances used are indicated on the chart. Seventy- five cubic centimeters of water were used in dissolving the different substances. The figures along the ordinate (0 to 780) indicate amounts of catalase measured in cubie centimeters of oxygen and those along the abscissae time in minutes. It may be seen that 2 grams of urea, 1 gram of sodium chloride and 15 ec. of water per kilo produced no increase m catalase in keeping with Lusk’s observation that amounts of these substances as small as these produced no increase in oxidation, while 10 grams of urea, 10 grams of sodium chloride and 150 ce. of water per kilo did produce an increase in catalase in keeping with the observations of Dahm and Steck, Tangl, Speck and Raeder, that large amounts of these substances produce an increase in oxidation. It may also be seen that triacetin is more effective in producing an increase in catalase than glycerine, this result being attributed to the presence of the acetic acid radicle’in the glycerine molecule. The second part of the paper is concerned with determining the mode of action of the substances already mentioned as well as several other substances in producing an increase in catalase. The animals used were dogs and the method for determining catalase was the same as that already described. The substances were dissolved in 200 ce. of dis- tilled water and were introduced at body temperature into the upper part of the small intestine. The catalase content of 0.5 ec. of blood taken from the liver, portal and jugular veins was determined before as well as at certain fixed intervals after the introduction of the materials. In figures 2, 3 and 4 the continuous line curves were con- structed from data obtained from the blood of the liver, the discon- tinuous line curves from the blood of the portal, and the dotted line curves from that of the jugular vein. In figure 2 it may be seen, as we have found before, that the intro- duction of glycocoll produced an increase in catalase as is indicated by CATALASE PRODUCTION 167 the increase in the amount of oxygen liberated by the blood. It may be seen further that this increase is greater, particularly during the first fifteen minutes, in the blood of the liver than in the blood of the jugular or portal veins. UREA SODIUM CHLORIDE WATER » TRIACETIN GLYCERINE | paint Fara | | | 4 ~ i i | : + oe + s fsialss | RIMM clagieN MIENIST celiac le lala 7 | 1 | | | +— ott) mF Me ee | (SS ee ee Ree Seles ea o 1 4 9 + is . ae : = Z | Fes Ball Dee | | 4 | | f | | fo} | | | | x | IR | = Ea a a 2 — | + | | | = US iL | ++ | | = £ | | 3 | Papal | 660, a eel fet 2 1 a = — | MES u ° | iS) — = | tT al as ‘al + T T jaa | | | a} al lest | ~ + + r + — wh & | | bal He >| | 690 ess | Jide jos te fe [sa5/ ml - | | | | | | eb 580 es] 58 ~ = o— | 583 560/ | if i LS FEL 5e0 laa 558) icalaal eal | {ai ad —o + 535. ——. t — = Si ths + [teal Bea ’ el | | | | | i 1 = | | | | S20 sis | ian, | ey i T - = al T 500 | | jie | ] vi 7 i met 1 == — , ~ a ] | | | | | | | | | | ~ Ss + . : + r - = ee} | g | | iL =e 1c 3 1 4 | | is | ie ml < fmeles|meliarli =| ima mie Spe a mS (ati x & ‘eal | ° | es pron fg rut ase ! a ie pe Base 2 | 4. —|—-— 4 leat [ ala | Ce ee eee Pilato Gehalel =] laa | lien a 7 my | 5 = | [ | | aI jeead ll ewe? | | Ale — TF | T Jean =) T T | T a | } . aa = | a 2 | Eee | ) te ale | | 1 | | ia | ea Al | | | | | ' = | | | | | | ile [ fee 280 290 ry — bi} oon a = | | Pel | | y 265 teil lind] | rl jreciteajealieatl eal ei ateg Ee ig pay ma EeO 1 +——+ 4 + =i}. —e eet [ —~ inal isl ihaee2 rae cee] [ea aw Smotenlantatael im leat j | T are fal ian |_| | HE ee | | [a | ame | | | | | | ire We | | lll | (el eel | | | a = — } mat ——-+ ok [| alii | |_| BME ie T | | T ] al | Teolealaw lea Pea | Lilliaalees le es | | {fe pe =e es iL eee ae | | 4 a +——}-_+ +--+ 4 BEimieigick | oa feist te alent a | ee | 0 15 30 45 60 75 90 105 i20 185 150 0 15 XS 45 60 75 30 0 15 30 45 6 © 15 30 45 60 75 90 105 120 0 15 30 45 60 75 90 TIME IN MINUTES Fig. 1. The effect of the introduction into the intestines of rabbits of the substances named in the chart on blood catalase. Under urea it may be seen as with glycocoll that this substance increases the catalase of the blood of the liver more rapidly than that of the portal and jugular veins. Under glycerine and triacetin it may be seen that while 5 grams per kilo of glycerine produced no increase in catalase, a similar amount of triacetin produced an in- crease thus showing that triacetin is more effective in this respect than BURGE W. E. 168 I eee “UIOA uvpnsnt ayy JO pooyq oy} Woay SoAINO OUI] poyyop oYyy puUB [e}10d oYy JO pooyg oYyy Woy SoUO SNONnUTyUODSIP oYy4 {1OAT] OY} JO POOT OY} WA} pouTeyqo BIVpP WOA, PoYONAYSUOD 910M SOAIND OUT] SNONUIZUOD OY], “oSvleyBo Poo]|q UO JABYD OY} UI pouvU sodURYSqns oY} JO SBoOp Jo souTysoJUT OY} OFUT UOTPONpOAzUT 9yy JO Jooyo oy, °% “SLT SALANIN NI GWIL 09 Gr O¢ GI O 09 Gy O€ Gi O Gol 06 Gl O9 Gy OF GI O Gel OZ! SO! 06 Gl O9 Sr OF GI O Gy o¢ Gl O 0 ee > ial nf Sia) Sle a 2 ae aa oe 2 SS Ss eS VS | w =e ee ha } 02 L_ se aa oe S ee ee oO | ce 3 EEE es | a Ov f > ie ga nm saa le og — Ne I a Ss eh x + 2 o : al is sl 09 a i Ninitel pss = 9 OL m ——}—_ (e/a ———+— if es E ye SS ES oye 2 2 og ale oO Seat : “re f 2 e ae : cope a — — 9) = 7 >] 9 aa Pa 96 |8 g io \i2 a | > (a ae a i bor | 00! vol in m a a (e) ae ca 901 eyes R a, a x : | a) tal | 601 Boa i i : Onl S - +— - ca et eal ie a o {si Zz { NIiavHOOVS NILAOVId ANISSAZDAI1SD vauyn OXYGEN CiG. OF IN MEASURED OF CATALASE AMOUNTS CATALASE PRODUCTION 169 glycerine. It may also be seen that 5 grams per kilo of saccharin produced an increase in catalase. It (10) had been found that 5 grams of sugar would not produce so large an increase in catalase as is here shown to be produced by this amount of saccharin, hence so far as WATER SODIUM CHLORIDE | ammonium CHLORIDE ley lay. | aa i i ie ; +——t. ee ee ee ee ee ee ee ee FERRIC CHLORIDE {SODIUM CITRATE — | hiv ent’ | | | | 165}— a a Ga a i rl =x{—— 160 E =| | 3 : 150 —} ce] 115 a J x 110 7 a x 7 'G7__ 0g B | Py 105 |—-+ aa = + +f t 7m 100 aN Z| YR eed bs eB a Ret toe ie Idea | es he | RSS 100; 100-5m | Pee ] at 95 at sitet er = ml a aol ss 30 T te “par + | 2 ss ee — + SS a a al i i | | 9 85 + + = a 165 80 i 75 70 3 65 T 60 55 ale: 45 40 35 Oo | sta | 0 15 © 45 & 75 30 O 15 30 45 60 75 9 105 O IS 30 45 60 O 15 30 45 60 75 O 15 30 45 60 Oo I 30 TIME In MINUTES Fig. 3. The effect of the introduction into the intestines of dogs of the sub- stances named in the chart on blood catalase. The continuous line curves were constructed from data obtained from the blood of the liver; the discontinuous ones from the blood of the portal and the dotted line curves from the blood of the jugular vein. catalase production is concerned saccharin is more effective than sugar. It would seem that in addition to being a sweetening agent, saccharin, although not oxidized itself to give rise to energy, as is the case with sugar, serves to stimulate the alimentary glands to an in- 170 W. E. BURGE TE TRAHYDRO- ADRENALIN |FRESH THYROD} - OLD THYROID T | | | 156 [ | 149 OXYGEN G ro) OF Ne cere 8 MEASURED Te ee eh oe od OF CATALASE AMOUNTS QO 15 30 45 60 0 15 30 0 15 30 45 60 75 90 105 oO 15 30 45 TIME IN MINUTES Fig. 4. The effect of the introduction into the intestines of dogs of the sub- stances named in the chart on blood catalase. The continuous line curves were constructed from data obtained from the blood of the liver; the discontinuous ones from the blood of the portal and the dotted line curves from the blood of the jugular vein. CATALASE PRODUCTION ie creased output of catalase thus facilitating the oxidation of the other food materials. In figure 3 it may be seen that the introduction of water, sodium chloride, ammonium chloride, sodium carbonate, ferric chloride and sodium citrate into the alimentary tract of dogs produced an increase in the catalase of the blood and that this increase was brought about by stimulating the alimentary glands, particularly the liver, to an increased output of this enzyme. The fact that these salts produce an increase in catalase and hence in oxidation may in some measure account for the beneficial effect of certain mineral waters which contain these substances in conspicuous amounts. In figure 4 is shown the effect of adrenalin, freshly prepared and old thyroid and tetrahydro-8-naphthylamin on catalase production. Three cubic centimeters of a 1: 1000 solution of adrenalin chloride were intro- duced into the portal vein in dog 3 and 5 ee. in dog 2.__In these two dogs the solutions were injected as quickly as could be conveniently done while in dog 1, 10 ce. of a 1: 1000 adrenalin chloride solution diluted to 50 ce. were injected at a rate of about 1 cc. per minute, thus requiring approximately 30 minutes for the injection. It may be seen that the adrenalin increased the catalase of the blood by stimulating the liver to an increased output of this enzyme. As a result of the work of Blum (11), Vosburgh and Richards (12), Dreyer (13), Oliver and Schaefer (14), Cannon and de la Paz (15) it is now believed that during combat the adrenals are stimulated to an increased output of adrenalin and that this produces a constriction of the small blood vessels of the abdominal viscera, thus increasing the blood supply to the heart, skeletal muscles and nervous system; that it hastens the coagulation of the blood and increases the output of sugar from the liver. It is evident that the result of diverting the blood from the abdominal viscera into the heart, skeletal muscles and nervous system during combat is to render conditions more favorable for the increased action of these organs; that the hastening of the co- agulation is to stop more quickly the bleeding from any superficial wound that may be inflicted, and that the flushing of the blood with sugar is to insure a plentiful supply of oxidizable material to the muscles. While the preceding hypothesis explains certain phases of adaptation of the organism for combat, it does not explain how the increased oxidation is brought about which gives rise to the energy for the fight. We had already found that the stimulation of the splanchnic nerves to the liver produced an increased output of catalase ey? W. E. BURGE from this organ, and had suggested that the increased oxidation during combat might be due to this increase in catalase. The fact that adrenalin stimulates the liver to an increased output of catalase sug- gests that this result may occur during combat when there is an increased output of adrenalin into the blood. It may be seen in the figure that the introduction into the alimentary tract of freshly prepared thyroid produced an increase in catalase while old thyroid did not. The old thyroid was a Parke-Davis prepa- ration which had been standing in the laboratory for four years, while the fresh thyroid was material recently purchased from this same company. The old material was fed to cats and found to have lost its virtue while the new material produced the characteristic effects, loss of weight, etc. The amount of the thyroid introduced into the intestines of the dogs was 1 gram per kilo dissolved in 200 ce. of water. We had already found that the feeding of thyroid to cats increased very greatly the catalase of the blood. The experiments described in this paper on the introduction of thyroid into the alimentary tract of dogs suggest that the increase in the catalase of the blood of animals fed with thyroid is due to the stimulation of the liver to an increased output of this enzyme. Winternitz (16) found that “the removal of the thyroid gland caused a drop in the catalase activity of the blood which was compensated if thyroid were fed” and that in hyperthy- reosis the catalase of the blood tends to increase while in hypothyreosis it assumes a lower level than normal. Becht (17), on the other hand, claims that thyroid feeding decreases the catalase of the blood. It should be mentioned also in this connection that Becht holds that narcotics slightly increase the catalase content of the blood, while we found that they produce a great decrease both in vivo and in vitro.! Under tetrahydro-8-naphthylamin it may be seen that the intro- duction into the intestine of 0.8 gram per kilo of this substance dis- solved in 200 ce. of water stimulated the liver very greatly to an in- creased output of catalase which is offered in explanation of the increased oxidation produced by this substance. 1 Owing to our proximity it has been suggested and even urged that Doctor Becht and myself carry out some joint experiments in an attempt to clear up the differences in our results. I am sorry to say that Doctor Becht seems to be unwilling to carry out such experiments. CATALASE PRODUCTION 173 SUMMARY The introduction into the alimentary tract of relatively small amounts of water (15 ec.), of sodium chloride (1 gm.), and of urea (2 gms.) per kilo, produces no increase in catalase in keeping with Lusk’s ob- servation that small amounts of these substances produce no increase in oxidation, while the introduction of large amounts of these sub- stances, 1500 ce. of water, 10 grams of urea per kilo and 10 grams of sodium chloride per kilo do produce an increase in catalase in keeping with the observations of Dahm and Steck, Tangl, Speck and Raeder, that large amounts of these substances produce an increase in oxidation. The injection of adrenalin into the portal vein stimulates the liver to an increased output of catalase. This fact suggests that the in- creased amount of adrenalin thrown into the circulation during com- bat may stimulate the liver to an increased output of catalase, and in this way aid in bringing about the increase in oxidation occurring during combat. Desiccated thyroid when introduced into the alimentary tract stimulates the liver to an increased output of catalase. This ob- servation suggests that the increase in the catalase of the blood which may be responsible for the increase in the respiratory exchange of an animal when fed with thyroid or in exophthalmie goiter is probably due to the stimulation of the liver to an increased output of catalase. BIBLIOGRAPHY (1) Daum anp Steck: Physiol. Gesellsch. zu Berlin Verhandl., 1909, xxxiv. (2) Tanex: Biochem. Zeitschr., 1911, xxxiv, 1. (3) RagpErR: Biochem. Zeitschr., 1915, lxix, 257. (4) Lonwy: Arch. f. d. gesammt. Physiol. d. Menschen, 1888, xliii, 515. (5) Speck: Policlinico II, 1892, xlii. (6) Lusk: Journ. Biol. Chem., 1912, xiii, 27. (7) Maenus-Levy: Berl. klin. Wochenschr., 1895, xxx, 650. (8) Grare: Deutsch. Arch. f. klin. Med., 1915, exviii, 1. (9) Buree: This Journal, 1919, xlviii, 133. (10) Burcre: Med. Rec., December, 1918. (11) Buum: Arch. gesammt. Physiol., 1902, xc, 617. (12) VospurGH AND Ricuarps: This Journal, 1903, ix, 29. (13) Dreyer: This Journal, 1899, ii, 283. (14) OLIveR AND ScHaEFER: Journ. Physiol., 1895, xviii, 230. (15) Cannon anp De La Paz: This Journal, 1911, xxviii, 64. (16) WinteRNITz: Arch. Int. Med., 1911, vii, 625. (17) Becut: This Journal, 1919, xlviii, 171. A NOTE ON THE QUESTION OF THE SECRETORY FUNCTION OF THE SYMPATHETIC INNERVATION TO THE THYROID GLAND C. A. MILLS From Laboratory of Pharmacology, University of Chicago and Biochemical Laboratory, University of Cincinnati Medical School Received for publication July 21, 1919 Considerable interest has attached itself to the question of the inner- vation of the thyroid gland, and as yet the conflicting evidence presented by different experimenters does not permit the drawing of definite conciusions. Asher and Flack (1), by stimulating the laryngeal fibers from the vagus, found the blood pressure response to adrenalin injection to be greatly increased. Asher (2) later decided that this increased effectiveness of adrenalin was due to the sensitizing action of the thyroid secretion on sympathetic endings following stimu- lation of the nerve fibers to the gland. Cannon and Cattell (3) demon- strated an electrical variation in the gland, which they interpreted as evidence of secretory activity following stimulation of the cervical sympathetic fibers, while stimulation of laryngeal fibers from the vagus was ineffective. Levy (4) observed an increased blood pressure re- sponse to adrenalin after stimulation of cervical sympathetic and after adrenalin injection, which he accepted as evidence of sensitization of sympathetic endings by the thyroid secretion following stimulation. Cannon, Ringer and Fitz (5), by fusion of the central end of the phrenic nerve to the peripheral portion of the cut cervical sympathetic trunk, obtained symptoms of sympathetic stimulation in the eye and in the ear vessels of rabbits, and also symptoms similar to those of hyper- thyroidism, such as rapid pulse, increased metabolism and hyper- excitability. These latter symptoms disappeared on removal of the thyroid on the side of the nerve suture, so the evidence was taken to indicate thyroid stimulation by the phrenic impulses. On the other hand Troell (6) failed to obtain any evidence of thyroid stimulation, either symptomatically or microscopically, on suturing the phrenic to the cervical sympathetic. Burget (7), and Marine, 174 SYMPATHETIC NERVOUS CONTROL OF THYROID FUNCTION 175 Rogoff and Stewart (8) also reported negative results by this same method. Manley and Marine (9) and others, have reported that thyroid transplants survive and function in various parts of the body, regardless of nerve supply, so that secretory nerves to the gland are at least not essential to its activity. From the above contradictory evidence presented it is evident that any new contribution should be welcome. The object of the experi- ments described in this paper was to determine whether the thyroid could be stimulated to greater activity through repeated injections of cocaine, the basis of the use of cocaine being the work of Froelich and Loewi (10), in which they demonstrate quite conclusively that cocaine sensitizes sympathetic nerve endings to the action of adrenalin and to sympathetic stimulation. EXPERIMENTAL The subcutaneous injection of 35 to 50 mgm. cocaine hydrochloride per kilo body weight, or of 10 mgm. cocaine hydrochloride accom- panied by 1 mgm. adrenalin per kilo body weight, calls forth in the rabbit symptoms of undoubted sympathetic stimulation, such as maximal dilatation of the pupils, slight excphthalmos, constriction of the ear vessels leaving the ears cold to the touch, relaxation and loss of tone in the intestines, and erection of the hair over the body. That it is either a sensitization of the sympathetic endings to the normal flow of impulses over these nerves, or else an increased central stimulation, ean be shown by cutting one sympathetic nerve in the neck and then noting the effect of cocaine on the pupils and ear vessels. Effects of the drug were seen only on the side having the intact nerve supply. Further, on frightening the animal, as by dropping from a distance of a foot or so onto the observation table, a slight dilatation of the pupil on the side of the intact nerve occurs within one-half second and dis- appears within one second after cessation of the frightening, thus indicating that the results obtained are not due to a sudden outpouring of epinephrin, but far more likely due to impulses traversing the sym- pathetic nerves. Having established to my satisfaction, then, that cocaine acts in the normal rabbit by increasing the effectiveness of the normally occurring sympathetic impulses, I next tried to determine the effects on the thyroid gland as evidenced by histological changes in its structure. Specimens of the gland were removed aseptically before and after the cocaine treatment and examined for changes in their microscopic 176 Cc. A. MILLS appearance. Ordinary aseptic precautions were observed in the injections. Since all references in the literature indicate that cocaine is only slowly eliminated, (Wiechowski (11), Grode (12) the treatment was begun with daily injections of sufficient strength to produce a maximal mydriasis of the -pupils lasting about 30 minutes, that is, about 10 mgm. of cocaine per kilo body weight. However, since the effects seemed to be over within an hour, the injections were increased to 3 or 4a day. In one series of five rabbits thus treated for 8 days there was observed no change whatever in the microscopic appearance of the thyroid glands. A second series of three rabbits was run, each animal receiving 5 to 10 injections a day for 11 days, and again no changes in the glands were observed. Neither did the rate of growth or general appearance and behavior of the animals indicate any lasting results from the use of the drug. Since the strength or effectiveness of the impulses over the sympa- thetic fibers should have been greatly augmented by the continued use of cocaine producing symptoms of thyroid hyperactivity and mor- phological changes in the gland, the evidence from these experiments contributes to the indication of a lack of secretory function of the sympathetic fibers to the thyroid gland. Note. The experimental part of this work was carried on in the Laboratory of Pharmacology of the University of Chicago at the suggestion, and under the direction, of Dr. A. L. Tatum, for whose aid and suggestions I desire to express my sincere thanks. BIBLIOGRAPHY (1) AsHer, LEon AND Fuack: Zeitschr. f. Biol., 1910, lv, 83. (2) AsHER AND Leon: Deutsch. Med. Wochenschr., 1916, xxxiv, 1. (3) Cannon, AND Carte: This Journal, 1916, xli, 58. (4) Levy: This Journal, 1916, xli, 492. (5) Cannon, BINGER AND Fitz: This Journal, 1914, xxxvi, 363. (6) Troeuy: Arch. Int. Med., 1916, xvii, 382. (7) Burecet: This Journal, 1917, xliv, 492. (8) Martine, Rocorr anp Stewart: This Journal, 1918; xlv, 268. (9) MantEey AND Marine: Proc. Soc. Exper. Biol. and Med., 1914, xii, 202. (10): FrRoELIcH AND Lorwt: Arch. f. Exper. Path., 1910, Ixii, 158. (11) Wrecnowskr: Arch. f. Exper. Path. u. Pharm., 1901, xlvi, 155. (12) Grove: Arch. f. Exper. Path., 1912, lxvu, 172. THE AMERICAN JOURNAL OF PHYSIOLOGY VOL. 50 NOVEMBER 1, 1919 No. 2 THE HYPERGLYCEMIA-PROVOKING ABILITY OF ASPHYXIAL BLOOD K. YAMAKAMI From Tohoku Imperial University, Japan Received for publication June 18, 1919 The role of adrenalin in the asphyxial hyperglycemia and glycosuria has been a question for a long time and is not yet decided. Pollak (1), in classifying the various experimental conditions in which hyperglycemia occurs, regarded asphyxial hyperglycemia as due to a stimulation of the sympathetic nervous system akin to that pro- duced by the injection of adrenalin. The experimental basis for this opinion lay in the fact that the asphyxial glycosuria and hyperglycemia could not be prevented by resecting the splanchinic nerves of both sides, though they were remarkably lowered in degree by this operation. Emil Starkenstein (2) carried out a number of experiments in which he proved histologically a diminution of the chromaffin substance in the medullary cells of the adrenals of asphyxiated rabbits. Extracts from the adrenals of these asphyxiated rabbits also exhibited a lessened power to raise the blood pressure when injected into normal rabbits. The histological staining abnormities as well as the decrease of the blood-pressure-raising substance could be prevented by cutting the splanchinic nerves before suffocation. He presumed upon this experi- mental ground that the asphyxial stimulation acts first upon the blood- sugar-controlling centre of the brain as in the case of Piqdre, and then this stimulation is transmitted along the sympathetic nerve to the suprerenal, causing a hypersecretion of suprarenin by exciting the medullary cells and this adrenalin, transferred into the circulation, gives rise to the increased sugar content of the blood. 177 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 2 178 K. YAMAKAMI Czubalski (3), supporting the view that hypersecretion of adrenalin occurs in asphyxia, proved that the rise of the blood pressure of animals does not occur if previous to suffocation the adrenals are removed. Borberg, Frederica and many others carried out similar experiments to prove the réle of the suprarenals or suprarenin in asphyxial glycosuria and hyperglycemia. But most of the methods of investigation em- ployed by these authors were indirect and did not permit a definite decision as to the réle of suprarenin. In order to solve the question whether the adrenals are involved in asphyxial hyperglycemia, it seems to be the wisest method to study this hyperglycemia in animals whose adrenals were removed entirely. This is, however, not an easy task for most animals live only a short time after the operation. It would not be safe, as Stewart has pointed out, to take the experimental results obtained from animals in such a condition that they are about to die, as evidence of an important réle of adrenalin in experimental diabetes. The only animal which can survive double adrenalectomy for long without any pathological symp- toms, is the rabbit. Consequently most of the experiments in this field have been performed upon this animal. Many investigators have stated that hyperglycemia or glycosuria do not occur after removal of the adrenals. The non-occurrence of Piqtre glycosuria after adrenalectomy was proved by Meyer (4), that of the glycosuria provoked by splanchnic ‘stimulation was proved by Thomas (5) and Macleod (6), that of the CO-glycosuria was proved by Starkenstein. Kahn, Starkenstein and others are said to have proved that various kinds of glycosuria do not manifest themselves in rabbits even one year after adrenalectomy (7). Macleod and Pearce (8) obtained similar results in regard to hyper- glycemia. Kahn has reported in his more recent paper the non-occurrence of hyperglycemia following diabetic puncture, CO, diuretin, emotion and salt, in rabbits after successful adrenalectomy (9). These results obtained by various authors might be considered as sufficient evidence to prove the indispensable intervention of the ad- renals or of adrenalin in all experimental diabetes and hyperglycemias, except the renal and pancreas diabetes, if other authors had not ob- tained the opposite results. Unfortunately many exceptional cases have been observed, in which adrenalectomy of both sides did not pre- vent the experimental glycosuria or hyperglycemia, especially in other animals than rabbits. For instance, Wertheimer and Battez (10) ASPHYXIAL HYPERGLYCEMIA 179 could obtain glycosuria in cats by Piqtire after adrenalectomy. Stark- enstein observed a remarkable glycosuria produced by stimulating the splanchnic nerves for a long time in rabbits deprived of their adrenals. Kahn comments upon the occasional occurrences of the glycosuria or hyperglycemia after adrenalectomy in the following words: Es wird also anzunehmen sein dass die auf dem direkten Wege erzeugte Rei- zung der sympatischen Nervenende in der Leber, weder vei der CO noch bei -Piqtre hoch gradig genug sei, um zur plétzlichen ueberstiirtzten Glykogen Mobil- isierung zu rithren. Diese notwendige Erregungsgrésse wird erst durch die Addition von Adrenalin zu Wege gebracht. Wenn aber durch irgend eine unbe- kannte Ursache, die Erregungsgrésse erbracht im direkten Wege gross genug sei, dann treten die Glykosurie oder Hyperglykaemie in nebennierenlosen oder Splanchinicus resezierten Tieren auf. There are too many presumptions which have not sufficient experi- mental foundation in the explanation offered by Kahn. But, in any case, this much seems to be true, that various experimental hyperglyce- mias can be made difficult to manifest themselves by extirpation of adrenals even in the presence of a sufficient deposit of liver glycogen. Besides there is one more possibility which may account for the hyper- glycemia after extirpation of the suprarenals, in favor of the adrenalin theory, and that is the probability of adrenalin secretion from other glands than the adrenals. The carotid glands, Zuckerkandls glands, Luschkas glands and others which contain chromaffin cells may produce and secrete suprarenin. The amount of suprarenin produced from this source may be very small under normal conditions of life. But it is quite conceivable that this amount may rise compensatorily when the main sources of adrenalin are removed and extraordinary stimulations are applied to the secretory nerves. Recent attempts to elucidate this problem have been made by two American investigators and their students, namely, Cannon and Stew- art. They tried to determine the increase of adrenalin in the blood of animals under experimental conditions directly by means of segments of uterus and intestine. And curious to say, their results were just opposite to each other, Cannon and his students having obtained posi- tive (11), Stewart and his collaborators having obtained negative results (12). Stewart worked on cats in which one adrenal was removed, and the nerve supply for the other was cut off. It is not surprising that he was able to provoke asphyxial or other experimental hyperglycemia in such animals because, as was cited before, the experimental hyper- 180 K. YAMAKAMI glycemia manifests itself sometimes even in cats deprived of both adrenals. But his negative results, in attempting to detect the epi- nephrin output from the remaining adrenal, upon which experimental basis he infers that the asphyxial hyperglycemia occurs without any intervention of adrenalin, may have been due to the exhaustion of epinephrin, because the methods employed by Stewart to obtain the blood sample included numerous procedures which have been shown by various investigators to cause experimental hyperglycemia. The fixation of animals upon the operation table, narcotization, laparotomy, ligation of a large artery and vein, etc., have all been reported as caus- ing a rise in blood sugar. And all of these hyperglycemias may be accompanied by hypersecretion of adrenalin, as many authors believe. Even if the asphyxial stimulation leads to hypersecretion from the adrenal glands, this secretory stimulation can not be very strong. The various manipulations involved in obtaining the asphyxial blood sample may very probably, therefore, have exhausted the adrenals to such a degree that no more epinephrin could be elicited from the glands by such a weak stimulation as asphyxia. Hypersecretion under these conditions might only have been expected in response to such an extra- ordinary stimulation as the direct massage of the glands or strong galvanization of the secretory nerve, which Stewart found to cause an increased adrenalin output. Indeed the entire lower part of the animal body with many abdominal organs was already in an asphyxial condition at the conclusion of the preparations for securing the blood sample, if I am not mistaken. And again, as stated above, the epi- nephrin may be produced and discharged into the circulation by other glands than the adrenals. There is too great a discrepancy between Stewart’s results and those of Cannon, who has proved a remarkable increase of adrenalin in the caval blood and sometimes even in the heart blood during asphyxia. Whether this discrepancy is due to the inaccuracy of the methods em- ployed for the estimation of epinephrin or due to the diversity of meth- ods of obtaining the blood sample, is a question which calls for further investigation. The results of the experiments which I have to report here seem to be rather in accordance with the adrenalin theory. I have undertaken this work with the object of ascertaining whether - the asphyxial blood has any appreciable ability of provoking hyper- glycemia when transferred into the circulation of other animals and have obtained a positive result. ASPHYXIAL HYPERGLYCEMIA 181 Lepine and Boulud (13) have proved the existence of a substance in asphyxial blood which can induce glycosuria when injected into animals. Lepine called this substance ““Leucomaines diabétogénes.”’ He extracted it from the asphyxial blood of dogs, and could cause a remarkable glyco- suria in guinea pigs by injecting this extract. According to his hypothe- sis, this substance is produced in the blood by the lack of oxygen, and causes the glycosuria by preventing the oxidation of sugar. But the Leu- comaines diabétogénes is said to be capable of inducing a glycosuria last- ing two to three days by only one subcutaneous injection. The as- phyxial glycosuria and hyperglycemia do not last so long generally. They usually disappear within five or six hours after the removal of the cause. Only the suffocation with CO-gas is reported to be sometimes followed by glycosuria lasting twenty or thirty hours, but this is prob- ably due to the CO poisoning and not to the asphyxia itself. Hence it is very doubtful whether this substance which Lepine separated from asphyxial blood has anything to do with asphyxial glycosuria. Besides no subsequent investigators have been able to confirm the existence of this substance. In my experiments rabbits were used as experimental animals. The blood was drawn by heart puncture from rabbits during deep asphyxia- tion and was injected into the auricular vein of other normal rabbits and then the blood sugar of the injected animals was determined twenty to thirty minutes after the injection. DESCRIPTION OF TECHNIQUE The hydroéxylamin method of Momose (14) was used for the determin- ation of blood sugar in the beginning of the work, but in the latter part I was obliged to use some other method because I could not obtain the apparatus for Momose’s method. I have chosen, therefore, Benedict’s method (15) in place of Momose’s. The principle of Momose’s method is to let a measured amount of a given sugar solution react with a boiling copper sulphate solution of a known amount larger than the amount of sugar to be reduced, in a atmosphere of ammonia gas in order to avoid the intervention of atmos- pheric oxygen and then to titrate the amount of copper left unreduced, by means of a standard solution of hydroéxylamine sulphate. The amount of sugar can then be calculated from the amount of the hydro- oxylamine solution used. This method gives a very fair result, though it is a little complicated and needs some training in practice. The reason 182 K. YAMAKAMI for employing hydroéxylamin, is that the reducing power of sugar in an alkaline solution is not strong and it decreases with the decreasing concentration of the metal salt to be reduced. If, therefore, the sugar solution were titrated up to the terminal reaction as in the Pavy’s method, we can not get a correct value, but a somewhat larger value is obtained. Hence, in this method, the terminal reaction is determined with hydrodéxylamine solution which has a strong reducing power. The blood protein is precipitated by colloidaliron. Momose obtained 0.1383 for the physiological percentage of blood sugar of rabbits. But I got a little lower value, probably owing to the precautions I took in obtain- ing the blood sample. I chose old rabbits and drew the blood from the auricular vein avoiding agitation of the animals and did not draw more than 6 ec. at one time from one rabbit. The blood was taken uni- formly 3 hours after the morning feeding. Some of the results are given below. RABBIT NUMBER DATE OF THE DETERMINATION PERCENTAGE OF SUGAR N.1 February 9 0.0960 N. 2 February 9 0.1036 INES February 9 0.1292 N. 4 February 11 0.1264 N.5 February 11 0.1132 INfoul February 13 0.1279 N. 4 February 14 0.1008 N.3 February 14 0.1270 N. 2 February 14 0.1124 Experiment 1. Healthy rabbits were suffocated by pressing the trachea with fingers, taking precautions not to press the carotids and vagus, and as soon as the animals were unconscious the pressing fingers were released and artificial respiration was applied. When the animals had become conscious, once more asphyxiation was brought about in the same way as before, and during this second asphyxia the blood was drawn by heart puncture with a sterile hypodermic needle, into about 0.01 gram of hirudin. (I must offer my heartiest thanks to Professor Tatum of Chicago University for his kindness in giving me the hirudin.) Asphyxial blood coagulates quite slowly, but it is liable to coagulate within the needle when we wish to reinject the blood into other animals unless an anticoagulant is used. The asphyxial blood thus taken was introduced immediately into the auricular vein of normal rabbits in which the percentage of blood sugar ASPHYXIAL HYPERGLYCEMIA 183 had been previously determined. A small portion of the asphyxial blood was left for the determination of sugar. The blood was drawn 20 to 30 minutes after the injection of the as- phyxial blood from the auricular vein of the injected animals and its percentage of sugar was estimated. In case the rabbits did not revive from the first suffocation, the asphyxial blood drawn from the dead animals was used. Many rabbits died when the blood was taken dur- ing the second asphyxiation. The results thus obtained are given below. WEIGHT AMOUNT |SUGAR or! SUGAR Mun | greene | B2FOmE | oF ctr | waeeren | TATE OF THE surrocareD nazpins | ATTER RABBITS BLOOD BLOOD TION grams per cent ce. per cent per cent 1 2905 0.121 | 11.5 0.266 | Alive 0.135 2 3202 0.090 | 14.0 0.146 | Alive 0.134 3 2650 0.147 | 12.0 0.279 | Alive 0.148 4 2895 0.128 | 17.0 0.320 | Died in second asphyxiation | 0.149 5 3430 OLS e020 0.208 | Alive 0.152 6 2250 0.105 | 15.0 0.217 | Alive 0.178 u 2280 O23 e520 0.265 | Died in second asphyxiation | 0.197 8 2545 O) IB} 1] al) 0.305 | Died in first asphyxiation 0.141 9 3160 0.105 | 18.0 0.196 | Alive 0.158 10 2750 ORs 720 0.190 | Died in first asphyxiation 0.204 11 2455 0.120} 19.0 0.283 | Died in second asphyxiation | 0.172 All animals used in the experiment were tested previously for the existence of isolysin in their blood because I feared that the hemolysis caused by isolysin might induce a kind of partial internal asphyxiation by destroying the red blood corpuscles of the injected animals, which in turn might provoke hyperglycemia, just in the same way as hyper- glycemia is believed to arise in CO-poisoning. In order to test for isolysin, a small amount of the blood sample was drawn from the rab- bits to be suffocated, the serum was separated and was mixed with the blood corpuscles taken from rabbits to be transfused, kept at 37°C. for one hour, and the result observed. As will be seen in the table, the blood sugar percentage of rabbits: injected with the asphyxial blood, showed an increase almost invari- ably twenty to thirty minutes after the injection. In no. 10. the in- crease was very remarkable. In this case the injected asphyxial blood had only 0.190 per cent sugar while the sugar percentage of the injected animal rose from 0.122 to 0.204. This might be considered as an error. 184 K. YAMAKAMI On the other hand it may be perhaps due to the greater sensibility of the injected animal than the suffocated for the substance causing dis- turbance of the sugar metabolism. Before concluding that this ability of provoking hyperglycemia is a special property of asphyxial blood, it is necessary to ascertain the influence of normal blood upon the sugar metabolism when introduced into the circulation. There are not many reports concerning the relationship between the sugar metabolism and the injection of protein. Henderson and Under- hill (16) reported that hyperglycemia is caused by peptone injection owing to the acapnia induced by peptone poisoning. Hugh Mac- Guigan (17) found, on the contrary, hypoglycemia in peptone poison- ing and generally in anaphylaxis. He writes in his report that “in making blood transfusion from one animal to another, they have noticed there is a general tendency for the blood sugar of the recipient to fall.”’ But unfortunately he does not furnish the experimental basis for this conclusion and I was therefore uncertain as to the dose of blood transfused and the manner of transfusion and whether his transfusion was iso- or heterotransfusion. It was necessary, accordingly, to under- take the next experiment as control for the experiment already described. Ls Experiment 2. Blood was drawn from the auricular vein of healthy rabbits very cautiously without causing their agitation. Five to six cubic centimeters of normal blood were obtained from one animal. About 0.01 gram of hirudin was added to each 10 ce. of the blood which was then injected into the auricular vein of other rabbits, in which the blood sugar percentage had been previously determined. The injected animals received, therefore, the normal blood drawn from 2 or 3 other rabbits. The so-called aderlass hyperglycemia can not occur by draw- ing 5 or 6 ce. blood from one animal (18, 19). Control experiment with normal blood suGAR 20 To 30 piaousen! |) Gnemaator ine, |) Sergi See ON grams per cent | cc. per cent 1 2795 0.112 15 0.112 2 3110 0.125 15 0.137 3 2560 0.122 15 0.137 4 2785 0.098 15 0.124 5 3310 0.102 15 0.090 6 2845 0.124 18 0.115 f 2405 0.135 18 0.135 8 2600 0.122 18 0.104 ASPHYXIAL HYPERGLYCEMIA 185 As will be seen from the table, I could not observe any distinct change of the blood sugar percentage caused by transfusion of the normal blood within the limits of dosage which I used. Such small variations as occurred in no. 4 or no. 9 may be caused by a slight technical failure of determination or by an agitation of the animals which can not be prevented. This experiment proves also that hirudin had no effect on blood sugar, at least within the dosage I employed. Thus I believe it certain that the ability of provoking hyperglycemia is a special property of asphyxial blood. It is a question as to what agent of asphyxial blood this property is due. I carried out a few experiments endeavoring to find out the prob- able agent. But I could not reach any definite conclusion. In the next experiment I have shown that the excess of sugar contained in the asphyxial blood is not responsible for this ability. Experiment 3. The injected asphyxial blood had somewhat large amount of sugar as shown in the table, some had 0.305 per cent and some had 0.320 per cent. In 18 cc. of such a blood there will be about 0.035 gram more glucose than the normal amount of sugar in that volume of blood, assuming the physiological percentage of sugar to be 0.12 per cent. If a rabbit is supposed to have 100 cc. blood (weight: blood = 20: 1), and 0.035 gram sugar was introduced into circulation, this will make 0.035 per cent and the total percentage of the blood sugar of animals injected with the asphyxial blood, therefore, ought to have been 0.155 per cent at most, if the excess of sugar in the asphyxial blood were the sole cause of the rise of sugar percentage in the injected animals. But such calculation as this was not at all applicable as a matter of fact. The increase of sugar found never corresponded to the amount contained in the asphyxial blood. This fact alone is obvious evidence that the hyperglycemia is caused by some other agent than the excess of sugar. But further evidence is advanced in experiment 3 in order to fully eliminate the possibility that the excess of sugar may have disturbed the sugar metabolism of the injected animals. Normal blood was drawn in the same way as described in the experi- ment 2, and 0.02 gram glucose was added to each 10 ce. of the blood, which was then injected slowly into the auricular vein of healthy rab- bits. If the excess of sugar in the injected blood were the main cause of the hyperglycemia obtained in experiment 1, this normal blood with added glucose should give a similar result. The sugar determination was done with blood taken 20 to 30 minutes after the injection. 186 K. YAMAKAMI The results are as follows: SUGAR BEFORE AMOUNT OF INJECTED WEIGHT OF RABBITS SAIS EREGES OOD SUGAR AFTER INJECTION grams per cent Cor per cent 3120 0.106 15 0.124 2875 0.130 15 0.135 2640 0.111 15 0.111 2800 0.128 15 0.142 2550 0.122 18 0.112 2925 0.137 18 0.133 As is shown in the table the result was not the same as that obtained with the asphyxial blood. The sugar percentage after injection was almost the same as before. This result agrees well with that of the experiment done by Thanhauser (20), who proved that the normal blood sugar percentage is restored within 15 minutes after the injection of 550 cc. of 7 per cent grape sugar solution into the vena mediana cubiti of man. Kleiner carried out similar experiments with animals and proved that a great part of the injected dextrose is transferred to the tissues and changed in polysaccharides very quickly (21). Experiment 4. It is a well-known fact that in various kinds of experi- mental as well as clinical diabetes, the acidity of blood increases. In the asphyxial blood the increase of acidity is partly due to the excess of CO2, but it has been proved by Araki (22) that lactic acid, oxalic acid, ete., are also responsible for it. This fact has been confirmed by sub- sequent investigators. On the other hand, it is a fact beyond doubt that weak acid has an accelerating influence upon the activity of diastase or glycogenolytic ferment (23). Arguing from these facts and supported by other ex- perimental data, many authors believe that the cause of asphyxial glycosuria or hyperglycemia is the excess of CO. in the blood (24) or the increase of acidity (25). The following experiment was, therefore, performed to investigate whether the hyperglycemia, provoked by the injection of the asphyxial blood, is not attributable to the acidity. According to my experiment, the H-ion concentration of the normal arterial blood, taken from the left ventricle of the rabbit’s heart, and that of the venous blood from the auricular vein, when determined by the gas chain method of Michaelis (26), is as follows: ASPHYXIAL HYPERGLYCEMIA 187 Arterial blood drawn by heart puncture MILLIVOLT OBTAINED MILLIVOLT OBTAINED PH eerie rr es SHADED SOLUTION WEES BEOOD: degrees C. 18 514.8 677.0 7.39 20 515.0 682.2 7.46 20 516.0 680 .0 7.42 20 516.0 678.0 7.40 19 515.5 686.0 7.54 Venous blood taken from auricular vein 20 516.6 668 .0 7.21 20 516.1 681.0 7.43 19 516.0 673.3 W232 20 516.0 675.5 7.34 20 516.5 672.5 7.30 The H-ion concentration of the asphyxial blood obtained in the same way as described in the experiment 1 and determined by the gas chain method, is as follows: MILLIVOLT OBTAINED ROOM TEMPERATURE Beetle Eee Beers chase! PH degrees C. 19 516.0 634.0 6.66 20 516.0 634.5 6.64 20 516.6 629.0 6.54 20 516.5 641.0 6.75 20 516.2 652.5 6.94 As mentioned above, the PH of the asphyxial blood was 6.54 to 6.94 while that of the physiological blood was 7.54 to 7.39 (arterial), 7.21 to 7.43 (venous). This acidity may have been the cause of the hyper- glycemia in the recipients in our experiment 1. I therefore neutralized the asphyxial blood with alkali carbonate solution, and repeated the experiment with this neutralized asphyxial blood, to investigate whether the hyperglycemia will fail to occur after neutralization of the blood. In order to neutralize the blood, the H-ion concentration was first determined by the indicator method and then NagCO; solution (10 per cent) was added to the saline solution used for the indicator method until the PH became 7.5 to 7.6, and then the amount necessary to be added to the asphyxial blood was calculated from the amount required. 188 K. YAMAKAMI The blood thus neutralized was injected into healthy rabbits. (1 wish to express here my gratitude for the kindness of Doctor Tashiro of Chicago University for lending me the apparatus for the determina- tion of H-ion concentration.) The results thus obtained are given below: WEIGHT OF RABBITS pier ti Nee SOON ee SUGAR AFTER INJECTION grams per cent ce: per cent 2815 0.135 15 0.172 2310 0.128 18 0.144 2570 0.096 15 0.180 2645 0.137 15 0.188 3105 0.119 7 lS 0.156 2760 0.131 15 0.145 2650 0.122 15 0.127 As cited in the table, I observed that the tendency to promote hyper- glycemia remains in the asphyxial blood even when the blood was neutralized. Evidently, therefore, the acidity of the asphyxial blood samples was not responsible for the effect induced. DISCUSSION It is shown in this work that asphyxial blood causes a rise in the sugar content of blood when introduced into the circulation of other animals. It can not be stated what products of asphyxia act as the primary cause of asphyxial hyperglycemia. The lack of oxygen may form a primary cause by leading to deficient oxidation of carbohydrates, as Claude Bernald, Dastre, Lepine, Terrey, Araki and others believe, or the excess of CO, and the increased acidity may be the primary cause, as Edie, Moore, Roaf, Macleod and others have suggested, or the blood sugar may increase owing to a decrease of the activity of tissue oxidase as Underhill thought, or the asphyxial glycosuria and hyperglycemia may be due to the emotional disturbance or the fear of death, as Bang and Stenstrém opine, because it is not possible to suffo- cate animals without causing fear of death. Or it is very possible that all of these factors may combine to give rise to asphyxial hyperglycemia. But however this may be, the asphyxial blood seems to possess in itself hyperglycemic ability. This ability is neither due to the excess of sugar contained in it, nor to its acidity. ASPHYXIAL HYPERGLYCEMIA 189 In order to explain this experimental fact, the hypothesis that adre- nalin is responsible is most acceptable because we do not know at present any other substance than adrenalin in the blood which can give rise to the enhanced sugar content, though, of course, we can not venture to claim that hyperadrenalinemia was proved by our experiment to exist in asphyxia, for some hitherto unknown agent may be discovered in the future to have been responsible for the effect obtained. SUMMARY 1. The transfusion of the normal blood from rabbits to rabbits has no remarkable effect upon the blood sugar percentage. 2. The transfusion of the asphyxial blood causes a rise in the sugar content of the blood of recipients. 3. The excess of sugar content in the asphyxial blood is not respon- sible for this increase of sugar percentage in the blood of the recipients, 4. The neutralization of the blood with NazCO: solution does not abolish this property of asphyxial blood. BIBLIOGRAPHY (1) Potuax: Arch. f. exper. Path. u. Pharm., 1909, lxi, 376. (2) STARKENSTEIN: Zeitschr. f. exper. Path. u. Therap., 1911, x, 78. (3) CzuBauskt: Zentralbl. f. Physiol., 1913, xxvii, 580. (4) Mryer: Compt. rend. Soc. Biol., 1908, lx, 1124. (5) GAUTRELET AND THoMAS: Compt. rend. Soc. Biol., 1909, Ixvii, 233. (6) Macterop: Proc. Soc. Exper. Biol. and Med., 1911, viii, 110. (7) Kaun: Pfliiger’s Arch., 1912, exlvi, 579. (8) MactEop AND PEearce: This Journal, 1912, xxiv, 419. (9) Kaun: Pfliiger’s Arch., 1917, exevi. (10) WERTHEIMER AND Barrez: Arch. intern. d. Physiol., 1910, ix, 363. (11) Cannon: This Journal, 1914, xxxili, 356. CANNON AND Hoskins: Ibid., 1911, xxviii, 274. CANNON AND DE LA Paz: Ibid., 1911, xxviii, 64. CANNON, SHOHL AND WriGut: Ibid., 1911, xxix, 280. (12) Stewart AND Rocorr: This Journal, 1917, xliv, 543. STEWART AND Rocorr: Journ. Exper. Med., 1912, xv, 547. Stewart: Journ. Exper. Med., 1911, xiv. 377. (13) Leprnr AnD Boutup: Compt. rend. d’Acad. d. Sci., 1902, exxxiv, 582, 1341. (14) Momose: Tokio Igakukaizassni, B. xxix, 112. (15) Lewis anp Benepict: Journ. Biol. Chem., 1915, xi, 26. (16) HENDERSON AND UNDERHILL: This Journal, 1911, xxviii, 281. (17) McGuiean: Journ. Laby. and Clin. Med., 1918, iii, 335. (18) SaEenxK: Pfliiger’s Arch., 1894, lvii, 553. (19) Rose: Arch. f. exper. Path. u. Pharm., 1903, 1 15. 190 K. YAMAKAMI (20 THANHAUSER: Miinch. Med. Wochenschr., xxxv, 2155. (21) Kuerner: Journ. Exper. Med., 1916, xxii, 507. (22) Araxt: Zeitschr. f. phys. Chemie, 1891, xv, 351. (23) ScatERBECK: Skand. Arch. f. Physiol. 1892, iii, 344. Deter: Zeitschr. f. Physiol. Chemie, 1883, vii, 1. CHITTENDEN AND GrIswoLD: Amer. Chem. Jour., 1881, iii, 305. (24) Epi, Moors anp Roar: Biochem. Journ., 1911, v, 325. Epi: Biochem. Journ., 1906, i, 455. (25) Macteop: This Journal, 1909, xxiii, 278. (26) Micuaruis: Die Wasserstoffionen Conzentration, Berlin, 1914. UREA EXCRETION AFTER SUPRARENALECTOMY GEORGE BEVIER anp A. E. SHEVKY From the Medical Division of the Stanford University Medical School, San Francisco Received for publication July 23, 1919 The rate of urea excretion has been shown to be primarily a function of the concentration of urea in the blood (1). But if we divide the urea excretion per hour by the amount of urea in 100 cc. of blood (Addis’ ratio, (2) ) and thereby get an expression of the rate of excre- tion per unit concentration, we find that, in rabbits, this rate of excre- tion per unit concentration presents a rather wide range of variability —in other words, that for any given concentration of urea in the blood the rate of excretion may be either rapid or slow. This variability must be accounted for by assuming the operation of factors other than the concentration of urea in the blood (3) and the quantity of function- ing renal tissue (4). It has further been shown that the subcutaneous injection of epinephrin has the same effect as those factors which increase the rate of urea excretion for a given blood concentration (5), and that the subcutaneous injection of pituitrin affects this ratio by depressing it (6), that is, the rate of urea excretion at any given blood urea concentration is accelerated after the injection of epinephrin and depressed after the injection of pituitrin. By mixing epinephrin and pituitrin in doses of varying proportions it is possible to get a modified effect of either one, or the doses may be so balanced that the ratio is not affected in either direction (7). These facts have suggested the hypothesis that an epinephrin-pitui- trin balance may exist in the blood which can alter the rate of renal activity in the handling of urea. One way in which the existence of such an epinephrin-pituitrin balance can be investigated is by a double suprarenalectomy which should leave the pituitary effect unopposed. Similarly, removal of the hypophysis cerebri should give an unopposed epinephrin effect. The results of a few not altogether satisfactory experiments on the effect of suprarenalectomy were referred to in a previous communication (7). At that time it was not possible to 191 192 GEORGE BEVIER AND A. E. SHEVKY obtain more data, but we have recently been able to return to the problem and in this paper present the results of more numerous and better planned experiments. METHODS Male rabbits were used and the procedure was the same as described previously (1) with the exception that no stomach tube was introduced. Briefly, it consisted of catheterizing the rabbits, which had been kept in the laboratory without food or water since the previous afternoon, at a definite time (9 a.m.) and then collecting the urine by catheteriza- tion at the end of the first, second, third and fifth hours. At the mid- dle of each interval 1 cc. samples of blood were obtained from the ear veins. The urea determinations were made with Marshall’s urease method, using for the urine the titration method with modifications as detailed by Addis and Watanabe (8) and for the blood the aeration method with the refinements introduced by Barnett (9). Using this technique the excretion of urea per hour for the various periods, and the corresponding blood urea concentrations were deter- mined, first, for a group of normal animals. By dividing the number of milligrams of urea excreted per hour by the number of milligrams of urea in 100 ce. of blood during the same hour, we get the ‘‘ratio”’ (2) or the rate of excretion per unit blood concentration. These values have also been tabulated in our data. Bilateral lumbar incisions were then made on some of the animals, exposing and manipulating the suprarenals, but not actually removing them, to ascertain any effect of the operation itself, or of the anesthetic, on our ratio curve. Subsequently the suprarenal glands of these rabbits were removed through lumbar incisions under ether anesthesia. At first we at- tempted to remove both glands on the same day and immediately follow the operation by a determination of the rate of urea excretion in the same manner as we had done on the normal animals. These ani- mals often died in an extremely exhausted condition within a few hours and often during the course of the experiment. Often kidney function was almost completely depressed in these dying animals. Evidently such a condition of shock would be accompanied by renal disturbances not representative of the true effect of suprarenal removal. After considerable experimenting we concluded that the removal of one gland at a time with an interval of several days between operations and a rest of a day after the final operation before we commenced the UREA EXCRETION AFTER SUPRARENALECTOMY 193 procedure of collecting the urine and blood, gave the most dependable results, coinciding with the findings of others (10), (11). A large per- centage of our animals survived, especially after we became more expert in the technique of the operation. The rabbits were under an- esthesia from 20 to 25 minutes. On the day following the final opera- tion they appeared to be about as lively and vigorous as the normal animals. The rate of excretion was determined by our procedure then, and on several subsequent occasions. THE SUPRARENAL GLANDS IN THE RABBIT Since tissues similar to those found within the suprarenal capsules are found elsewhere in the bodies of some animals, the question arises as to just what we remove when we excise these capsules from the rabbit. The comparative anatomy has been discussed quite fully, and also some of the general results of suprarenalectomy, by Biedl (12) and earlier by Tizzoni (13). Both the cortical and the medullary tissues of the suprarenal may occur separately outside of the capsules in the rabbit as-in nearly all of the animals having definite, separate, isolated glands. Medullary or chromaffine cells are found in the ganglia of the abdominal sympathetic and in the carotid ganglion. Stilling (14) observed that extirpation of one gland in the rabbit was followed by great hypertrophy of the other gland and of any remnants of the glands which happened to be left at the time of operation. Also that acces- sory suprarenals are frequently found after ablation of one gland, probably due to a marked proliferation of the isolated patches of this tissue which occur along the vena cava and in other parts. Fulk and Mac'eod (15) have shown that retroperitoneal chromaphil tissue is the same as that of the suprarenal capsules, and that extracts of it have the same reactions as the medullary tissue. DATA In figure 1 we have plotted curves (from table 1) showing the aver- age excretion of urea in milligrams, the average concentration of urea per 100 cc. of blood, and the average ratio for each of the periods of our experiment, for a series of twenty normal animals before operation. As has been previously mentioned, the ratio = Urea excreted per hour in mgm. Mem. of urea in 100 ce. of blood. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 2 194 GEORGE BEVIER AND A. E. SHEVKY TABLE 1 Normal animals UREA EXCRETED PER A I} Cc. D HOUR IN GRAMS FOR THE UREA IN 100 cc. OF BLOO RATIO FOUR PERIODS IN GRAMS RABBIT a aes Ca a Ist | 2d | 3d | and | ist | 2d | 3d | and | Ist 2d 3d herd hour | hour | hour | 5th | hour | hour | hour| 5th | hour | hour | hour | 1 hours hours ours 93 |0.020/0.039|0.058 0.075|0.045|0.043)0.051/0.058) 0.45) 0.91) 1 121 = |0.032)0.038)0 .017)|0 .042\0 .033/0.032)0.030/0.035) 0.91) 1.19) 0 121 =|0.026/0.032/0 .027/0 .029|0 .033)0..030/0.030/0.027; 0.80) 1.07) O 122 = |0.013)0.013/0 .021/0 .028/0 .020/0 .023/0.021/0.022| 0.67) 0.58) 1. 0 2 0 NSN 127 = |0.084)0.112/0.116|0.123)0. 105)0.143}0.145|0.148) 0.79) 0.80 134 |0.008/0.059/0 .066)0 .069)0 .030)0 .033)0 .030)0 .036 137. ~—-|0..023/0.013/0.010/0 .037|0 .032)0 .032/0 .035,0.036 138 = |0.026|0.024/0 .038)0 .035/0 .042)0 .042/0 .042/0.045 139 —|0.024/0 .023)0 .028)0 .030)0 .034/0 .035/0 .040)0 .039 142. = (0.005)0.007|0 .022/0 .024/0 .031|0.027/0 .057\0.041 51 = |0.009/0 .029}0.035/0 0510 .057,0 .048)0 .042/0 .051 51 = |0.046/0 .076)0.085/0.. 103\0 .039/0 .048/0 .048,0 .040 52 |0.023/0.023)0.013)0 .046'0 .049/0 .035)0 .052)/0.046 52 = |0.018/0.026)0.054/0 .086)0 .033,0 .037)/0.036/0 .039 © 0 w® & Ze\ eel wus .69} 0.41 2g 15} 0.60) 0.84 17) 2:59) ee : 0.24 53| 0.72) 1.51 for) ~) j=) for) ~I i=) 3 eg ae ee kone rene SSssssq S79 SO GrGs="O (Ore Ororore) Go iS lor) oO lor) Or 53 |0.031/0.032|0.044/0 .060)0 .027/0.037|0.043/0 .045 14, 0.88) 1.03) 1.34 56 = |0.029/0 .027|0.021/0 .039)0 .033,0 .038/0 .040/0 .030 89} 0.71) 0.53 30 58 = (0.024/0.028/0 .031/0 .039|0 .038/0 .033/0 .028)0 .035 63} 0.86; 1.12) 1.12 60 |0.004/0.004/0.015)0 .017|0.037/0 .034/0 .036)0 .036 12} 0.13) 0.42) 0.47 61 |0.007/0.004/0.016/0.015,0.022/0.024/0 .027|0.038 34| 0.19} 0.58} 0.41 56} 0.38) 0.64) 0.50 62 |0.013/0.009)0.013)0.015)0 .024/0 .024/0.021'0.030 0.465)0.618)0.730\0 .963/0.764/0.798)/0 .855|0.877| 12.08} 14.93] 17.41} 22.75 Average |0.023)0.031/0.037/0.048/0 .038/0.04010.043)0.044) 0.61) 0.75| 0.87) 1.14 Probableverror! sot tee. ee ee. eS +0.19)0.27|+0.33/--0.38 It will be observed that the ratio becomes higher—that the rate of excretion of urea per unit concentration of urea in the blood increases —as the experiment progresses, being highest in the last hour. It will be convenient for us, in this discussion, to refer to this tendency for the curve to become higher in the successive periods, as the “epi- nephrin effect’’ since it was found to be most marked after the injection of epinephrin (5). Opposite the curves for normals are drawn similar curves represent- ing the averages of seventeen experiments after both glands had been removed. The ratio curve here does not rise, that is, we do not find UREA EXCRETION AFTER SUPRARENALECTOMY 195 the “epinephrin effect” that occurs in normal animals. The data for this curve are tabulated in table 2. Figures 3 and 4 are taken from papers already referred to (3), (5) and are introduced for comparison. They show the effects, respectively, of x -20 ~20 ie gy SR Ratio 8 xt f Bf J00— .075- o—o—o—_O lood Urea N .O50— = — —o ~o50% oe oe s o25— ih Be Reza ee SB 4 oe so Normals After epinephrin Mise aioe after Peat aniiony Big. 1 Fig. 2 Fig. 3 Fig. 4 N N ~~ < 8 $5 na ‘ g S 1090 — —1o g RN —10 % O75 - —o75 O50 — —as50 After control operation. After control operation Same or next day. 6to7 days Fig. 5 Fig. 6 injecting, subcutaneously, 0.25 ec. Parke, Davis & Company adrenalin at the beginning of each hour, and 0.25 ec. of the same company’s pituitrin. The ratio curve rises most markedly after epinephrin and is depressed after pituitrin. 196 GEORGE BEVIER AND A. E. SHEVKY Figure 5 gives the curves for the average of six experiments (table 3) performed on the same day as the control operation when the glands were exposed and manipulated but not removed, and figure 6 shows the results of determinations made on some of these same animals on the sixth and seventh days after the control operation. Evidently TABLE 2 A compilation of all observations made after both suprarenal glands had been re- moved. They were made on the day following the removal of the final gland or later. sf , IN ! : Al TRE: 1 avGlureD ue HOUR oNee. on BSC. m a TLO RABBIT 4th 4th ef [ad | ad end |. Wet, | 2d. | aac eee) Sat | dd) | dial eens hour | hour | hour | 5th |} hour | hour | hour} 5th hour hour hour ih ours hours One 134 0.034/0 .029/0 .035/0 .039/0 .043/0 .042/0.045|0.047| 0.78) 0.68) 0.77) 0.82 134 0.026|0 .024/0 .032|Lost |0.024/0.024/0.026)0.027) 1.10} 1.00} 1.25 138 0.033/0.040/0 .035/0 .047|0.036/0.048/0.045|0.039| 0.92) 0.83) 0.80) 1.20 51 0 .037|0 .040)0 .036|0 .047|0 .036)0.036)0.036)0.040) 1.03) 1.10) 1.00) 1.15 52 ().010/0.019)0 .009|0 .008/0 .043\0 .042/0.045)0.042| 0.24) 0.45) 0.20) 0.18 58 0.0100 .021/0.013)0.019|0.030/0.037|0.040\0.038) 0.25) 0.57) 0.32) 0.50 60 0.012/0.007|0.001/0 .011/0.057|0.056)0.051/0.048]} 0.21) 0.12) 0.02) 0.23 61 0.056/0 .052/0 .056/0 .049|0 .045/0 .045|0.036|0.042} 1.27) 1.15) 1.53] 1.16 62 Lost|0.031)/0.032)/0 .035)/0 .045/0 .042|0.039)0 .042 0.75} 0.81) 0.83 51 0.049]0 .041/0 .060)0 .063)/0 .047/0 .043)0.048|0.051)| 1.04) 0.96) 1.25) 1.23 52 0.063/0 .055/0 .075|0 .068)/0 .076/0.081|0.085/0.092| 0.82} 0.68} 0.88} 0.74 58 0.010}0.008/0 .001}0 .001/0.019/0 .031/0.025/0.029) 0.55) 0.26) 0.04) 0.04 60 0 .024/0.017/0 .019|0 .014/0 .037/0.042/0.038/0.039| 0.65) 0.41} 0.51) 0.37 61 0.010/0.011)0 .009\0 .008/0 .032/0 .030)0.029\0.027| 0.33) 0.35) 0.31) 0.29 62 0.016/0 .020/0 .007|0 .001/0 .072|0 .078|0.082/0.096| 0.23) 0.23) 0.08} 0.02 134 0.029|0 .047)|0 .040|0 .065|0 .028/0 .024/0.027|0.034| 1.05) 1.96} 1.50} 1.90 138 0 .029/0 .035/0 .035/0 .053|0 .042/0 .042/0.045|0.042| 0.68) 0.84) 0.77) 1.27 0.448/0.497/0.495/0.528/0.712/0.743)/0.742/0.775| 11.15) 12.34) 12.04] 11.93 Average|0.028/0.029|0.029|0 .033/0 .042|/0.044/0.044/0.045| 0.70) 0.73) 0.71) 0.74 Normal.|0.023/0.031\0 .037|0 .048/0.038/0.040\0.043/0.044| 0.61) 0.75) 0.87) 1.14 WitterencesbetweeleraliOss eee ero Tees. | eee ere +0.09)—0.02|—0.16)—0.40 any differences in the ratio curves from the normal, due to the anes- thetic or to the operation per se, lie entirely within the range of vari- ability of the series and are not measureable. We are, then, justified in interpreting the differences in the curves before and after supra- renalectomy as due to the loss of the glands. UREA EXCRETION AFTER SUPRARENALECTOMY 197 TABLE 3 These data were obtained on the same day as the control operations, when lumbar incisions were made and the glands exposed and manipulated but not removed. The purpose is to ascertain the effect of the anesthetic and the operation per se, as differentiated from the removal of the suprarenals. Sele er namo SEE 4th 4th 4th 1st 2d 3d and 1st 2d 3d and Ist 2d 3d and hour | hour | hour |} 5th | hour | hour} hour | 5th | hour | hour } hour | 5th hours hours hours 93 0 .007)/0 .007|0 .024/0 .054/0 .045|0.045/0.075|0.060) 0.15) 0.15) 0.32) 0.89 116 0.049)/0 .043/0 .068/0 .067/0 .043)0.041/0 .042/0.039} 1.13) 1.06) 1.62) 1.72 117 0.046/0 .051|0 .062/0.049/0 .030\0 .028)0 .030/0.043) 1.51) 1.81) 2.08) 1.14 121 0.010/0.021|0 .035)/0 .050/0 .036\0 .038)0.045/0.044) 0.29) 0.56) 0.77) 1.13 122 0 .037/0.038/0.043/0.099/0.100|0.100/0.108/0.110) 0.37} 0.38) 0.40) 0.90 134 0.020)0 .023/0 .048)0 .072/0.012)0.009)0.021/0.034) 1.65} 2.53) 2.24] 2.10 Average..... 0.028/0 .031|0 .047/0 .065)0 .044/0 .043/0.053)/0.055) 0.85} 1.08} 1.24] 1.32 TABLE 4 Observations made on the sixth and seventh days after the control operation. In both this table and table 3 it will be noted hat the values of the ratio for the succes- sive periods continually increase after the manner of normal animals GRAMS OF UREA EXCRETED GRAMS OF UREA IN 100 ATTO PER HOUR CC. OF BLOOD ‘ NEE 4th 4th 4th Ist 2d 3d and ist 2d 3d and Ist 2d 3d and hour | hour |. hour | 5th | hour | hour | hour} 5th | hour } hour } hour | 5th hours hours hours 93 0.012/0.014)0 .044/0 .065)0 .033/0 .033/0.036)/0.033) 0.36) 0 1 116 0.016)0.021)0.019)0.025|0.020)0.015)0.017/0.019) 0.81) 1.42) 1.13) 1.20 117 0.044)0 .044)0 .052\0 .055)0.024/0 .026/0 .027|0.026] 1.84] 1 1 Average.....|0.024/0.027|0.038/0.047|0 .026/0 .025|0 .027/0.026| 1.00) 1.19} 1.43) 1.73 Nine observations on rabbits after the removal of the right gland and before the removal of the left one, gave averages shown in figure 7 and table 5. The ratio curve presents a distinct upward tendency but is slightly less pronounced than in the normal animals. These findings are introduced merely for the sake of completeness; they seem to indi- cate that the remaining gland is ample to supply the needs of the body. Figures 8 and 9 are constructed, respectively, from observations made from 24 to 48 hours after the removal of the final gland, and 198 GEORGE BEVIER AND A. E. SHEVKY TABLE 5 These observations were made on the same day or the next day after the removal of the right suprarenal gland and before the excision of the left gland, in order to ascertain the renal behavior of an animal with only one suprarenal. If there is any effect it is too small to be definite UREA EXCRETED PER UREA IN 100 cc. OF BLOOD, HOUR IN GRAMS IN GRAMS ALO TA Sp 4th 4th 4th Ist 2d 3d and ist 2d 3d and ist 2d 3d and hour | hour} hour! 5th | hour | hour | hour | 5th | hour | hour |} hour | 5th hours hours hours 127 0.002/0.006|0.014/0 .014)0 .046)0 .046\0 .054,0.032) 0.05) 0.14) 0.25) 0.32 137 0.002|/0.004| Lost|0.014/0.018|0.008)0.023,0.039) 0.08) 0.56 0.36 138 0.009)0.009'0 .024\0 .017)\0.060\0 .054\0.064,0.075, 0.15) 0.17) 0.37) 0.24 139 0.003)0.020/0 .014/0 .024/0 .039/0 .036)0.038/0.049) 0.08) 0.54) 0.37) 0.50 141 0.002/0.003)0 .007/0 .018)0 .030)/0 .033\0 .033/0.037| 0.07) 0.09) 0.23) 0.46 143 0.027|0 .006)0 .030/0 .035)0 .039/0 .042\0.047/0 066) 0.70) 0.14) 0.64) 0.53 51 0.005|0.043)0 0300 .051\0.015,0.025/0 .042/0.048) 0.30) 1.78) 0.95) 1.07 52 0.019|0.019)0 .030 0 .050\0 .033/0.046/0.049/0.056) 0.58) 0.42) 0.62) 0.89 62 0.035/0 .044/0.041|0 .051|0.039/0 .040/0 .042|0 .036) 0.88) 1.07) 0.99) 1.40 Average... .0.012)0.017/0 .025)0 .030/0 .035/0.037/0 04410 .049 0.32) 0.54) 0.55) 0.64 Ratio. C= Coma See —o50 250 — —0.5 i | rag gel <— 24 +o 48 hours after final 5 to 8 days after Unilateral — suprarenalectomy. operation final operation. Fig. 7 Fig. 8 Fig. 9 from 5 to 8 days after the last operation. There is no difference be- tween these curves, indicating that if there is a readjustment of the “balance” after the loss of the suprarenals, it does not occur very soon. Figure 10 is a graphical summary comparing the present results with some previously obtained. The base line represents the average ratio during the first hour for a group of normal animals. It is the UREA EXCRETION AFTER SUPRARENALECTOMY 199 average of observations on fifty-seven rabbits and its value is 0.69. The rest of the curve is obtained by computing the percentage of this quantity by which the ratio is increased or decreased during the follow- ing hours of the experiment and under the varying conditions of it. In the curve for ‘‘normals’”’ we have the percentage by which the ratio was increased after the first hour. The curve for ‘epinephrin” was Ratio after epinephrin.. =< an prarenalectom y: nae Ath. eek hour 0° ttuitrin. -50 Fig. 10. Chart showing the percentage increase or decrease in the average “‘ratio” for the various periods of the experiments and under the various con- ditions listed, compared with the average ‘“‘ratio’”’ for the first period obtained from 57 normal animals. See table 8. obtained from data recorded by Addis, Barnett and Shevky (5) and shows the marked increase obtained after the subcutaneous injection of 0.25 cc. Parke, Davis & Company adrenalin at the beginning of each hour, in percentage of the initial ratio for the first hour for animals under normal conditions. It will be noted that there was an increase in the average ratio of about 80 per cent as the result of the first injec- 200 GEORGE BEVIER AND A. E. TABLE 6 SHEVKY In this table we have collected those observations which were made from twenty-four - to forty-eight hours after the excision of the final suprarenal capsule. It may be compared with the next table in whi h we have collected observations made from _ five to eight days after the removal of the last gland UREA EXCRETED PER HOUR IN GRAMS UREA IN 100 CC. OF BLOOD RABBIT ; 4th 4th Ist 2d 3d and Ist 2d 3d and hour | hour |} hour | 5th | hour | hour | hour | 5th hours ours 134 0.034/0.029|0 .035/0 .039)0 .043)0 .042|/0 .045)0 .047 134 0.027|0.024|0.033} Lost|0.024/0.024/0.026)0.027 138 0 .033/0.040)0 .035/0 .047)/0 .036,0.048)0 .045/0.039 51 0.037)\0 .040)0 .036/0 .047|0 .036)/0 .036)0.036,0 .040 52 0.010,0.019}0.010}0 .008'0.043 0 .042/0.045/0.042 58 0.010|0.021/0.013)0.019)0 .039)0 .037|0 .040/0.038 60 0.012/0.001)0.001)0.011)0.057|0 .056)0.051/0 .048 61 0.056)/0 .052/0 .056)0 .049'0 .044/0 .045/0 .036)0 .042 62 Lost/0.032)0.032/0.035)0 .045|0 .042/0 .039/0 .042 0.219)0.258/0. 251/0. 255)0 .367|0.372|0. 8363/0. 365 Average.... .|0.027/0.029/0.028/0.0 °2)0.041/0.041/0 .040)0 .041 TABLE 7 RATIO 4th Ist 2d 3d and our | hour | hour | 5th hours 0.78) 0.68] 0.77} 0.82 1.10} 1.00} 1.25 0.92] 0.83) 0.80) 1.20 1.03} 1.10) 1/00) 1:15 0.24] 0.45) 0.20) 0.18 0.25] 0.57) 0.32) 0.50 0.21) 0.12) 0-02) 0223 1.27) 1.15) bs | eG 0.75} 0.81] 0.83 5.80} 6.65) 6.70) 6.07 0.73) 0.75) 0.74| 0.76 Data obtained from five to eight days after the excision of the last suprarenal capsule UREA EXCRETED PER UREA IN 100 CC. OF BLOOD, HOUR IN GRAMS IN GRAMS Ae) RABBIT 4th 4th | 4th 1st 2d 3d and Ist 2d 3d and | Ist 2d 3d and hour | hour} hour} 5th | hour | hour} hour} 5th | hour |} hour |} hour |} 5th hours hours hours 51 0.049|0.041)0 .060/0 .063/0 .047\0.043/0.048/0.051} 1.04) 0.96) 1.25) 1.23 52 0 .062)\0 .055)0 .075|0 .068/0 .076/0.081/0.085/0.092} 0.82) 0.68) 0.88) 0.74 58 0.011)0.008)0.001)0.001)0.019}0.031)0.C25|0.029) 0.55} 0.26) 0.04] 0.04 60 0 .024/0.017/0 .020/0.014)0.037|0 .042/0.038/0.039) 0.65) 0.41) 0.51) 0.37 61 0.010/0.0:0)0.009/0 .008)0 .032/0 .030)0.029|0.027| 0.33] 0.35} 0.31) 0.29 62 0 .016)0.020)0 .007|0 .002/0 .072/0.078}0.082|0.096| 0.23} 0.23) 0.08) 0.02 , 34 0 .029/0 .047|0 .041/0 .064/0.028/0 .024/0.027|0.034| 1.05] 1.96) 1.50) 1.90 0.201)0.198)0.213/0.220/0.311/0.329)0.334/0.368) 4.67) 4.83} 4.57) 4.55 Average.....|0.0290.029)0 .030/0.032/0.044/0.047\0.048/0.052|) 0.67) 0.69) 0.65} 0.66 UREA EXCRETION AFTER SUPRARENALECTOMY 201 tion of adrenalin and that this became greater as the hours passed. The curve for pituitrin shows that the subcutaneous injection of pitui- trin at the beginning of each hour depressed the rate of excretion per unit blood concentration. Similarly our results after suprarenalec- tomy show that the progressive increase detected in normal animals with intact glands, or after the injection of epinephrin, does not occur. It might be noted that the urea concentration in the blood remains fairly constant under the various conditions of the experiments. It was only in the moribund animals after simultaneous double supra- TABLE 8 HOURS 1 2 3 4to5 Normals: average of 57 animals.......... 0.69 1.01 1.13 issu Percentage increase over first hour... 00 46.0% | 64.0% | 90.0% After epinephrin (28 animals)............ 1.04 1.86 BPA 2.52 Percentage increase over normals....| 80.0% | 170.0% | 220.0% | 265.0% After pituitrin, (9 animals)........- 2.5.4. 0.25 0.52 0.38 0.43 Percentage decrease............:...-. —49.0%-|—25.0% |—45.0% |—38.0% After suprarenalectomy (17 animals).....| 0.70 0.72 0.70 0.71 Percentage increase................. 1.0% 4.0% 1.0% 3.0% The normal ratio for the first hour of the experiment for a group of fifty-seven rabbits is 0.69. In the second hour it is 1.01 or an increase of 46 per cent and so on for the rest of the periods. After a subcutaneous injection of epinephrin we get an average ratio of 1.04 for the first period or an increase of 80 per cent over the first period excretion for normal animals. renalectomy that we found an increase in the level of blood urea similar to that described by Marshall and Davis in cats (16). When one gland was removed at a time the rabbits remained more normal. DISCUSSION Normal rabbits present a progressive increase in the rate of urea excretion during the consecutive periods of our experiment, so that at the end the rate was nearly twice as great as it was in the beginning, in spite of the fact that the urea concentration in the blood remained practically constant. This means, of course, that the kidneys did not have more to do, but performed what they did have to do at a more 202 GEORGE BEVIER AND A. E. SHEVKY rapid rate. The conditions of the experiment were the same at the end as at the beginning except that the animals had undergone con- siderable handling and discomfort. The progressive increase found in normal animals may be greatly accentuated by the subcutaneous injection of epinephrin, and may be prevented, or even a progressive decrease may be obtained, by the injection of pituitrin. It has been shown that injections of epinephrin and pituitrin in varying proportions may affect the rate of urea exere- tion in mutually antagonistic directions and that each may neutralize the effect of the other. And on the basis of these data the hypothesis was advanced that a possible balance between the secretions of the suprarenals and of the hypophysis, in the blood, may be a factor in determining the state of renal activity (7). The fact that the removal of the suprarenal glands affected the form of the ratio curve in a manner remarkably similar to that produced by the subcutaneous injection of pituitrin (figs. 2 and 4) strongly suggests the existence of such a balance—the effect of suprarenalectomy being an unbalanced pituitary effect. Marshall and Davis (16) have observed in cats a similar decrease in the rate of excretion of urea, creatinin and chlorides, with which we, of course, agree. Motzfeldt (17) found that the extract or secretion of the posterior lobe of the hypophysis had a strong antidiuretic effect on rabbits, which was most pronounced after 2 to 4 hours. Rees (18) noted that pituitary extracts delay diuresis after ingested water, for 7 or 8 hours, but do not alter the 24-hour volume. The depression recorded after ablation of the glands is less marked than after the injection of an optimum dose of pituitrm. This is as might be expected, for the excision of the suprarenal capsules has only removed most, not all, of the medullary tissue, as previously pointed out. Furthermore, there has been no stimulus to call forth a maximum pituitrin effect, and only the normal amount of pituitary secretion is probably present in the blood. This causes, however, a definite “pi- tuitary effect” for it is not balanced by the full normal suprarenal secretion. - | We attribute the progressive increase in the rate of excretion, in spite of constant blood urea concentration, which we found in all nor- mal rabbits during the successive periods of the experiments, to a gradual increase in the rate of secretion of epinephrin from the supra- renal glands. After the removal of the glands such an increase of epinephrin in the blood, of course, could not occur. UREA EXCRETION AFTER SUPRARENALECTOMY 203 CONCLUSIONS 1. The removal of the suprarenal glands in rabbits is followed by a depression of the rate of urea excretion by the kidneys. 2. The form of the curve obtained by plotting the ratio between the urea excreted per hour and the concentration of urea in the blood, for the various intervals of the experiment, is modified after suprarenal- ectomy in a manner strikingly like that obtained by the subcutaneous injection of optimum doses of pituitrin, and in a manner contrary to that obtained after the injection of epinephrin. 3. It is suggested that these findings support the hypothesis that an epinephrin-pituitrin balance exists in the blood which may regulate the rate of kidney function, the results obtained after suprarenalectomy exhibiting a pituitary effect unopposed by the normal secretion of the suprarenals. We wish to acknowledge our indebtedness to Dr. Thomas Addis for his valuable suggestions and for his kindly interest, which has always been most stimulating. BIBLIOGRAPHY. (1) Appis, BARNETT AND SHEvkKy: This Journal, 1918, xlvi, 1. (2) Appts: Journ. Urology, 1917, i, 268. (3) Appis, BARNETT AND SHEvKY: This Journal, 1918, xlvi, 22. (4) Appis, SHevKy AND Bevier: This Journal, 1918, xlvi, 11. (5) Appis, BARNETT AND SHEvKyY: This Journal, 1918, xlvi, 39. (6) Appis, Foster AND Barnett: This Journal, 1918, xlvi, 52. (7) App1is, SHEvVKY AND Bevier: This Journal, 1918, xlvi, 129. (8) Appis AND WaTANABE: Journ. Biol. Chem., 1916, xxvii, 250. (9) Barnet: Journ. Biol. Chem., 1916, xxix, 459. (10) Lanators: Arch. d. Physiol. normale et pathol., 1893, v, 488. (11) Exuiorr: Journ. Physiol., 1914, xlix, 38. (12) Brepu: Internal secretory organs (transl. by Forster), 1913, 138. (13) Tizzont: Ziegler’s Beitr. z. pathol. Anat., 1889, vi, 3. (14) Sriuuine: Virchow’s Arch., 1889, exviil, 569. (15) Futx anp Macteop: This Journal, 1916, xl, 21. (16) Marsuauy anp Davis: Journ. Pharm. Exper. Therap., 1916, vili, 525. (17) Morzretpt: Journ. Exper. Med., 1917, xxv, 153. (18) Rees: This Journal, 1918, xlv, 471. POSTURE-SENSE CONDUCTION PATHS IN THE SPINAL CORD A PRELIMINARY REPORT EUGENE S. MAY anp JOHN A. LARSON From the Rudolph Speckels Physiological Laboratory of the University of California Received for publication July 24, 1919 Exact knowledge is lacking as to how impulses mediating posture- sense are conducted in the spinal cord. These, with such other afferent impulses as muscle-sense, deep-sensibility and others of the sensation- complex are supposed to travel upward, without decussation, in the dorsal and lateral columns of the spinal cord. Decussation of these fibers takes place in the medulla superior to the pyramidal decussation. To obtain more definite information as to the manner of transmission of posture-sense impulses in the spinal cord, we have applied to our problem the animal behavior method described by O. Kalischer' in his experiments on audition. By means of a strong ‘‘hunger-motif” this investigator trained dogs to discriminate between various tones, per- mitting them to take food only when the correct or ‘‘training-stimulus”’ was presented. Dogs were trained daily over a period of weeks until they perfectly discriminated the ‘‘feeding-stimulus” from any other. This ‘“‘feeding-stimulus” was always the same tone sounded on the organ. After they had learned the ‘‘lesson’”’ the animals were operated and an area of the cerebral cortex was destroyed. After complete recovery from the shock of the operation the dogs were again critically tested to determine their ability to perfectly differentiate the “‘feeding- stimulus” from other stimuli. If discrimination by the animal was still perfect, conclusion was drawn that the center fer reception of the impulse was not destroyed. Likewise the converse was held true: that lost or confused discrimination indicated destruction of a specific center. In our experiments we applied the same principle to tracts in the spinal cord. A dog was trained to take food only when the right hind 1Sitzungsb. d. Konigl. Preus. Akad. d. Wissensch., 1907, x. 204 PATHS OF POSTURE-SENSE IN SPINAL CORD 205 leg was held in a certain position,—that of rigid extension backward; and to refuse food when the same foot was held in any other position. To eliminate possible habit formation by the dog to the stimulus of pressure on the leg, in both phases of the training (extension and flex- ion) the dog’s foot was subjected to an equal pressure by the hand of the operator. Unconscious cues and helps were carefully eliminated. In fact, however, due to the extreme hunger of the animal, it would have been almost impossible to distract his attention from the prob- lem. Grown dogs of all breeds were used but the sharp-nosed type served our purpose best. The dogs were never petted, spoken to nor allowed to commingle. In this way was developed a state of lonesome- ness and eagerness for companionship which made them tractable to training and eager for work. The exact procedure was as follows: The dogs were placed in clean cages and allowed no food for some days. Then by means of the strong ‘“hunger-motif”’ developed they were taught to leave their cages, to mount three steps to the experiment table, to take a certain position and to wait there until fed. Then the dog’s right hind leg was grasped by the operator and extended rigidly backward. A cube of cooked meat was then placed before the dog which he was allowed to take during this phase of the training. This backward extension furnished the “‘feeding-stimulus” for this animal. Next, the foot was placed in position of rest by the operator, and another cube of meat was offered the dog. During this phase of training the dog was not allowed to seize the particle of food. Only during the first days of the training was it necessary for the operator to interpose his hand be- tween the dog’s muzzle and the cube of meat. The dog soon learned when to take food and when to leave it. Punishment was never given for mistakes. In our experience such treatment of the dog rendered him unfit for training. As part of the training the animals were taught to return to their cages after feeding. Each daily lesson lasted about eight or ten minutes during which time about fifty equal-sized cubes of meat were fed to the dog. Great care was taken not to impair the dog’s “hunger-motif,’”’ either by over-feeding at the daily lesson or by - feeding between lessons. The dogs remained healthy during the experiment but became somewhat emaciated. The following precautions were taken in our training experiments: 1. Dogs were never petted, spoken to nor punished throughout the course of the training. 206 EUGENE S. MAY AND JOHN A. LARSON 2. Dogs were never over-fed nor fed at any other time than training time. In this way a strong ‘“‘hunger-motif’’ was maintained. This is the key to the experiment and eliminates such distracting factors as inattention, indifference or fatigue. 3. Duration of time of stimuli was the same and the order of pre- sentation varied by daily rearrangement. This precaution was taken to prevent rhythmic habit formation which might occur if the order of presentation of stimuli were left to chance. The dogs were considered perfectly trained after they had been taught to differentiate without error the “feeding-stimulus” from all others. They were tested with the “feeding-stimulus” and other stimuli fifteen to thirty times at the daily lesson over a period of two or three weeks. After they were found perfect, the spinal cord was hemisected on the right side about the level of the first thoracic vertebra. Laminectomy was done under ether anesthesia and the cord carefully and completely exposed. Then the dura was incised and the wound packed for a few minutes. After a dry field was secured the cord was carefully hemisected. After recovery from the shock of the operation, usually the second ~ or third day, discrimination tests were again undertaken, similar to those used during the training of the-animal. These tests were carried out over a period of from three to six weeks and were very satisfactory because of the prompt and decided responses of the animals to the stimuli employed. Careful notes on the behavior of the animals were made and will form the basis of full protocols in a later paper. At the end of six weeks the dogs were killed, the cords removed and the gross hemisection noted. The cords were then preserved in Miiller’s fluid for histological study of the degenerations. Marchi’s method will be used. SUMMARY Dog 1. Trained to accept food with right hind leg rigidly extended. Right hemisection of cord about level of first thoracic vertebra. Re- sponses to posture-tests in right hind leg prompt and. decided. Motor paralysis right hind leg complete. Pain sense lost in right hind leg. Dog 2. Trained like dog 1. Right hemisection of cord about the level of the last thoracic vertebra; and two months later another right hemisection of cord about the level of the first lumbar vertebra. Re- sponses to posture-tests in right hind leg prompt and decided. Motor paralysis right hind leg complete. Pain sense lost in right hind leg. PATHS OF POSTURE-SENSE IN SPINAL CORD 207 Dog 3. Trained like dogs 1 and 2. Right hemisection of cord at level of first thoracic vertebra and two months later left hemisection at level of first lumbar vertebra. Responses to posture-tests in right hind leg prompt and decided. Motor paralysis of both hind legs complete. | It was suggested that perhaps the animal was acting upon cutaneous stimuli alone; to eliminate this factor we blocked the cutaneous nerves of the right hind extremity of one animal by injecting a 2.0 per cent solution of cocaine into the skin close to the trunk. Pain sense was then tested with a red hot wire and the animal made no response. The response was unmistakable when the left side was tested. CONCLUSIONS 1. That decussation of part of the fibers mediating posture-sense impulses occurs within the cord. 2. That some of the impulses mediating posture-sense probably travel back and forth across the cord at different levels by short asso- ciation fibers. STUDIES ON THE REGULATION OF THE BLOOD DIASTASE B. FUJIMOTO From the Forensic-Medical Institute of the Imperial University at Tokyo Received for publication July 25, 1919 Wohlgemuth (1) reported that the diastase content of the blood is very stable and not influenced by feeding or administration of certain drugs (adrenalin, morphine, etc.), which have a marked effect upon its sugar content. The investigations reported in this paper were carried out in order to determine, if possible, how the blood diastase is regulated. Clere and Loeper (2) and Gould and Carlson (3) observed that the ligation of the pancreatic ducts is followed by an increased diastatic activity in the blood serum. They assumed this to be absorbed amylopsin. Otten and Galloway (4) and King (5) found, on the other hand, that the blood diastase sinks rapidly after complete pancrea- tectomy. The pancreas is, therefore, regarded as the chief source of the blood diastase. Wohlgemuth (6) stated that the diastatic activity of the blood serum in the portal vein is stronger than that in the hepatic vein. Schlesinger (7) also found that the diastatic power of the blood serum in the pan- creatic vein was two or three times stronger than that in the mesenteric vein or the peripheral blood vessel in some cases. Wohlgemuth, in discussing this report of Schlesinger, states that some difference in the diastatic power can be found between the blood in the portal and peripheral veins but not between the blood in the portal and pan- creatic veins. Thus it may be said that the diastatic substance passes from the pancreas into the liver and is there mixed with the blood and lymph, its output into the blood being regulated by the liver. The diastase content of the blood in the portal vein compared with that in the peripheral blood vessel. We have compared the diastase content of the blood in the portal vein with that in the peripheral blood vessel (table 1). In our experiments guinea pigs were always employed. For the estimation of diastase, we have employed Wohlgemuth’s 208 REGULATION OF BLOOD DIASTASE 209 method as modified by Inoue (8). The results of the digestion were seen after incubating for 30 minutes in a water bath at 38°C. In table 1 we see that the diastase content of the blood in the portal vein varies considerably and that sometimes it is larger than that in the peripheral blood vessel. These results confirm the reports of: Wohlgemuth and Schlesinger. Hence it can be said with certainty that the diastase content of the blood is regulated by the liver. The influence of hepatotoxin upon the diastase content of the blood. Next we have undertaken to injure the liver-cells by parenteral admin- istration of hepatotoxin to disturb the regulation of diastase in the liver. TABLE 1 The diastase content of the blood serum from the portal vein and carotid artery 30 DIASTASE D—~ GUINEA PIG 38 NUMBER a REMARKS Carotid Portal vein 21 185 150 55 125 150 After feeding 56 250 185 After feeding (V. mesenter. 250) 57 215 215 58 125 150 59 125 125 To a series of test tubes there were added increasing amounts of the blood serum and a constant dose of the amylum solution. In our experiments the amounts of the blood serum were so graduated that the diastatic activity in each test tube, when the amylum in it is completely digested, shows each 75, 85, 95, 105, 125, 150, 165, 185, 215, 250, 300. Though Karsner and Aub (9) have brought forth contradictory findings against the view of Delezenne (10), who had affirmed the organ specificity of hepatotoxin, I have recently proved (not yet pub- lished) that hepatotoxin acts specifically on the liver-cells and destroys their normal function. Hence the hypothesis that, if the liver is truly a regulator of the blood diastase, the latter may be influenced by the administration of the hepatotoxin. The hepatotoxin used in our experiments was obtained by immu- nizing rabbits with emulsions of the liver of the guinea pig. The in- traperitoneal injections were repeated twice. Various doses of hepato- toxin thus obtained were injected intraperitoneally in guinea pigs and the diastase content of the blood was examined repeatedly. The blood was always obtained from the ear. The results are shown in table 2 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 2 210 B. FUJIMOTO As we see in table 2, the diastase content of the blood sinks rapidly after injection of the hepatotoxin and then rises in a few days to its normal value. From these results we are able to say that the liver cells which regu- late the diastase content in the blood were intoxicated by hepatotoxin. The effect of pancreatotoxin upon the diastase content of the blood. The foregoing experiments were accompanied by the following tests in order to see if other organ toxins can also influence the diastatic activ- ity of the blood. Pancreatotoxin was prepared by immunizing rab- bits with emulsions of the pancreas of the guinea pig. The adminis- tration of the pancreatotoxin may cause a degeneration of the pan- TABLE 2 The diastatic activity of the blood in the peripheral blood vessel before and after the injection of hepatotoxin 30’ AMOUNT DIASTATIC ACTIVITY Dage OF HEPATO- GUINEA aK TOXIN IN- Fata ee WEIGHT | | Sapam Before Day after injection of hepatotoxin TONEAL | injec- CAVITY tion eS 1 Bea 3 4 5 6 ese grams 5 630 3 165 105 } 125) 125 150 29 570 3 165 | 125] 105 165 31 500 3 150 | 125) 105 125 t. 600 4 185 | 150} 125 150 165 9 415 4 165 | 125) 105 165 10 530 a 165 125 150 150 47 730 8 105 80 | 105 105 125 46 590 10 150 95 | 105 105 165 creatic cells. Hence supposing that the pancreatotoxin would influence the production of diastase in the pancreas, we have examined the diastatic activity of the blood after the injection of this toxin (see table 3). As we can see in table 3, no appreciable change in the diastase con- tent of the blood in the peripheral blood vessel was observed after injection of the pancreatotoxin, except in two guinea pigs, which re- ceived large doses (8 or 10 ec.). Such a large amount of this toxin was near to the lethal dose for guinea pigs. Therefore, it is assumed that the decrease in the diastase content in these two animals was probably caused by an intoxication of the liver cells. REGULATION OF BLOOD DIASTASE 211 TABLE 3 The diastatic activity of the peripheral blood before and after injection of pancreatotoxin AMOUNT OF DIASTATIC ACTIVITY Dae PANCREA- GUINEA popy |TOTOXININ- Taek SPA WEIGHT et at ae Bearers Day after injection of pancreatotoxin TONEAL injec- CAVITY tion pas 2 Sey 4 5 6 grams 3 560 3 185 185 185 185 4 370 3 215 185 215 185 2 390 2 185 185 185 215 185 18 710 8 125 95 125 150 150 19 640 10 150 105 105 150 165 In our experiments it was not decided whether the production of diastase in the pancreas was influenced by pancreatotoxin or not. But we cab say now that the production or mobilization of diastase in the pancreas was not so much affected that the liver could not regulate the diastase content of the blood in the peripheral blood vessel, and also that the regulating action of the liver was not influenced by a dose of 3 or 4 cc. of this toxin. Several experiments for control were undertaken with neurotoxin and the blood serum of normal rabbits. The results are shown in table 4. TABLE 4 The diastatic activity of the peripheral blood before and after injection of the neuro- toxin and the blood serum of normal rabbits | AMOUNT 30’ OF NEURO- brasratic acrrviry D3os TOXIN OR GUINEA SLL | SSS PIG BODY | smRUM IN- Day after injection of the neurotoxin or the blood serum NumBeER | “®IGHT | J ocrep IN- Before of norma] rabbits TO PERI- inj ec- TONE ion [a 2 3 4 5 6 grams 30 510 Bie 150 165 34 630 on 150 165 150 35 580 ie 215 215 215 36 490 8* | 165 . 125 | 165 58 580 3t 150 125 125 125 59 520 3t | 125 | 150 | 125 | 125 | * Neurotoxin. + Normal serum. 212 ‘B. FUJIMOTO As we see in the above table, 3 to 4 cc. of the neurotoxin or the blood serum of normal rabbits caused no change in the diastase content in the blood, while the same dose of hepatotoxin markedly affected it. Here we have also exceptions which may, however, be explained in the same manner as in the cases of pancreatotoxin. It can be now concluded that the hepatotoxin attacked the liver and affected its regulating power over the blood diastase. The effect of the injection of adrenalin upon the blood diastase of guinea pigs, to which hepatotoxin was intraperitoneally injected. Starkenstein (1), Allen (12) and Watanabe (14) found no marked change in the TABLE 5 The diastatic activity of the blood before and after injection of adrenalin or morphine DIASTATIC ACTIVITY GUINEA PIGS Piha aR ered Bopy weIcHT | “MOUNT pitta ete DEENAULEY iBétare injection 2 hours afer HEPATOTOXIN eos 30’ 38 38° number grams Adrenalin, 0.1 mgm. 250 250 11 430 Adrenalin, 0.2 mgm. 250 250 Adrenalin, 0.4 mgm. 185 165 Adrenalin, 0.15 mgm. 185 185 12 470 Adrenalin, 0.25 mgm. 215 215 Adrenalin, 0.5 mgm. 165 150 3 560 Morphine, 0.02 gram 185 185 1 500 Morphine, 0.02 gram 250 250 diastase content of the blood after subcutaneous or intravenous injec- tions of adrenalin or morphine. As we have proved in the foregoing experiments, the hepatotoxin can injure the liver, which regulates the blood diastase. Supposing, therefore, that adrenalin or morphine might affect the diastase content of the blood of the guinea pigs, which were treated with hepatotoxin, these drugs were subcutaneously in- jected in them (see table 5). But the diastase content of the blood was never affected. B The effect of Taka-diastase administered intraperitoneally upon the blood diastase. The injection of Taka-diastase might cause an increase in the diastatic activity of the blood of guinea pigs, especially after they REGULATION OF BLOOD DIASTASE 213 have been treated with hepatotoxin. In the following experiments a large dose of Taka-diastase was administered intraperitoneally, and the diastatic activity of the blood was examined (see table 6). I was surprised to see such a marked decrease in the diastatic activ- ity of the blood in spite of the injection of such a considerable quan- tity of. diastatic substance. In 1917 Richard Weil (14) proved that peptone markedly affects the liver. Taka-diastase contains peptone, besides diastatic ferment. TABLE 6 The diastatic activity of the blood of guinea pigs, which were treated with hepato- toxin, before and after injection of Taka-diastase ( 30' AMOUNT DIASTATIC ACTIVITY Dag) AFTER INJECTION OF TAKA-DIASTASE GUINEA OS ae wre '| noo | BEAEBARE , oe WEIGHT | INTO PERI- Before Birnates Day TONEAL | injec- CEES EAU NG ogy I ee) Nae 1 2 3 4 5 =i grams ne Pee es |) aes |e ol 500 0.05 125 95 btay il ealpaay |) A I SSO) 41* 860 0.05 150 105 | 105 12 al2 aie Zot loO 44* 710 0.05 165 125 OS: 105) 125) e550. 45* 730 0.5 125 125 85 75 | 105 | 105 | 150 47* 710 0.05 150 | 105 95 95 85 Obs LOS Oba at25 15t 610 0.05 150 95 Sd) Zea SON 50 e165 197 640 0.05 165) | eloOR Zon ELZ5 LOST Z om ee2p 49t 550 0.05 165 125 | 105 165 | 165 50t 550 0.05 150 125 | 105 165 | 185 * Treated with hepatotoxin. { Treated with pancreatotoxin. t Not treated. The latter may be eliminated by heating at 100°C. Hence, if Taka- diastase is heated at 100°C. there will remain the cocto-stable peptone. In the following experiments we have injected the heated Taka-diastase in normal guinea pigs in order to see if it affects the diastatic activity of the blood. Pure peptone dissolved in aqua destillata was also in- jected in guinea pigs intraperitoneally for control. The results are shown in table 7. We see now from the table 7 that the heated Taka-diastase and pep- tone produced similarly a marked change in the blood diastase. The 214 ‘ B. FUJIMOTO curves of the diastase content in the blood following injections of hepatotoxin, Taka-diastase and peptone are almost the same. As such a decrease in the diastase content of the blood may be caused by 3 or 4 ce. of hepatotoxin, but not by the same dose of the other organ toxin, we can explain this decrease by assuming that the change in the blood diastase after injections of Taka-diastase and peptone was caused by the intoxication of the liver cells. TABLE 7 The diastatic activity of the blood before and after injection of heated Taka-diastase or peptone 30’ DIASTATIC ACTIVITY Doge GUIN- eee 2 oie | ee Day after injection of heated Taka Num. |W CHT CAVITY Before stage te ol Sane BER injec- a tion | +3 | Set 2) 98). 4 ane grams 29 570 | Heated Taka-diastase, 0.05 | 125 | 95} 75 | 95/125! 150 30 520 | Heated Taka-diastase, 0.05 | 125 | 105 | 95 | 105) 125 55 490 | Peptone, 0.05 150 | 105 | 95 | 95}105)125)125) 125 56 620 | Peptone, 0.05 125 | 105 | 95 | 105) 105 | 125 | 125 57 610 | Peptone, 0.05 150 | 105 | 105 | 105) 125) 125 | 150 } ) SUMMARY 1. The diastase content of the blood in the portal vein varies, while that in peripheral blood vessel remains constant. 2. Hepatotoxin administered intraperitoneally causes a marked decrease in the diastase content of the peripheral blood, while the same dose of pancreatotoxin or neurotoxin has no effect upon it. 3. Even the intraperitoneal injection of a large dose of Taka-diastase does not cause an increase in the blood diastase; on the contrary, it is followed by a marked decrease in the diastase content of the blood. 4. The diastatic activity of the blood is considerably weakened after the injection of heated Taka-diastase or peptoné, which intoxicate the liver cells. 5. It seems sure that the diastase is regulated by the action of the liver cells. I wish to express my thanks to Prof. Dr. K. Katayama and Prof. Dr. 8. Mita for their kind advice. REGULATION OF BLOOD DIASTASE 215 BIBLIOGRAPHY (1) WoHLGEMUTH: Biochem. Zeitschr., 1909, xxi, 381. (2) CLierc anp LorprEr: Compt. Rend. Soc. Biol., 1911, lxxi, 75. (3) GouLtp anD Cartson: This Journal, 1911, xxix, 165. (4) Orrmn anp Gattoway: This Journal, 1910, xxvi, 347. (5) Kine: This Journal, 1914, xxxv, 301. (6) WouucGemouTH: Verhandl. d. 25 Kongr. f. inn. Med., 1908, 500. (7) ScHLESINGER: Verhandl. d. 25 Kongr. f. inn. Med., 1908, 505. (8) Fustmoto: This Journal, 1918, xlvii, 342. (9) KarRsNER AND Avs: Journ. Med. Research, 1913, xxviii, 377. (10) DrLEZzENNE: Semaine Med., 1900, xx, 290. (11) STaARKENSTEIN: Zeitschr. f. exper. Path. u. Therap., 1912, x, 78. (12) Atten: Glykosuria and diabetes, Boston, 1913, 117. (13) WaranaBE: This Journal, 1917, xlv, 30. (14) Wert: Journ. Immunol., 1917, ii, 525 THE CHANGES IN THE CONTENT OF HEMOGLOBIN AND ERYTHROCYTES OF THE BLOOD IN MAN DURING SHORT EXPOSURES TO LOW OXYGEN HAROLD W. GREGG, B. R. LUTZ anp EDWARD C. SCHNEIDER From the Medical Research Laboratory of the Air Service, Mineola, New York Received for publication August 7, 1919 The compensatory changes that occur in the blood of man and animals living under low oxygen tension have held the interest of many investigators. As long ago as 1878 Paul Bert (1) predicted that the blood of those living at high altitudes would be found to have a greater oxygen capacity than the blood of similar individuals living at lower levels, and he further suggested that the cause of the increase would be found to be the decrease in the partial pressure of the oxygen in the air respired. Since that time it has been clearly proved that a decrease in the partial pressure of oxygen of the respired air, regardless of the method of reduction, causes, if the reduction is great enough and the exposure continued through a sufficient interval of time, an in- crease in the erythrocytes and hemoglobin per unit volume of the blood. This increase has been found to be gradual, requiring from three to five and even more weeks to reach its maximal value (2), (8). — The time required for the increase in erythrocytes and hemoglobin to be first manifest has received some consideration. Campbell and Hoagland (4) carried rabbits to the summit of Pike’s Peak, from an altitude of 6000 feet to one of 14,110, and found that in the ascent the number of red corpuscles had made an average increase of 9 per cent. Abderhalden (5), working with rabbits and rats, found an increase within a few hours. Ehrlich and Lazarus (6) state that the increase occurs immediately when considerable altitudes are reached. Douglas, Haldane, Henderson and Schneider (7) found in four men, several hours after their arrival on Pike’s Peak, a slight increase in hemoglobin that varied for these individuals from 0.9 to 3.9 per cent. Schneider and Havens (2) were unable to demonstrate a clearly defined increase during the first seven hours spent on Pike’s Peak, but within twenty-four hours there was a marked increase in the number of red 216 BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 217 corpuscles and the percentage of hemoglobin. The increase occurred earliest and was most rapid in physically fit men. Dallwig, Kolls and Loevenhart (3) observed that animals kept at normal atmospheric pressure but under low oxygen showed a definite increase inthe blood counts at the time of the first observations, viz., after two or three days. Six rabbits living at 352 mm. Hg. pressure required as much as twenty- four to forty-eight hours for the increase to become definite. All of the above observations were made after hours or even days of exposure to the effects of high altitudes and deficiency of oxygen. In order that the aviator may benefit by blood compensatory changes they would have to occur during exposures of thirty minutes to three hours. We have, therefore, investigated the blood changes that occur in men subjected to a lowered barometric pressure in a low pressure chamber and to low oxygen, 10 per cent, for intervals not exceeding two hours. In a preliminary report by one of us (8) it was shown that in short exposures at least 25 per cent of all men examined had a well-defined increase in the percentage of hemoglobin. Corbett and Bazett (9) using a low oxygen method conclude that after about half an hour some degree of blood concentration may occur. Blood for the estimation of hemoglobin was obtained in the usual manner by pricking the finger. In a few experiments with the Dreyer Nitrogen Apparatus it was taken from the lobe of the ear. In many cases the blood from the finger was compared with blood taken with- out stasis from a prominent vein in the forearm. Blood was obtained from the vein with a 10 cc. Record syringe and put in a short tube containing a little sodium oxalate. After a thorough stirring a few drops were taken with a pipette and put on a watch glass from which the sample was immediately taken into the blood pipette. In the earlier experiments the Gower-Haldane carbon monoxide method (10) was used. Each sample was matched at once with the standard in the low pressure chamber by the aid of a white background and a ‘‘daylight” electric lamp. It was found more convenient, because of difficulty with the carbon monoxide supply, to dilute the blood in small test tubes containing 0.4 per cent ammonia. These samples were later transferred with proper rinsing to the Gower-Haldane graduated tube, saturated with carbon monoxide, and diluted further. The last sample taken in the experiment was always received directly in the Gower-Haldane tube. 218 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER Several experiments were carried out using the Palmer method (11). It was found difficult to keep the 1 per cent standard carbon monoxide blood, therefore the normal samples were usually considered as 100 per cent and the later samples were matched against them with the Duboseq colorimeter. The most convenient method was found to be a modification of the Palmer method which consisted in diluting the blood in 5 ee. of N/10 hydrochloric acid, instead of 0.4 per cent am- monia. The diluting fluid was measured in small test tubes which were stoppered with cotton and taken into the low pressure chamber. The blood was rinsed immediately into the diluting fluid from the 0.05 ec. pipette. The samples were matched at the end of the experi- ment as in the Palmer method. Blood for the erythrocyte counts was taken by pricking the finger so that a free flow was obtained. The same Thoma mixing pipette was used for all comparative counts, and the blood was diluted in Hayem’s solution. Two drops were put in a Levy double counting chamber with Neubauer ruling. Experiments in the low pressure chamber. These experiments con- stitute the major part of this study. Throughout the work we adopted the plan of lowering the barometric pressure within the chamber at arate that would be comparable to ascending in the air at the rate of 1000 feet per minute. The following altitudes were employed: 425, 395 and 380 mm. Hg., which are the pressures ordinarily encountered at 15,000, 17,000 and 18,000 feet respectively. The desired pressure having been” attained it was then maintained for a time during which several samples of blood were taken and other observations made. From 30 to 100 minutes was the usual exposure to the lowered pressure. A total of forty-five experiments was made upon thirty-five men, five of whom served as subjects two to five times. In fifteen of the experiments the hemoglobin determinations were made on blood from a vein of the arm as well as from the peripheral vessels of the finger. Also in fifteen other experiments the erythrocytes were counted as a check against the final hemoglobin determination. The data from these fifteen cases are given in table 2. The results for the thirty exam- inations in which the hemoglobin changes only were considered are collected in table 1. The blood changes were definite in thirty-five, or in 78 per cent, of the experiments. In only ten, or 22 per cent, was the period of ex- posure to the lowered barometric pressure too brief or not sufficiently low to cause the blood response. Of the eight cases held at a pressure BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 219 NAME AND DATE AS EH. May 21 Caw. D. May 23 iW: G. May 24 W. B.M. May 27 W. O. K. June 6, 1918 Bas. V. June 26, . 1918 TABLE 1 Low barometric pressure and hemoglobin HEMO- MINUTES GLOBIN 15,000 feet 0 94 15 94 55 97 85 ° 98 0 104 15 104 58 105 75 106 0 96 15 108 25 106 35 106 45 105 65 105 0 104 15 102 25 104 35 104 45 104 55 104 75 110 0 106 25 106 45 106 65 107 83 107 0-V 106 93-V - 107 17,000 feet 0 92 25 93 40 94 54 94 0-V 98 INCREASE IN YER CENT 4.3 WE) 9.4 5.8 0.0 0.0 2.2 2.0 NAME AND DATE MINUTES HEMO- GLOBIN 18,000 feet N. E. F. 0 June 7, 25 1918 45 Re Sans: 0 June 10, 25 1918 35 Ne Gab: 0 June 11, 25 1918 35 dala (oad te 0 June 14, 25 1918 43 94 96 96 97 100 95 99 102 104 104 106 107 106 102 104 107 110 110 110 112 112 111 112 107 111 105 105 111 110 105 109 112 112 114 114 114 108 110 INCREASE IN PER CENT 6.9 4.2 3.9 2.0 4.7 3.7 4.8 3.8 1.8 io 220 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER TABLE 1—Continued INCREASE INCREASE wane 42 | servonne | Emon | The |] AME 2 | aeocones | SMO | a B. M. L. 0 96 WiC Wes 0 98 June 17, 26 96 July 16, 64 102 1918 40 99 1918 78 104 6.1 56 106 (5; a= 104 8.3 biG. 0 100 July 19, 40 - 100 ASW ena, 0 99 1918 68 102 June 18, 26 100 80 103 3.0 1918 43 98 55 100 1 eS 0 108 65 100 July 26, 45 108 75 98 1918 63 108 | . 0.0 85 100 0.0 0-V 98 E. W. B. 0 98 86-V 100 2:3 June 30, 25 98 1918 40 98 Dae eks AVE 0 103 60 100 July 21, | 25 102 80 102 1918 35 102 92 104 6.1 45 104 55 107 I. M. 0 108 65 106 3.0 July 1, 79 111 2.8 0-V 104 1918 0-V 107 68-V 106 1.9 80-V 110 2.8 iG: CW: 0 98 H. W. B. 0 106 June 25, 27 94 July 2, 58 110 3.8 1918 41 94 1918 0-V 108 50 97 60-V 112 37 82 100 2.0 0-V 97 B. F. 0 106 83-V 99 Py July 2, 50 106 1918 66 109 Pose W. B. M. 0 104 June 28, 25 103 1 B gel a 2 0° 98 1918 40 104 July 2, 40 97 55 107 1918 60 97 72 110 5.8 88 97 0.0 0-V 106 0-V 100 75-V 108 1.9 89-V 100 0.0 A. F. H. 0 94 15 led lle 0 92 July 15, 40 95 July 5, 41 89 1918 60 95 1918 59 90 0.0 78 98 A430 0-V 90 61-V 89 0.0 BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 221 TABLE 1—Concluded INCREASE INCREASE meee (laoreet | ae WEARER | Leen | SOS | ANG ite dele 0 99 E. C. S. 0 100 July 8, 40 100 April 29, 29 108 1918 60 104 1919 Sf/ 107 86 | 104 5.0 UU 109 92 107 7.0 12 [Sh 1B 0 104 July 9, 36 102 INGE Be 0 100 1918 60 110 1919 42 108 85 109 4.8 75 110 76 110 10.0 V = Blood from vein. corresponding to 15,000 feet three, or 38 per cent, failed to compensate, while among the five at 17,000 feet one, or 20 per cent, and among the thirty-two at 18,000 feet six, or 19 per cent, did not respond. There were six men, 13 per cent, who showed a well-defined increase in the hemoglobin within the first 26 minutes of the experiment. This included the time allowed for ascent which varied from 15 to 18 min- utes. The majority of men require between 45 and 60 minutes for the increase to be sufficient to be detected by the methods used. There was a marked parallelism between the blood from the finger and blood from the vein, but in almost every case that from the vein was found to contain slightly less hemoglobin than that from the finger. In the fifteen experiments in which the erythrocytes were counted there was an increase per cubic millimeter in each case in which the hemoglobin gave evidence of concentration. In five of these experi- ments no blood change occurred. The relation between the increase in the number of erythrocytes and hemoglobin shows a greater increase in erythrocytes than hemoglobin in seven of the ten positive cases. W. B. M. served five times as a subject, once at 425 and four times at 380 mm. Hg. His hemoglobin percentage increases were 5.8, 4.8, 5.8, 0 and 8.8 respectively. A. F. H. was examined four times, one at 425 and three at 380 mm. respectively, with 4.3, 5, 4.3.and 0 per cent increases in hemoglobin. W. H. G. in two times showed 9.4 and 6.1 per cent, and B. R. L. in two experiments showed 8.4 and 8 per cent rises in hemoglobin. W. O. K. at 425 mm. failed to show a response, but at 395 mm. Hg. had an increase of 7 per cent in hemoglobin. The surprising feature of these repeated cases is the fact that, when a defi- H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER 222 ee ee ee SSSCoONMNMD N re 000‘82‘¢ 000‘0zg‘¢ 000186 ‘F 000‘009‘¢ 000‘F02‘F 000‘Z10'¢ 000‘0809 000‘0z2‘S 000‘¢28‘¢ 000‘E86‘F 000‘996‘F 000‘F90'S 000‘08F F 000‘898‘¢ 000‘9Sz‘¢ bl — COMANSHNOOCOOHNA re SHONWMOSO & bo = ° quad Lad fe) arom | 0B yy SALANOUHLAUAL 000‘F2z‘¢ 000‘FL2'F 000‘2S2‘F 000‘729°F 000986 'F 000‘080‘¢ 000‘Z8z‘¢ 000‘9T2‘¢ 000‘898‘F 000‘00z‘¢ 000‘808'F 000‘0FF'F 000‘Z62'F 000‘0ZI°S 000‘8F0'S DS 1) 09S Oe OO OOS HOODOROMSOONDNHNODW S ° quan sad [BULION gsuvyy Oct FOr Pied jo pus ty ‘OINUIUT GB OOF QOOT JO OFBI OY} 4B OPVUL SBM YUODS" oY L, 000‘ST 000‘ST 000‘8T 000‘8T 000‘ST 000‘ST 000‘ST 000‘8T 000‘2T 000‘21 000‘2T 000‘2T 000‘ST 000‘ST 000‘ST qaaf O8& Oss O8é OSE OSE OSE OSE O8& G6E S6E S6E G6E SCV SGP GCP “UU OOT 08 OOT €8 16 G8 LOT SPT FOL 9¢ 66 69 G6 68 86 cOl 98 cOL 16 98 c8 cs &8 16 Gol 19 0or 09 OOT GL saynuvue [BULION ING NIgOTOONYH HIDNG@T aqoOLinty avossaud 6I6L ‘ZI AVIN 6161 ‘Og [udy SI6I ‘Lz toqui000q, SI6I ‘¢ ysnsny SI6I ‘Tg Ate SI6T ‘Os AIL SI6I ‘er Aine SI6T ‘Or Ale 6I6L ‘FI Arenuee 6161 ‘ET Avenue 6I6T ‘9 Arvnuee SIGI ‘TE Joquie00qy SI6I ‘ET taquies0q, SIGI ‘OT taquie00q7 SIGI ‘6 rTequrie00q Haz Om Seeing = OHa nasa oe Sa< 40 yaiee Se ee Le oe moa O wi A Yoo BHORM aitya GANVN Boi Wii ee ee ee i eS SS ee ee waquipyo aunssasd mop ur Uurgopbowaey pun sajhioowy hig ¢ WTaViL BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 223 nite response does occur, the total increase is so often approximately the same in an individual. Just why there was failure in the blood response in the two men who ordinarily reacted well to low oxygen is not indicated in our data. It is evident that there are physiological conditions under which an individual may not react in equal degree every time he encounters a given barometric pressure. We believe that when the blood fails to give the increase in red corpuscles and hemoglobin under the low oxygen of low barometric pressures that a heavier burden is thrown upon the respiration and the circulation of the blood. Experiments at normal atmospheric pressure and 10 per cent oxygen. In these experiments the subject breathed atmospheric air diluted with nitrogen by the Dreyer Nitrogen Low Oxygen apparatus (12). Start- ing with undiluted air,’20.96 per cent oxygen, the subject breathing through a mask, the nitrogen was gradually added in greater and greater proportion so that by the end of 20 minutes the mixture con- tained only 10 per cent oxygen. Ten per cent oxygen at 760 mm. Hg. pressure corresponds in oxygen partial pressure to an altitude of ap- proximately 19,000 feet. The subject of the experiment was held at this level of oxygen for from 30 to 90 minutes, thus he was kept under low oxygen for a period of from 50 to as much as 112 minutes. Only seven men weré examined for the hemoglobin changes by this method, four of them gave a positive response (see table 3). In three of the men the increase in hemoglobin had already begun when the first sample of blood was taken, 20 to 26 minutes, which was soon after 10 per cent oxygen was reached. Two of the men, W. O. K. and E. A. R., were also tested in the low pressure chamber. W. O. K. showed an 8 per cent increase in hemo- globin under 10 per cent oxygen and 7 per cent under 395 mm. baro- metric pressure. E. A. R. with the low oxygen gave 5 per cent increase in hemoglobin and with the low pressure 380 mm. Hg., gave 3.2 per cent. -The results obtained by the two methods of subjecting men to low oxygen—lowered barometric pressure and lowered oxygen percentage— show that an increase in hemoglobin and the red corpuscles of the blood may result during short exposures. About 60 per cent of the men sub- jected to 10 per cent oxygen and 78 per cent of those subjected to low barometric pressure showed within from 15 to 90 minutes clearly defined increases in hemoglobin that ranged between 1.8 and 10 per cent. From the data presented it appears that the majority of men make the blood compensation rather quickly. Some delay is usually present in this response to low oxygen, but the lag is surprisingly short. 224 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER We have not investigated the mechanism by means of which these quick and early blood changes occur, when the organism is subjected to lowered partial pressure of oxygen. Views have differed as to the mechanism by which the marked increases in erythrocytes and hemo- TABLE 3 Normal atmospheric pressure, oxygen 10 per cent > INCREASE INCREASE NAME AND DATE| MINUTES Saat pee NAME AND DATE} MINUTES oan pee PAS 0 98 GLA: 0 115 May 25, 20 101 June 1, 22 112 1918 37 99 1918 40 113 57 99 0.0 55 116 | 65 114 Gabor: 0 100 a 114 May 28, 18 100 90 116 1918 34 100 100 113 0.0 46 100 70 100 WO Ske 0 100 82 104 4.0 June 3, 26 102 1918 39 106 C.N. 0 109 50 108 8.0 May 29, 20 107 1918 40 107 BvAs Rig ta 100 75 106 June 4, 20 102 104 106 1918 40 103 112 106 0.0 55 103 74 105 Wi-A” B: 0 104 81 105 5.0 May 31, 22 106 0-V 90 1918 41 106 80-V 102 13.3 56 108 71 112 81 112 V = Blood from vein. globin occur during residence at a high altitude. Schneider and Havens (2) have given their opinion of the changes in the blood on adaptation to high altitudes as follows: A rapid increase in the number of red corpuscles and percentage of hemo- globin in the blood of the peripheral vessels occurs during the first two to four days of residence at the high altitude, then follows a more gradual increase for BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 225 about three weeks. The initial rapid increase is brought about in part by throw- ing into the systemic circulation a large number of red corpuscles that under ordinary circumstances at low altitudes are side-tracked and inactive, and in part by a concentration resulting from a loss of fluid in the blood. The more gradual increase in red corpuscles and hemoglobin extending over several weeks is brought about by the increased activity of the blood-forming centers so that there results a large increase in the total number of corpuscles and amount of hemoglobin. The question which presents itself here is whether the early increase in hemoglobin and red corpuscles, such as we obtained within the short space of an hour, is to be attributed to concentration of the blood, i.e., a reduction in the total blood volume, or to changes in the dis- tribution of the erythrocytes. A suddenly increased production of hemoglobin and erythrocytes by the bone marrow is improbable. That the increase is not caused by an increased evaporation of water from the body is indicated by the conditions of experimentation and the fact that in some men the change is already well developed within a 15 to 20 minute period and that perspiration is not noticeably in- creased. It has been claimed that all aviators engaged on long patrols at 17,000 feet, or over, complain of over-filling of the bladder. Birley (13) looks upon this as confirmatory of the theory that one factor in the reaction of the organism to lowered barometric pressure is a con- centration of the blood at the expense of the plasma. Against this theory of a polyuria we urge that the time, 15 to 20 minutes as seen in a few cases, is too short. If this were the mechanism, all subjects should be conscious of the filling of the bladder. The majority of our subjects have not been conscious of an increased action of the kidneys. In fact we are inclined to believe that only the nervousness experi- enced during the first time or so spent in the low pressure chamber gives rise to the sensation of an increased sensation of urine. Several observations made, in and out of the low pressure chamber in this laboratory, on the secretion of urine fail to confirm the polyuria theory. The increased production of urine during flights at high altitudes finds an explanation in the action of cold. Against the likelihood of this increased urine formation being evidenced in blood concentration we have the studies of Bogert, Underhill and Mendel (15) in which they introduced large volumes of fluids without appreciably affecting the unit blood content of hemoglobin. Another possibility is that lowered oxygen tension changes the prop- erty of the muscles so that they absorb a larger volume of water and THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 2 226 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER in sufficient quantity to reduce the blood volume. We know of no experimental proof for this view. The percentage of increase in hemoglobin and erythrocytes observed in these short exposures to low oxygen is within the limits of those ob- served after various forms of physical exertion. Schneider and Havens (14) found that exercise increased the hemoglobin to from 3.5 to 11 per cent and the number of red corpuscles per cubic millimeter to from 3.2 to 22 per cent. They held that this increase was the result of throwing into the systemic circulation a large number of erythro- cytes that under ordinary circumstances are side-tracked and inactive. The same explanation might be advanced for the low oxygen com- pensatory blood changes. A decision as to the value of the concen- tration theory and the theory of the’ dormant supply of erythrocytes cannot be made at this time. The physiological significance of an increase in erythrocytes and hemoglobin during exposure to low oxygen is that a unit volume of blood can carry for a given oxygen pressure more oxygen than nor- mally. The supposition is that the aviator whose blood concentrates will, other things being equal, tolerate high altitudes more comfort- ably and more efficiently than the man who does not react with an increase in erythrocytes and hemoglobin. SUMMARY 1. Low oxygen tension was produced by lowering the barometric pressure in different experiments to 380, 395 and 425 mm. Hg., and by replacing oxygen by nitrogen gradually until 10 per cent oxygen was reached. The subjects were maintained at the low oxygen tensions from periods varying from 30 minutes to 145 minutes. 2. Blood for the estimation of hemoglobin was taken from a finger and a vein. The determinations were made by the Gower-Haldane carbon monoxide method, by the Palmer method, and by a modified Palmer method using hydrochloric acid. 3. An increase in hemoglobin was obtained under reduced barometric pressure in 78 per cent of all examinations made. The majority of the men required between 40 and 60 minutes for the increase to be- come definite, 13 per cent showed a well defined increase within 26 minutes. In the experiments with 10 per cent oxygen 57 per cent gave the increase in hemoglobin. BLOOD CHANGES FROM SHORT EXPOSURES TO LOW OXYGEN 227 4. In fifteen cases in which the erythrocytes and hemoglobin were determined corresponding changes occurred in both, 66 per cent were positive. The erythrocyte increase ranged between 3.8 and 20 per cent, the hemoglobin between 3.2 and 9.8 per cent. 5. In the several experiments on the same individual, the increase in the hemoglobin was approximately the same each time. 6. The blood concentration theory and the theory of the dormant supply of erythrocytes are briefly contrasted. BIBLIOGRAPHY (1) Bert: La Pression Barometrique, 1878, 1108. (2) ScHNEIDER AND Havens: This Journal, 1915, xxxvi, 380. (3) Datuwie, Kouts anp LorveNnnART: This Journal, 1915, xxxix, 70. (4) CaMPpBELL AND HoaGcuanp: Amer. Journ. Med. Sci., 1901, exxii, 654. (5) ABDERHALDEN: Zeitschr. f. Biol., 1902, xliii, 125. (6) Enrutcu anp Lazarus: Anaemia, Nothnagel’s Encyclopedia, Philadelphia and London, 1905, 22. (7) Dovuatas, HatpANE, HENDERSON AND ScHNEIDER: Phil. Trans. Roy. Soc., London, 1913, Series B, ciii, 271. (8) ScunerIpER: Journ. Amer. Med. Assoc., 1918, Ixil. (9) Corsertr anv Bazerr: Repts. Air Medical Investigation Committee, Lon- don, no. 5, November 14, 1918. (10) Hatpane: Journ. Physiol., 1900, xxvi, 497. (11) Paumer: Journ. Biochem., 1918, xxxiii, 119. (12) Dreyer: Repts. Air Medical Investigation Committee, London, no. 2, March 23, 1918, 8. (13) Brrtey: Repts. Air Medical Investigation Committee, London, no. 2, 1918, 5. (14) ScoNErDER AND Havens: This Journal, 1915, xxxvi, 239. (15) Bogert, UNDERHILL AND MenpDEL: This Journal, 1916, xli, 189. CIRCULATORY RESPONSES, TO LOW OXYGEN TENSIONS BRENTON R. LUTZ anp EDWARD C. SCHNEIDER From the Medical Research Laboratory of the Air Service, Mineola, New York Received for publication August 7, 1919 ~ , It is a well established fact that residence at high altitudes exerts a profound influence on the human body. Because of the recent devel- opment of aviation, which has made rapid ascents to very high alti- tudes possible, the knowledge of the effects of short exposures to the influence of altitude assumes practical importance. The adaptive reactions to altitude observed in men who take up residence at a high altitude develop rather slowly and are of a fairly permanent char- acter (1). The aviator does not remain long enough at a high alti- tude to benefit from slow adaptive physiological changes. If he toler- ates and does well, he must depend upon rapid compensatory changes to provide the oxygen needed by the tissues. That the body is capa- ble of responding to abrupt and great changes in atmospheric pres- sure has been proved by studies made in this laboratory (2). Among the physiological responses made to low oxygen tensions are those of the circulatory mechanism. The object of this paper is to present observations on the pulse rate and the arterial blood pressures made upon men who were subjected to low oxygen tension produced in three ways, by low barometric pressures in a low pressure chamber, by low percentage of oxygen caused by rebreathing under normal atmospheric pressure, and by diluting the respired air with increasing amounts of nitrogen. The low pressure chamber and its control has been briefly described elsewhere (3). The rebreathing experiments were made with the Henderson-Pierce rebreathing machine (3). The Dreyer method (4) was used to dilute atmospheric air with nitrogen which was delivered to the subject by means of an American model of a Tissot gas mask. Two types of experiments have been carried on. In one the oxygen tension was gradually reduced until the mental condition of the sub- ject showed that he was no longer able to compensate, or until syn- cope appeared. In the other group of experiments the oxygen tension 228 CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS 229 was reduced at a rate corresponding to an ascent of 1000 feet per minute until a desired level,—that is, 15,000, 17,000, 18,000 feet,— was reached, after which the level was maintained for from 30 to 90 or more minutes. Throughout all experiments the subjects were seated, and it was the rule to count the pulse rate for half a minute during each minute of the experiment. Usually the count was begun at 45 on the seconds dial of a stop-watch and continued to 15 and then recorded as though taken on the minute. In the remaining portion of the minute the blood pressures were taken. To keep such a record requires close attention and extensive experience. During each experiment one man was held responsible for all these determinations and was relieved of the necessity of watching the condition of the subject and arrange- ments of experimentation. The arterial blood pressures were determined by the auscultatory method with the aid of a Tycos sphygmomanometer which was ad- justed over the brachial artery of the left arm. A Bowles stethoscope with special arm band was used. The systolic pressure was read at the beginning of the first phase and the diastolic pressure was measured at the fourth phase, that is, at the dulling of the intense sounds of the third phase. In many of the low pressure chamber experiments, in which the pump was run continuously, the diastolic determinations had to be omitted because of noise. In the selection of subjects an effort was made to secure men who had not been doing physical work during the hour previous to the experiment. Before any observations were made the subject was allowed to sit quietly for a while, after which a number of preliminary determinations of the pulse rate and arterial blood pressures were made to establish the so-called normal. Occasionally the first time a subject appeared for the low oxygen test he showed some degree of anxiety or excitement in a slightly rapid pulse or increased systolic ' pressure. After a few moments of tactful conversation this nervous- ness was usually overcome. In much of our work we have used men accustomed to being subjects, and in these the excitement effect is often absent, or if present almost negligible. In all experiments in which an attempt is made to detect the earliest effects of low oxygen, this psychic factor has to be considered. Unquestionably it often masks the onset of heart rate acceleration due to low oxygen. Earlier work on heart rate during exposures to low oxygen tension. When the oxygen tension of the respired air is decreased, the blood for 230 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER a time may be less completely saturated with oxygen than when air of normal composition and pressure is breathed. During such a con- dition the tissues would very likely be inadequately supplied with oxygen. If during this period the blood contains less oxygen than normally, and the rate of blood flow through the capillaries is increased, the tissues will be provided with the oxygen demanded for their ac- tivity. More blood flowing to the tissues, even though it contains a lessened amount of oxygen, results to some extent in maintaining the oxygen tension in the tissues. Throughout our experimental work with low oxygen we have as- sumed that an increase in the rate of the heart beat, the arterial pres- sures being maintained within normal limits, meant an increase in the per-minute output of the heart. Under ordinary circumstances an increase in the pulse rate during exercise is recognized as satisfactory evidence that the output of the heart has increased and the flow of the blood has accelerated. The idea that an increase in the heart rate is a method of compen- sating for lack of oxygen is by no means new. Finkler (5) in 1875 induced anemia in dogs by bleeding and found that the decrease in the oxygen content of the blood may stimulate both the heart and respiration to greater activity. These, as Lusk (6) has pointed out, are efforts of compensation for the decrease in oxygen, although nothing resembling asphyxia is present. Kohler (7) in 1877 interfered with the respiration of rabbits by compressing the trachea by means of a lead wire tied around it. This was followed by compensation by means of increased respiration and heart activity so that there was no lack of oxygen in the animals. In these animals the heart hypertrophied. Experimental studies on men living at a high altitude have seemed to prove an increased rate of blood flow. Schneider and Sisco (8), on Pike’s Peak by use of Stewart’s hand-colorimeters, concluded that “‘the rate of blood flow in the hands of six men examined was increased .- by an amount varying from 30 to 70 per cent.’’ The in- crease in the rate of flow has been associated in part with an augmented rate of heart beat and a fall in the venous pressure, also in part with a dilatation of the arterioles. Kuhn (9) working on Monte Rosa dem- onstrated by calculations made from determinations of the oxygen capacity of the blood, the total oxygen consumption, and the pulse’ rate, that the heart rate responds to the influence of lowered baro- metric pressure by increasing its output per minute. CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS 231 Hasselbach and Lindhard (10) working with three men in a pneu- matic chamber failed to prove an increased blood flow with the ni- trous-oxide method. It should be noted, however, that their pressure changes were too small to produce profound change. They main- tained pressures of 589 to 514 mm. Hg. (6800 to 10,400 feet) for five to seven days and attained these pressures very slowly. Under these conditions other compensatory changes might have been sufficient to meet the call for oxygen. Our knowledge of the influence of high altitudes on circulation has been secured chiefly from men living at high altitudes on mountains. Of all the circulatory changes due to diminished barometric pressure, the acceleration of the heart rate has been most studied. Mountain ascents, even when made passively by railway car or automobile are slow, 8000 feet in an hour and a half or longer, when compared with altitude flights in an aeroplane. It has been shown by studies on Pike’s Peak (11), (14,110 feet), that the pulse rate does not accelerate im- mediately on arrival at the summit. It accelerates gradually in those who ascend passively by train and remain well, and requires several days to reach the maximum rates. In men who become mountain or altitude sick the augmentation comes on earlier and is greater than in those who remain well. Later the rate returns to the normal for the particular altitude. In men fatigued by walking to the summit the high altitude heart rate is usually established within a few hours. Changes in heart rate during a gradual decrease in oxygen tension. Seventeen men served as subjects for a series of examinations in which the action of a steadily decreasing barometric pressure was compared with that of a steady decrease in the oxygen percentage by the re- breathing method. In this work it was customary to give the rebreath- ing test first, and on a later day the low pressure test. From the rebreathing data, in which the final oxygen percentage and the dura-- tion of the test in minutes were recorded, we calculated the barometric pressure that was equivalent in oxygen tension to the final rebreathing oxygen per cent, and then determined the rate at which the pressure should be lowered in the low pressure chamber. It was thus possible to reproduce with a fair degree of accuracy in the low pressure chamber test the oxygen tension changes experienced during rebreathing exam- ination. In two pairs of experiments (see G. F. H. and F. D., table 1) the rebreathing test was prolonged by introducing into the tank of the apparatus a continuous flow of oxygen. This, in effect, made the altitude ascent about three times slower than that of the average test. In about 70 per cent of the tests the men were carried down in oxygen. 232 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER until they became inefficient as judged by the psychologist or the fail- ure of the compensatory mechanism. The response of the heart during rebreathing tests of 25 to 30 minutes duration has been described by Schneider (12). He reported that in men under a gradually decreasing oxygen supply the heart rate soon began to accelerate, at first by a slight increase of from one to three beats, and later by a very marked acceleration when the oxygen had fallen to between 13 and 9 per cent. The heart rate was shown to accelerate in a few men as early as 17.5 per cent of oxygen (5000 feet) while 12 per cent of all cases examined began to respond between 15.5 and 14.9 per cent oxygen, (8000 to 9000 feet). In our series of low pressure and low oxygen percentage comparisons the similarity of the circulatory responses made by the individual to the two conditions was most striking. The data have been analyzed and tabulated in table 1. It was impossible to have the subject in exactly the same condition for each test because the two types of tests were separated by at least 24 hours and in one case by 18 days. Fur- thermore the degree of apprehension in the subjects differed for the tests. Some men dreaded going into the low pressure chamber and others disliked the mouthpiece and nose clip of the rebreather. The apprehension was of course registered in a quickened pulse rate or increased systolic pressure when the normals of each were compared with the determinations made at the start of the actual experiment. When the psychic factor was not in evidence, the pulse rate, in each of the two kinds of low oxygen experiments, maintained for a short time the normal or pre-experiment rate and then gradually began to accelerate. In the majority of cases there was at first a slow increase in rate, but when the oxygen tension had fallen to that corresponding to from 15 to 10 per cent oxygen at normal atmospheric pressure, it gave way to a more rapid rate of acceleration. The beginning of a pulse rate acceleration has been determined as to time and pressure, or oxygen per cent. Three of the men showed in both the low pressure chamber and under low oxygen per cent a definite gradual acceleration which began with the first change in pres- sure or oxygen per cent at the beginning of the experiment. In the seventeen comparisons the latest onset in the acceleration occurred at 15.2 per cent oxygen (8400 feet). We believe that the evidence proves the heart to be responsive to a slight decrease in oxygen in the air respired if the psychic acceleration is satisfactorily eliminated, as it has been in a great many of our cases. Fourteen times in this series of CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS Zao comparisons, or in 41 per cent of the tests, the acceleration began between 17 and 15 per cent oxygen, that is, between altitudes of 5800 and 8800 feet. All men in this series showed an increase in heart rate before an oxygen percentage corresponding to 9000 feet was reached. Omitting from our calculations the cases that gave an immediate acceleration there were nine cases, or 26 per cent, that responded with an increase in heart rate at 18 or more per cent of oxygen (4000 feet or less). It is generally found that men living at moderately high altitudes, 6000 to 9500 feet, to which they are acclimated, do not show an aug- mentation in the rate of heart beat. The reaction shown in our tests is an immediate compensation to low oxygen, and as will be seen later is not necessarily a permanent change which would be maintained so long as the particular oxygen tension was held. The normal pulse rates and the maximum rates, which occurred when the oxygen tension was lowest, are recorded in table 1. In several of the runs the final heart rate was only 16 or 18 beats per minute above the preliminary or normal rate. In other men the acceleration was 45 and in one case 57 beats per minute. The smaller acceleration usually occurred in men who reached only 9 or 10 per cent oxygen, while the greater increase occurred with 6 and 7 per cent oxygen. The per- centage of acceleration brings out more clearly the comparative dif- ferences; thus R. M. B. at 10 per cent oxygen had an acceleration of 19 per cent, and G. F. H. at 6.3 per cent oxygen showed a 79.3 per cent increase. Those who would account for the circulatory change at reduced atmospheric pressure apart from decreased oxygen tension, would find it difficult to explain the parallelism observed in the responses of the heart to the two methods of experimentation employed in these pairs of tests, under low barometric pressure and low oxygen caused by rebreathing at normal atmospheric pressure. The heart rate re- sponded at so nearly the same percentage of oxygen in eleven of the seventeen pairs of experiments that we are justified in speaking of them as duplicate responses. In F. D. the acceleration began at 17 and 16.9 per cent, in D. T. R. at 15.5 and 15.8 per cent, in R. M. B. at 15.4 and 16 per cent, in C. N. at 15.7 and 15.5 per cent and in G. M. at 19.2 and 20 per cent for the low pressure chamber and the rebreathing experiments respectively. The plotted curves (see fig. 1) showing the relationship between low pressure and low oxygen changes likewise indicate that the same cause must be operating in the two methods of experimentation. ee ee ee ee ee 6IT SII 8c1 SCHNEIDER SIT L&T 821 62T SPI iva! LUTZ AND EDWARD C. 9¢1 StI KO) BRENTON R. uoIsudy, OLS SoS O19 oss 8T O19 LI 069 € 91 009 61 1940 -w018q 10 yuoo 10g rat at -90u0 JW a0U0 YW 6 a0u0 YW 61 Vv GI a0u0 VW a0u0 YW G (o-) oynuryA, NOILVUATANNV JO DNINNIDGE ZL L0¥ osvol0Uul 4ue0 10g asTod 18 18 96 88 v8 G6 PS 08 OL v8 ulsog ¢' 79 808 b Of 8 8°19 G66 yc9 | 38 06g G8Z GrSs, Lee 6°19 Org 8°09 | 08 GOL G9E orgg, | SZ 8°02 8&& 9502, S36 T'29 GE 64h | &°9 169 O&€ $09 | ¢°8 1940 uoIsuay, | -wloreq 10 quedo 19g N@ADAXO 'IVNIG UaaNVHO ‘d ‘1 UO ONI -HLVaGUaae 81/t2/F 81/02/F 8I/8I/F 81/6 /F 81/Z1/F 81/6 /F 8I/L / 81/9 /s 8I/Fr /€ 81/¢ /8 81/8 /€ 8I/F /S 81/8 /§ 8t/¢ /s 81/9 -/¢ St/F /8 aLva GAVN eS Eee eee pua ay, qyun hponpoi6 paonpas worsua, Uabhxo 234 ‘syuauiwadxa aquoys ainssaid no) pun Buryznaigas fo uosrundwoy T ATaV iL 235 ' CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS OFT cSt TI LIT SIT ial LET SFT ical LIT CGT 6&T a0u0 4V 9 IT IT 8 v1 6 Or al IT 200 IW 9 a0u0 VW 90U0 VW a0u0 VW a0u0 4V 06 06 99 99 GL 81/62/F S1/82/F 8I/8I/F ST/I1/F 81/S2/F 81/ST/F 81/9T/F 8I/Z1/F SI/T /¢ 81/08/F S1/61/F 8I/L1/F SI/L1/F 81/01/F ST/ST/F 81/01/F 81/22/F 8I/F /¥ 236 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER The difference between the normal and maximum rates expressed in percentage increase also shows that the pulse response took place in about equal degree when equal oxygen tensions were reached. The discrepancies for G. F. H., 1 and 2, and G. W. D. in table 1 are explained by failure to reach the same lower limit in each experiment. The wide difference shown by I. M. and F. L. D. cannot be explained by our data. The parallelism in response is best shown by C. H., 33.7 and 33.3 per cent, C. N., 31.1 and 34.4 per cent, and G. M. with a 36.4 per cent total acceleration in the low pressure and low oxygen by the re- breathing method. The difficulty in reproducing exactly conditions 440 oD | Syste (LP ¢.) cg ea Panne eee ee ee Pie 126 S 1 Systoue (reer) BO S oS | DIASSIAS P/TOLSWD Fig. 1. F. D. Comparison of the rebreathing and low pressure chamber methods. Blood pressures and pulse rate taken every minute. This case illus- trates the close correspondence in pulse response, although the experiments were made on different days. See table 1. in experimentation and securing subjects who are exactly the same physically on two different days makes the parallelism here reported all the more striking. It is evidence for the theory that lack of oxygen, or decrease in oxygen tension, is the cause of the heart rate response under the two very different procedures. Two sets of comparative experiments were conducted on G. F. H., and F. D. (see table 1), one of moderate length for each, and one ex- tending over 85 and 90 minutes. In each set the oxygen and the pres- sure were gradually decreased throughout the period of experimentation. Rebreathing experiments carried on in this laboratory have shown that CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS 2a when the oxygen is lowered rapidly, the subject compensates to a lower percentage than is possible when the rate of decrease in the oxygen is slower. The following cases illustrate the point. G. F. H. in a re- breathing experiment of 24 minutes compensated to 6.3 oxygen, but in one of 85 minutes he reached only 8.5 per cent. FF. D. in 36 minutes compensated to 7.3 per cent, and in 90 minutes to 8 per cent oxygen. Unfortunately the low pressure chamber experiments for each were terminated for other reasons than failure of physiological compensations. The data obtained from seven comparative experiments with the - Dreyer nitrogen dilution method of giving low oxygen and the rebreath- ing method are given in table 2. Unfortunately in this series the attempt was not made to reproduce exactly in rate and low percentage the conditions of low oxygen experienced in rebreathing by the subject. The two examinations were never made on a man during the same day. The air breathed was under normal atmospheric pressure. In both we deal with a gradual decrease in oxygen percentage, that is to say, partial pressure of oxygen. The results show the similarity that was to be expected. The total acceleration and the plotted curves of the gradual increase in pulse rate corresponded in all of the comparisons made. The comparative sets of experiments on R. 8. 8., J. B. H., L. S. L. and P. S. B. show a very satisfactory parallelism. The onset of the acceleration was delayed longer in several of the experiments of this series than in any of the low pressure chamber series. P.S. B. in both the Dreyer and the rebreathing method showed no heart rate response until 11.3 per cent oxygen was reached, approximately 16,000 feet. Changes in the heart rate while a low oxygen level is maintained. The majority of these experiments were conducted in the low pressure chamber. The barometric pressure was lowered to 425, 395 or 380 mm. Hg. (15,000, 17,000 and 18,000 feet) at the rate of 1000 feet per minute and held at that pressure for periods varying from 30 to 130 minutes, the pulse rate being taken every minute during the entire period. We selected these pressures because most men would stand them without discomfort or noticeable loss in efficiency. In fifty cases the average pulse rate at 760 mm. was 74 per minute which ‘points to a lack of anxiety in the subjects. The maximum rate for the men taken to 425 mm. (15,000 feet) was 89, and for 40 men at 395 and 380 mm. it was 94. The increase in rate at 425 mm. Hg. ranged between 5 and 19 beats. The percentage of acceleration ranged between 5.8 and 30. The average percentage acceleration in rate was BRENTON R. LUTZ AND EDWARD C. SCHNEIDER 238 ‘ast oryoAsd Aq pornosqg y 98 8 | 1Z ICG ZOT SL b 6F G9 ¢'0oe | “N Tid \ Se gsa 98 £11 02 0°02 OOT 08 ¢° 8g Be), 0°82 "IqoxyT | 8T/9 /¢ * Gcpy | 82r 88 1°02 £6 G9 “Iqoy | 8I/Fz/S \ Rae ‘a-1‘¥ OIT G FI FI aS FIL 98 z0S 9°9 0°82 | ‘N ‘Ifa | 81/82/¢ ZOl ysl 91 898 FOL 9 SIF Gg OFZe NIG SIEST/Si79 \ it Ee “MEO Cor 8° SI FI gg OIT 8s 9° 1¢ 89 0'8z "Iqey | 8T/9 /¢ £01 GSI bee 9° OF $8 69 Zz 99 9°8 0'8z | ‘N ‘Iida | 81/T12/¢ \ Rt eA: ong T aa 0°61 F 28 +6 89 6°99 (ee) 0° Sz “Iqey | 8I/ST/E +8 0'IT GG 8°02 18 all 8° Lg 92 0°62 9 Se NGG s281/12/S \ eds LR ee nae S01 9ST cl Z'61 £6 82 Z'6¢ Se) 0°82 "Iqoy | 8T/L1/¢ StI C61 if ere +6 02 1°69 16 8°62 “Iqey | SI/Fz/S \ eee gg -y eou0 yy | FS 96 a 1°69 16 O62 |) SNUG: |SS0/12/s 98 11 02 0 IF OI SL 9° SF 09 o'og | ‘N ‘Ilda | 8I/It/¢ \ Ae MN ae WT LET 0ST 9 0'0¢ LIT 82 82g 92 G8 “Iqey | SI/Ez/E ae) ‘O oynUryAy eae pug ulsog | worsuay, | 7000 10d | ga toNniw | NEDOULIN uorsuay, | 3ue0 Jeg NOILVUAIANV JO DNINNIOGH as tod N@DAXO IVNIA NI aqgLoatid ONI aLvya GWVN HLONGT -HLVaUday a ee a ee re SS SS ee pua ay jyun fyonposb poonpat Worsua} uabhxQ ‘syuauriadxa ua 6 ATAVL Bonu pamnyyp pun Buryjna.ga. ay) fo uosrundutoy) CIRCULATORY RESPONSES TO LOW OXYGEN TENSIONS 239 18.7. The increase in rate at 380 mm. (18,000 feet) ranged between 6 and 45 beats. The percentage augmentation varied between 6.7 and 59, with an average increase in rate of 26.3 per cent above the normal at 760 mm. The maximum pulse rate usually did not occur simultaneously with the arrival of the desired altitude. The lag was somewhat longer in the men taken to 380 mm., than in those at the lower altitude. Intwo cases in which the pressure was only 425 mm., the maximum rate was attained during the ascent. In the other men held at this level the lag varied between 2 and 7 minutes. The average lag at this pressure %, PULSE DIASTOLIC PRESSURE & a s x 6 e 4 Y pase ES 7 70 cz 20 25 Jo 35 40 45 Fig. 2. K. D. Taken by the rebreathing method to 12.5 per cent oxygen (13800 feet) in 18 minutes and maintained at that level. Blood pressures and pulse rate taken every minute. This chart illustrates the return of the circula- tory factors, while the low oxygen is being maintained, toward their original values. See table 5. was 2.8 minutes. In all of the experiments in which a pressure of 380 mm. was reached and then maintained the average delay in the ap- pearance of the maximum pulse rate was 6.6 per minutes. Every man showed some lag. In two the delay was only for 1 minute, but in several it was as long as 20 minutes, and in one the heart rate continued to increase gradually for 26 minutes, accelerating during this period from a rate of 90 to one of 108 beats per minute. We also carried nine men by the Dreyer method to 10 per cent oxygen (19,400 feet) in 20 minutes and then held them at that level. In these cases a definite lag in the pulse rate in reaching its maximum 240 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER was observed. The shortest time taken to reach the maximum after the oxygen level was established was 3 minutes, the longest 38 minutes. The average lag was 14 minutes. These results compared with those obtained in the low pressure chamber seem to indicate that the lower the partial pressure in the air respired the slower will be the pulse in reaching its maximum rate. Observations on the development of cyanosis are suggestive of the fact that the available oxygen is gradually decreasing within the blood during this period in pulse lag. Some cyanosis has been observed during the holding period for each level studied, but it was most con- spicuous in the low pressure chamber at 380 mm. Hg. (18,000 feet). The cyanosis comes on slowly and, like the pulse rate, requires some minutes to reach its greatest degree. In tables 3 and 4 have been tabulated the circulatory data obtained in forty experiments in the low pressure chamber in which the pressures were gradually reduced and then maintained at some time at 425, 395 and 380 mm. Hg. These data show that in many men the pulse rate does not maintain its maximum during the holding period at low pres- sure. Corbett and Bazett (13) on subjecting men to a constant per cent of low oxygen, observed for the pulse rate that ‘‘as adaptation takes place it tends to fall to a slightly lower level.” We find three types of pulse rate reaction during the holding period. These are: a, a definite and gradual decrease in rate after a brief period of maintained maximum; b, maintenance of the maximum rate; and c, a steady but slow rise in rate throughout the entire holding period. Our forty are distributed as follows: Type a, 29; b, 9; and c, 2. The amount of fall that occurs with adaptation is an individual matter. In some it is slight, only two to four beats, but in others the rate may return very nearly to normal. Other methods of subjecting to low oxygen gave similar results. During the holding period at 10 per cent oxygen with the Dreyer method nine men (see table 5) reacted as follows: five had a gradual retardation after maintaining the maximum pu se rate for a short time, three maintained the maximum rate and one gave a steady rise to the end. In one long experiment with the rebreathing antannas after the oxygen had been reduced to 13 per cent (12,400 feet) in 18 minutes, this level was maintained for 60 minutes by admitting pure oxygen into the reservoir. The pulse rate accelerated to 115 beats per minute from a normal of 92 when the low level of oxygen was reached. The rate retarded gradually to 92 during the following 80 minutes. We shall ey 99 |Z [02 92 FOT/90T/ZTT | 801) |78 [8 [8 [98 [62 |92 [Pz Itz | GI/9 /¢ -|---- °° SC Ma x FS [9¢ |g¢ |F9 89 9IT/9TT/OZT|ZaT SIT) |26 {$6 |86 |ZOTIF6 |68 |F8 lP8 | 6I/oe/p |--°-: NOLS 8¢ 8¢ |z9 G9 /9GI} | FET|S2T SOT} |S0T/SOT/OOT/86 |F8 Jos joz joz | st/6z/p |g og 96 JOOT/OOT/OTT|O6 |8Z [69 |69 | 8I/2e/a |e Ia n 9 |99 |09 z9 G9 |ZOT/86 |Z6 COT/OOTCOT/FOTIES |Z8 /FS |98 |06 {88 [G8 jes |F8 | 8I/¢ /g fro TM a a 09 |99 89 FIT|PIT 9ZT 08 /92 [28 |89 [29 [29 [09 | 8I/I8/2 | -z Wa a 8S [29 |Z4 02 |z9 89 96 |SOT/SOT/SOT/ZIT) —|Z11 06 |48 /98 |98 |T8 |T8 j82 | 8I/og/2 |: € ‘Hav a FL |0L OL FOL|FOL OL 86 |96 |96 /82 |€24 [02 |99 | 8I/9%/2 | Beater & FL |9L |82 82 zs OOT/OOT/ZOT| —00T VIT| 66 |96 [26 |96 [86 /86 [€6 [06 | 8I/6T/Z |-°°--:* ETO) 4 Z 09 |F9 [FF cg Z COT/FOT/POT|COT/FOT/90T| |FOT] {06 |z8 J98 |ss lez |zz \69 [99 | st/or/2 Ye 2 Moe c ol OL OL SOT/ZOL/POTIOOT]86 [86 lOOT|90T} {zz ez |T8 |sz zz Iez \69 [69 | St/et/, [ess Guinea. Vy 9 GL |89 [ZL [92 CL SIT/OTTSITFET SIT St Toztjozt} 9g |ss |z6 |o6 Iss loz lez Iez | st/et/, [oc pcs en At | © #9 |F9 |0Z [89 Ig9 02 /OTT/FITOTTOTTOZTFIT|SOTOTLOLE|F6 jg6 /96 86 loot\z6 los Izg {gy BIOL 2 P= sey sn AR E 82 |0L |89 |F9 06 GIT|ZIT|SIT|SI1 OTT) 8 [28 /68 |68 |T8 /62 [82 |e2 | S8I/6 /2 | ggg 53 99 OL 56 VOI! |P8 |F8 [28 |88 |28 [eZ Ie2 |e2 | SI/8/z2 foc ECE NT a ° 8h |8F £9 PITZI1 SOI G6 [26 JOOTIE6 [68 /€8 JOS | S8I/¢ /2 fo Wwe 2 S 82 08 PITOTTOTTSTTPITITZIIFIT| 88 Jo6 [28 |28 loz Izz \zz leo | st/e /2 Pee a a G¢ |09 OL - OST/9ET|FZT|FZT/9ZT|OST OTIPITIECTIOOLIPS (82 \82 | 8I/2 /2 foc faq 8 ie G8 |98 /98 /86 |06 {28 [82 [92 | 8I/T /2 Joo IT ig S OL |82-|F9 |29 Itz PL JOPL/OST/OST|FET/FFI OcT| P78 /e8 |98 |88 [Ps 18 Jos jez | st/og/9 fg Meg 8 eB 8h |8P |8¢ ito | {09 PIL|STLOZTOZT OOT) {28 {82 [82 |82 02 |F9 [29 [09 | Bt/ez/9 | MOD 5 fs 9F |9F |8F |zg 89 GOL|ZOT|ZOT|ZOT GOT] 88/28 |06 |68 84 [OL |69 89 | St/zt/9 [oT Ig E re 09 |9¢ |F9 /8z 92 STTOTTOITOZTFer/eet/OcT|SIt] | tor/ToTloot|s6 |g6 96 |e6 leo | gt/st/9 he SEAN ee 5 PS |9G |8¢ [zz z8 SIT|GIT|SIT|SIT SIT 84/82 |8Z |89 |29 /29 |29 | SI/zt/9 | TW aM 3 a 9 |29 |89 |99 OL |GTI/ZILOLLZT1 211 SITIZ6 /68 |98 |88 06 |Z8 |84 FZ [ZZ | SI/II/9 | GOWN 4 S GP |Gh ISb Ifo iF9 OL OOT|FOT|OLLFIL|FIT SIT) [88 |S6 |06 jP6 j08 /924 |b2 [02 | St/oI/9 | gg ay 8 2 FS |8¢ |s¢ jog cL |0L OOT|FOT|ZOT|OOT GITPIT) | 04 |92 64 |28 [62 |Gz Jez lez | 8/2 /9 |--- ‘fH ON 4 . $6 | 92 | 99] 98} ee] er] or}¢ |o $6 | $2 | 99) ce] ez} st} or!¢ |o | ¢6 Sh) S95), 98) Se. et | Or | 9.0 aI == 2 [a a ** oynuUlyl SOE O8é S6E OF OS? 09¢ 9¢9 092 cues“ Sa LA WOU Suazy) Wi auinjod fisojourd say ‘Play Pup sani gy wr (mu Ose) 12af O00'8T 07 Wayn7 spoalgngy ‘aajawouds WasanTT ay) Yyun LAQUuenyd ainssald NO] ay) UL SpuaUMrsad xy $ WIAVL 294 ALVEOLAR AIR AND RESPIRATORY VOLUME AT LOW OXYGEN 295 fore investigated. The per-minute volume of breathing was measured with a Larsen spirometer in men reduced to pressures of 395 mm., 380 mm., 365 mm. (17,000, 18,000 and 19,000 feet respectively). The data are shown in table 5. The subjects all showed a considerable increase in breathing varying between 1.8 and 9.3 liters, or 34 and 103 per cent. The average amount breathed per minute at 760 mm. was 7.49 liters. Just as soon as the reduction started the average figure went to 7.94 liters, due no doubt to anxiety of some of the subjects. By the third minute it had fallen to 7.61 liters. The readings thereafter showed a progressive increase until at the 19th minute the average figure was 11.59 liters, an increase of 54.7 per cent. It will be seen both from the average figures and from the individual cases that the onset of increased breathing started usually between the fourth and sixth minutes or between 656 and 605 mm., that is, 4000 and 6000 feet. This confirms the alveolar air findings reported above. The onset of increased breathing due to low oxygen produced by the rebreathing method was reported by Schneider (3) to begin in some cases as early as 16 per cent oxygen, corresponding to about 7000 feet or 580 mm., and by Ellis (23) before 17.5 per cent oxygen was reached. Schneider (3) found in the rebreathing test that the rate remained unchanged for many men but the majority increased the rate by 2 to 4 breaths per minute. The depth of breathing he found increased from 20 to 128 per cent when at 8.5 to 6 per cent oxygen. The respiratory volume during maintained low barometric pressure. In the majority of alveolar air determinations the lowest carbon dioxide figure was found shortly after the reduced barometric pressure was attained. Thereafter the carbon dioxide level either rose slowly or was maintained. Experiments in which the volume per minute of breath- ing was measured during a reduction of pressure to 380 mm. at the usual rate and during the following 48 to 84 minutes of the maintained low barometric pressure are tabulated in table 6. Control experiments at 760 mm. using the mask and meter are also shown. The figures given are the three-minute averages in liters. All of the eleven subjects showed an increase in lung ventilation usually most marked within 10 minutes after 380 mm. was reached. Seven showed a reduction of ventilation thereafter continuing until the end of the experiment. Two showed a reduction followed by a terminal rise, which in the case of A. F. H. was very marked. Two cases showed a very slow rise which tended to be maintained until the end. The usual type of response is shown in figure 6, in which the cases of I. M., W. H. G., 296 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER P.S. B. and E. A. R. are plotted. In three subjects taken to 380 mm. at the same rate and held from 59 to 81 minutes the Larsen spirometer was used. Two showed this type of response and in one the increase in respiration continued until the end. : The typical response seen in nine cases out of fourteen in which the volume of breathing was measured, corresponds to that observed in six of the nine cases in which the alveolar air tensions were followed under similar conditions. In one case the subject was taken to 12.5 per cent oxygen in 17 minutes by the rebreathing method and held at = % Su sy =| wiS Olsen DAROwETER. |. 2 Hh Se soe. el eee 10> | q ty - eS 7? 5B. VOLUME qj 8 T BAR VOLVME at Ne WHO G VOLUME fee We i SLM. Volume Siig Lis / A A La\P.$.B, \RATE 4 \ 4 em W GRATER -~A oc v Senta \ / 3 ei 166 fom V\ A / i x A Ue z [2 * Bact ae \ Be 1 v-~ x iy VN ued, y s | Oe +4 7 MINOT. Psa Ld Ww Se ee a a ee ae an a 6 5 G ¢ 12 15 18 2124 2730 33 56 3942 45-46 51 54 51 Go U3 6b G4 7275 18 GL 84 OT y G3 4 49 d02 Fig. 6. The respiratory per-minute volume of four cases in liters, taken to 380 mm. (18,000 feet) in 18 minutes and maintained at that level. Note the decrease in ventilation after the maximum is reached. This corresponds to the alveolar CO, tensions under similar conditions shown in figure 5. that level for 68 minutes. The response was similar to the typical low pressure chamber experiment. His per-minute ventilation increased from 8.3 liters to 10.4 liters at the 14th minute. It held a level at about 9.2 liters from that time until the 40th minute, when it gradually fell to 8.4 liters at the 84th minute. Several subjects taken to 10 per cent oxygen in from 17 to 20 minutes by the Dreyer method showed similar responses. In these cases the ventilation was indicated by a Fitz pneumograph and the average amplitude times the rate per minute was taken as a figure to indicate the per-minute ventilation. I. M. increased from 90 to 252 at the 17th Td j2°9/9°9 |Z'9 |8°9 |F'9 |2°9 |F'9 IT OF ZI¢-912 9129102 |o-z 9°L16'9 (69/9 9IF "9102 09/6 S16 OT 9IT 91° 9/99" °° “ST/9T/ 2 “AO CM 11 (6610'S [28 |2'8 |Z°2 |2°8 |L'8 [2's 82/6 2/28/16 0 OT/8 8166 8°8/9 6/2 S/T “OL/8 8/8 GE 8/8 "SiS LO LIS 2)" “ST/e1/2 “wv cm I 61/8 61/S 9T/S STIS STF IT/0'6/2'8le'8!9 2182/2 OT|0'6 $°6/6°8 |G 6/2 8/E'6)2'6 |S SL LIF LIS 216-219 2182 ee SH ST 28° Ea eV 6°9)2°9/0 '9/F 9 |6°S/8°9 |9°9 [89 |9°9 |0°z |¢°9 |¢-9/8 912 z/8'918:919°9 94 |O' L192 |]°2/8' LIS 9IT'8 0°24/6°9/9'9/T 2/8 °919°9/2°S]"* “ST/O1/2 “D HM 4°89 8/96 |2'SIF'S8 |€"Z |E'8 |8Z |¢°8 |9°8 l9'sl6 sle‘sie'si6 ‘slo's 0'8 |9'8S/F'8 |T'8/I'8ig"8i0'6 TSiS'2/6 216 OT 2/89/29) 81/6 /2 “a “Sa (0°96 F8'S jE OTL \€'2 |8"2 188 I¢"2 |O'z Ie-z16 zit’ sis ‘sie‘sit6 68 /6 6G OL 6/0 6/F 6/16 |F Siz 916 FF SIS 9129 2 eS Se) 2a Ea y 9°9 |9°9 |F'9 /8°9/0'S/6'S/8'S/9 Gie'9 16'9 lo-clb'9 IT 9lz 910-9129 SZ S6'SI9 FE 'SIL SIP Sl" -ST/¢ /2 ‘We EL LIE L102 |TLI0'S |€'2 |9'2 [e°2 lez eZ 6°98 IE LIF 216 °216'9 8°24 |F LZ|P'S [2° ZIT 8/9 ZIe'6 SL8 LIE LE SL z/o° 2 28 S22 ue sa 68 |€'8 |L'8 |S 8/7 6/6 8/8 ZIT 8/26 |2°8 |8°SI6'8 IT 616 Sir 6186 6 8/9'8/T 86 29 2/2 SiG Sl St/e |Z a a 46/4 6\F 6/9'6 |6'8/Z'6 |0°6 |Z'8 |8°6 |9'8 IT'S 68/6 8/2 2/9 SIT 8/0'6 IF'8 162126 9° 8/€°60 8/0 T1/% 6/9 Sie Sig 84° S/S GTS] °° “ST/08/9 “aM ‘a i LGiS°9 16'S 12°9 |2°9 |@°9 |2'¢ 1989991 'S/4°916'9 69 /SZIT'L 16 LIE 216 OF 2 DLP OS 'Sh SIT Sh PGS): ST/T [2 WIS PIEY PUD sojnunu gy] ur (“mau OBE) 12af OOO'ST 07 way) spoalany “saquipyo aunssaud mop ay) ur SpuaUlrisad xy 43/6 8/9 80 OT/8 6/G'S |¢°8 |F'8 ies le's |z°8 €°8/8 8/2 ST 6S SIPS |Z°6 |F'8|0'6 |2°8/8'S|0 6/28 0°6/& 6/3 '6/Z 6/6 6/26 SS -SU ie ot oe 6°9/9°9/0° L1G "212°S16'9 |F'9 |G LIF'9 |2°9/0 OIF 9/89 6°9/2°S/6'9/Z 8/9 L1G 9 STON ee OH Ak L'Y |O'S/F FP FZ FL 48°F ITF 16 Fe'F [8 FO FS Fe'g G FO G/0'S|9 FZ GI8'¢ 1" “S1/ 81/9 gobucha,®) 06 | 48) 78 | 18 | 82) o2 | ce | 69 | 99 | 9 | 09 | zg] 49} 19] earl or| ce | oe | oe] ee | og 2cl ae! tz BE Tle tal Ge 60) tee lino) WU 094 1D SJOujUoy ‘anuru wad siny ur uoynudsoy “burrs ootang -aaqow pup yspu sob yjzun syuawrsodag 9 HIAV4L 298 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER minute and then fell to 154 at the 90th minute. This is a picture similar to that of I. M. in table 6 and figure 6. C. L. 8S. in a similar experiment went from 192 to 360 at the 17th minute and then fell to 128 at the 95th minute. The rate of breathing was reported in seven cases during the reduc- tion of pressure and the holding period in the low pressure chamber. In four eases it fell from two to five breaths per minute as the experi- ment proceeded. The rates of W. H. G. and P. 8S. B. are plotted in figure 6. One showed no change in rate. Two showed an increase, one, N. E. B., from 13 per minute to 15 at 380 mm. and then to 19 at the 82nd minute when the low pressure was maintained. The other, A. F. H., showed no increase until the 58th minute when the rate per minute started to rise from 17 to 38 at the 77th minute. The per- minute volume of breathing increased markedly, as will be seen in table 6. The tidal air has never been observed to decrease in the low pressure chamber. The majority of subjects responded to the low oxygen exposures by deep slow breathing although frequently Cheyne-Stokes breathing has been observed. DISCUSSION The relation of respiration to low oxygen tension presented in this paper is, in a general way, in accord with most of the literature. The early response to decreased oxygen tension and the tendency of the breathing to return toward the normal during maintained low oxygen which we find under the conditions of our experiments, may appear at first sight to be contrary to the views which have been presented by Haldane and others. Haldane and Smith (24) in 1893 found marked hyperpnoea when the oxygen was reduced to 12 per cent, the carbon dioxide being removed. They write the fact that any hyperpnoea should have been caused by a reduction of oxygen to 12 per cent may seem at first sight to be hardly consistent with our former conclusions that hyperpnoea caused by vitiated air is entirely due to carbon dioxide. They explain that in the former carbon dioxide and low oxygen experiments the increased supply of oxygen brought about by the carbon dioxide hyperpnoea prevented an extra hyperpnoea due to want of oxygen from developing. Haldane and Poulton (2) in 1908 reported experiments in which the subjects reduced 25 liters of air from 9 ALVEOLAR AIR AND RESPIRATORY VOLUME AT LOW OXYGEN 299 or 10 per cent oxygen to 4 or 5 per cent in less than 10 minutes. They found marked hyperpnoea which they believed was not due to the direct effect of oxygen want but to lowering of the threshold of the respiratory center to carbon dioxide which has not had time to escape. In their experiments, however, the alveolar carbon dioxide fell to between 3.2 and 4 per cent. In another group of experiments the oxygen per cent in the inspired air was reduced to about 9 per cent in from 15 to 23 minutes. In these ‘no noticeable hyperpnoea”’ is reported, although the alveolar carbon dioxide fell to between 3.9 and 4.3 per cent, which indicates that a considerable increase in ventilation must have occurred. Haldane, Meakins and Priestley (25) in 1919 _ conclude, from exposures to low oxygen of about 10 per cent, lasting about 6 minutes, that the first result of diminution in the percentage of oxygen is an increase in the depth of respiration owing to a lowering of the threshold-exciting value of carbon dioxide. This is followed by a period of periodic breathing due to the much quicker action of want of oxygen as compared with that of increase of carbon dioxide. Further reduction of the oxygen percentage showed the periodicity replaced by a very rapid shallow breathing. They write, ‘‘Want of oxygen in the inspired air causes shallow breathing which in turn in- tensifies the anoxhemia.’’ The point of view taken by these authors is that after the first period during which the threshold is lowered to carbon dioxide, oxygen want acts as a paralyzing agent on the respi- ratory center. We believe that one is not justified in drawing too general conclusions regarding the effects of oxygen want from experiments of extreme degree and short duration. We shall show in a later paper the quick respiratory and circulatory responses to the breathing of pure nitrogen. At the other extreme is the well-known ascent of the Duke of the Abruzzi in the Himalayas to 24,580 feet. In quick extreme anoxhemia, respi- ratory and circulatory factors respond quickly and to their greatest capacity. If the exposure to low oxygen is slow and long-continued, other factors have time to assist in the compensation. We feel, there- fore, that the rate of exposure as well as the degree is an important condition when considering the effects of oxygen want. The lack of recognition of this fact brings about confusion. We have never seen a case of shallow breathing and only two cases of increased rate under the conditions of our experiments. They are, however, quite different from those of Haldane, Meakins and Priestley. The decrease in the respiratory per-minute volume, which occurs in our experiments after THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 300 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER the preliminary increase with the reduction of pressure, takes place during exposures of from 30 to 120 minutes rather than during exposures of from 6 to 10 minutes. We do not believe it to be a sign of failing respiratory center, but an indication of improvement in condition, as will be pointed out in another paper. SUMMARY 1. Twenty-four men were taken to 352 mm. pressure in a low pressure chamber at a rate equivalent to an ascent of 1000 feet per minute. In these cases the average alveolar oxygen tension fell 66 per cent, and the alveolar carbon dioxide fell 24 per cent. 2. The average carbon dioxide tension was definitely lowered at 656 mm. (4000 feet) which indicates that the onset of increased breathing had occurred. 3. Alveolar tensions taken during a reduction of pressure to 380 mm. (18,000 feet) at the usual rate, and during the subsequent 30 to 120 minutes while the low pressure level was maintained, showed that after the preliminary fall in carbon dioxide tension there was a tendency for this tension to rise for a time although it remained low during the holding period. After 760 mm. had been reached again within 20 minutes, the carbon dioxide had not recovered its former level in the majority of cases. 4. The lowest carbon dioxide tension occurred about 5 minutes after 380 mm. was reached, when the reduction was equivalent to an ascent of 1000 feet per minute. In some cases this latent period did not occur and the maximum breathing was coincident with the arrival at 380 mm. 5. Tensions taken while the pressure was maintained at 428 mm. (15,000 feet) did not show the same profound effects. The carbon dioxide tension did not fall so far and maintained a level. 6. Both oxygen and carbon dioxide alveolar tensions responded quickly to rapid successive reductions of barometric pressure to 428 mm. 7. The per-minute volume of breathing was determined for each minute during a reduction of pressure at the usual rate to 395, 380 and 365 mm. The majority of cases showed a definite increase in venti- lation taking place between 656 and 605 mm. (4000 and 6000 feet). This final increase amounted to an average of 54.7 per cent. Individual cases varied from 34 to 103 per cent increase. 8. The per-minute volume of breathing was determined during a reduction of pressure to 380 mm. at the usual rate, and during a period ALVEOLAR AIR AND RESPIRATORY VOLUME AT LOW OXYGEN 301 of from 48 to 84 minutes while the low level was maintained. In nine out of fourteen cases the maximum ventilation occurred within 10 minutes after 380 mm. was reached. Following this period there was a distinct falling off in the per-minute volume. These cases correspond to the six cases out of nine in which the alveolar carbon dioxide showed arise after the preliminary fall. 9. The decrease in the per-minute volume of breathing after the first maximum value, as 380 mm. was reached, was also found in cases in which the low oxygen tension was produced by the rebreathing method and by the Dreyer nitrogen dilution method. 10. The partial return of the respiration toward the normal is believed to indicate a temporary improvement in condition. 11. Alveolar air tensions taken during a reduction of the oxygen partial pressure by the rebreathing method or the nitrogen dilution method corresponded to those taken under reduced barometric pressure. BIBLIOGRAPHY (1) Greee, Lurz, anp ScuneipEr: This Journal, 1919, 1, 216. (2) HaLpANE AND Poutron: Journ. Physiol., 1908, xxxvii, 390. (3) ScHNEIDER: Journ. Amer. Med. Assoc., 1918, Ixxi, 1382. ~ (4) Lurz: Proc. Amer. Physiol. Soc., This Journal, 1919, xlix, 119. (5) Lozrwy: Untersuchungen iiber die Respiration, u. s. w., 1915, Berlin. (6) Lozwy, Lorwy anv Zuntz: Pfliiger’s Arch., 1897, lxvi, 486. (7) Boycorr anp Haupane: Journ. Physiol., 1908, xxxvii, 355. (8) HALDANE AND PriestLey: Journ. Physiol., 1905, xxxii, 225. (9) Warp: Journ. Physiol., 1908, xxxvii, 378. (10) Dovuauas, HaLpANn, HENDERSON AND SCHNEIDER: Phil. Trans., Royal Soe. London, 1913, B, ceiii, 271. (11) FirzGrraup: Phil. Trans., Royal Soc. London, 1913, B, cciii, 319. (12) HasseLBatcu aNp LinpHarp: Biochem. Zeitschr., 1915, Ixviii, 265. (13) Mosso: Life of man on the high Alps, London, 1898, 1. (14) Manual of the Medical Research Laboratory, War Department, Air Service, 1918, 212. (15) Henprerson anp Morariss: Journ. Biol. Chem., 1917, xxxi, 217. (16) Henprerson: Journ. Biol. Chem., 1918, xxxiii, 31. (17) Pearce: This Journal, 1917, xliii, 73. (18) Pearce: This Journal, 1917, xliv, 369. (19) Pearce: This Journal, 1917, xliv, 391. (20) Larsen: To be published. (21) Dreyer: Repts. Air Medical Investigation Committee, England, 1918, no. 2, 8. (22) ScuneipER: This Journal, 1913, xxxii, 295. (23) Exuis: Proc. Amer. Physiol. Soc., This Journal, 1919, xlix, 119. (24) HaLDANE AND SmituH: Journ. Pathol. and Bacteriol., 1893, i, 168. (25) Hatpanr, MEAKINS AND PrigesTLEY: Journ. Physiol., 1919, lii, 420. COMPENSATORY REACTIONS TO LOW OXYGEN HAROLD W. GREGG, BRENTON R. LUTZ ann EDWARD C. SCHNEIDER From the Medical Research Laboratory of the Air Service, Mineola, New York Received for publication August 27, 1919 In earlier papers we have dealt separately with the blood, circulatory and respiratory changes induced by short periods of exposure to lowered oxygen tensions. It was shown that men responded with definite adaptive physiological changes when subjected to gradually decreasing oxygen partial pressures which reached values between 76 and 51 mm. Hg., corresponding to barometric pressures of from 360 to 240 mm. (19,200 to 29,000 feet), and also when kept for from 30 to 130 minutes at oxygen partial pressures of from 88 to 80 mm., corresponding to barometric pressures of from 425 to 380 mm. (15,000 to 18,000 feet). In approximately 78 per cent of all men examined the erythrocytes and hemoglobin increased in a unit volume of blood. This increase did not occur immediately but usually required between 40 and 60 minutes to become definite. About 13 per cent of all cases showed a well-defined increase in hemoglobin within 26 minutes (1). The heart responded to slight changes in oxygen tension by an accel- eration in the rate of beat. Some men gave the first response at an oxygen partial pressure of 137 mm., barometric pressure 656 mm. (4000 feet) ; but in the majority the acceleration began between oxygen partial pressures of from 113 to 128 mm., barometric pressures of from 610 to 542 mm. (6000 to 8800 feet). Evidence of an increased rate of blood flow was found in the acceleration of the heart rate, and in a fall in the diastolic blood pressure which resulted in an augmented pulse pressure. When a constant level of oxygen was maintained, the heart reached the maximum rate after a lapse of a period of variable length. It continued at the maximum rate for some time after which the rate retarded some- what. The evidence indicated that a marked and progressive increase in the rate of blood flow occurred during the reduction and early holding period, after which there followed a period of more or less constant rate of flow. Later in many subjects, as shown by the heart retardation and the rise in the diastolic pressure, the flow of blood in some degree approached the normal rate (2). 302 COMPENSATORY REACTIONS TO LOW OXYGEN 303 The per-minute volume of breathing showed a definite increase between 656 and 605 mm. (4000 and 6000 feet). In the majority of cases the maximum ventilation occurred within 10 minutes after 380 mm. was reached. Following this period there was a distinct falling off in the per-minute volume (8). In the present paper we propose to consider the relative values of the compensatory reactions to low oxygen tensions. Men differ in sensitiveness to lowered oxygen and in the power to make physiological adaptations which will, from a decreased supply, provide sufficient oxygen to maintain tissue and body efficiency. In some there is an immediate or at least an early response to a decrease in oxygen, in others the response occurs much later and may be less adequate. Some men make excellent compensations to low oxygen tensions while others show insufficient compensations at only moderately low oxygen. Indi- viduals differ also in the use of the several ways of responding to the decrease in oxygen. The majority of men appear to make a well- balanced use of the three mechanisms for supplying oxygen. The ventilation of the lungs, the rate of blood flow and the percentage of red corpuscles and hemoglobin are definitely increased. Some meet the new condition largely by increased respiration and others depend almost entirely upon an increased blood flow. In many individuals, during the early period of exposure to a decreasing oxygen, the burden of compensation is borne wholly by the circulatory and respiratory mechanisms, but later the blood changes relieve one or both of these mechanisms from a part of the burden. Our data show an interdepend- ence and an interplay of the adaptive mechanisms when a subject is held under a constant low oxygen tension. Schneider (4) has reported . briefly several cases in which the interplay was present. The majority of the experiments which have been presented in part in our earlier papers were conducted in the low pressure chamber. The barometric pressure was lowered to 425, 395 or 380 mm. (15,000, 17,000 or 18,000 feet) at the rate of 1000 feet per minute, and held at that pressure for periods varying from 30 to 130 minutes. In a smaller number of experiments the subject breathed atmospheric air diluted with nitrogen by the Dreyer method. Starting with undiluted air, 20.96 per cent oxygen, the nitrogen was added gradually in greater and greater proportion, so that at the end of 20 minutes the mixture contained only 10 per cent oxygen. This percentage of oxygen was then maintained for from 30 to 90 minutes. Thus the subject was kept under low oxygen for a period of from 50 to 112 minutes. 304 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER In the low pressure experiments the observers were given oxygen by means of a tube held in the mouth. It was, therefore, necessary to determine whether oxygen accumulated within the chamber during the period of experimentation. In the majority of experiments samples of air were taken 3 to 5 times during the experiment, and later analyzed for oxygen and carbon dioxide. The exhaustion pump was kept work- ing continuously throughout an experiment so that sufficient ventilation was maintained to prevent an accumulation of carbon dioxide. Often there was some accumulation of oxygen, but it was found to reach quickly a constant level. With such data a corresponding correction for altitude was sometimes made. We have, however, many experi- ments in which no accumulation occurred, and we have usually omitted the correction when the accumulation was slight and the oxygen per- centage remained constant during the holding period. We are satisfied that the interpretation of our data is not vitiated by this accumulation. As shown in our earlier papers, the effects upon the blood, circulation and respiration were the same under the three methods used for provid- ing low oxygen tensions. Since this was found to be the case, we have demanded only a constant oxygen tension during the holding period. THE LOW PRESSURE CHAMBER EXPERIMENTS The interplay of the three adaptive responses has been studied in forty-seven experiments. For convenience of discussion we have divided the reactions observed during the period at which the barometric pressure remained constant into four groups: a, Cases in which the pulse retarded after maintaining a high rate for a period of variable length, and the hemoglobin percentage of the blood increased; b, cases in which the pulse maintained the new level after an increase in rate, and the percentage of hemoglobin increased; c, cases in which the pulse rate remained constant and the hemoglobin did not increase; d, a few cases in which the pulse rate retarded and the hemoglobin did not increase. The variations in respiration have been determined for each of the zroups. : a. Retardation of the pulse rate during the holding period with an in- crease in hemoglobin. There were twenty-six cases in which an increase of hemoglobin seemed to favor the heart and sometimes the respiration. The beneficial cardiac effect, as we interpret the data, was manifested by a slowing of the pulse rate and frequently by a decrease in the blood COMPENSATORY REACTIONS TO LOW OXYGEN 305 flow shown by a rise in the diastolic pressure and a corresponding de- crease in the pulse pressure. No two cases were exactly the same. The interdependence of the three compensatory responses can best be shown by a detailed study of a few individual cases. N. E. F. June 7, 1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 83 misGrece se. -.3-..2| to 75 | 75 79 82 79 76 72 DWSvOllehaasn as -a5s =| LILA 112 100 | 102 104 | 100 BRON Cr <0. 5-22) 00 72 56 58 58 54 Pulse pressure...... 44 40 AE 44 46 46 Auveolar Op: ..-..2. 106 65.2 | 41.5 | 37.4) 37.0! 33.6) 32.3 Alveolar COz.......| 39.1 35.8 | 30.0 | 28.7) 32.0) 34.4) 33.8 Hemoglobin........| 94 96 96 97 100 This subject was in good condition up to the 85th minute when blood was drawn from a vein. It will be observed that during the period of ascent the pulse rate accelerated and the alveolar carbon dioxide tension fell. Thus the burden of compensation to decreasing oxygen was at first borne by the circulation and respiration. The pulse rate reached its maximum four minutes after the barometric pressure of 380 mm. was attained. About this time, the 25th minute, the blood flow as judged from the pulse rate and pulse pressure reached its maximum. Coin- cident with this the per-minute ventilation of the lungs was greatest as indicated by the carbon dioxide which at this time was only 28.7 mm. The circulatory conditions remained about the same during the next 10 minutes, but the breathing, as Judged from the carbon dioxide, was lessened. Both the circulation and the ventilation of the lungs fell off from this time up to about the 60th minute, after which they remained constant to the end of the experiment. It should be noted that the hemoglobin had begun to increase at the 25th minute and continued until the close of the experiment. Coincident with this increase in hemoglobin there was a retardation in the pulse rate and a decrease in the breathing. The interplay of compensatory factors for this case is shown graphically in figure 1. 306 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER SYSTOLIC 100 BAROMETER 360%.K. go é 60 y s N ~ ~~ 7? Us ibs DASTOLIC 2 fey 60 Le Pa HAEMOGLOBIN — — 17 aE Fahl Lite 9S 60 | / ALVEOLAR O~ 40 MLVEOLAR Coz 30 TIME 1h MINUTES 20 0 10 20 Jo 40 Jo 60 Jo i) Fig. 1. N. E. F., June 7, 1918. Taken to 380 mm. (18,000 feet) in 18 minutes in the low pressure chamber and maintained at that level. This case illustrates the inter-relation of pulse rate, respiration and oxygen-carrying-capacity of the blood. R. S. S. June 6, 1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 45 75 Pulses: no. nelson vad-beace eae tOc 74 | 76. |necSO |194: tal e900 ill CCE OAV Gian on55 bn o9 ooleaeae a|| dlls 114 | 114 | 110 | 106 | 100 Diastolics epee ee. eal 64 | 54 48 44 42 Pulse pressure............| 48 50 | 60 62 62 58 AlveolatiOsspaeeen eee soles BYU apiey| yA By AlveolatiCOrnesma neces 38.8 32.7 Sold] o2edleeotes Hemoglobin... 2.2205. 2.2 2/102 104 | 104 | 106 | 106 COMPENSATORY REACTIONS TO LOW OXYGEN 307 In this subject there was a progressive increase in the rate of blood flow, as shown by the pulse rate and pulse pressure, which reached the maximum at the 45th minute. The pulse rate reached its maximum 10 minutes earlier. We believe that this illustrates that the pulse rate alone did not determine the maximum compensation in circulation. The fall in diastolic pressure with the resulting increase in pulse pressure is considered evidence of vasodilatation in the systemic circulation. Judging by the decrease in pulse rate and in pulse pressure, the rate of blood flow began to slow at about the 58th minute. The respiration attained its maximum soon after a pressure of 880 mm. was reached, and then maintained a fairly constant per-minute volume of ventilation until the end of the experiment. The hemoglobin showed a slight increase at the 25th minute and reached its maximum concentration at about the 50th minute. In this experiment the circulation seems to have been favored by the concentration in hemoglobin while the increase in respiration was maintained throughout. B.M.L. June 11,1918. Barometric pressure 380 mm. MINUTE ' 0 5 10 15 25 35 55 75 Eseries fee Ae] 0 6851 69 70 78 89 91 87 85 ShySHOING - tstwina ae a amein pO 102 102 102 | 102 DIASHONG: .<%-l2s seh sics|),, 68 52 48 46 46 Pulse pressure. ...:.... 34 50 54 56 56 mivgecolar Op... o:...2.-.|. 109 30-8) | 32.0/) 3178 31.9} 32.4 mivemiad COs... oo... ss 37 36.6 | 32.9) 33.0 31.1) 30.3 Hemoglobin............ 96° 96 | 40th, 99| 106 | 104 In this case the systolic pressureremained constant throughout while in the cases of N. E. F. and R.S.8. it fell. The increase in pulse pres- sure, asin the case of R.S8.8., is again definite and is determined wholly by a fall in the diastolic pressure. The rate of blood flow reached its maximum at about the 35th minute, which was approximately the time of maximum pulse rate. From this time the pulse rate fell gradu- ally about 9 per cent, while the pulse pressure remained high and increased slightly, with the result that the rate of blood flow was presum- ably reduced. The respiration increased early and then maintained a level until the 50th minute, after which it increased gradually until the end of the experiment. The hemoglobin did not begin to increase until between the 26th and 40th minutes. The pulse rate increased 308 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER until the 32nd minute. This suggests a relationship between the hemoglobin and the pulse rate. The increase in breathing is also a factor that may have permitted a slowing of the heart rate. K. O. N. April 30, 1919. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 Puilsea. vee. sale 84 | 84 89 94 102 98 95 92 Systolic ose naee ase eA 122) 120) | AGS eet. Digstolicaweeecaan tae ae 68 64 58 56 54 PulsSeppressuneseeree sce 48 58 62 60 62 MINUTE 0 10 15 18 20 32 36 42 | 49 | 59 | 74 Respiration volume..| 5.3/5.5} 5.7 |5.9) 6.6 [5.7 6.0 |5.2/5.4/5.5)/5.6 Hemoglobin......... 100 27th, 101 52d, 106 74th, 106 In this case we determined the per-minute volume of breathing iv liters and took the alveolar air occasionally to compare with the volume. The blood flow and respiration each reached the maximum at once on arriving at 380 mm. The pulse rate then held until the 35th minute when it fell slowly until the fall was 12 per cent at the end. The respiratory volume fell slowly until the 42nd minute, after which it held at a volume slightly above the normal ventilation. The hemo- globin was just beginning to concentrate at the 27th minute. It reached its maximum by the 52nd minute. In this: case the circulatory and respiratory mechanisms seemed to have been relieved somewhat by the increase in hemoglobin. In figure 2 the data for P. 8. B., July 7, 1918, has been plotted. The pulse rate and diastolic pressure changes indicate that the blood flow reached its maximum during the early part of the holding period, and also that toward the end it decreased markedly. The decrease occurred when the hemoglobin had increased. The respiration was not benefited by the increase in hemoglobin. The remaining twenty-one cases in this group show in a similar manner the relationship between the increase in hemoglobin and the retardation in the pulse rate. We believe that this group of cases represents the usual reaction when the transition from normal oxygen tension to low oxygen is made gradually and at a moderately rapid rate. COMPENSATORY REACTIONS TO LOW OXYGEN 309 i54N 442 RESPIRATION RATE 348 124% VEN Ro. 40 : 94n SYSTOLIC ede JYITOLS AS 0 S ES BS Ss bad NS “I awl s x » s ] cS ¢ N an // K TiN Ros ba) a A re > , fet we Vasey Who pane s | / \ \ / n vy eat rw eer WS7Ad WASTOLIC 60 TIME. O Fs to 1S. 20 2S OT SO 4. 50 5§ GO oF a IM MINUTES 2 3 4 < a Fig. 2. P. S. B., July 9, 1918. Taken to 380 mm. (18,000 feet) in 18 minutes in the low pressure chamber. This case illustrates the interplay of compensatory factors, particularly the blood flow and the oxygen carrying capacity. b. Maximum pulse rate maintained with an increase in hemoglobin. There were nine cases in this group, four of which are presented here. G.C. W. June 25, 1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 Paniseee ei 2s aces. 3) (GO 62 64 70 78 78 78 82 SONIC: t/s).12.-1.f22) 542.5 LOO, 120 | 120 | 118 | 114 Miastolies of .7..,......| 60 64 58 48 48 Pulse pressure.........| 40 56 62 70 66 UVeO TOs... .... 0220! , 9S<0 36.2} 29.7) 30.4) 31.8 Alveolar COz........... Bie! 29:6) 32.7) 32.3) 23.1 Hemoglobin............| 98 | 82d, 100 In this case first the pulse, then the systolic pressure, and then the diastolic pressure each in turn aided in maintaining an increased rate of blood flow. The subject appeared to be compensating satisfactorily but his reactions seemed to be insufficient in that the circulatory and respiratory reactions continued to increase even toward the end of the 310 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER experiment. The pulse rate was higher after the 70th minute than at any time before. The respiration increased gradually but not so much as in the average case, until the 45th minute, when a marked increase in ventilation took place lowering the carbon dioxide from 31.2 mm. to 22.4 mm. The fact that the respiration increased markedly without affecting the pulse rate shows that the demand for oxygen was not sufficiently cared for. The increase in hemoglobin was slight and not in evidence until the end. W.C.W. July 16, 1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 Pulse mere. 66 | 69 194: 75 88 86 82 90 Systolic 34 eea.ckt aoe a.) 9, LOF 106 | 104} 102| 104] 104] 102 Diastolic tee 70 52 44 64 60 Pulse*pressures?...<...-7|/ 34 50 60 40 42 MINUTE 0 5 10 | 27 30 36 | 39 | 54 | 75 | 84 Respiration volume...........| 6.6) 6.2 |6.2/6.9| 6.9 | |7.6/7.0|7.416.5|/7.1 Hemoglobine.. ---. -asce ves s|90. ||.64th, 102 78th, 104 The heart rate at the 20th minute was 90. It then varied markedly for the next 40 minutes but at the 65th minute it reached 90 once more and showed a tendency to go higher, reaching 95 at times. The respira- tion reached its maximum per-minute volume at about the 36th minute and maintained it until the 54th minute, after which it decreased slightly. The hemoglobin had increased definitely by the 64th minute. In this experiment, if the hemoglobin exerted any sparing action it was shown in respiration. Compensation seemed to be somewhat inadequate. E.C.S. April 24,1919. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 95 Pulse................. 76| 76 | 80 | 84 | 98 | 100| 108] 105| 102 Systolic............| 108 128 | 134 126 Diastolic-eeeeeeeeee 62 62 58 58 Pulse pressure...... 46 66 76 68 0 17 25 Respiration volume........ 5:8! 7.4 |8.5| 8.5. 17-4). eBoy de 2iGe Glam Hemoglobin...............|100 | 29th, 108 57th, 107 77th, 109 COMPENSATORY REACTIONS TO LOW OXYGEN 311 The rate of blood flow undoubtedly increased markedly up to the 35th minute. The pulse rate rose early to a first high point (100), then held on a plateau until the 39th minute, after which it again accelerated until the 46th minute, when it held more or less constant until the end. A slight lowering appeared at the 95th minute. The respiration increased during the ascent and attained its maximum at the 23rd minute. It held this level until the 36th minute, after which the per-minute volume fell and maintained a new level until the 88th minute. During the interval from the 36th to the 42nd minutes, while the respiration was being reduced, the pulse rate rose to its second high point (105). The hemoglobin showed concentration at the 29th minute. In this experiment the respiration seems to have been spared by the increase in hemoglobin. An interplay between circulation and respiration was present during the middle period, from the 36th to the 46th minutes. N. E. B. May 6, 1919. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 Rab lstereteey cei? isc: ts chicas aiid CAR A 76 79 -| 86 85 85 84 Bron re te.. S bs. tay LOS 112 | 106] 104 EUS toss 2 cia att ey 76 70 70 66 IPulsespressure::. 3...-..-: 32 42 36 38 MINUTE 0 8 14 18 23 71 82 Respiration volume....... 5.0} 5.4 5.9] 6.0 6.6 se || Gets Hemogiobm. .:.......... | 100- | 42d, 108 76th, 110 The pulse rate and blood flow reached the maximum together about the 21st minute. The per-minute volume of breathing had increased by the 23rd minute, to the volume which was maintained throughout the holding period. The increase in hemoglobin did not favor either the circulation or respiration. We believe that some members of this group failed to show an inter- play between the hemoglobin, circulation and respiration because they were too near their critical low oxygen limit. The compensations were just able to meet the demand of the tissues for oxygen. c. Maximum pulse rate maintained without an increase in hemoglobin. There were eight cases in this group. Several of the experiments will 312 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER be discussed in the study of repeated cases. It was to be expected that a failure in compensation by one mechanism might cause the others to hold a constant level when they had responded sufficiently to meet the demands of the body for oxygen. The following case is typical of the group. A.W. L. June 18,1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 Pulse 2.2 ek. SE Es) “93 93 96 98 98 | 100 | 101) 101 Systoliceneb Aeek ch eat 118 | 120.| 122 |) 124 | 1205) 1169)" LIGA eas IDIER Kol Om ee eae coeas ome 1 ch 78 64 56 60 iP.ulsegpressures era a 42 42 52 60 58 MINUTE 0 10 15 25 45 55 65 75 85 Alveolar Og. ......... 101.3) 54.5 | 31.3 | 32.0 | 29.6 | 30.3 | 30.1 | 30.5 | 34.0 Alveolar CO........| 42.7) 40.5 | 34.8 | 34.5 | 34.4 | 33.4 | 34.4 | 33.5 | 30.7 The rate of blood flow appeared to have reached a maximum about the 55th minute and then maintained the new level. The respiration also, after reaching its maximum value at the 25th minute, remained fairly constant until about the 80th minute, when it increased once more. There was no evidence of interplay of compensatory mechan- isms throughout this experiment. d. Retardation in the pulse rate dur oe the holding period with no in- crease in hemoglobin. There were four cases in this group and the data for these is given below. F. C. P. December 8, 1918. Barometer 428 mm. MINUTE 0 5 10 15 25 35 42 50 75 Pulse...ote.. coe. | 72 |74 |78 | 80 77 198 75 eee Alveolar Oo......... 97.1 58.1 | 44.2 | 44.6 38.6 | 49.3 | 38.6 Alveolar CO}.......| 39.7 36.7 | 32.8 | 30.3 31.9 | 25.5 | 31.4 This experiment was conducted at a barometric pressure of 428 mm. (15,000 feet). The breathing as shown by the alveolar carbon dioxide was variable. The carbon dioxide tension was lowest at the 25th and COMPENSATORY REACTIONS TO LOW OXYGEN 313 50th minutes. The high alveolar oxygen tension at the 50th minute was sufficient to account for the falling off in pulse rate. C. P. C. December 13, 1918. Barometric pressure 428 mm. MINUTE 0 5) 10 15 25 35 45 58 POS 2 vic Gee ea fet 88 94 98 95 92 88 miveolan (Op... ......2....-| 96.0) | 84.5 | 56.05) 41.6 48 . 4 Alveolar CO; .............| 32.2 | 92.6 | 23.3 | 20.4 18.7 The respiration increased throughout the entire period which prob- ably accounts for th~ slowing of the pulse rate. F. D., January 13, 1919, was subjected to a barometric pressure of 395 mm. (17,000 feet) in an experiment which lasted 100 minutes. The hemoglobin did not increase. The pulse rate accelerated from 72 to 99 by the 15th minute. It held this rate for three minutes, was 90 at the 25th minute and 78 at the 70th minute, where it remained until the close of the experiment. The respiration was not measured, but the observer and the subject noticed that the subject’s breathing in- creased and became labored at the 25th minute, and remained so until the end. H. J. M. June 20, 1918. Barcmetric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 63 enineeererre = ote leg = 83" hag" b98-) 100i" 597 | 92) 100 TSHVSUOING OS ape ee peed cea Pel O fs) 112 | 114 WiastoliCr...) eso scdeew es 64. 48 48 Pulse pressure.::.....4....| 44 64 66 MINUTE Respiration volume Hemoglobin... A definite fall in the pulse rate began at the 46th minute and lasted until the 56th minute, after which it gradually accelerated again. There is nothing that accounts for this fall in pulse rate. The arterial pres- sures were not taken often enough to make an interpretation of the blood flow changes. 314 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER We believe that our data accounts, in three out of four of the cases, for the decrease in pulse rate that occurred during the holding period. The interplay in these cases was between the respiratory and circulatory compensations to low oxygen. REPEATED EXPERIMENTS ON ONE INDIVIDUAL Five men served as subjects from two to five times each. The data in four cases are complete enough to make comparisons worth while. W. H.G. May 24, 1918. Barometric pressure 428 mm. MINUTE 0 5 10 15 25 35 55 RUS Ae We teeter name we aice ik « 76 82 85 94 93 91 90 Systoli@. cs. c..e-ccaeeee-s--+-| 100, 102°) W020) 102) 5 Ageky SiO Dias toliceeee mam cers ees cies ore 66 60 62 66 68 Pulse pLESSUTE mack hee eran ies ars: 34 42 42 38 32 MINUTE 0 10 15 25 35 45 65 AlveolariOsesaer ccna ck tenis 106.8) 72.5 | 50.8} 45.2) 48.8) 43.7) 46.7 Alveolar COs:.......2+2+2s.-....| 07.0] 60.0 | 30:4), 31-7) 2879) Soiinezones Hemoglobin: .o. tee oeeaceee = |) eo0 103 | 106 | 106 | 105 | 105 W.H.G. July 10, 1918. Barometric pressure 380 mm. MINUTE 0 5 10 15 25 35 55 75 95 Puls: sin... ee ts 78 82 90 97 | 100 98 96 93 94 Systolie?)..../%../.) 1105) 110) 108) 1044) °*120)} : 116 |) 116%) tee Diastolic’. .-.n4see|) wo 68 68 70 64 64 Pulse pressure...... 40 52 48 46 50 44 MINUTE 0 6 12 | 21 27 39 | 42 60 75 | 102 Respiration volume.....| 5.7/ 6.3 |6.6/8.1| 7.8 {7.6/6.6} 6.5 (6.8/6.7 Hemoglobinu: tee: 4:45- 98 | 46th, 101 65th, 101 95th, 104 In both experiments the blood flow, as shown by the pulse rate and the pulse pressure, reached its maximum immediately after the low barometric pressure was attained. An increase in hemoglobin was COMPENSATORY REACTIONS TO LOW OXYGEN SD observed in each experiment when the pulse rate began to retard. The respiration, in the experiment in which the barometric pressure was 428 mm., increased during the ascent, then maintained a level. In the experiment at 380 mm. the respiratory per-minute volume was increased from 5.7 to 8.1 liters during the ascent. It then decreased slowly to 6.6 liters at the 42nd minute, after which it remained constant. In the first experiment the increase in hemoglobin spared the circulation, in the second both circulation and respiration shared the gain. B. R. L. was taken twice to 3880 mm. The results are tabulated below. B. R. L. August 5, 1918 MINUTE 0 5 10 15 25 35 55 75 95 LZ 84 85 85 88 90 86 84 82 83 Systolic.......|104 102° {100 102 92 98 102 Diastolic...... 62 62 60 66 64 Pulse __ pres- URES Saree 42 40 32 32 38 Alveolar Oz... .|107.3 75.0| 44.1 |42.2) 44.0 |48.0} 50.0 | 43.2 Alveolar CO:..| 37.3 30.9} 27.4 (29.6, 25.0 |22.9| 19.8 | 23.4 Hemoglobin...|107 | 45th, 111 60th, 112 100th, 112 142d, 116 B. R. L. May 12, 1919 MINUTE 0 5 10 15 22 25 35 55 66 IPUISERPE arse ote oe SDSS 94 94 86 86 | 88 88 | 82 SVSuOlWCe ess. «sea! LLO 110 112 WiatstoliGss..2 <: 4 sa2: 79 70 68 Pulse pressure...... 40 40 | At MINUTE 0 5 23 33 46 59 VvEolar Obj. ca... .2 snes) L044! 9020) | 4913 48.3 51.6 52.0 Ailigeolari COs... ss. sda =oae 39.1; 33.6 | 31.4 19.5 18.3 18.3 Hemoglobin ..-.........:..| 100 - | 25th, 106 43d, 106 | 88th, 107) 90th, 107 The two experiments, while separated by nine months, were quite similar and unusual in several respects. In both the pulse reached its maximum rate quickly and returned to normal or subnormal before the close of the experiment. The respiratory increase was more marked THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 316 H. W. GREGG, B. R. LUTZ AND E. C. SCHNEIDER than in the usual case, in that the carbon dioxide instead of falling to the average figure of 31 mm., reached 19.8 mm. in the first and 18.3 mm. in the second experiment. kh oO an o ro) Z = ae} xq > |4 = ms co a Hn |e eel] sec. mm. | decil. sec. mm. | decil. sec. mm. | decil, ON osbel ely 4 0|}4.3] 14 3 0 | 2.5) 24 5 Nitrogen on Nitrogen on Nitrogen on 1 | 4404) 15 Sella) PLaCSeS.|, 16 4/ 2.80} 1] 4.0) 15 4 | 3.36 2;4.0) 15 Bo] 2e24| 24.2.) 14 4 | 3.90) 2 | 3.3) 18 5 | 2.80 3 | 4.0] 15 4) 3.36] 3) 4.6] 138 6 | 5.60) 3 | 2.6) 23 5 | 1.68 4/4.0} 15 3 | 5.60} 4) 4.5] 138 SAle2 SO Aaa tl Peg, 5 | 5.04 5 4e2n 14 6 | 5.60) 5 | 7.5 8 7 | 5-04) 5 | 2.8) 21 5 | 3.90 Gu 452 Gl ele 9 | 6.72) 6| 4.2 | 14 9 | 6.16} 6 | 3.0} 20 6 | 3.90 | 4305) 15 ON Giai2|) ot) 4.85), 12 9 |. 7.841 7 | 3.0) 20 7 | 5.04 S| 402 ews 10) G272| 58) 4:69) 15%) bE We Say S25 Olee20 6 | 5.60 Or SxSal BG) | e129 7e84)) 94.3") 14 9 | 8.40! 9 | 2.8) 21 6 | 4.48 10 | 3.8| 16] 138 | 8.96} 10} 4.4} 14) 11 | 7.84) 10 | 2.7) 22 6 | 3.36 11 | 3:6) V7 |) 13-7284) 11 | 4.4) 14) 1298240) a aaa e2 6 | 6.16 12|3.8| 16] 16 /11.20) 12|} 4.3) 14 | 14 |12.89) 12 | 4.2) 14 7 | S308 13 | 3.6 | 17] 17 |10.10) 13) 4.38 | 14) 14 /11.20) 13 | 2.5) 24 8 | 4.12 1413.2] 19] 18 |10.10) 14/ 3.8] 16] 15 115.68] 14 | 2.5) 24 9 | 5.60 15 | 3.4] 18 | 18 |11.20) 15 | 3.6) 17 | 20 |14.57| 15 | 2.2), 27) On oate 16 | 3.0) 20) 19 St GZ ee 17 |) 2.6%) 923. | 207 )11220 Nitrogen off V7 |2.2). 27-5) a Gee 18 | 2.8 | 22) 21 1105310 —S— a 18 2 ee TOW 2A S255 | pe2eel2e20| AGe 3 .5e) soles 19 | 2.2) 277s 2a Ga 175), 229; | Ziey 20) | 25:21) 277 ea Sa Gai: Nitrogen off 18; 3.0; 20) 15 21 | 2.2) 27 | 13>) 78 T9NS20)\7 20%)" 1s 22) 2.1), 28°) Msn Gai 20) 2.4) 25 | 23 20K 2298) 21 9 2i 2:4 | 25 | 25 21%) 3:05) :20 7 Nitrogen off Do We2eee |) 22o\' E19 DON Dati |. 22 4 23 | 24 || 25 | 12 23 N224-| 25 4 23422227 aaa DAN 24 oval okie, 2A 3:05) 20 3 2A 252) 927 |e: 2) rove | tone ao P| PAA Perl ||. 5) 26 | 3.4) 18] 10 26 | 2.4, 25 | 15 27 | 3.438) 18 8 27 | 228i) QE 28 | 4.0 15 5 28 | 2.8) 21 12 29) | 3.65) 17 5 | QO a2, 21 taal eles 30 | 3.8] 16 5 30) 22) a ee 31 | 4.0 15 5 SH alictsy|) Gis) 5 32 | 4.0 Woy |) te} | | REACTION OF MEDULLARY CENTERS TO LOW OXYGEN TABLE 3—Concluded 337 of respirations Number HSANAoORWN He w.B ' 5 E Me eee | oe ee a) a5 S, 4 fon is} sec. mm 4.3 14 3 Nitrogen on Bez) 39 | 4 2b) (221.5 a-h || 16) 4 4.2 | 14] 5 a0} 20) |) 5 2.5] 24] 5 Beso = 2% Pech |p Pal) Sp Petey | a) eM i | 24: | 7 ZO) 20 | ¢ At le Say f 2.4) 25] 9 Zen 205) 8 Beene oat & Nitrogen off Pall eee ld aad Per ile faa aaa Zrilwiie2on| 7 2.3 | 26| 8 Dene, 20) | © iG | or | 5 ZeGH “23: |) 5 2.4) 25] 5 Den Pao: | 6 257 \\ 7-0 ee) 2.5) 24) 7 Bete eene | 6 Zt) 29°). 6 ae 2h | 6 UI 5 Ua SO Ne |) .3 Cc. M. L. Sg # 2 E HS 5B eles] B | es| 3 Ss Sk oO an © > 14 4 oo q decil sec. mm. OFN226)11 223 5 Nitrogen on 1/3.2)| 19 4 3-90) 42h | Bat ie 22 4 6.16] 3] 3.3] 18 4 4.48} 4)3.3] 18 4 4.48] 5 | 3.6} 17 6 5.60} 6|3.3 | 18 ia 4.48) 7|3.2 | 19 7 6.16) <8 )|-322):/ 919) 10 6G272| °95)3-2)))) 219) |) 10 6.72) 10 ; 2.8 | 26 ts 7.20) 1 | 226 | 23 7 (M6074| 35 |) Bee || 2 9 7.84] 13°] 2.6 | 238 9 7.84, 14 | 2.5 | 24 9 7.84) 15) 2.0 129) |) 43 Nitrogen off 16} 2.0) 30} 138 WPA P4a0) |) 0) Weed 18 | 2.0} 30 7 19 | 2.5] 24 9 20 | 2.8 | 21 6 21 | 3.4 | 18 5 22'|-3.1 |’ 19 5 23 | 3.6 | 17 5 24|)3.4] 18 5 4.48 6.16 3.90 4.48 5.60 8.40 7.27 7.84 11.20 10.80 5.60 10.80 10.80 8.95 8.40 of respirations Number OONIoair WN KH K. O. N ' fB: g o a Q, ~ ~ i | 28 | & oO ar o Hy | & q sec. mm. 320 16 tf Nitrogen on 3.6) 17 6 3.6} 17 7 3.2) 19 10 2.9) 21 14 2.9) 21 15 3.2) 19] 19 3.9) 15] 20 3.5, 17 | 19 3.2) 19 | 19 2.9; 21) 20 2.2) 27 | 20 2.2| 2¢ |) 20 Nitrogen off 2.5) 24 2.5) 24 2.3) 26 2.4) 25 2.4) 25 2.8) 21 Qel| 22 2.8) 21! 2.7| 22 6.72 6.16 8.40 9.52 14.00 14.60 26.30 22.40 16.80 24.60 19.10 338 BRENTON R. LUTZ AND EDWARD C. SCHNEIDER medullary centers, the shorter latent period at the return to air or oxygen is explained in part by the increased blood flow. : When anoxhemia was produced by breathing nitrogen, the latent period for change in volume of breathing as estimated by the height of the respiratory curve ranged between 4 and 35 seconds. The average was 14.5 seconds. About 32 per cent of all cases have a latent period of 10 seconds or less, while in 29 per cent it was between 10 and 15 seconds. In the determinations of the volume of each breath by the Larsen spirometer the latent period ranged between 4 and 25 seconds, averaging 12.4 seconds. In almost all cases the latent period as esti- mated by this method was slightly less than that determined by the height of the respiratory curves. The volume of breathing diminished quickly when the subject was restored to atmospheric air. The latent period as determined by the height of the pneumograph curve ranged between 3 and 34 seconds, averaging 6.9 seconds. A large number, 89.3 per cent, had a latent period of 10 seconds or less. This would indicate that the respiratory center, with respect to the volume of each breath responds to oxygen changes slightly earlier than the cardiac center. The rate of breathing is increased by anoxhemia when the fall n oxygen has become marked. On administering nitrogen the latent period for the increase in the rate of breathing ranged between 8 and 80 seconds, averaging 35.5 seconds. The normal rate of breathing ranged between 8 and 26 breaths per minute, and at the height of anoxhemia it ranged between 11 and 46 breaths. The rate of breathing decreased more rapidly when the anoxhemia was alleviated than it increased during the withdrawal of oxygen. The latent period for rate on the return to air ranged between 3 and 31 seconds, averaging 9.5 seconds. The respiratory stimulation due to low oxygen with respect to rate and depth passed away completely within from 10 to 53 seconds after returning to atmospheric air. In a few cases which were kept at 380 mm. in the low pressure chamber we have determined the volume change when oxygen was administered. The per-minute volume of breathing was recorded in these experiments and the data are given in the table. It will be observed that the per- minute volume of breathing was much reduced even during the first minute of oxygen administration. Comparisons of the increased rate due to sciatic stimulation with the effects of adrenalin (quantitated as base) injected intravenously indicate that the range of reflex adrenal secretion lies 4 At the same time Cannon reported (Science, May, 1917, xlv, 463) that the denervated heart revealed an increase of adrenal secretion after sciatic stimula- tion or asphyxiation, and promised a full report of the experiments in this Journal. Absence from the United States for nearly two years has unavoidably delayed the complete paper until this time. °> Many years before Levy’s observation, Hunt noted (this Journal, 1899, i, 444) that stimulation of a sensory nerve would cause cardiac acceleration after all cardiac nerves were divided, and that the same result followed stimulating. the peripheral splanchnic. It differed from the acceleration following sympa- thetic stimulation by beginning slowly, i.e., about ten seconds after the start of stimulation. It is interesting to observe that this period is almost exactly that required for the distribution of adrenal secretion by the circulating blood. "07BI JO VSBaINUI ON “OT :g ‘spuodes Ye UOTPR[NWTYS O1YBIDg Sg ‘(8G :% “Yo pay spuv[s [eusrpy) ‘oynurU 1od $ZZ 0} SgT UIOAy 0y841 JO asvO. Ul “13 :% ‘spuodes Og uoTyR[NUITYs orperng ‘py “‘Spuodos G ‘ST@A 10} UI ouly, “lojJoUIOUBU oUuBIqUIOUL ‘Q1Boy p29} BALOUOp oq} jo }BOq oy} jo Sploo9y “II ‘Sly GSl Sa ee Sees rece oe CANNON B. WwW. A HH ayy 408 ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION TABLE 1 409 Examples of increased rate of the denervated heart on sciatic stimulation DATE Marchi2ip Marche2ar-neerence AprilsG: o.... 7701 87 (aes ana SS BO PSEMURC Ns Pons ce caperstochas cestrane eesti. i i..'k. > cles ZL eo eo SS eee Oe COCO SCIATIC STIMULATION RATE BEFORE RATE AFTER INCREASE PER MINUTE 2.58 30 seconds 220 264 44 3.04 30 seconds 220 256 36 Solo 15 seconds 212 240 28 3.16 | Splanchnics cut 3.25 30 seconds 184 192 8 3.28 15 seconds 184 192 8 2.16 30 seconds 176 228 52 PAPA 30 seconds 188 224 44 2.52 | Adrenal glands tied 3.07 15 seconds 152 156 4 3.10 30 seconds 152 152 0 PSA 30 seconds 216 240 24 2.39 30 seconds 212 236 24 3.12 30 seconds 184 208 24 SoD 30 seconds 196 216 20 2.50 30 seconds 200 236 36 BABY 30 seconds 200 224 24. 2.56 30 seconds 200 244 44 3.01 30 seconds 200 240 40 3.05 | Cerebrum removed 3.19 30 seconds 208 236 28 1.01 45 seconds 144 180 36 1.05 55 seconds 152 184 32 1.14 55 seconds 156 180 24 1.20 55 seconds 156 180 24 1153333 50 seconds 162 192 30 1.54 | Right adrenal tied, left splanchnic cut 2.06 45 seconds 164 164 0 10.35 30 seconds 156 180 24 10.43 40 seconds 156 180 24 10.47 45 seconds 156 180 24 11.03 | Both adrenals removed, 100 ec. gum-salt solution 11.07 40 seconds 174 174 0 11.10 | 40 seconds | 138 138 0 410 W. B. CANNON between 0.001 and 0.005 mgm. per k. per minute—1.e., from five to twenty-five times the amount regarded by Stewart and Rogoff as the normal output. Reflex increase of the cardiac rate does not occur if the adrenal glands are removed (see fig. 1 and table 1). Asphyzia. Asphyxiation of the cat with the heart completely denervated will cause a noteworthy increase in the heart rate (see fig. 2 and table 2), an effect not seen after adrenalectomy. The figures in table 2 illustrate another point mentioned in 1917, viz., that an indica- tion of adrenal secretion may be obtained from the denervated gland if asphyxiais prolonged. Inthe experiment of February 24, for example, asphyxia for 20 seconds, though previously effective, caused no change after severance of thesplanchnics. In that of February 27, asphyxia of 60 seconds caused no change after splanchnic section; and in that of February 28, though asphyxia of 35 seconds had been highly effective before the splachnics were cut, thereafter asphyxia of 45 seconds in- creased the heart rate only 4 beats per minute, whereas asphyxia of 90 seconds caused an increase of 68 beats a minute. Similar differences are observed in the experiment of March 21. Unfortunately these observations were not checked by final proof that cutting the splanchnics completely denervated the glands, though the marked drop in pulse rate may be regarded as testimony to that conclusion. The results are in agreement, however, with evidence adduced by Czubalski (24) that asphyxia, if sufficiently prolonged, may have a direct stimulating action on the adrenal medulla, and perhaps on other chromaffine tissue as well. In 1917 Cannon described another method of demonstrating adrenal secretion, which consists in cutting all the nerves in the gastro-intestinal mesentery, tying all the limb arteries and the carotids, and thus leaving the circulation confined chiefly to the splanchnic area which, however, is denervated (23). Under these circumstances it is not uncommon for asphyxia to cause a slight rise of pressure after an interval of 40 to 60 seconds and a very considerably greater rise as soon as respiration begins again; these results do not occur if the adrenal glands are excluded (see fig. 3). Emotional excitement. The completely denervated heart can be used as an indicator of adrenal secretion in testing the influence of emotional excitement quite as well as in testing the influence of sensory stimulation and asphyxia. It is only necessary to take somewhat greater pains in order to keep animals in normal condition after opera- tion. To denervate the heart, the stellate ganglia are first removed ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION Aj] Fig. 2. mercury manometer. Original size. A, Asphyxia 60 seconds, 3:55. Increase of rate from 188 to 22 Beginning and end of records of the beat of the denervated heart, Time intervals, 5 seconds, 220 per minute. (Adrenal glands tied off, 4: 45). B, Asphyxia 60 seconds, 4:50. No increase of rate. | 3lood pressure rose from 92 to 124 mm. Hg. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, No. 3 412 W. B. CANNON TABLE 2 Examples of increased rate of the denervated heart on asphyxiation. (In the first five cases the abdomen had been opened) DATE TIME ASPHYXIA eee RATE AFTER eee ; 80 seconds 240 256 16 JABURTY Zoe peer sao Adrenals removed ( 90 seconds 192 192 0 2.50 60 seconds 162 202 40 2.58 60 seconds 158 200 42 February 17 3.01 | Veins tied both sides of adrenal glands ae 3.04 60seconds | 166 | 206 | 40 3.10 | Adrenal glands tied off completely 3.18 60 seconds 146 146 0 3.39 40 seconds 204 256 52 Fébegary 24 3.38 20 seconds 216 228 12 Se aes 2) Soc eer ee 4.10 | Splanchnies cut in thorax 4.15 20 seconds 204 204 0 3.28 90 seconds 208 236 28 ERebruarye27-0. 4b eee 3.41 | Splanchnics cut in thorax || 3.43 60 seconds 188 188 0 11.39 35 seconds 180 212 32 12.08 | Splanchniecs cut in thorax 12.10 45 seconds 164 168 4 Pebruary: 28.2.3 .02:4> & yo aaado 90 seconds 156 224 68 12.18 90 seconds 172 224 52 12.25 | Veins tied both sides of adrenal glands 12.31 90 seconds 168 180 12 2.51 60 seconds 196 212 16 March 16h: eee 3.05 45 seconds 208 228 20 (| 3.59 | 90 seconds 188 224 36 3.09 45 seconds 220) 240 20 3.20 | Splanchnies cut in thorax Marehi2icy - cee. cees eee 3 30 Op eomeae 180 212 39 BB 45 seconds 176 188 12 (| 2.24 | 60 seconds 192 236 44 2.34 | Abdomen opened 192 204 12 aaa | 2.53 | Adrenal glands tied off completely 3.20 60 seconds 152 | 156 | | iA SECRETION INAL ADRI INDICATOR OF iD HEART AS ISOLATE . oluyouelds oy} 3 1} ‘JUOSQ’ sosll YOG ‘soynulu Zz BIxAydsy ‘Fg od [eixAydse 104j8 Uledse puv JNU [ JO pus 4¥ ainssoid Jo osIy “soynurM Z BIxAydsy ‘F ‘SpUOddS OE ‘S[VAIOJUI OUT], “OZIS [VUIBIIQ, “BolE ALIUOP PUB SOLo}IB OVI]. puw UBIAB[OGnS ‘prjoIvd oY} ButAy 1aqJw plode1 ainsseid poorg “e “31g f ‘urd ¢6 32 ‘OT [ady ‘aynurur aad SYB9q 1ZZ 0} BL Java ‘poztoxo |wuTUy ‘q7 a] agra Ss > rund OF :z0T Judy aynurur rad syeoq 21% 981 4Avay “WI[wd [BUOY on COL [dy “ure Og :6 poyotduoo [vaoutar [vuoeIpy) ‘g [dy ‘oynurur aad syveq geez ayer qtvoy ‘poztoxe [eunruy ‘g : < “8 [udy ‘oynutur sad syvaq 21z 0784 javoy ‘Wye peuTUYy ‘Py ‘ aie Ree . ‘puodes QT/T yreut syop [BUS of} plOoo1 out} VY} OAOGY “puodas OOT/T ‘s[/@Adoqy Ur OUNTT, ‘yABOY po}BArouap oY} JO SurvISOIpIBIOLOII| “fF “Sry W. B. CANNON ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 415 under ether with aseptic precautions; later the right vagus nerve is severed below the recurrent laryngeal branch; and still later, the left vagus nerve is cut in the neck. The heart.is thus wholly disconnected from the central nervous system and any agency causing an increase in the heart rate must exert its influence through the blood stream. In figure 4 are presented electrocardiographic records of the heart rate in a cat, operated upon as above described. The records show that with the adrenal glands normally innervated the rate was 217 per minute when the animal was calm, and 255 when excited. And after the adrenal glands were removed the rate when calm was 217 and when excited was 221. The results obtained with the isolated heart used as an indicator of adrenal secretion thus confirm in every respect the results obtained eight years ago by the catheter method. Care in assuring tsolation of the adrenal glands. If the splanchnic nerves are severed or if the adrenal gland is removed on one side and the splanchnic fibers are cut on the other, as Stewart and Rogoff have noted, adrenal secretion may be isolated from nervous control in most cases, but there is not absolute certainty that this procedure will wholly eliminate nervous influences (29). For example, in one case, after the heart was wholly denervated, sciatic stimulation for one minute in- creased the rate from 220 to 264 beats per minute. The splanchnic nerves were then isolated in the thorax and cut. In two minutes the heart rate had dropped down to 192 beats per minute. Sciatic stimu- lation now increased the rate to 204, i.e., a rise of 12 beats per minute. Similar observations have been made on animals with denervated heart that have been kept alive and observed under excitement. In one such case there was an increase of 42 beats per minute, although the left adrenal gland had been removed and the right splanchnic cut in the abdomen on the previous day. After removal of the right gland excitement had no effect. In another instance in which a similar oper- ation had been performed there was an increase of approximately 28 beats a minute during excitement, an increase which disappeared as soon as the remaining adrenal gland was excised and the animal allowed to recover from etherization. It is possible, therefore, that other fibers than those contained in the splanchnic supply, or that occasionally, perhaps, a crossing of fibers from one splanchnic supply to the gland of the other side of the body, may be present in the cat and may thus lead to erroneous conclusions. 416 W. B. CANNON It has been assumed that by tying the adrenal veins at their junction with the inferior cava and the lumbar veins as they approach the adrenal glands, all possibility of an entrance of adrenal secretion into the blood stream has been excluded (10). That this may be a reasonable assump- tion in most cases was shown by Flint’s studies of the blood supply of the cortex and medulla, which brought out the fact that the vessels of the two parts of the gland are separate. He reported, however, that as a variation from the usual condition, anastomosis may be present between the branches of the venous tree in the adrenal medulla and the venous plexus of the capsule (25). Under these circumstances the blood might flow from the medulla to the venous plexus which normally empties into the renal, phrenic and lumbar veins, into the venae comites of the suprarenal arteries, and to a less degree into the other veins. Further evidence of a direct vascular connection between the suprarenal gland and the veins of the kidney has been reported by Cow (26) who obtained an adrenal-like effect with blood taken from the kidney capsule of a cat. In several instances in the course of the observations reported in this paper adrenal effects were seen after tying the lumbo-adrenal veins on both sides of the glands. In one case, after these veins had been thus tied, asphyxia caused the heart rate to increase from 166 to 206 beats a minute. The glands were then tied off completely, whereupon asphyxia had no effect (see table 2, February 17). In another instance, after the lumbo-adrenal veins had been carefully tied, injection of adrenalin into the vein as it crossed the gland caused a high rise of blood pressure, and in still another instance in which the veins were tied, splanchnic stimulation for 30 seconds caused the heart rate to rise from 172 to 248 beats per minute. From these observations it is clear that conclusions based on results obtained when only the lumbo-adrenal veins are tied may lead to erro- neous conclusions. The only absolutely safe method is that of excluding the glands from any possible action in the body by removing them or completely tying them off. The gradual rise of pulse rate on repeated stimulation. A fact com- monly noted in the course of the present experimentation was a gradual rise of the pulse rate with the lapse of time and with repeated indirect stimulation of the adrenal glands. In one instance an animal anesthe- tized with urethane had, after denervation of the heart, a pulse of 176 beats per minute. Repeated sciatic stimulation and asphyxia were accompanied by temporary increases of the pulse above the basal rate, ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 417 varying from 16 to 52 beats per minute. As these stimulations recurred, however, the basal rate gradually rose from 176 to 204. On tying off the adrenal glands completely, the rate fell to 152. The increase of rate and its persistence at a progressively higher level with repeated stimulation are possibly facts of considerable importance in relation to the interaction of the endocrine glands, and deserve further examination. A fall of pulse rate on sciatic stimulation. A curious fact noted in a number of instances after the abdomen had been opened was that sciatic stimulation, instead of causing an increase in the rate of the denervated heart, actually resulted in a slower beat. In one such case sciatic stimulation for 30 seconds reduced the rate from 216 to 212 beats per minute; subsequent stimulation for 45 seconds lowered it from 216 to 204, and still later from 236 to 216. In another instance sciatic stimulation lowered the rate from 216 to 212 and later from 232 to 212. No important changes of blood pressure preceded the altered rate. The significance of these effects is difficult to perceive. In the cases mentioned asphyxia caused a marked increase in the heart. rate. DEFENSE OF THE ISOLATED HEART AS AN INDICATOR OF ADRENAL SECRETION In a recent paper on hyperglycemia, Stewart and Rogoff have inci- dentally offered four different arguments opposed to the conclusion that effects seen in the denervated heart are satisfactory proof of increased adrenal secretion (27). These arguments are as follows: 1. They state that there is nothing strange about an increase in the rate of the denervated heart when the central end of the sciatic or the peripheral end of the splanchnic nerve is stimulated—‘‘it is obviously dependent upon the better blood flow through the coronary vessels.”’ For evidence they cite Guthrie and Pike as having shown that in the perfused mammalian heart the rate could be made decidedly faster by raising the pressure of the perfusion fluid. In the experiments cited, however, Guthrie and Pike were using the excised heart; they definitely declare that the denervated heart in situ (the preparation described in this paper) does not follow the law of the excised heart as regards pressure changes. After complete denervation of the heart, they report, ‘there is either no change in the pulse rate (with variation of pressure), or an increase in rate with a fall in pressure, or a decrease in rate with rise in pressure.”’ In so far as these observations testify that variations of arte- 418 W. B. CANNON rial pressure have no effect on the rate of the denervated heart, they are in accord with the earlier observations of Martin (28) and the more recent studies of Knowlton and Starling (29) who found that, between 20 and 200 mm. Hg., pressure changes did not change the rate. Frequently in the course of the work here reported arterial pressure has been raised 30 to 40 mm. Hg., after adrenal influence had been excluded, with no increase whatever in the rate of cardiac pulsation (cf. fig. 2). This concordant evidence wholly contradicts the first argument which Stewart and Rogoff have offered to account for the faster rate of the denervated heart when used as an indicator of adrenal secretion. 2. The second argument offered by them is that the rise of blood pressure, by increasing the rate of blood flow through the denervated heart or other organ, increases the amount of adrenin passing in unit time, and the sensitive denervated area responds to increase in the amount even if no change takes place in the rate of adrenal secretion. In presenting this argument the critics have not considered that with a rise of pressure the blood would pass more rapidly through the adrenal vessels (see p. 406); and therefore, on the basis of their own views of unvarying adrenal secretion, the higher the pressure the more diluted the adrenin—a condition which renders their argument unsound. It is not necessary, however, to rely on argument. The simple experiment of preventing a rise of pressure may be tried. In figure 5 a record is presented of an increase in rate of the denervated heart from 144 to 180, 1.e., a rise of 36 beats per minute, as a consequence of sciatic stimulation. The pressure rose about 58 mm. Hg. When the rate had fallen to 148 per minute the sciatic was again stimulated, but the pressure was prevented from rising by compression of the flexible thorax between the fingers and thumb. This is a procedure which, in the absence of the adrenal glands, is not attended by a faster beat of the denervated heart. The rate increased, however, from 148 to 184 beats per minute, a rise of 36 beats, as before. The more rapid rate developed during stimulation cannot be due to more adrenin contained in a larger volume-flow through the coronary arteries, for during stim- ulation the arterial pressure was not allowed to rise and augment the coronary flow. Furthermore, when the pressure was allowed to rise (50 mm.) the rapid rate, developed when the pressure was held down, did not become more rapid. The only explanation which affords a reasonable account of the faster rate is that there is something delivered to the heart through the blood stream which excites it to greater speed. Adrenin will do this. The fact that the faster rate disappears after ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 419 gat al fh. ) uth ih hy nee el) Da Ay Lo ggg igh Fig. 5. Beginning and end of records of the beat of the denervated heart, mercury manometer. Original size. Time intervals, 5 seconds. A, Sciatic stimulation 45 seconds, 1:01. Increase of rate from 144 to 180 per minute. B, Sciatic stimulation 55 seconds (1:05). Increase of rate from 148 to 184 though pressure-rise checked by thoracic compression. No further increase with rise of pressure. ‘oynuret sod syvoq QT 107] ‘ROT “[]Bs oAnssord a10Joq 0}VL JAI ‘urxAydse Jo spuooas Gg jo puny *) ‘ajnurur tod syvoq ZT oRA QavoFT “eIxAYdse Jo spuds GR JO Suruuisog ‘g ‘spuodes G ‘S[BALOJUT OUI], ‘V pre BF ‘agnutut tad syvoq ZL] ‘upeusIpe JO woTP OU SNOUSABAPUT WLO}TUN snonuty 0} ZE] Worl posBotoUr sny} o}VI JavayT ‘oynuTUT tod “UISUL 80'0 -u0d & YY pus AUOZD9[VUdIpY 109j8 ‘(LoyoUTOURUT Aan 9.101) ‘ }AVY PoPVALoUp oY} JO ¥BVoG OY} JO pLOVOY “9 “SLY CANNON B. Wie 420 ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 421 removal of the adrenal glands, although the rise of pressure still follows splanchnic stimulation (see fig. 1), is proof that this is the agent which is acting. Experimental test, therefore, denies the validity of Stewart and Rogoff’s second argument. 3. Their third argument is directed against the use of any organ in the body as an indicator of adrenal secretion when asphyxia is employed as a stimulus, because asphyxia may be expected to alter the reactivity of the test object to adrenin, making it, for example, more sensitive. ““We never supposed,”’ they declare, ‘‘that it was possible to use in one observation an asphyxiated test object and in the comparison observation the same object with unobstructed respiration, or to assume that if there was any difference in reactions, it must be due to a dif- ference in the rate of output of epinephrin; the condition of the test object itself being of no moment.” Again this is argument and not experiment. Experiment has shown that increase of carbon dioxide causes a decrease, not an increase, in the rate of the denervated heart, and that nevertheless, adrenin, if superadded, produces a faster beat (30). As shown in figure 6, when, after removal of the adrenal glands, adrenalin (1: 200,000) is allowed to run (1 ce. in 15 seconds) steadily into a vein, asphyxia does not cause an increase of rate. The stream of adrenalin raised the rate from 132 to 172; after asphyxia had prevailed for 50 seconds the rate dropped to 168; and as the asphyxial state con- tinued the rate became slower, dropping to 156 with a fall of pressure. A higher rate was possible, for the heart was obviously not beating at top speed, and yet there was no increase of rate at any stage in the develop- mentof asphyxia. Clearly the asphyxial condition did not render the test object more sensitive to the steady inflow of adrenalin. In the exper- iment illustrated in figure 2, an asphyxia lasting one minute caused an increase in the rate of the denervated heart of 32 beats a minute when the adrenal glands were connected with the circulation, but when these glands were completely tied off asphyxia for the same length of time caused no increase. The test object was in both cases subjected to identical periods of asphyxiation. Since asphyxia in the absence of the adrenal glands had no effect on the rate, whereas asphyxia with the adrenal glands present caused the characteristic acceleration which attends adrenal activity, the conclusion is warranted that the differ- ential element in the complex, namely, the possibility of adrenal secre- tion, is the occasion for the typical adrenal effect. It should be remem- bered that Anrep (11) likewise obtained no effect of asphyxia alone, 422 W. B. CANNON i.e., no contraction of the denervated limb, in the absence of the adrenal glands; indeed, toward the end of the asphyxial period there was dila- tation of the vessels ascribable to the direct action of the asphyxial blood on the vessel walls. This was in marked contrast to the asphyxial effect seen when the Adrenal gland was present; then even a large rise of general arterial pressure, more than 50 mm. Hg., was insufficient to distend the vessels of the denervated limb, which were held contracted, according to Anrep’s evidence, by secreted and circulating adrenin. Stewart and Rogoff’s third argument, therefore, has no experimental warrant. Fig. 7. Record of the beat of the denervated heart, (mercury manometer) in an animal with limb and carotid arteries tied and all mesenteric nerves severed. Enlarged one-sixth. Time intervals, 5 seconds. Asphyxia for 35 seconds increased the heart rate from 180 to 212 beats per min- ute, with no noteworthy previous change in blood pressure. 4. Their fourth argument is that afferent stimulation by constricting the splanchnic vessels lessens the blood flow through the liver; in con- sequence the adrenal secretion contained in the cava blood is less diluted (i.e., more concentrated) than normal and therefore has more stimulating power. Again it is not necessary to rely on argument. In figure 7 is presented a record of the beats of a denervated heart in an animal in which all the nerves of the mesentery were entirely severed and the animal then asphyxiated. The rate before asphyxia was 180 ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 423 beats per minute. This was increased by asphyxia 32 beats per minute. There is no possibility under these circumstances of any greater con- centration of secreted adrenin because of failure of blood to pass through a constricted splanchnic area, for the nerves which would cause constric- tion of these vessels had been previously cut. Furthermore, the pressure did not fall, i.e., the flow was not made slower during the asphyxiated period. The effect must be ascribed to greater concentration of adrenin in blood delivered to the heart, due to an increased secretion of the adrenal medulla. The fourth argument, therefore, like the first three, fails to stand experimental test. An observation having an important bearing on all four arguments and, indeed, on all conclusions arising from use of the “‘cava pocket”’ is that reported above (see p. 415) in giving evidence of adrenal activity at times of emotional excitement. As figure 4 shows, the rate of the denervated heart in an animal resting quietly with the adrenal glands intact was 217 beats a minute. When these glands were removed, there was not at any time a reduction of the rate. Recently Stewart and Rogoff (30) have testified that the ‘‘steady spontaneous discharge”’ from the adrenal glands in their experiments—an amount estimated as not more than 0.0002 mgm. per k. per minute—is sufficient to affect the heart. If with nerves intact there were in natural conditions the constant secretion which they declare to be ‘‘normal,”’ removal of the glands should have been followed by a slower pulse. That the pulse did not fall below the “quiet”? rate after adrenalectomy obviously permits the inference that in calm and peaceful existence there is no secretion from the adrenal glands sufficient to influence the response of an extremely sensitive indicator. In that case any attempt to explain the increased heart rate by greater delivery of adrenin or by greater concentration of it in the blood, due solely to shifts of the circu- lation, would be not at all pertinent. The only factor which Knowlton and Starling found effective in causing prompt alteration of rate of the isolated heart was change of temperature. In order to increase the rate 40 beats per minute, how- ever, the temperature of blood entering the heart had to be raised about 7°C. (29). It is inconceivable that the effects recorded above are due to the delivery of warmer blood to the heart. From the foregoing facts the conclusion is warranted that the expla- nations offered by Stewart and Rogoff to account for adrenal effects on the basis of greater flow or altered distribution of the blood have no experimental support. | 474 W. B. CANNON CRITICISM OF METHODS YIELDING NEGATIVE EVIDENCE A review of the previous sections of this paper reveals unanimous agreement among investigators, with the exception of Stewart and Rogoff, that painful stimulation, asphyxia and emotional excitement evoke adrenal secretion. Nevertheless, the care with which Stewart and Rogoff conducted their experiments, the quantitative methods which they employed and the variety of their experiments have led to their results being given a considerable degree of credence. As pre- viously stated, the discrepancy between their conclusions and those reached by all other investigators naturally raises the question as to whether some difference in the methods employed would not account for the difference in the results. Since Stewart and Rogoff are alone in their contentions, it is perhaps reasonable to inquire whether the peculiar method which they employed, rather than the various methods used by others, may not have features which would account for the discrepant results. The method of Stewart and Rogoff. Stewart and Rogoff obtained evidence of adrenal secretion by the use of a ‘‘pocket’’ in the inferior vena cava (32). This pocket was made by clamping the vena cava immediately above the iliacs, then clamping the renal veins, emptying the cava segment by stripping it upwards, and placing a clamp on the vessel above the entrance of the lumbo-adrenal veins. Any small branches of the cava segment were tied. The pocket thus formed was allowed to fill with blood from the adrenal veins, and the blood was either allowed to pass into the general circulation by removal of the clamp on the inferior cava, or was withdrawn and tested outside the body on preparations of rabbit uterus and intestine. The arrange- ment was modified in the “permanent pocket’? by tying splanchnic vessels and shutting off the blood flow in the hind quarters. Experi- ments performed under these conditions revealed a spontaneous libera- tion of adrenin. In one of their early papers Stewart and Rogoff state (32) that they are “not able to decide definitely whether this liberation is a normal physiological process merely unveiled by the experiments, or an ab- normal process dependent upon the necessary conditions of the obser- vations,—anesthesia, unavoidable excitation of afferent nerves, etc.’ They mention, however, the relative constancy of the amount secreted as in favor of the former hypothesis. Later they suggest that the extensive operation required by their procedure may have produced so ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 425 great a spontaneous discharge that no detectable increase could be produced, and they admit that Tscheboksaroff’s failure to obtain increased adrenal secretion on sensory stimulation may have been due to the severe operative procedure which he employed (33). This -earlier caution regarding their method they seem to have gradually -abandoned, for later they mention (7) the spontaneous secretion as being the “normal output of epinephrin” and state (34) that after section of the spinal cord the secretion has all the characters of ‘normal secretion,” and they repeatedly allude (9) to the amounts of adrenin found in the pocket as being the normal amounts. In recent papers (27), (35) they refer to their results as constituting ‘a striking illustra- tion of the fact dwelt upon in previous papers that the output of -epinephrin is relatively stable and not easily influenced experimentally,” and they speak of the secretion occurring at a relatively constant rate as the “naturally secreted epinephrin of the organism.’ Rogoff (36) -goes so far as to declare “it has been established beyond doubt that the adrenal glands continuously secrete a certain normal amount of ‘epinephrin.”’ Before this view can be admitted, the effect of opening the abdominal cavity, clamping off the inferior cava, and repeatedly manipulating the abdominal contents, either in pressing blood out of the inferior cava or withdrawing it by syringe, must be examined. Fully twenty years ago Bayliss and Starling called attention to the profound effect which opening the abdominal cavity has on the intestines in causing them to become absolutely motionless. Local stimulation then pro- vokes no response or only local contraction. If both splanchnic nerves are divided, the intestines within a short time commence to contract rhythmically and show the usual local reflexes. In order to study intestinal movements with the abdomen opened, they had to section both splanchnic nerves, or destroy the spinal cord, or excise the abdomi- mal ganglia. “These facts,” they state (37), “suggest that in the intact animal, at any rate under the conditions of our experiment, tonic or reflex influences are continually descending the splanchnic nerves and inhibiting the activity of the intestines.” The observations of Bayliss and Starling may be confirmed by any one who will study gastro- intestinal movements in the opened abdomen. Even if there is slight indication of activity at any time with the splanchnies intact, the least stimulation applied to the intestine, even a gentle handling of the gut, suffices to produce a reflex inhibition of its entire extent. These well- established facts make an interesting commentary on the use of the 426 W. B. CANNON cava pocket as a mode of obtaining evidence of ‘‘normal” or “natural” secretion. There is no doubt that secretion from the adrenal medulla is subject to impulses delivered by the splanchnic nerves, and there is no doubt that opening the abdominal cavity under anesthesia results in a discharge of impulses along these nerves. The adrenal glands, therefore, are continuously and abnormally stimulated if the abdomen is opened. The conclusion that must be drawn is that the pocket method is incapable of yielding any reliable evidence regarding the ‘normal’ secretion of these glands. The isolated heart yields pertinent testimony as to the discharge of impulses along splanchnic pathways under experimental conditions. An examination of the cases summarized in tables 1 and 2 reveals that, after section of the splanchnic nerves or exclusion of the adrenal glands, there is a drop in the heart rate—in some instances 40, 44 and even 48 beats per minute. The most reasonable explanation for this result is that in these experiments splanchnic impulses were continu- ously stimulating the glands to activity and thus making the heart beat faster than it otherwise would. Quite apart from these effects, evidence exists in the inhibitory influence of anesthesia on gastro- intestinal movements that anesthesia alone can arouse splanchnic im- pulses (cf. also Elliott, loc. cit.). Thus the “steady spontaneous dis- charge’’ from the adrenal glands, described by Stewart and Rogoff as “normal,” is confirmed and explained. But one needs only to compare the drop in heart rate after adrenalectomy in acute experimental conditions (see figs. 1 and 2) with the absence of a drop after adrenal- ectomy in the non-anesthetized animal (see fig. 4) to realize how ab- normal is the so-called ‘“‘normal’’ secretion which occurs during operation. Stewart and Rogoff, after considering the possibility that their “extensive operation”? may have caused so great a secretion of the adrenal glands that asphyxia, for example, could not evoke a detectable increase, became convinced that this suspicion was not well founded because they noted, on stimulating the cut splanchnic nerve directly, evidence of a decidedly greater rate of secretion (38). -Obviously, when a nerve is cut and then stimulated, an unusual effect may be due to liberation of material accumulated during the inactivity which followed denervation. Furthermore, because direct stimulation of a nerve, or central excitation by strychnine, will produce certain results, that is not proof that reflex stimulation, done under anesthesia, should produce the same results. For example, there is a marked difference between the intensity of muscular response caused by direct stimulation of the ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION A427 sciatic nerve and that which may be induced by reflex stimulation. Again, an abdominal operation which arouses continuous activity in the splanchnic nerves might readily interfere with splanchnic reflexes. One of the methods employed by Cannon for recording graphically the effect of secreted adrenin in the body was that of denervating the mesentery, as described above (see p. 410). This method required opening the abdomen. It yielded constant results so far as the belated influence of asphyxia was concerned, but was commonly disappointing as a means of demonstrating the early influence of asphyxia; and in the entire series of cases with opened abdomen there was only one in which sensory stimulation caused any effect ascribable to adrenal secretion. For example, in the first five cases of table 2, the abdomen had been opened, and in these instances, though asphyxia was effective, sciatic stimula- tion yielded no response whatever. From this evidence it is clear that, either because the opening of the abdomen produces a secretion unsur- passable by reflex stimulation, or because that operation abolishes abdominal reflexes, the influence of sensory stimulation on the adrenal glands is not manifested. There is little wonder, therefore, that Stewart and Rogoff, who alone have employed the pocket method, with its attendant severe abdominal operation and repeated manipulation of the abdominal contents, failed to obtain the positive results which have been obtained by all other observers. The foregoing facts and considerations warrant the conclusion that although the work of Stewart and Rogoff was admirably quantitative in character, it was done under experimental conditions which could not afford information regarding the normal secretion of the adrenal glands or the natural conditions which affect that secretion. This con- clusion applies to all inferences as to the nature of adrenal activity which they have based upon employment of the pocket method. The method of Gley and Quinquaud. In the paper by Gley and Quinquaud previously mentioned (10), evidence is adduced to prove that adrenal secretion has nothing to do with the efficacy of sym- pathetic nervous impulses as they affect the smooth muscles of blood vessels, a conclusion well supported by the previous observations of Hoskins and McClure (39). Gley and Quinquaud removed blood from the inferior cava immediately above the opening of the subhepatic veins, and again from the right or left ventricle, in each case after splanchnic stimulation. The blood thus obtained was injected in 20 cc. amounts into other dogs weighing from 4 to nearly 10 kilos. Only the blood which was taken from directly above the opening of the THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 428 Ww. B. CANNON adrenal veins caused any rise of pressure in the dog injected. They conclude, therefore, that the adrenin present in adrenal blood after splanchnic stimulation is found neither in the blood of the vena cava above the subhepatic veins nor in the blood of the heart. In drawing this conclusion Gley and Quinquaud seem to have dis- regarded the fact that they were, in the first place, taking only a small portion of the secreted adrenin, which had already been diluted by the blood of the donor, and were then injecting this small portion into the blood stream of another dog, where it would be diluted to a much greater degree. Gley and Quinquaud declare categorically that secreted adrenin is not carried by the circulation to the organs on which it acts, and that, if present at all, it is present in a quantity altogether minimal and insufficient to exercise its action. This declaration again is made with- out due regard to evidence already in the literature. The observations on the denervated limb, on the denervated kidney, on the denervated salivary gland and on the denervated heart, quoted or described above, clearly demonstrate that adrenal secretion may be stimulated by painful impulses, by asphyxia and by emotional excitement, and that the sub- stance secreted under these circumstances not only is carried to the structures on which it acts, but produces on these structures pronounced physiological effects. Until this evidence is definitely proved to be unworthy of acceptation, the conclusion which Gley and Quinquaud have drawn must be regarded as quite unjustified. INTERPRETATION OF THE FUNCTION OF THE ADRENAL MEDULLA With the disappearance of the view that the adrenal glands produce some substance which neutralizes toxic material developed in the body, there have been left two theories to account for the réle played by the adrenal medulla in the bodily economy. These are the tonus theory and the emergency theory. The tonus theory, which has been advocated in the past (40) and still receives attention, holds that the function of the secreted adrenin is to maintain the sympathetic endings in a state of responsiveness to nervous stimulation or in a condition of moderate activity or tone. This view has definitely lost ground in the course of relatively recent investigations. A number of investigators have called attention to the depressive effect of small doses of adrenalin (39), (41). If the smallest dose which will have any influence whatever on the blood ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 429 vessels induces relaxation of the vessels, it is difficult to understand how the function of the secreted adrenin could be that of maintaining a state of tonic contraction. Furthermore, as has been repeatedly noted (42), double adrenalectomy does not for some time cause the fall of arterial pressure which naturally would be expected if continued adrenin secretion were needed to keep the pressure up; and also stimu- lation of the splanchnic nerves induces the same rise of pressure after adrenalectomy as before (10). From these results the conclusion has been drawn by Hoskins and McClure and by Gley and Quinquaud that the tonus theory is without adequate experimental support. The emergency theory was presented by Cannon on the basis of studies of adrenal secretion following stimulation of afferent nerves, asphyxia and emotional excitement. In the papers bearing upon this theory emphasis was repeatedly laid upon the association between adrenal activity and the activity of the sympathetic division of the autonomic system in such emergencies. Nowhere has the statement been made that secreted adrenin has a function separate from that of the nerve impulses, except to increase the irritability of fatigued muscles (3) and to speed the coagulation of the blood (4). The idea originally suggesting these studies on adrenal secretion was that changes in the viscera originally induced by nervous impulses might be con- tinued by circulating adrenin (43, p. 40). No claim has ever been made that there is at any stage a primacy of adrenin in the production of physiological or psychological changes seen during strong emotion. In spite of the foregoing facts authors have written as if Cannon had been attempting to support the idea that emotional experiences were dependent upon circulating adrenin. Thus Stewart and Rogoff report (32), as if the matter had been questioned, that all signs of fright can be elicited by administering morphine to a cat with one adrenal removed and the other denervated.’ Rogoff points out (386) that the secretion of sufficient adrenin to produce symptoms of fright would be impossible, again as if any claim had been made that these symptoms were due to secreted adrenin. Gley and Quinquaud declare that the 6 Stewart and Rogoff noted dilatation of the pupil of the ‘‘denervated eye”’ when animals became frightened, though one adrenal was removed and the nerves to the other sectioned. Cannon and de la Paz tried this method of testing for adrenal secretion but could not persuade themselves that an eye still innervated by the third cranial nerve was really ‘‘denervated’’ and interpreted the prompt dilatation of the pupil in a paroxysm of rage as due to central inhibition of the still active constrictor muscles (see Cannon: Bodily Changes in Pain, Hunger, Fear and Rage, New York, 1915, p. 35). 430 WwW. B. CANNON persistent irritability of the nerves after the adrenal glands have been removed is opposed to the explanation which Cannon has given to ex- periments on the adrenin origin of emotions (10). Indeed, Cannon has been definitely charged with assuming that the reaction to fear and other emotional states is dependent on hypersecretion of adrenin (44). Careful reading of his work gives no support for these interpretations. The concept of an emotion may be expressed either in psychological terms of subjective experience or in physiological terms of bodily change. Cannon’s observations lend no support to the idea that adrenal secre- tion is essential to the subjective experience of strong emotion. Adre- nin has its effect peripherally, on outlying viscera. An assumption that subjective feeling depends on circulating adrenin involves, there- fore, supporting the view that emotion as a psychological state is the consequence of visceral changes. Cannon has, in fact, definitely argued against this view (43, p. 275). If the critics of the emergency theory conceive emotion as bodily change, they will find in Cannon’s consideration of the interrelations of emotions the point emphasized that it is the sympathetic division of the autonomic system which is the primary agency in mobilizing the bodily forces in times of great fear or rage (43, p. 268). To assume that secreted adrenin is necessary for the changes which occur under such conditions implies an acceptance of the tonus theory. This view has not been held by Cannon and receives no support in any observation he has reported. The only suggestion which he has offered (48, p. 64) that might be construed into support of such a view is that adrenal secretion given forth into the blood stream during excitement is a substance capable of inducing or augmenting the nervous influences which bring about the very changes in the viscera that accompany excitement. Naturally, this suggestion should be considered in con- junction with others; e.g., “it is possible that disturbances in the realm of the sympathetic are automatically augmented and prolonged through chemical effects of the adrenal secretion”’ (48, p. 38), and “‘the changes originally induced in the digestive organs by nervous impulses might be continued by circulating adrenin’” (43, p. 40). These suggestions imply codperation of chemical and nervous factors, but not a depen- dence of the nervous factors on the chemical. The possibility has been recognized (43, p. 65) that in times of emo- tional stress there may be coéperation of secreted adrenin with the prod- ucts of other endocrine glands simultaneously excited, which might render the adrenin much more effective than it would be by itself. ISOLATED HEART AS INDICATOR OF ADRENAL SECRETION 431 This is a possibility which should be kept in mind in connection with the emergency theory of adrenal secretion. Until this possibility has been tested, however, there is no need of going further than the facts will warrant in appreciating the codperative character of secreted adrenin and sympathetic nervous impulses. Thus far no reliable evidence has been brought out by any investi- gator that there is any secretion whatever of the adrenal glands under quiet, peaceful conditions. Results reported in this paper present the first indication that under such conditions there is no adrenal secretion or a secretion so slight as not to affect the denervated heart, an extremely sensitive indicator. Stewart and Rogoff have shown that the cat and the dog will live normally for weeks with one adrenal excised and the other denervated, an operation which results in no demonstrable flow of adrenin from the adrenal vein (45). These observations prove that adrenal secretion is not a necessity, at least in times of serene existence. Adrenin is secreted, however, in times of great emotional stress and under circumstances which cause pain or asphyxia. As stated at the beginning, the function of the adrenal medulla is to be looked for under conditions which rouse it to action. Excitement, pain or asphyxia are, in natural existence, commonly associated with violent struggle for self-preservation. Under such circumstances, as has been emphasized in the presentation of the emergency theory, the operation of the sympathetic division of the autonomic system together with the aid which adrenin affords will muster the resources of the organism in such a way as to be of greatest service to such organs as are absolutely essential for combat, flight or pursuit. It appears, therefore, that the emergency theory of the adrenal medulla is the only one which thus far has any experimental support. It is a pleasure to express my thanks to Mr. H. F. Pierce for help in the early experiments above reported, and to Dr. Alexander Forbes for making the electrocardiograms. ; BIBLIOGRAPHY (1) CANNON AND DELA Paz: This Journal, 1911, xxviii, 64. (2) Cannon AnpD Hoskins: This Journal, 1911, xxix, 274. (3) Cannon anpD Nice: This Journal, 1913, xxxii, 44. Also Gruser: Ibid., 1914, xxxili, 335. (4) CaNNON AND MENDENHALL: This Journal, 1914, xxxiv, 243 and 251. (5) Cannon: This Journal, 1914, xxxill, 356. (6) Stewart AND Roaorr: This Journal, 1917, xliv, 149. 432 W. B. CANNON (7) Sr—ewaRT AND Rocorr: Journ. Exper. Med., 1917, xxvi, 637. (8) Stewart AND Rocorr: Journ. Pharm. Exper. Therap., 1917, x, 49. (9) StEwaRT AND Rocorr: This Journal, 1917, xliv, 548. (10) GLtey anp Qurnquaup: Journ. d. Physiol. et d. Path. Gen., 1918, xvii, 807. (11) Anrep: Journ. Physiol., 1912, xlv, 307. (12) PEARLMAN AND VINCENT: Endocrinology, 1919, iii, 126. (13) Levy: Heart, 1913, iv, 342. (14) FLorovsxy: Bull. Acad. Imp. d. Sci., Petrograd, 1917, ix, 119. (15) Iramr: Journ. Physiol., 1912, xlv, 338. (16) Gasser AnD Meek: This Journal, 1914, xxxiv, 63. (17) GiEY AND QuiNquatp: Compt. rend. Soc. d. Biol., 1917, Ixxx, 16. (18) Hrrcnines, SLOAN AND AusTIN: Cleveland Med. Journ., 1913, xii, 686. (19) ReprieLp: Journ. Exper. Zodl., 1918, xxvi, 295. (20) Lamson: Journ. Pharm. Exper. Therap., 1915, vil, 212. (21) Burton-Opitz AND Epwarps: This Journal, 1917, xliui, 408. (22) Brmpu: Arch. f. d. gesammt. Physiol., 1897, Ixvii, 456. (23) Cannon: Science, 1917, xlv, 463. (24) Czusatsxi: Zentralbl. f. Physiol., 1913, xxvii, 580. (25) Fuinr: Johns Hopkins Hosp. Repts., 1900, ix, 171. (26) Cow: Journ. Physiol., 1914, xlviii, 446. (27) SteEWART AND Rocorr: This Journal, 1918, xlvi, 92. (28) Martin: Studies, Biol. Lab., Johns Hopkins Univ., 1882, ii, 213. (29) KNOWLTON AND STarRuinG: Journ. Physiol., 1912, xliv, 206. (30) Parrerson: Proc. Roy. Soc., London, 1915, lxxxviii, B, 380. (31) Stewart AND Roaorr: Journ. Pharm. Exper. Therap., 1919, xiii, 397. (82) Stewart AND Rocorr: Journ. Pharm. Exper. Therap., 1915, viii, 479. (33) TscHEeBoxsarorF: Arch. f. d. gesammt. Physiol., 1910, exxxvii, 59. (84) Stewart AND Rocorr: Journ. Exper. Med., 1917, xxvi, 618. (85) STEWART AND Roaorr: This Journal, 1919, xlviii, 22. (36) Rocorr: Journ. Lab. and Clin. Med., 1918, iii, 214. (37) Baytiss AND STaRLina: Journ. Physiol., 1899, xxiv, 122. (38) Stewart AND Rocorr: Proce. Soc. Exper. Biol. and Med., 1917, xiv, 77. (39) Hoskins anp McCuvre: Arch. Int. Med., 1912, x, 343. (40) See Exuiorr: Journ. Physiol., 1904, xxxi, 20; 1914, xlix, 38; Brepu: Innere Sekretion, Berlin, 1913, ii, 30. (41) CANNON AND Lyman: This Journal, 1913, xxxi, 376. (42) Lewanpowsky: Zeitschr. f. Klin. Med., 1899, xxxvii, 535; Camus AND LaNG.Lotis: Compt. rend. Soc. d. Biol., 1900, li, 210; Hoskins aNpD WHEELON: This Journal, 1914, xxxiv, 172. (43) Cannon: Bodily changes in pain, hunger, fear and rage, New York, 1915. (44) See Journ. Amer. Med. Assoc., 1919, Ixxili, 192. (45) Srewarr AND Rocorr: Journ. Pharm. Exper. Therap., 1917, x, 1; This Journal, 1919, xlviii, 397. EFFECT OF WORK AND HEAT ON THE HYDROGEN ION CONCENTRATION OF THE SWEAT G. A. TALBERT From the Physiological Laboratory of the School of Hygiene and Public Health, Johns Hopkins University Received for publication October 2, 1919 INTRODUCTION ' Early in the year 1918 my attention was attracted to some work that had been done several years ago demonstrating, incidentally, how muscular fatigue might affect the reaction of the urine. The fact that the results were conflicting, and particularly that all the investigators used methods of titration, which we now recognize as quite unreliable for obtaining total acidity, led me to think that this might prove an interesting field for research. Furthermore, inasmuch as the literature seemed to be silent as to whether exercise in any way affected the reaction of the sweat, it seemed to be desirable to make determinations on the urine and sweat simultaneously, with the subject under the same conditions, with the view first, of establishing more definitely the results of the earlier observers as far as the urine was concerned; second, to ascertain if any reaction changes occur in the perspiration; and third, whether these changes, if they occur, show either a supplementary or a compensatory relationship with the urine. With these ideas in mind I made a few preliminary tests in Ripon College chemical laboratory through the courtesy of Dean Barber and Professor Barker. However, the problem was not attacked in an in- tensive manner until I had the privilege of pursuing the work during the school year of 1918-1919 in the physiological laboratory of Professor Howell in the School of Hygiene and Public Health of the Johns Hop- kins University. It therefore seems most fitting that I should here acknowledge my deep gratitude to him for his helpful suggestions and above all the inspiration that I received from his kindly interest. In the same manner I wish to thank Doctor Spaeth of this laboratory for assistance with the experimental technique. 433 434 G. A. TALBERT It is my purpose in this article to consider simply the perspiration, and to discuss subsequently the results obtained from the urine. We have in the past been quite willing to assume that the principal function of the sweat glands is for the regulation of body temperature, and not much emphasis has been laid upon the significance of the com- position of the excretion. Before entering upon the discussion of the problem some of the views that have been held in the past as to the reaction of the sweat, more or less conflicting, may be stated briefly. Foster (1) states that ‘“‘Sweat from a well washed skin is alkaline. It is only when mixed with sebum that it is acid. Horses’ sweat is said to be always acid.” Smith (2) finds horses’ sweat strongly alkaline. Moriggia (3) states that in herbivora the reaction is generally alkaline, while in carnivora it is acid. Gaube (4) speaks of the human sweat as acid, while that of the horse, cow, dog, cat and hog is alkaline. Cer- tainly omnivora, herbivora and carnivora are found in this last list. In Landois’ text (5) we find the statement that “‘swine sweat (?) on the snout, cattle about the mouth (?), while goats, rabbits, rats, mice and dogs do not sweat at all.”’ Aron has stated that monkeys have no sweat glands but Shaklee (20) in later work reports that this is an error, “the skin seems everywhere provided with well-developed sweat glands.” METHODS As has been suggested above, the criticism that may be made upon earlier work is concerned with the uncertainty of the data obtained by titration. In the observations which follow, use was made of the colori- metric and the gas chain methods in order to obtain the total acidity. As to the former, I followed mainly the method of Henderson and Palmer (6), which is an adaptation of the Sérensen method, while for the gas chain measurement I used a Leeds and Northrop’s potenti- ometer with Clark’s (7) hydrogen electrode shaker. It was hardly thought necessary to use a constant temperature chamber, but a ther- mometer was kept in the potassium chlorid vessel and therewith the corrections were made. For the colorimetric method it may be stated briefly that there are required five standard solutions made up of monopotassium phosphate and disodium phosphate, mixed in such proportions that each solution will possess a distinct and definite pH value. In like manner, acetic acid and sodium acetate are so mixed as to give six standard solutions of EFFECT OF WORK AND HEAT ON PH OF SWEAT 435 definite pH values. The former mixture is for use on the alkaline side with values ranging fron 8.7 pH to about 7.0 pH. The latter reaches from 7.0 pH to 4.7 pH. These standards are as a rule made fresh once a week, but never kept longer than two weeks. When ready for tests, 4 ec. of each standard was taken and diluted with distilled water up to the markin a 100 ce. flask, and a similar dilution was taken for the sweat. An equal amount of indicator added to the same volume of the diluted standard and the diluted sweat and the colors thereby matched gave a fair evaluation for the reaction of the latter even where interpolation was necessary. As to the indicators, phenolphthalein can be used for values between 8.7 pH and 8.0 pH. Neutral red, however, is very useful between 8.0 pH to even below the neutral point 7.0 pH. Methyl red is excellent on the acid side from 5.7 pH to 4.7 pH, and is not affected by proteins, while p.nitrolphenol covers a wider range and is an excellent indicator between 6.7 pH and 4.7 pH. Sodium alizarine sulphonate covers the entire field from 8.0 pH to 4.7 pH, thereby meeting all of the variations that I have found in the perspiration. There were, at times, in use the indicators that are recommended by Clark and Lubs (8), which were quite satisfactory. In the main the determinations were made by the colorimetric method, while the potentiometer was used for checking up the standard solutions. Unfortunately the gas chain was not brought into use as often as one might wish, because at times tedious delays were caused by not being able to get the services of a mechanician. I am aware that objections may be urged against placing too much reliance upon absolute evaluation by the colorimetric method. ‘‘Off- shades”’ did arise at times, especially with the use of sodium alizarine sulphonate. In that case determinations had to be made with other indicators or the test would have to be rejected. The protein and salt errors have been long recognized by Sorensen and others. As to the former, I have a feeling that the per cent in the sweat is so low that its effects are quite negligible. As to the latter, I have not that same confidence, for the work of Viale (9) shows variations in the salt content. Of late some have laid considerable emphasis upon the variations due to the carbon dioxid factor. However, after making due allowance for these liabilities to error it must not be overlooked that Lubs and Clark (10) and others have obtained some remarkable agreements in the two methods. 436 G. A. TALBERT Granting that there might be some slight errors in the absolute, there can hardly be any errors in the relative values, as the samples to be compared were tested almost simultaneously. My first tests were made on volunteers from the Baltimore Central Y. M. C. A. and upon a few of the workmen about the laboratory. These were hardly more than preliminary try-outs, as was ultimately proven. In some cases the skin was cleansed with water or alcohol, while in others not at all. The samples which were collected at the Y. M. C. A. were really the combination effect of heat and work, for the subjects, after playing basket-ball or volley-ball, as the case might be, came into the hot-room where the samples were obtained. One important point revealed by these tests was that the sweat contains a greater concentration of hydrogen ions than one would naturally suppose. However, the futility of the volunteer plan soon became apparent. It was evident that in order to make the experiments worth while it was necessary to have reliable and well selected subjects who would be willing to come to the laboratory, where it would be possible to have better controls. This end was realized by obtaining as subjects some medical students, who had a much better appreciation of the importance of the work and were willing to codperate in a most excellent manner. The few subjects that I used, mainly, were in perfect health, as they had passed the tests as donors of blood for transfusion cases. After undertaking the work in the laboratory my first thought was to ascer- tain if exercise produced any changes in the hydrogen ion concentration of the sweat. Consequently there arose the suggestion of a control. My idea was to obtain heat-sweat first and then secure work-sweat immediately afterwards, with the feeling that if any change took place in the latter it would be as in the case of the urine in the elimination of more acid as a result of work. The subjects were stripped and the parts from which the sweat was to be taken, viz., face, chest and abdomen, were first washed with soap followed by cleansing with water, ether and alcohol in-the order named. In the application of the last two liquids dental napkins were used. The subjects were then placed in a hospital sweat-cabinet with a fair amount of moisture, starting with a temperature of about 30°C. and finally increasing it to 40° or 45°C. The heating lasted from fifteen to twenty-five minutes, according to the subject and the number of samples to be obtained. All samples of sweat were collected by means of lipless specimen tubes. EFFECT OF WORK AND HEAT ON PH OF SWEAT 437 After this procedure, the parts being controlled as before, the subject was placed on a stationary bicycle where he worked for fifteen to twenty- five minutes. The samples of work-sweat and heat-sweat were then tested as soon as possible and comparisons were made. ri The most striking thing about the data, in the securing of which there were sixteen observations upon six different individuals, was that in each instance the heat-sweat was of a greater hydrogen ion concen- tration than the work-sweat. (See table 1, series A.) TABLE 1 HEAT PRECEDING WORK SERIES A WORK PRECEDING HEAT SERIES B Subjects PH eat DH work Difference Subjects pH pork: pH lbat- Difference G. 5.4 5.9 0.5 G. 5.8 5.4 0.4 G. 0) 5.65 0.15 G. 5.8 5.5 0.3 G. 5.25 5.8 0.55 G. sf 5.5 0.2 Ve 5.8 6.4 0.4 G: 5.9 5.2 ORs Jj. 6.4 6.6 0.2 G. 5.9 5.15 0.75 Ve 5.4 7.0 1.6 Ale 5.6 5.6 0.0 EK. Uo Ug 0.3 do 6.1 5.5 0.6 E. Una! 7.4 0.3 E. 7.4 5.3 2.1 E: 6.0 7.4 1.4 F. 5.8 5.2 0.6 E. 5.6 7.4 1.8 C 5.45 4.7 0.75 E. ond 6.5 0.8 S. 6.15 5.5 0.65 E. Sai 6.5 0.8 PSS E. 5.6 6.8 1.2 6 5.96 5.32 |Av. dif. 0.64 B: 6.2 7.4 2 A. 5.6 6.0 0.4 M. 5.1 5.9 0.8 5.22 6.63 |Av. dif. 1.41 or) Fearing that the first excretions might be more acid due to the sebum or other causes, I reversed the process by producing work-sweat first. I made eleven observations with the use of six individuals, with the result that the heat-sweat was still of higher acidity. (Note table 1, series B.) I next adopted the plan for a few experiments upon subject G. of pro- ducing sweat in the morning by work and by heat in the afternoon, and then reversed this process, with no particular differences in the results. All of the remainder of the experiments were performed on two indi- viduals simultaneously, one producing heat-sweat and the other work- 438 G. A. TALBERT TABLE 2 TESTS | SUBJECTS {DE Epes os TESTS | SUBJECTS gpEe ees i G. 5.9 5.5 1 E. Uae Coe 2 G 5aG5i |) 25e4 2 E. 7.4 Coll 3 G. 5.8 Se20 3 107 a5) We 4 G. 5.8 525 4 E. 7.4 Uo 5 G. 5.7 Bee 5 E. 7.4 6.0 6 G. 5.9 5.0 6 E. 6.5 5.6 @ G. 5.8 §.25 at E. 6.5 50 8 G. Bul 55 8 E. 6.8 5.7 9 G. 5.6 5.4 9 E. 7.4 5.6 10 G: 5.8 5.4 10 E. 5.8 11 G. 5.8 5.4 11 E. eo 12 G. 6.0 5.85 13 G. 5.6 | 5.6 7.15 | 6.24 14 G. 5.6 5.6 15 G. 5.8 5.4 16 G. 5.8 1 realy 6.4 5.8 17 G. acl 2 Al 6.6 6.4 18 G. 5.35 3 Je 5.6 5.4 19 G. DOD 4 Abe 6.1 6.4 20 r 5.9 5 Je 5.8 ram 6 I. Nore RB | “Ossal 6.18 | 5.838 1 M. jail Sal 2 M. 5.4 5.4 3 M. 5.95 | 5.85 4 M. 5.65 || 5.26 5 M. 5.8 5.6 6 M. 5.8 5.4 ai M. 6.4 5.8 5.84 |— 5.54, (05380 TABLE 3 OBSERVATION SUBJECTS pH work OBSERVATION SUBJECTS pH Hear 20 G. he 15 G. 5.42 i M. 5.84 i M. 5.54 9 EH. 7.15 11 1Bp 6.24 4 AI. 6.18 6 di 5.83 13 De 6.19 14 X. 5.61 Total 53 Av. 6.22 Total 53 Av. 5.73 —_ EFFECT OF WORK AND HEAT ON PH OF SWEKAT 439 TABLE 4 SERIES A HEAT-SWEAT pH SERIES B WORK-SWEAT pH Subjects | Firstsample oe Third sample Subjects First sample Fecond G. 5.5 5.6 G. 5.7 5.8 G. 5.5 5.65 G. 5.6 5 65 G. 5.2 5.45 G. 5.8 5.9 G. 5.12 5.2 (Ce 6.0 5.8 G. 5.0 5.2 G. 5.6 5.55 G. 5.25 5.3 5.5 G. 5.8 5.6 G. G55) 5.55 5.6 G. 5.7 5.9 G. 5.4 5.55 HeOo G. 5.9 5.8 G 5.85 5.15 G. 5.35 5.3 G. 5.6 5.45 G. 5.7 5.75 G. 5.6 5.3 G. 5.9 5.95 G. 5.4 5.45 G. 5.9 5.6 G. 5.4 5.5 (Gy 5.65 5.7 G. 3.0 5.6 G. 5.8 5.75 G. 5.25 5.2 M. 5.4 5.15 5.5 _M. I), 5) 5.8 M. 8.0* 5.2 5.25 M. 5.95 5.15 M. 5.85 5.8 M. 5.65 5.9 M. 5.6 5.35 M. 5.8 5.85 M. 5.4 5.25 M. 5.8 5.4 M. 5.8 5.6 5.8 M. 6.4 6.3 M. 5.1 5.2 5.4 M. 5.9 5.8 M. 5.8 6.0 E. 5.9 pC E. 5.4 5.25 F. 4.85 4.7 ea Ss. 5.35 ed 5.8 J. 5.2 5.35 5.3 O. 5.9 Sat * This unusual reading may be due to the fact that the subject had served as a donor in a blood transfusion three hours before the experiment. 440 G. A. TALBERT sweat, while the next day the performance was reversed. The grand totals from these experiments are given for four of the subjects in table 2. Table 3 gives averages of the four principal subjects, while X. repre- sents the average of several individuals taken together where not more than one or two tests were made on each person. In table 4, series A, are found twenty-eight experiments in which two samples of heat-sweat were taken, while in series B of the same table we have twenty-two experiments in which two samples of work-sweat were taken. It is a remarkable coincidence that just 50 per cent of the second samples increased in acidity in both kinds of experiments. There were ten tests in which three samples were taken successively as a result of heat. Comparing the third with the first sample we now find that eight are less acid, while one is more acid and one remains the same. In comparing the third with the second sample we again find that eight are less acid, one is more so, while one remains the same. The skin was cleansed as before after each collection, and an interval of five minutes elapsed between the collection of the successive samples. CONCLUSIONS In my conclusions I wish to emphasize that in many cases the changes were exceedingly small, yet the distinction was always obvious. Further- more it is to be remembered that the work experiments were not long- enduring and fatiguing tests like those that have been reported in the past on the urine. On the contrary the subjects as a rule had per- formed their part of the experiments within one hour. While the time was short, the experiments, especially from exercise, were rather intense. In a majority of cases these tests took place late in the forenoon, rang- ing from two to five hours after a meal. The work reveals the following facts: First, that sweat caused from either work or heat is acid, probably always so in perfect health, and the degree of acidity is greater than we have heretofore believed. Second, in a continued secretion of sweat the reaction does not remain entirely constant. A second sample may show a slight increase or decrease in acidity, while a third sample shows practically in all cases a small but distinct diminution compared with the first sample. Third, the sweat caused by external heat is always more acid than that caused by muscular work. EFFECT OF WORK AND HEAT ON PH OF SWEAT 441 Many authors have assumed that the secretion of sweat is normally alkaline and that the acid reaction actually shown in many cases is due to admixture with sebaceous secretion. But Francois-Franck (11), Kitt- steiner (12) and others, state that the sweat from the palm of the hand is acid, although this portion of the skin is devoid of sebaceous glands. The observations reported in this paper also throw doubt upon this expla- nation of the acidity of sweat as usually collected, since the precautions taken to cleanse the skin before collecting the sample should have been sufficient to remove any deposit of sebaceous material. The immediate cause of the acidity of sweat has not been determined satisfactorily. Halliburton (13) states that the sweat, like the urine, contains acid phosphates, but this explanation has not been corroborated by satis- factory analyses. Others have assumed the presence of volatile fatty acids in the secretion, and some observations of my own tend to support this view. In a number of cases the insensible perspiration was col- lected by fixing a finger suitably in a glass chamber, so that the vapor would condense upon the walls of the vessel. When diluted and com- pared with distilled water this condensate gave always an acid reaction. Aubert (14), Rohig (15), Schierbeck (16), Fubini and Ronchi (17), have emphasized the importance of carbon dioxid. The amount of carbon dioxid excreted varies with exercise and especially with external temper- ature. Schierbeck, for example, found that at a temperature of 29.8°C. there was an elimination of 8.9 grams of carbon dioxid in twenty-four hours, while at a temperature of 38.4°C. the amount excreted in the same period was 29.5 grams. The larger excretion of carbon dioxid would increase the acidity of the sweat as secreted, but presumably this factor did not enter into the reactions as determined by the method described in this paper. In these determinations the diluted speci- ‘mens were exposed freely to atmospheric air, and presumably took on a corresponding tension of carbon dioxid. The statement of Heuss (19), “‘Der schweiss reagirt in der Ruhe normaler weise sauer bei profuser secretion (Pilocarpin, Schwitzbiider) kann er neutral ja sogar alkalisch werden,’ I could hardly support unreservedly. So far I have not tested any sweat produced by drugs, so I have nothing to offer on that point. However, when I have fol- lowed work with heat I have found that the sweat from the latter is not only more acid, but more profuse. So the question of the profuseness, so often referred to in the litera- ture, does not in itself offer a satisfactory explanation of differences in reaction. 442 G. A. TALBERT The fact that heat-sweat shows uniformly a higher hydrogen ion concentration than work-sweat is surprising and contrary to expecta- tion. In muscular work there is a large increase in the acid products of metabolism, and the output of acid in the sweat can be understood as part of the mechanism for preserving the acid-base equilibrium of the body. In heat-sweat we have heretofore regarded the secretion and evaporation. of the water as an important means of controlling the body-temperature, the so-called physical regulation of the heat- equilibrium of the body. The fact that this heat-sweat is acid may be looked upon as a demonstration that the sweat in man under ordi- nary dietary conditions is normally acid, and that this secretion, like that of the urine, helps to maintain the acid-base equilibrium of the organism. But why the heat-sweat should exhibit a greater acidity than the work-sweat is not clear. It is hoped that further study of this problem may prove not only of physiological, but of therapeutical value. BIBLIOGRAPHY (1) Foster: Textbook of physiology, 1890, 560. (2) Smrrxa: Journ. Physiol., 1890, xi, 497. (3) Morteeta: Moleschott’s Untersuch. z. Naturlehre, xi. (4) GausE: Memoires Soe. d. Biol., 1891, 115. (5) Lanpois: Textbook of human physiology, 10th ed., 1905, 534. (6) HENDERSON AND Patmer: Journ. Biol. Chem., 1913, xiii, 393. (7) Cuarxk: Journ. Biol. Chem., 1915, xxii, 478. (8) CLARK AND Luss: Journ. Bacteriol., 1917, ii, 1, 109, 191. (9) Viaute: Arch. Ital. d. Biol., 1913, lix, 269. (10) Luss anp Cuark: Wash. Acad. Sci., 1915, v, 609. (11) Francois-Francxk: Dict. Encyl. d. Sci. Med. Paris, 1884, Ser. 3, xili, 51, Sueur. (12) Kirrsterner: Arch. f. Hygiene, 1911, Ixxiii, 275; 1913, Ixxvili, 275. (13) Haturpurton: Kirk’s Hand-book of physiology, 1907, 592. (14) AuBERT: Pfliiger’s Arch., 1872, vi, 539. (15) Réute: Deutsch. Klinik, 1872, 209. (16) ScutreRBeEcK: Arch. f. Physiol., 1893, 116. (17) Fusrnt AND Roncur: Moleschott’s Untersuch. z. Naturlehre, xii. (18) Cramer: Arch. f. Hygiene, 1890, x, 231. (19) Heuss: Monatschr. f. Prack. Dermatol., 1903, xiv, 10. (20) SHAKLEE: Philippine Journ. Sci., 1917, xii, B. 1. EFFECT OF PHYSICAL TRAINING AND PRACTICE ON THE PULSE RATE AND BLOOD PRESSURES DURING ACTIVITY AND DURING REST, WITH A NOTE ON CERTAIN ACUTE INFECTIONS AND ON THE DISTRESS RESULTING FROM EXERCISE! PERCY M. DAWSON From the Physiological Laboratory of the University of Wisconsin Received for publication October 4, 1919 INTRODUCTION There are two methods of conducting an inquiry into the effects of physical exercise. On the one hand one may make a study of a very large number of subjects and average the results. The objection to this method is that it assumes that the numerous irrelevant variables, of which no account is taken, cancel one another so that the result indi- cates the action of a single common factor (exercise) and of its variations. On the other hand one may make an intensive study of a very few subjects. The objection here is that we are prone to assume that what is true of these few subjects is true of all potential subjects, whereas the cases studied may be exceptional ones. To those who do not forget the limitations of these methods both are useful, suggestive and devoid of dangers. In the present instance the writer has chosen the second of the two, and has made an intensive study of a single subject, namely himself. THE SUBJECT The subject’s history in so far as it has any possible bearing on the present research is as follows: 1892-1895 intercollegiate athletics (Lacrosse) ; 1896-1905 in winter, excepting two months in mid-winter, vigorous exercise (one to two mile run several times a week) ; 1896-1907 in summer very strenuous mountain climbing sometimes involving feats of endurance. Since these dates systematic exercise has been 1A preliminary report of this work was made at the annual meeting of the American Physiological Society, December 1916 (5). 443 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 444 PERCY M. DAWSON moderate and only for relatively short periods. Alcohol rarely and then in moderation, since 1907 not at all; no tobacco; coffee has been used occasionally. Between 1893 and 1911 acute dilatation occurred on four occasions. Recovery was always complete and prompt. Age 41 years in 1914 when these experiments were begun. A partial physical examination of the subject made on April 27, 1916, gave the following data of possible relevancy to this research: Height 182.3 em. weight 66.5 kilos (on June 7, 1916, 67.5 kilos); heart silhouette area 119.9 cm.? which is too small for age (43 years) and height, but normal for body weight according to Bardeen’s tables. The estimated systolic out-put was 41 percent (Bardeen). A brief discussion and skiagraph of this heart has already been published (1). METHOD OF MAKING OBSERVATIONS The blood pressures were determined by means of the auscultatory method. Whenever the routine determinations were taken during rest (tables A, B and C) from five to twelve (usually about seven) separate observations were made of both systolic and diastolic pressures. This precaution eliminated variations due to respiration and possible Mayer’s waves.’ If the first systolic readings were markedly higher than those which followed, these first readings (with corresponding diastolic readings) were rejected on the ground that their exceptionally high level was probably the result of previous exertion so that they were not comparable with the subsequent readings. In regard to this matter a considerable amount of care has always to be exercised since the systolic pressure may be influenced by activities which appear quite inconsiderable, as going up or down stairs, going from room to room, or even fetching and arranging the necessary apparatus. The satis- factory readings were averaged and it is these averages which are used in the compilation of the accompanying tables (tables A, Band C). All the calculations of the averages were performed twice’ independently to eliminate errors. The counting of the pulse rate (one minute period) always followed immediately after the blood pressure determinations when the latter were made. When for any reason they were omitted, the pulse was counted until its rate became constant. 2 The writer must again protest against the use of the term ““Traube Herring wave’’ to designate vasomotor waves which are not synchronous with the respi- ratory movements (6), (7), (9). 3 J have much pleasure in thanking my daughter Emily for performing the tedious task of making one of the two series of calculations. EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 445 While making these observations the subject remained comfortably seated. In no case was there demonstrable any psychical variation of either pulse or pressure. As a matter of fact there were usually no emotions accompanying the observations except occasionally an intense impatience and desire to have done with the readings. This state of mind appeared not to affect the readings at all. The observations of pulse and blood pressure fall naturally into two categories: the daily routine observations and the special observations made in connection with the special physical tests and exercises. The routine observations may be divided into four sets according to the time of day at which they were taken. These times were a, imme- diately after rising (about 6 a.m.); 6, just before lunch (12 noon); c, late afternoon (4: 30 to 6 p.m.); d, before retiring (9:30 to 11:30, usually 10 p.m.). These four sets are referred to in this paper as “morning,’’ “noon,” “afternoon” and “evening” respectively. These routine examinations show the changes in the resting pressure and pulse rate which were induced by systematic exercise. In addition to these there was, as stated above, another category of observations, namely those which were made in connection with specific exercises, during and after riding the -cycle ergometer and those made after the three-mile runs. In both cases the observations made before the exercises were part of the daily ‘‘routine”’ observations. These observations may be designated as ‘‘special’’ to distinguish them from the ‘‘routine”’ observations. MODES OF EXERCISE Series I. From December, 1913, to April 1, 1914, the subject took no exercise whatever. Soon after the latter date, however, he began to take exercise at first gently, later with some vigor. The routine obser- ‘vations are given in table A. The dates of occurrence of the exercise are seen in the accompanying calendar (1914). The exercise generally consisted in alternately walking and running for a distance of about three miles. As the strength of the subject improved he soon began to run the entire distance. It should be noted that owing to the age and prudence of the subject, the run was at a leisurely pace so that it required about thirty minutes for its accomplishment. There was no attempt made to shorten the time. The run ended in a short sprint which increased in vigor as the condition of the subject improved, but which did not appreciably shorten the time of the whole run. Both 446 PERCY M. DAWSON before and after the period of systematic exercise (three-mile run) special tests were made by means of the bicycle ergometer. On these occasions the blood pressures and pulse rate were determined before, during‘ and after the ride. The pulse rate during the test was obtained in part by palpation, in part from a record of the carotid pulse obtained by means of tambours with air transmission. No observations were made during the run, but only before and after. Those of the former period were in reality a part of the daily routine examinations already referred to. The method of securing the observations after exercise was the following: On returning from the run the subject at once dropped into an arm- chair and in about one-half minute was beginning to make blood pressure determinations. These were made in rapid succession for several minutes. Meanwhile an assistant (the subject’s wife, to whose assiduity the writer is much indebted) adjusted the cuff of the Erlanger apparatus to the subject’s ankle and in this manner obtained a graphic record of the pulse rate. The desirability of making a graphic record of the pulse rate depends upon the fact that after exercise the rate of the pulse falls from its maximum too rapidly to be estimated by means of the watch and palpating finger. One must therefore measure the duration of single beats or of small groups of beats. These facts are gathered together in the accompanying calendar (1914). Calendar 1914 All exercise ceased early in December. April 1. First ride on cycle ergometer. Special observations were made during and after ride. April 2. Daily routine observations of pressure and pulse rate begun. April 5 to April 20. Series of runs and walks in preparation for three mile runs. Special observations made only after runs. April 22 to May 27. A series of three-mile runs, 14 in 35 days. Special obser-- vations made only after runs. May 29 and June 8. Second and third rides on the cycle ergometer respec- tively. Special observations made during and after ride. June 2. Daily routine observation discontinued. “Series IZ. In 1915 the observations were begun on March 5 and ended May 5. From early in December to March 13, the subject took no exercise whatever. He then began a systematic daily performance of 4 During the test the blood pressure observations were made by Professor Eyster, assisted by Miss Cantril. The latter embodied a presentation of a part of this work in her Master’s thesis (2). EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 447 J. P. Miiller’s exercises in a moderated form (8). Later the exercises became more vigorous, but never reached in severity and tempo the standard set by Miilier. ; No special tests were made with the bicycle ergometer or in other ways, but routine observations were made of blood pressures and pulse rate from March 5 to April 19 (table B). These facts are gathered together in the accompanying calendar (1915). Calendar 1916 All exercise ceased early in December. March 5. Daily routine observations of blood pressures and pulse rate begun. March 13 to April 13. Modified Miiller’s exercises. April 19. Daily routine observations discontinued. Series III. In 1916 observations were begun on March 30 and con- tinued until June 17 (table C). From December 1, 1915, to April 14, 1916, the subject had gone without exercise. On April 14 systematic exercises were begun. They consisted in riding upon the bicycle ergom- eter. On May 7 this exercise was discontinued. On June 6 and 7 two tests were made by means of the bicycle ergometer. Here, besides the daily routine observations of the pressures and pulse rate, the pressures (but not the pulse rate) were determined immediately after the rides upon the cycle ergometer in a manner similar to that described in the case of the three-mile runs undertaken in 1914. It should be noted in passing that the observations in 1916 were affected but not interrupted by two ‘‘colds,’’ one at the beginning and one near the end of the period of observation. These facts are gathered together in the accompanying calendar (1916). Calendar 1916 All exercise ceased early in December. March 30. Daily routine observations begun. April 3. First acute infection begins. April 14 to May 7. Six rides on cycle ergometer followed by special determi- nations of blood pressures. May 9 to June 3. Tennis. June 6 and June 7. Two rides on eycle ergometer followed by special deter- mination of blood pressures. June 8. Second acute infection. June 17. Daily routine observations discontinued. 448 PERCY M. DAWSON ABBREVIATIONS Since it is highly important that the reader should always have in mind just what is implied in the reference to, let us say, the series of observations performed in 1914, namely, that in this year the form of exercise taken consisted in a three-mile run, the writer has resorted to the following way of designating the three series of experiments, namely: 14-run, ’15-gym, 16’-cy-ten, where the suffixes run, gym (gymnastics), cy (cycle ergometer) and ten (tennis) signify the variety or varieties of exercises peculiar to the series mentioned. RESULTS I Effect of training on the resting pulse rate. The effect of training upon the pulse of the subject while at ‘‘rest,” that is to say, while doing only the ordinary work of the day and being therefore quite uninfluenced by any active physical exercise, was, with a single exception, to cause a decrease in the rate amounting to 3 to 9 beats per minute. (See table 1 and fig. 1.) The exception referred to is the rate on rising of series 15-gym where there is a slight increase in rate (one beat per minute). The rise is too small to lie beyond the limits of error but it is significant that in the morning observations of this series there was no decrease in rate. The exceptional nature of this result is probably due to the fact that in this series the exercise was exceptionally light (modified Miiller’s system). TABLE 1 Effect of training on the resting pulse rate NUM- PULSE SERIES ba ottcte TIME OF DAY DATES (BEFORE) ae DATES (AFTER) VATIONS 5 5 a|< ra aan 9 | Noon April 2-25 54|-51) May 10—June 2 ay <— ie 3 | Afternoon | April 2-4 61) 52) May 31—June 2 enti 10 | Onrising | March 30—-April 10* | 58) 52) May 29-June 7 See aire WENoon April 1-17} 58| 50/ May 29-June 7 Hien 6 | Onrising | March 6-12 55| 56} April 12-17** ye ee 7 | Noon March 5-12 61| 55| April 8-19 EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 449 Table 1. The table shows that in the first column the series (date) and variety of exercise, given in order of the decreasing severity of the exercise; in the second, the number of observations from which the averages were obtained; in the third, the time of day at which the obser- vations were made; fourth, the period (dates) during which observations were made prior to or early in the beginning of training; fifth and sizth, the average pulse rates corresponding to the dates given in columns four and seven respectively; seventh, the period (dates) during which the observations were made late in or subsequent to the period of training. * Omitting April 3 and 4 on account of sickness. ** Omitting April 13 when exercise immediately preceded the routine observations. +tOmit- i UWE Md Afternoon Morning ree eR Noon Evemr3. PR. i dal unt A Fig. 1. Effect of training upon the resting pulse rate. The data presented in this figure are from series ’16-cy-ten, table C. The lines in this figure represent pulse rates above or below 50 beats per minute. The lines are gathered into four groups corresponding to the time of day at which the observations were made. The observations in each group are arranged in order of time from left to right and wherever an observation is omitted a space has been left. Where the rate was just 50 per minute, a small stroke has been placed crossing the ‘‘50- line.’’ Note the gradual decrease in the pulse rate with a slight increase at the end, the latter marking a corresponding decrease in the amount of physical exercise. ting April 3, 4, and 5 on account of sickness. Evening observations which were influenced by previous periods of exercise, have not been used in making the averages. Note the decrease in the pulse rate after training. Effect of training on the resting blood pressures. The effect of training on the resting blood pressures systolic and diastolic, is neither striking 450 PERCY M. DAWSON nor constant at least not with exercises of the moderation employed in these experiments. Some changes have been noted of small amount and uncertain significance. For example, the behavior of the diastolic pressure in ’14-run and ’16-cy-ten is such that whether it falls or rises (the direction of the change depending upon the time of day at which the observations were made) there results an approach to 80 mm. (Table 2 and figs. 2 and 3). In this respect the noon observations of TABLE 2 Effect of training on the resting blood pressures AVERAGE BLOOD PRESSURES NUMBER —————————————— SERIES OF OBSER-| TIME OF DAY |DATES(BEFORE)] Before | After | DATES (AFTER) VATIONS Se) DADS iS: 6 | Noon April 2-13/121) 90/116] 88} May 23- 214-1 eee June 30 \ 3 | Afternoon | April 2-4 |123] 97/120] 90] May 27-29 ( 10 On rising | April 6-15/102} 76106) 77) June 8-17 iecy en 10 | Noon April 6-18/115) 85/115) 82! June 8-17 i ie 5 Afternoon | April 510/115) 84/117) 82) June 6-11 7 Evening April 4-17/112| 78/116) 78) June 8-17 te Gee 6 On rising | March 6-12/111} 85/110) 80} April 12-17 Bae a | 10 | Noon March 5-15|112| 84/120| 91) April 8-19 16 is SS ————~_—_——- Vw —_~_—_ —__—_ Morning Noon Afternoon Evening BP Fig. 2. Effect of training on the resting blood pressures. The numerals at the extreme left represent mm. Hg. The vertical lines are the pulse pressures, the upper end of each being at the systolic level, the lower at the diastolic level. The numerals placed below designate by dates the series of experiments to which the two lines immediately above each belong. Of each pair of vertical lines the one to the left represents the pressure before training; that to the right, the pressures after training. The series (date numerals) are arranged in order of decreasing severity of the exercise. Note the absence of any conspicuous or constant effect of training wpon the systolic and diastolic pressures, at least when the exercise is of the moderation represented by these experiments. «Possibly the diastolic pressure tends to approach 80 mm. EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 451 ’15-gym differ from the other two series but whether this was due to the difference in kind or in severity of the exercise or to some other cause, is purely conjectural. It is also to be noted that the change in the relation of the systolic to the diastolic pressure in all three series was usually such as to increase the pulse pressure (vide infra). Table 2. The table shows in the first column the series (date) and the variety of exercise, given in the order of the decreasing severity of the exercise; in the second, the number of observations from which the averages were obtained; third, the time of day at which the observations were made; fourth, the period (dates) during which observations were made prior to or early in the beginning of the training; fifth and sizth, the average systolic and diastolic pressures corresponding to the dates in columns four and seven respectively; seventh, the period (dates) during which observations were made late in or subsequent to the period ihe duly - AEM | i Noon ede Morning Afternoon 14! 6 ¥ Evening Fig. 3. Development of the effect of training upon the blood pressures. In as much as the observations are more complete for ’16-cy-ten than for either of the other series, it has seemed justifiable to devote to their presentation a sepa- rate figure. This figure is based upon the entire number of observations com- prised in the series in question (table C). In this figure the vertical lines are the pulse pressure, the upper end of each being at the systolic level, the lower end at the diastolic level. The horizontal lines of reference represent 120 mm. Hg. (upper line) and 80 mm. Hg. (lower line) respectively. The vertical lines are divided into four groups corresponding to the observations made at differ- ent times of the day. The changes in the pressures are appreciated best when the observer looks at the chart from the side and foreshortens it by holding it at an obtuse angle to the line of vision. Note the gradual developments of the changes in the blood pressures, of which figure 2 gives only the final outcome. One may also obtain from this chart some idea of the character of the diurnal variations in the blood pressures, to which special reference will be made later. 452 PERCY M. DAWSON of training. The evening observations which were influenced by pre- vious periods of exercise, have not been used in making the averages. Effect of training on the resting pulse pressures. The effect of training upon the pulse pressure during rest is almost always in the direction of an increase. Sometimes this increase is small (3 mm.), at others larger (6 mm.), but in only one instance was there a decrease (3 mm.) namely in ’14-run, noon. (Table 3 and also fig. 4.) The writer has been unable to assign a plausible reason for the exception mentioned above. Numer- ically it was due, at least in part, to a period of high systolic pressure during which the series ’14-run was begun to a period of high diastolic pressure during the last few days of the series. 40 3o ao Morning Noon Afternoon Evening PP 1919-16 Fig. 4. Effect of training on the resting pulse pressure. The values used in this figure are the same as those presented in table 3. The numerals at the ex- treme left and right represent mm. Hg. The vertical lines are the pulse pres- sures. The numerals placed below indicate the series (dates) to which the two lines immediately above belong. Of each pair of vertical lines, the one to the left represents the pulse pressures before training and that to the right the pulse pressures after training. The date numerals are arranged in order of the sever- ity of the exercise from left to right. Note that in almost every case there is an increase in the pulse pressure as the result of training, but that the increase is often quite small. Effect of training on the product P. R. xX P.P. during rest. The product of the pulse rate times the pulse pressure is of interest since it is possibly (3), (4) an index of “minute volume”’ (output of the heart per minute). It has been seen above that as a result of training the pulse rate falls while the pulse pressure rises in value. The product would consequently tend to remain unchanged and would be entirely unchanged if the variations in rate and pulse pressure were proportional as well as being in the opposite direction. This is, however, not the case. The product may be found to have risen or fallen so that the compensation (if we may speak of such) is not exact but sometimes falls short and sometimes EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 453 TABLE 3 Effect of training on the pulse pressure and on the product P. P. X P. P. during rest PULSE RATE X PULSE PRESSURE SERIES TIME OF DAY Before After POR. Pork Pos Bars Waser Noon 55 >< Sl = 1705 1512 = 54 X 28 Beniareg ere i. 5" FF Afternoon 61 X 26 = 1580 1740 = 58 x 30 On rising 58 < 26 = 1508 1653 = 57 X 29 ; Noon 58 xX 30 = 1740 1815 = 55 X 33 lG=ey lenin wishes es Afternoon 60 X 31 = 1860 1991 = 57 X 35 || Evening 62* X 34 = 2108 2356 = 62 X 38 Meee {| On rising 55 X 26 = 1400 1680 = 56 X 30 SS a es ca ae Noon 60 X 28 = 1680 1595 = 55 x 29 is excessive. An increase in the product was of more frequent occur- rence than a decrease. These differences are very slight and the pre- ponderance of the increase over the decrease may not be significant. (Table 3, and fig. 5.) Morning Noon Atternoen Evening PPx PR Fig. 5. Effect of training on the product of the pulse rate and pulse pressure P. R. X P. P. during rest. This product is of interest from the fact that it may represent the cardiac output per minute (minute volume). The data from which this figure was constructed are those presented in table 3. The numerals at the extreme left and right represent mm. Hg. The vertical lines represent the prod- uct P. R. x P. P. The numerals placed below indicate the series (dates) of experiments to which the two lines immediately above belong. Of each pair of vertical lines the one to the left represents the product before training and that to the right the product after training. The series (date numerals) are arranged in order of the decreasing severity of the exercise from left to right. Note that in almost every case there is an increase in the product as the result of training although this increase is often quite small. 454 PERCY M. DAWSON There is in the product under consideration a diurnal variation of the nature of an increase throughout the day. This variation is apparently not affected by training. Table 3. The table shows in the first column the series (date) and the variety of exercise, given in the order of the severity of the exercise; second, the time of day at which the observations were made; third, the pulse rates corresponding to the pulse pressures in the next column; fourth, the pulse pressure before or early in the period of training, derived in great measure from the data given in table 2; fifth, the product of columns three and four; sizth, the product of columns seven and eight, placed next to column five for convenience of comparison; seventh, pulse rates corresponding to pulse pressure in the last column; ezghth, the pulse pressure after or late in the period of training, derived in great measure from the data given in table 2. *From April 2-11. Note that the pulse pressure and the values of P. R. X P. P. are usually greater “after” than “before.” Table 4. The data are from ’16-cy-ten. This table shows in the first column the dates upon which the observations were made. On these dates there was no afternoon exercise or other unusual event to influence the evening values; the second to fifth columns are pulse rates and require no explanation. The figures are divided into two sets (April 6 to May 2and May 6 to 31). The first set occurs in the earlier part of the period of training, the second set in the latter part. Below these columns are placed the averages. Note that the pulse rate shows no striking change which might be attributed to training in the character of the diurnal varia- tions but that the extent of the variations is slightly increased. Effect of training on the diurnal variations of the resting pulse rate. ‘The effect of training upon the resting pulse rate is not of the same mag- nitude for all times of the day (table 4). The most pronounced slowing is that of the pulse rates for noon and late afternoon, namely, five and four beats respectively as compared with three beats per minute in the morning and two in the evening. An unequally distributed change of this kind would naturally alter the form of the diurnal pulse curve. This alteration is, however, so small that it may lie entirely within the limits of error. The curve is altered also in its extent, for if the comparison be made between the amount of the average daily variation before and after training, the greater variation is found to occur after training. The increase in variation is, however, very small, amounting to only 1.3 beats and may well lie within the limits of error or if real be too small to be significant. vt EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 45: TABLE 4 Effect of training on the diurnal variations of the pulse rate during rest DATE MORNING NOON AFTERNOON EVENING Meio ei te clon dady & 58 58 61 62 /\jovetl LAUT? Fee crab oraiee claretors 56 58 56 56 EO TNC ee 50 58 64 60 PaO Ne 8 6 ec tec cvs « 54 66 64 65 SPRUE OU etl cir vare pers a vate via pis) 3s 56 66 60 70 ANG Eel 92 ee a nA 54 56 59 62 Miao tte Lai aly 63 62 62 64 Average (before)............ 58 61 63 63 AN sty game OME AS eos, ce Sesiveveiieee ove! Save 60 61 61 64 hie 2 a ae 5A 59 61 61 Loy 55 55 62 62 av Te 54 57 60 64 Lia 20a a 59 60 59 61 MMiginy: 209). sds seca See eet ane 48 47 49 56 IVa vameilemtearpechs | Rite coke oak 53 50 53 56 Average Guiben ete Abdio bs BD 56 57 61 Effect of training on the diurnal variations of the blood pressures during rest. The diurnal variations of blood pressures in one series of obser- vations (’16-cy-ten) are also shown in figure 3. From this series there have been selected all the complete diurnal cycles which have not been interfered with in respect to the evening observations by a period of exercise preceding the latter or by any other disturbing influence. When this was done the number of such cycles was found to be fourteen. The variations may be regarded from two points of view. First, one may consider the separate daily curves to note any variation in the form of these curves; and second, one may compare all the absolute values of the pressures at different periods of the day. (For example, we may compare all the noon pressures). If we examine the daily curves (fig. 6) it is found that the systolic pressures show seven different rhythms (table 5) of which two predom- inate, namely, a, a rise to a maximum at noon followed by a fall in the afternoon and evening; and b, the noon maximum is sustained until afternoon and the fall is not observed until evening. 456 PERCY M. DAWSON Table 5. The data from which this table is derived are those which have formed the basisof figure 6. Theyrepresent observations made on fourteen days of the series ’16-cy-ten. The table shows in the first column the number of instances in which the rhythm, which imme- diately follows, has occurred; in the second, the character of the change which took place between the morning and the noon observations; the third column, between the noon and the afternoon observations; the fourth, between the afternoon and the evening observations. Note that the usual systolic rhythms are ‘‘up-no change-up” and “up-uwp-down,” while the usual diastolic rhythms are ‘‘uwp-down-down”’ and “‘up-wp-down.” On the other hand the changes in the diastolic pressure are more * constant.. Here only three rhythms occur (table 5) one of which appears but once. The remaining two are, a, a rise to noon and then a fall; b, a rise to afternoon and then a fall. Fig. 6.. The daily variations in systolic, diastolic and pulse pressure during training. The observations from which this figure has been constructed were made on fourteen days of the series ’16-cy-ten. The numerals at the ex- treme left and right represent mm. Hg. Vertical bars are pulse pressures, upper end of each being at systolic level, lower at the diastolic. Dotted bars repre- sent morning pressures. Note increase in diurnal variations of systolic pressure, decrease in diurnal variations of diastolic pressure, and that the character of the daily rhythms is unchanged. The pulse pressure corresponding to the systolic and diastolic pres- sures just mentioned show daily rhythms which are remarkably variable. In these fourteen days there are twelve different rhythms of which two are repeated (making the total of fourteen cycles). There seems to be no evidence in these qualitative data that training produces any effect upon the inter-relations of the four points on the EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 457 diurnal curves of systolic, diastolic or pulse pressure. In other words, we do not find that one form of diurnal curve predominates before training while another predominates afterwards. The effect of training upon the absolute values of the blood pressures observed at different times of the day has already been referred to. By way of recapitulation it may be said that (omitting ’15-gym) the morning pressures are raised, the noon lowered while the afternoon and evening pulse pressures are increased. Moreover there is a gradual reduction of the amount of diurnal variation of the blood pressures (as shown in fig. 6). The amount of this reduction of the diurnal variations may be ex- pressed as follows. If we observe the values of the systolic pressure TABLE 5 Diurnal rhythms in systolic and diastolic pressure NUMBER OF MORNING TO NOON TO AFTERNOON TO INSTANCES NOON AFTERNOON EVENING 4 Rise No change | Fall 4 Rise Rise Fall 1 Rise Fall Rise Systolic pressure.......... 1 Rise Fall No change 1 Rise Rise Rise 1 Rise No change } Rise 1 Fall Rise Fall 8 Rise Fall Fall Diastolic pressure......... 5 Rise Rise Fall 1 Rise Fall Rise for one day, and subtract the minimum systolic pressure from the maximum systolic pressure we get a figure indicative of the difference between the extremes of systolic variation on this day. If the values obtained for several days selected from the period preceding or early in training be compared with the values similarly selected from a period later in or subsequent to the training, then the latter are found to be less than the former by 20 per cent. If the values of the diastolic pressure be treated in the same manner, the variations after training are found to have decreased by 50 per cent. Finally if the difference between the maximum and minimum pulse pressure be considered, one finds a decrease of 10 per cent only, which the writer does not think is large enough to be safely beyond the limits of error. Consequently, although there is as has been shown an increase in the pulse pressure 458 PERCY M. DAWSON after training, yet the diurnal fluctuations in the size of the pulse pressure have not been positively shown to vary. The product of P.R. x P.P. shows a diurnal variation of the nature of an increase throughout the day. The character of this variation seems not to be affected by training. RESULTS II Effect of training and practice on the reaction to exercise of blood pressures and pulse rate. The reaction to exercise depends upon two factors. If a subject be tested on a bicycle ergometer and then runs several miles every few days for several weeks, and if he be finally tested again with the ergometer, the change in his reaction is due to an improvement in his general physical condition which is the result of this systematic exercise. This state or condition we have called ‘‘training.”” If on the other hand the reaction of the subject to the first run be compared with the reaction of the subject to the last run, then the change is attributable to two factors: first, to what has been called above ‘‘training,’’ and second, to ‘‘practice.”’ The latter is in all probability a condition of neuromuscular adaptation to a certain special sort of exercise, in the case cited, running. The reaction to exercise is modified by these two factors. If the subject is in training or has practiced he reacts in one way ;if not, he reacts differently. It is these differences which constitute the topic of the present section. That practice and training are quite different states of efficiency is shown by such observations as the following: A young woman who was an expert swimmer and possessed great physical strength, was seized with an acute dilatation of the heart on her first attempt at mountain climbing, the tax not being a severe one. By beginning again with great caution after a few days of rest, and progressively increasing the severity of the climbing, she became at the end of a few weeks as efficient in this form of exertion as in those to which she had already been inured, namely, rowing and swimming. Here the subject though trained was not practiced. And the result when practice had been added to train- ing was quite different from the effect when the condition was one of training only. It is not possible to separate with entire satisfaction the effects of practice from those of training and practice combined, but it is readily possible to separate the effect of training from this combi- nation as has been intimated. It is exceedingly important to remember that throughout the present study of the reaction to exercise before and after training, the amount of EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 459 exercise is not a constant factor. In the beginning of the experiments it was thought desirable to keep the amount of work done as nearly con- stant as possible. That would have made the results obtained show the effect of practice and training on the circulatory reaction to a fixed amount of work. But in the enthusiasm and delight of physical activity, it was found unbearably irksome to restrain the trained body. In every eycle ride the subject when once underway proceeded forthwith to pedal joyously. In the case of the three-mile run the natural beauty of the environment proved a sufficient distraction to permit an almost uniform rate of progression, but the approaching of home called forth an inevitable sprint which increased in vigor with training and practice. What these experiments show is the effect of training and practice upon the reaction of the pulse rate and blood pressures of a subject who (within the limits imposed by prudence) performs as vigorously as possible on every cycle ride and three-mile finish. With the recently improved facilities of this laboratory it is practically certain that studies will soon be made of the reaction to a standardized task, but it cannot be too frequently reiterated that the present article is not such a study. It was found that the same qualitative results were obtained whether the observations were made during the latter part of the period of exercise or immediately after exercise. But quantitatively the changes noted after exercise were, as might be expected, just a little less in extent and the condition continued to approach normal with the lapse of time. In some of the instances described in this section the observations were made before, during and after the test exercises (cycle ergometer), while in others they were made only before and after exercise (cycle ergometer and three-mile run). The reaction of the blood pressures and pulse rate to exercise is well known. It consists in a rise in the systolic and pulse pressures and to a much less extent of the diastolic pressure. The pulse rate is also markedly increased. Effect of training only. The effect produced by training on the circu- latory reaction (fig. 7) is as follows: a, the systolic pressure rises more rapidly and much higher than is the case in the untrained individual; b, the diastolic pressure often returns to normal before the cessation of the exercise; c, the pulse pressure is enormously increased. Before training the rise in pulse pressure was from 38 (normal) to 62, after training from 37 to 114. d, The product P.R. x P.P. is also greatly increased (before training the rise was from 2660 (normal) to 6820, after THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 460 PERCY M. DAWSON training from 2220 to 12540). e, The change in the pulse rate is but little affected; f, moreover with less effort more mechanical work is done. Effect of practice only. Nosystematic attempt was made to dissociate training from practice so that a separate study of the latter might be made. There was, however, an incidental observation which is worth recording. In the series ’16-cy-ten, it was found that the blood pressure reaction after the. second ride was remarkably different from that following the first ride. This change in the reaction was of the nature of an increase in the systolic pressure and pulse pressure and subjectively of a decrease in the discomfort. It seems more improbable that a single ride should have put the subject into a condition of training than that he should have experienced 200 2co £80. it 180. April 2. 19/4 e May 29 19/5 A coin Tie Fig. 7. Effect of training on the reaction of the pulse rate and blood pressures to cycle riding. The numerals in vertical columns represent mm. Hg. The vertical lines are the pulse pressures, the upper end of each being at the systolic level, the lower at the diastolic level. The letters B, D and A stand for ‘‘before,”’ ‘‘during’”’ and ‘‘after’’ respectively. April 1: immediately after B, the cycle ride began and observations were made at frequent intervals both during the ride, D and afterward, A. Here kilos x revolutions = 7 X .3276 = 23,849 and P. R. X PP. = 110 X 60 = 6820, during exercise. May 29: as before, but S is fifty-five minutes later than the end of A. Here kilos X revolutions = 7 X 4423 = 30,961 and P. R. X P. P. = 110 X 114 = 12,540. June 8: B is as before but of the pressures recorded during exercise a single pair (that giving the maxi- mum pulse pressure) is here drawn, D, and only the last three pairs of observa- tions from the period after exercise are given, A. Note that after training there is an increase in the reaction to exercise in respect to the systolic and pulse pressures but a decrease in the diastolic pressure. EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 461 the beneficial effects of a little practice. But however this may be, the fact remains that a single ride produced a pronounced effect. (Table 6 and fig. 8). TABLE 6 Effect of practice and of training and practice on the reaction of the blood pressures to exercise (cycle riding) ' = n BEFORE EXERCISE AFTER EXERCISE DATE R 2% e 2 ; ; ; a Bel ce bell lee) xe) Pe exces a 2) Ala }A) A | i alo | A] A | a ay April 14................] 6 [3900] 23,400 | 25/114] 78] 36) 66] 2376/120| 74] 46 (5060) | (5520) (5980 ) April 16................|:6 |5399] 32,394 | 30/116] 77| 39] 65] 2535/140| 80] 60 (6600) | (7200) | (7800) April 29................| 8 |6113] 48,904 | 30/111) 84] 27| 56] 1512/146| 76] 70 (7700) | (8400) | (9100) May 3......-.........| 8 [5790] 46,3820 | 30/112) 90) 32) 66 2112}140 76| 64) (7040) | (7680) (8320) April 14.1916. April 16. April 29. May 3. Fig. 8. Effect of practice and training on the reaction of the blood pressures to cycle riding. The data corresponding to this figure are shown in table 6. They are from the series ’16-cy-ten. The numerals placed in the vertical col- umns represent mm. Hg. The vertical lines are the pulse pressures, the upper end of each being at the systolic level, the lower at the diastolic level. The figure shows observations taken before B and immediately after A, riding the bicycle ergometer. At F, there was a transitory feeling of faintness; at N,an equally transitory feeling of nausea. The duration of the rides were in three cases thirty minutes, in one case (April 14) twenty-five minutes. Kilos X reyo- lutions = (April 14) 23,400; (April 16) = 32,394; (April 29) = 48,904; (May 3) = 46,320. Note that after training the heart is capable of a much greater reaction without untoward symptoms. 462 PERCY M. DAWSON Table 6. The data are from the series ’16-cy-ten and the table cor- responds to figure 8. In the first column are given the dates of the experiments; in the second, the weight of the brake, in kilos; third, the number of revolutions of the cycle ergometer; fourth, the product of the weight times the revolutions; fifth, the duration of the ride; szzth to tenth inclusive, blood pressures, pulse rate and their derivatives, taken before exercise; eleventh to fourteenth inclusive, the blood pressures, pulse rate and their derivatives observed after exercise with the excep- tion to be mentioned immediately, namely, that as the P.R. was not observed, the numerals in parentheses were calculated after assuming arate of 110, 120 or 130 beats per minute, and are presented as suggestive merely. Note that the practice and training exaggerate the increase in systolic pressure and pulse pressure and assuming that the difference in increase of the pulse rate was not more than twenty beats (cf. table 9) that the product P.R. X P.P. is also increased. TABLE 7 Effect of training and practice on the reaction of the blood pressures and pulse rate to exercise (three-mile run) BEFORE EXERCISE AFTER EXERCISE DATE Ss. | D. | PP. | BR. |*s3%! Sie Dea) BP pees Aprile] eee. 128 | 90 38 58 | 2204 | 138); 90 48 100 | 4800 May 8.:.25: TT | kes) 30 55 | 2815 | 164 | 80 84 96 | 8064 May 14...... 110 | 838 27 65 | 1701 | 168 | 76 92 108 | 9936 Combined effect of training and practice. These effects were studied in two series of experiments. In the first (’14-run) observations were made of both pulse rate and blood pressures while in the second (’16-cy) the blood pressure only was determined. In both, the observations were made before and again immediately after exercise. In respect to the blood pressures, the results were essentially the same in both series and were in accord with those obtained by training alone. In the first series (’14-run) a, there was an increase in the rise of the systolic pressure; b, the change in diastolic pressure was not markedly affected; c, the pulse pressure was enormously increased; d, the product P.R X P.P. was also greatly increased. Also the mechanical work done was greater with apparently less exertion. e, The pulse rate was decreased (cf. table 9). Unfortunately the negative phases cannot be compared in this series because the late observations were not made at the same interval of time after the cycle rides. (Table 7, fig. 9). EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 463 Table 7. The data are from the series ’14-run and the table corre- sponds to figure 9. The tablerequiresno further explanation. Note that the training and practice exaggerate the increase in systolic pressure, pulse pressure and the product P.R. * P.P. while the change in the pulse rate is unaffected. In the second series (’16-cy-ten) the first exercise of the season was a cycle ride taking place on April 14. (Table 8, fig. 10). On April 29 the ninth cycle ride took place. On June 6 and 7 came the tenth and eleventh rides respectively. Between the first and ninth rides cycle Inheval 10 min ra J Inkrval 20 min. Interval 45 min Uy Ald yA Bouya BA ; B A April 17 1979. May & May I¢ Fig. 9. Effect of training and practice on the reaction of the blood pressure to the three-mile run. The numerical data corresponding to this figure are shown in table 7. They are from series ’14-run. The numerals in vertical col- umns represent mm. Hg. The vertical lines are the pulse pressures, the upper end of each being at the systolic level while the lower end is at the diastolic level. B indicates observations made immediately before and A, immediately after exercise. Here P. R. X P. P. (April 17) = 4800; (May 8) = 8064; (May 14) = 9936. The observations made after the exercise are in each case interrupted for a few minutes as indicated in the figure. Note that after training there is an increase in the reaction to exercise in respect to the systolic and pulse pressures. riding was the only form of exercise. Between the ninth and tenth rides there intervened several weeks of tennis playing. Consequently in the first period the cycle ergometer tested the effect of both training and practice while in the second period it tested only that of training. The second period, however, was not particularly significant because the subject had just passed through a course of exercise which had already removed most of the effects of the prolonged inactivity which had preceded the series. On comparing the effect of the first cycle ride with that of the ninth, one finds the same results as in the first series with respect to the blood pressure. Moreover in comparing the effect 464 PERCY M. DAWSON of the ninth ride with that of the tenth and eleventh, one finds that the changes in the reaction to exercise are no greater after nine rides and several weeks of tennis than after nine rides alone. In the former case the negative phase is not quite so conspicuous as in the latter which suggests that the result of the further training may have been to permit an increase of this phase. This conclusion is nevertheless insecurely based since the negative phase, properly so called, would not be expected to occur until some time subsequent to the last observations made in each of the riding tests in question. TABLE 8 Effect of practice and training on the reaction of the blood pressure to exercise (cycle riding) 5 rE a BEFORE EXERCISE AFTER EXERCISE DATE g 36 x 5 eae 5 pers St BE ad ol 8 ellie tee Var ec calter lieee April 14................| 6 |8900) 23,400 | 25/114, 78 36) 66| 2376/120| 74! 46] (060) (5520) | | (5980) April 29................| 8 |6113) 48,904 | 30/111) 84) 27) 56) 1215/146) 76) 70) (7700) | (8400) (9100) ANTI YS dots olin ommoces.cli== Ifo — | —|117) 80! 37| 53] 1961|144) 70) 74) (8140) (S880) (9620) June 7....2...:.......| 8 |7007| 56,056 | 30/115) 78 37| 56) 2072|144) 80} 64! (7040) os (7680 ) | (8320) Table 8. The data from ’16-cy-ten and the table corresponds to figure 10. In the first column are given the dates of the experiments; in the second, the weight of the break; third, the number of revolutions of the cycle-ergometer; fourth, the product of the weight times the revo- lutions; fifth, the duration of the ride; szxth to tenth inclusive, blood pressures, pulse rate and their derivatives, taken before the exercise; eleventh to sixteenth inclusive, the blood pressures, pulse rate and their derivatives observed after the exercise, the exception to be mentioned immediately, namely, since the P.R. was not observed, the numerals in parentheses were calculated on the assumption of a rate of 110, 120 or 130 beats per minute, and are presented as suggestive merely. Note that the practice and training exaggerate the increase in systolic pressure and pulse pressure, and assuming that the increase of the pulse rate is in- ou EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 46 creased not more than twenty beats (see table 9) that the product P.R.XP.P. zs also increased. From the subjective side the effect of training alone and of training and practice combined is to enable the heart to work more vigorously without untoward symptoms. Although before training or practice the pulse pressure (systolic output?) is relatively small during (and just after) exercise, the distress is considerable; later, after training or practice, although the pulse pressure is very much greater, the individual suffers no subjective embarrassment whatever. BOA BLA B BUA April 14. 1916 April 29. June 6. June 7. Fig. 10. Effect of practice and of training and practice (?) on the reaction of the blood pressure to cycle riding. The data corresponding to this figure are shown in table 8. They are from series ’16-cy-ten. The numerals placed in the vertical columns represent mm. Hg. The vertical lines are the pulse pressures, the upper end of each being at the systolic level, the lower at the diastolic level. The observations were made before, B, and immediately after, A, riding the bicycle ergometer. Here kilos X revolutions (April 14) = 23,400; (April 29) = 48,904; (June 7) = 56,056. The duration of the exercise was twenty-five min- utes in one case (April 14), thirty minutes in the other two. Note that after training there is an increase in the reaction to exercise in respect to the systolic and pulse pressures. Effect of training and practice on the pulse rate after exercise and on its return to normal. A few observations have been made with reference to the effect of practice and training a, upon the increase of the pulse rate due to exercise, and b, upon the rapidity of the return of the pulse rate to normal after exercise has ceased. a. If the relevant data from series ’14-run be divided into two parts, the first of which represents the results obtained early in the period of training and practice, and the second the data obtained later in this period, it is possible to make certain comparisons (table 9). If one compares the maximum rate after exercise during the first half period with the corresponding rate during the second half period, one finds that the former exceeds the latter (150 beats per minute as compared with 126). Again if one compares the average rate during the first half 466 PERCY M. DAWSON period with the average rate during the second half period, one finds that once more the former exceeds the latter (119 and 127 respectively). It is quite probable that these differences would have been greater but for certain counteracting influences, namely a, that the final sprint was more vigorous as the condition of the runner improved, and b, that the pulse record was probably begun a trifle earlier (that is, with less loss of time) as the assistant became more deft in adjusting and inflating the cuff upon the ankle. It is possible that the fact that during the last two days (May 26 and 27) the weather was unusually hot may also have tended to raise the pulse. The resting pulse of this subject is faster in warm weather than in cool, and it may be that the pulse in reacting to exercise rises higher during warm than during cold weather, but this has not been ascertained. Table 9. The data are from series ’14-run and represent the pulse after a three-mile run. During the two hours referred to in column eight, the subject bathed and dined. The values given (pulse rates per minute) were calculated from the rates for 5 to 10 seconds, consequently all errors are highly magnified. This is why columns two and three, and four and five have been averaged together (average II). The values used in making the averages are enclosed in parenthesis. The value 132* is so high that it arouses distrust as to its accuracy, it is therefore omitted from one average and included in the other (*). Note that the averages are lower “immediately after” running in the trained and practiced subject, but that at the end of two minutes this difference has considerably decreased. b. After exercise the pulse rate falls at first rapidly and then more slowly. Even at the end of five hours it was still above normal as can be seen from the following values, namely, 65.5 (62 if we omit the warm days May 26 and 27). Inseries ’14-run, the pulse rate at the time of the retiring was 65.5 (62 if we omit May 26 and 27, when the weather was unusually warm) on days on which the subject exercised, while on the resting days, the corresponding figure was only 56. This tardiness in the return of the pulse to normal is seen again in series ’16-cy-ten. Here the average pulse rate on retiring was 68 on days of cycle riding and on days of resting 62. (Table C). For comparing the readiness with which the pulse rate returned to normal during the first part of the period of training with the corre- sponding values for the second part, one turns again to series ’14-run. Here observations made two minutes after cessation of exercise were divided into two sets corresponding to the first part of the period of EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 467 training, and to the second part respectively. Such a comparison (table 9) shows that the difference between the averages of the rates during these two half-periods has decreased below what it was immedi- ately after exercise. In other words, the higher pulse rate has fallen more rapidly than the lower although neither had of course as yet reached the normal level. TABLE 9 Effect of training and practice upon the return of the pulse rates to normal after exercise AAS SAS i 3 : pie | tenuis ined 1 MINUTE MINUTES ee ones rape gob deee RUN Negro. iii: doe 100 80 PTORUN LS? 25). enn 5's (102) (96) (90) (80) JA) 0) C1 77 rr (150) | (150) | (114) (92) 88 Joel iS) ae (132) | (132) | (108) (94) May NR eres ee (150) | (114) | (132)*) (92) Avis Pee (120) | (102) | (108) | (96) él 56 sta BOE Sache! ti: (120) | (108) | (108) (90) 90 58 Mine ts ih S052 ft 108 90 96 82 72 63 ULES eo i 96 102 90 96 86 64 Mayes. oh 5 wb. (126) | (102) | (114) (90) 90 88 70 63 IMIG Sr (108) | (108) | (102) (96) 96 92 78 64 Rien 20: sk 5, (120) | (108) | (102) (96) 76 90 77 70 IMiivar Zot S cit: (108) | (114) | (102) (96) 90 88 78 66 EAD 35s 2.45 eae (126), |*.44)5 |, G20)).) -@4)..)) 102 108 80 73 1M 7 rs 126 120 120 102 98 76 71 Average I........ WO) a7 ieee tO, | 105 ; ; ; Average II........ 123 if ee EN EIN 97 Average I........ 117 109 108 92 Care 3 Average IT ....... 113 100 Mla yits, 2onclasive Turning the attention to the evening pulse rate, that is the rate at about five hours after the cessation of the exercise, one finds noteworthy data in series ’16-cy-ten. If this series be divided into two parts and comparison be made between the average evening pulse rate during the first part (April 14 to May 15 inclusive) with the average evening pulse rate during the second part (May 16 to June 3 inclusive) selecting in both cases the days upon which exercise was performed, the former is 468 PERCY M. DAWSON found to be 68.7, the latter 61.4. From this one would conclude either that the effect of training and practice favored the return of the pulse rate to normal, or that the difference of rate of return to normal resulted from the difference in the form of exercise, namely, cycle riding in the first case and tennis in the second.’ To the writer, the latter alternative seems to be the less important factor.® RESULTS III Effect of training on infection. The effect of training on the course of an acute naso-pharyngeal infection was noted during the period of observation in 1916. The subject was attacked with the malady in question on April 3 and again on June 8. Both attacks began with nasal and pharyngeal inflammation and marked constitutional symptoms. Then, after a period of rapid convalescence, a laryngitis set in with a return of the constitutional symptoms. This in turn gave place to complete recovery. On perusal of the blood pressure readings (see table C and fig. 3) the writer cannot see that the blood pressure has been affected at all by either infection, even after the data were charted with reference to the ~ diurnal changes and subjected to close scrutiny no effect could be de- tected. The pulse rate however shows a great increase during the first attack, but a relatively small one during the second (see fig. 11). It should be noted that the symptoms in both cases (both local and constitutional) occurred before the change in pulse rate. The interest in these attacks lies in the facts, a, that the observations. of blood pressure and pulse rate were begun a considerable time before the infection occurred and were continued for some time after and the abnormal phenomena are therefore carefully controlled; and b, that the infections in question were very similar to each other and also to other infections not infrequently experienced by the subject, which run in every case a perfectly definite and predicable course, independent to a 5 Unfortunately, in the first part of series ’14-run observations were rarely made after the end of two minutes, so that a conclusive comparison cannot be made between the evening pulses of the first and second series. This is the more to be regretted since the few values obtained seemed to contradict the conclusions drawn from series ’16-cy-ten. 6 Indeed since it is often found that prolonged exercise favors a retarded return of the pulse to normal, one might expect that, after one and a half to two. hours of tennis, the pulse would return to normal more slowly than after the thirty-minute cycle rides. EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 469 great degree of such differences in the weather conditions as distinguish early April from early June in Madison in 1916. It seems probable therefore that the only variable factor and hence the cause of any difference between these two attacks is the fact that the subject was in bad physical condition during the first attack while the contrary was the case during the second attack. Effect of acute infection on the response of the pulse rate to exercise. It is of interest and also is germane to the present topic to note at this point the effect of an acute infection upon the response of the pulse rate to exercise in a case of acute naso-pharyngeal affection. The writer has long been familiar with the feeling that the heart beats more rapidly | 60 =) <— Before After > Fig. 11. Effect of training upon the pulse rate during an acute infection. The figures in the ventrical column signify pulse beats per minute. The ventri- cal lines show the pulse rate in series ’16-cy-ten. They are arranged in the order of time. Wherever the observer failed to determine the pulse rate at the ap- pointed hour, a blank space has been left. The observations fall into two groups (1) early in the series (2) late in the series; in other words ‘‘before’’ and ‘‘after’’ training. The maximum pulse rate ‘‘before”’ occurs on the evening of the first day (April 3) upon which symptoms of the first infection were observed; that ‘‘after’’ on the evening of the first day (June 8) of the symptoms of the second infection. Note that the maximum before training is much higher than that after training. on running up stairs when one has a ‘“‘cold”’ than under normal circum- stances. No careful observations were made until May, 1919. At that time the writer was engaged in a series of experiments which involved turning by hand the wheel of a cycle ergometer having a four kilo break. The mode of procedure was to work for ten minutes and rest for five minutes alternately for the space of an hour. During each five minutes of rest, the pulse was counted. The counting began at exactly fifteen 470 PERCY M. DAWSON seconds after the cessation of the exercise and was continued every other quarter-minute until eight counts were made. Thus four minutes elapsed and the remaining minute of rest was devoted to reading the ergometer and in getting ready for the next ten minutes of work. In the course of these experiments an acute naso-pharyngeal in- fection occurred and two tests were performed while the subject was “under the weather’? (May 17 and 19). On comparing the effect of the muscular exertion upon the heart rate, we find a, that the heart rate is much higher during the infection than before while the work done is on the average less; b, that the heart rate is no less during the infection than afterward while the work done is much less. The rise in the heart rate after the period of infection as compared with that before infection is the result of practice which permits a great increase TABLE 10 Effect of acute infection on the response of the heart rate to exercise PULSE RATE DATE ee After each heat Padget el Sere lat 2d 3d 4th Marys =5-p1 919 steep ee 55 21-15 | 24-16 | 27-17 | 27-19 | 5610 May a7 i919 ree eee 56 29-19 | 28-20 | 30-20 | 30-19 eed Average May2l9 st O19 Naren eee 53 30-19 | 34-20 | 32-21 | 32-22 | 5992) 5574 June 9 IOI Re eee 59 30-20 | 34-22 | 32-21 | 34-23 | 7098 in the number of revolutions accompanied by an increase in the heart rate without however any untoward effects. But the infection causes increase in the heart rate without permitting any corresponding increase in the amount of work done, and the exertion in this case distresses the subject to a considerable extent. Table 10. In the first column is shown the date of the experiment; in the second, the pulse rate per minute before exercise; third, the pulse rate after the first ten minutes of work, the first figure representing the first reading, the second figure the last reading in the first four minutes following the first heat; fourth, fifth and sixth observations after the second, third and fourth heats respectively ;in the seventh, the total num- ber of revolutionsinall four heats. May 5and June9are before and after the infection respectively. May 17 and 19 are during the infection. Note that during infection the reaction of the pulse rate to exercise is increased while the number of revolutions (amount of work done) is not increased. ~- EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 471 Exercise and distress. A few notes were made of the feeling of dis- tress which occasionally occurred during exercise (’14-run). Here there seemed to be no discoverable connection between these symptoms and the extent of the disturbance of the pulse rate or blood pressure sub- sequently obtained (at the end of the run) nor between the symptoms and the return of the pulse rate toward normal. For example, on May 22 and April 17 following distress, the pulse rate was 120 and 100 respectively while no untoward symptoms occurred on May 1, 2 and 8 when the pulse rate at the end of the run was 150, 120 and 96 respectively. Following exercise there may be a feeling of distress which consists of giddiness or nausea or both. This does not necessarily occur during the negative phase. It may occur some minutes after the exercise has ceased. It was of momentary duration in every case except after the first bicycle ride in 1914. In the latter case the sequence of events was as follows: TIME s. D. P.R. 5.08 p.m. cessation of exercise 5.19 116 90 90 Began to feel faint 5.20 118 90 9.21 116 92 Kneeling with head touching floor Feeling better, head still low 5.25 | 118 90 52 Head raised, still kneeling 5.26 118 88 88 5.27 116 88 80 5.28 116 88 126 Sitting on bicycle again 5.30 115 88 5.35 116 90 126 Subject lies down 5.43 108 80 76 recumbent 5.48 106 80 82 recumbent Subject changes clothes partly 5.57 118 76 76 recumbent Finished dressing 6.05 124 80 74 Dinner with staff of medical school followed by discussion 10.00 115 | 75 67 recumbent 10.00 118 | 80 84 standing 472 PERCY M. DAWSON A satisfactory interpretation of this attack must await more exten- sive observation and the following explanation is little more than a con- jecture. It appears to the writer as if the primary factors were cardiac. A sudden drop in the pulse rate (from 90 to 52) causes cerebral and systemic anemia, a powerful, peripheral constriction restores the blood pressure (reading B.P. 118-90, P.R. 52). Then the peripheral constriction all the time compensating the changes in pulse rate, the latter rises, runs beyond the mark, falls again, rises again, and finally becomes tranquil. It is, of course, possible to con- ceive that the vascular change (constriction) is primary and is com- pensated by the changes in heart rate. But. such a supposition seems unwarranted since it implies that the cerebral vessels share in the general constriction to a degree that renders the brain anemic, an exploit on the part of cerebral vasomotors which at the present time seems incredible. Lastly it might be urged that perhaps the factor of peripheral resist- ance remains constant and that the factor which varies inversely with the heart rate (thus keeping the pressure constant) is the force (output) of the heart beat. But the writer is still addicted to the belief that such variations in systolic output are indicated by change in the pulse pressure and, since this does not occur, he is loath to entertain this explanation. During the vicissitudes of the experiment Professor Eyster made the determinations for which I am indebted. In 1916, distress was sometimes observed after cycle rides. On one occasion, April 14, fleeting faintness was felt and on the other, May 3, the symptom was an equally transitory nausea. On neither occasion did these symptoms last longer than the time required for a single pair of blood pressure readings. The faintness was not asso- ciated with any decrease in pulse pressure though the nausea may have been (fig. 10). Unfortunately, the pulse rate was not deter- mined in these particular experiments. The symptoms on these two occasions differ from those in 1914, not only in being less severe but also in their time relations to the cessation of the exercise. Those of ’16-cy-ten were within five minutes of the cessation of exercise (cycle) while that in ’14-run was fifteen minutes after cessation of the exercise {cycle test). EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 473 SUMMARY OF RESULTS The following summary is submitted without further discussion. Credit to previous investigators together with comment and criticism is reserved for a separate communication. I. The effect of training upon the resting pulse rate, blood pressures and their derivatives was as follows: a. The pulse rate was slowed especially the noon and afternoon pulses. This altered slightly the form of the diurnal pulse curve. ‘The extent of the diurnal variation was increased. b. The diurnal variations of the systolic pressure were increased. c. The diastolic pressure approached the daily mean while its diurnal variations were much decreased. d. The pulse pressure was usually increased and its diurnal varia- tions were somewhat decreased. e. The forms of the daily curves of blood pressures were not obvi- ously altered. f. The product of the pulse rate times the pulse pressure was usu- ally increased but the character of its daily variations was unchanged. Il. The effect of training or of practice or of both upon the cardio- vascular reactions to exercise was as follows: g. When the trained or practiced individual engaged in physical exercise he naturally accomplished more work with less apparent exertion and less subjective distress. This was in spite of the fact that the systolic pressure rose higher, the pulse pressure increased enormously and with it sometimes the product P.R. x P.P. h. The diastolic pressure sometimes returned earlier to normal, L.e., while the exercise was still in progress. 7. The pulse rate however was sometimes less affected than in the untrained subject and probably reached normal before that of the latter. III. The following interesting, miscellaneous observations were also made: j. An acute infection (naso-pharyngitis) caused an increase in the pulse rate but no change in the blood pressures. In the trained sub- ject the change in pulse rate was much less pronounced. k. During acute infection (naso-pharyngitis) exercise caused a greater increase in pulse rate than with the normal subject and the amount of work accomplished was less. 474 PERCY M. DAWSON l. The feeling of distress which occurred during exercise showed no relation to the heart rate and blood pressure determined at the cessa- tion of the exercise. When distress followed exercise it had no relation to the blood pressure present at the time but the heart rate was found to be greatly decreased. As already stated in the introduction (q.v.), these conclusions are subject to certain reservations. REFERENCES (1) BarpEEN: Amer. Journ. Anat., 1918, xxiii, 423. (2) Canrrit: Thesis for degree of M.S., University of Wisconsin, 1914. (3) Dawson: Brit. Med. Journ., 1906, 11, 996. (4) Dawson AND GorHAM: Journ. Exper. Med., 1908, x, 484. (5) Dawson: This Journal, 1917, xlii, 590. (6) Herne: Wiener Sitzungsb., 1869, Ix, 2, 829. (7) Freperica: Arch. f. Anat. u. Physiol., 1887, 351. (8) MtLiter: My system, Copenhagen, 1905. ‘9) Travupe: Centralbl. f. d. med. Wissensch., 1865, 881. EFFECT OF TRAINING ON BLOOD PRESSURE IN MAN 475 APPENDIX. ROUTINE OBSERVATIONS TABLE A "14-run MORNING NOON AFTERNOON| EVENING DATE Serr jae C a - REMARKS Gla} Gla lal a |e 2) a 2) a <2) = QD a Agni 1... | Cycle test April 2) .....: 116-88 |50)120-98 |59 101g aaa 120-104'55/128-100/61 April... 124-90 |67/122-92 |62 mpril 5... ... 122-92 |52 . 8 miles, walk and run PELLIG:...-. 8 miles, walk and run TU +>. - 8 miles, walk and run ppl S... .: 3 miles, walk and run April9..... 130-98 [58 8 miles, walk and run April 10.... Gymnastics April it... 12-mile walk April 12... .|118-86 126-92 |82 22-mile walk April 13.... 116-82 |50| Gymnastics April 14.... 54/124-90 56 114-90 64 April 15.... 54 2 3 miles, walk and run April 16.... 120-86 |64/120-82 (60 40). ao 128-90 [58 3 miles, walk and run April 18.... 3 miles, walk and run April 19.... Coryza April 22.... 3-mile run April 23... ... April 24.... 54 April 25.... 122-88 |56 April 26.... Apr 27 .:.. < 3-mile run * * * * * * * * April 30.... 3-mile run IMiaiy Us... 2 8 3-mile run Maia 2: <,. 128-88] 56) 3-mile run May 3......|111-85) 54 | | 96-74! 59) 13-mile walk May 4...... 112-81} 52 May 5 53 Miaiy: Gi)... 109-77 |53)110-77| 58} 3-mile run WV Metalic 115-80) 47 126-83 \55(115-81 63) 3-mile run IMSS a. 110-76) 55 113-78 |55 3-mile run May 9...... 120-94 |50/120-94 |52 May 10..... 114-87 |51 * * * * * * * * THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 3 476 PERCY M. DAWSON TABLE A—Concluded DATE MORNING NOON fa Mayalsecnc: May 14..... 107-85 May-15..... May 22 May 23 100-81 May 24 109-89 May 25 106-81 May 26..... 116-83 May 27...5.- 101-81 May 28 101-78 May 29 108-89 May 30 106-82 May 31 June 1 June 2 8.-D. 58 59 50)114-88 57|112-83 48/112-91 59/119-89 53 52/122-92 49 50/117-85 47 54 49 | P.R. AFTERNOON EVENING A ' mM 113-83 120-87 114-84 115-86 110-83 120-91 121-87 123-94 117-90 a 1 ie mM fe Ay REMARKS 59/105-75 63|108-85 53/105-84 57/130-90 54/122-84 119-90 62|108-79 64/103-83 57|115-89 52/119-95 53 53 53) 3-mile run 3-mile run 3-mile run 3-mile run 3-mile run 3-mile run | Cycle test eee Se TABLE B MORNING AFTERNOON EVENING DATE ————EEE ——————————— eee | | S.-D. 1D S.-D. PAR S.-D. Bane March 5.......... 118-84 | 60° | 111-85 | 59 NiamchrGenes. oc... 109-82 | 57 Marchi 70). ......: 118-80 67 March 8..........| 103-90 | 54 114-85 | 60 | 113-88] 59 Maneh-9. 4.2. 2 «.-|118588)| 56 110-79 | 63 March 10.........| 114-90 | 52 113-89 | 62 March ll......... 114-86 | 54 | 126-85] 61 | 118-91! 58 | 118-86] 62 Mareby12.. 2... 112-85 | 58 | 110-89] 60 March 13*........ 109-77 | 69 | 114-91 | 55 114-78 | 60 March 14*........| 105-76 | 70 | 112-77 | 59 114-78 | 60 March 15*........ LOR=77h i 772° WULOS=79) PEL. |1202900) 63 ~ 117-86 |. 164 March 16*........| 114-78 | 71 113-87 | 59 March 17*... ..9o 108-75 | 69 | 117-86 | 69 | 113-80] 64 | 116-86] 60 March 18*........ 110-76 | 70 108-80 | 58 * * * ok * * * * March 21*........ 105-70 | 75 114-76 | 70 March 22*........ FIZ=768). 17499) AAG-SSN G2" | BAE287 1) 62, 107-80) 66 March 23*........ 109-76 | 67 119-86 | 58 March 24*........ 68 | 115-89 | 53 March 25*........ 111-76 | 60 122-85 | 65 March 26*........ 116-84 | 66 | 116-92 | 60 March 27*........ 111-77 | 72 | 115-88 | 65 115-84 | 72 March 28......... 115-78 | 57 | 112-86 | 57 123-81 | 61 March 29*........ 115-84 | 68 | 119-89 | 58 118-80 | 61 March 30*........ 109-79 114-87 | 56 125-79 | 70 March 31*........ 1227491) 170 112-85 | 56 | 129-85 Aeeieie se eee ltee7al) San e7-85. |. 538 |MIS=85. 1 156. 20286) lias Apne? .. 1.900. Ps2784)) 65 | 24-e6 hi6st \2t=e5 |’ 60 April 3*...........] 111-79 | 69 | 118-87 | 62 | Pipe. |... | EB) Gt 17275. |) 167 Aprnl5*..... < MOTE AG| 259 Aprili27ey. =. 1. sate. eS relno 115-84; 64/114-80| 68) Cycle ride April 28..............|102-72] 54/109-81) 56/116-88) 59/100—70) 62 April 29..............|106-84| 53/112-85} 58/111-84] 56 Cycle ride April 30..............|107-83] 53/106-80) 59)113-80| 58}100—79) 56 Maylene. ae sces. = |LOO=79|956 é May 2................|L02-76| 63)112-86) 62)112-82) 62)116-81) 64 May 32................|106—77| 58)115-83) 62/123-91| 66/109-80| 69) Cycle ride May 4s 2).no.. o-es ee LOS T8 97 56 119-84] 58 May 5e..2.28. 2.06... s04=80)Sait15—90 113-84; 74) Tennis, walk May 6...............-|101-78] 60/116-84) 61/111-82) 61]126-85) 64 May 7................|L08-78} 57)109-81)} 55/111-77| 68/110-80) 72) Cycle ride MMayiS cnc were vere 104-81| 54/111-84) 59,113-81) 61)112-80) 61 EFFECT OF DATE WOR re ila? 0). og eee eee AC oe Waive irae. ene6 2. cas IVT cs 8h. castes <8 “Ste be Re green eerie NW Ney TIGR So a ieee Msn yg Gata ce oe LOS SLCC eee ee pene As Maye LS 8k ae Rees IMaiygl9 82 is aes oth:. Mang 20 2 pe See |: 1 S24 OR ean [si TS 524 Oa i an aa IMIG 775 es ie near ee DN? 0 ee ee RPMS aa ee Mer saan tia IN [eine 2G ere rs eh chclalaee. 14 S070) (ne er ee IM NP A ee ae POTD ee ciate se aoe | #. J. He: 130 25 Breath held, inspiratory 1G |) Ao eedi, del 112 | 120 20 Breath held, expiratory a2) |B. Co By 109711307 106-118 65 Valsalva severe 13:4) PC) Ee 100, 96, 100 15, 40, 60 | Valsalva severe The figures 100, 96, 100 and 15, 40, 60 given for experiment 13 indicate succes- sive determinations of systolic pressure and the times at which they were respec- ively made in a single Valsalva effort. The same method of presentation is followed in several of the other experiments in this table. Frequently only the total duration of the effort is noted although several determinations of systolic pressure were made, e.g., experiment 7. Note that during the effort there is no rise in systolic pressure comparable to that described by Bruck, e.g., 160 to 200 mm. released, the blood pressure rose to 168 mm. Hg. From this point it gradually returned to normal.* The writers are therefore still uncertain how to obtain the secondary rise described by Bruck, although a suggestion is offered below (p. 505). 4The writers wish to acknowledge the assistance of Dr. Clarence J. Brown in these animal experiments. CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 487 The pulse rate During the height of an energetic Valsalva experiment, the radial pulse becomes imperceptible (Valsalva) and upon auscultation the heart sounds are found to be very indistinct or even inaudible (EK. H. Weber). It is therefore often impossible to determine the heart rate by either palpation or auscultation and consequently the string-gal- vanometer was resorted to. The electrodes (large lamp wicks satu- rated with physiological salt solution) were bound about the forearms. A short control record was made and then at a word of command the subject began to strain and simultaneously a mark was placed upon the record. When the subject desired to cease from his effort, he did so with a loud vocal sound and the observer at once recorded upon the electrocardiogram. The electrocardiographic clock-work continued to run for a short time after the cessation of the effort. It was then stopped but after an interval of thirty seconds was started again for the purpose of making a final observation. ; aoe ees sy da MA Fs he Se \ : NAR pat ue Fig. 1. Effect of inflating lungs of dog with compressed air and of abdominal compression. The uppermost line shows the respiratory rate and amplitude as recorded by means of a drum-shaped pneumograph fastened about the body of the animal by an inelastic band. On inspiration the heads of the drum were pulled out. The pneumograph was connected with a recording tambour. Down strokes are inspirations. The extensive variations of this line which occur dur- ing the period of inflation of the lungs are merely mechanical, due to the manip- ulations of the experiments. The second line records the blood pressure, the cannula being in the carotid artery, the recording apparatus being the mercury manometer. The third line is the chronographic record (seconds). a, inflation begins; b, abdominal compression; c, respiratory efforts;d, inflation ends. Note the effect of inflation (rise of pressure followed by fall), effect of compression of abdomen repeated three times (rise but not above normal), onset of dyspnea, cessa- tion of inflation with a slow rise of pressure above normal. The electrocardiogram showed the following sequence of events: (1) a preliminary quickening;’? (2) an initial slowing; (3) an accele- ®> It should be remarked that the writers attribute no difference in meaning to these two terms (quickening and acceleration). The contrast is made merely for the sake of avoiding confusion. 488 PERCY M. DAWSON AND PAUL C. HODGES ration; (4) a subsequent slowing which follows the cessation of the . effort. Of these changes the subsequent slowing is the most constant; the initial slowing the least so. Thirteen experiments were performed on seven subjects. ; 1. Preliminary quickening. This fairly constant phenomenon is attributable to the deep inspiration which precedes the Valsalva experi- ment (cf. p. 482). It is the usual quickening which the heart nor- mally exhibits as the accompaniment of inspiration. It differs only TABLE 3 Changes in heart rate resulting from Valsalva experiment svnsner ed peewee ee Doe TOUS Ae ek Absent Absent 80 170 TOS 2 se Pa ae Absent Absent 50 Absent IV St@T roster racmete sh ise ene: 88 117 88 125 cde dia deol Ibe pees de cecee 81 Absent 50 185 PC SHod gest llis. sro 80 133 73 140 Teepe SAG ated cee 80 123 Absent 143 ASCOT] S'e<: IS eee eee te oe as 87 Absent Absent 143 UBB Dy Reet prety tatters are 88 114 Absent 132 (1) TI Pera ehe dee 85 Salil? 56 156 Menninger, alleen sa ? 120 55 155 Cy CP cee 88 123 66 142 Wass SMe Fy. te OR, Nae ey 80 116 64 133 1H beeen eee ee Ao MEU 70 Absent 56 122 I Asverapei2a 50 Sota ee 82.7 120 63.8 145 TA Viet OC seein acrueere troche es 85 112 72 142 Thirteen experiments on seven subjects have been recorded. In the first column are shown the names of the subjects and the designations of the different experiments performed upon them; in the second, the percentage of the prelim- inary quickening, the normal heart cycle being regarded as 100; in the third to fifth columns, the initial slowing, acceleration, and subsequent slowing are simi- larly dealt with. For (1) see page 505. No change of less than 10 per cent is regarded as exceeding the limits of normal variation and consequently suchare designated ‘‘absent’”’ in the table. The interrogation mark indicates that the electrocardiogram was imperfect so that the particular point designated could not be determined. The first average is that of experiments in which a change took place; the second that of all thirteen observations, ‘‘absent’’ being reckoned as 100. Note relative constancy of subsequent slowing and variability of prelim- inary quickening, of initial slowing and of acceleration. a= CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 489 in being more pronounced just as the inspiration itself is more pro- nounced than usual. This quickening was obvious in ten experiments out of twelve (table 3). The maximum reduction in the length of the cardiac cycle was to 70 per cent of its normal value. 2. Initial slowing. In eight cases out of thirteen there occurred either simultaneously with the beginning of the effort, or very soon after its beginning, a cardiac slowing which was sometimes very marked, the maximum increase in the cardiac cycle being 33 per cent (figs. 2 and 3). The duration of the slowing was short in every case, its great- est length being ten beats. The increase in rate which followed was often abrupt, giving a sharp angle to the charted curve (fig. 3). Since 150 A A Fig. 2. The ‘‘typical”’ effect of the Valsalva experiment upon the heart rate. Each dot represents a single heart beat. Time is represented by distances along the abscissa; the numerals arranged in vertical column on the left represent the duration of the cardiac cycle in hundredths of a second; those in vertical column on the right represent the number of beats per minute corresponding to the figures in the first column. Of the wedges pointing upward that to the left indicates the beginning of effort, that to the right the end of the effort. Vertical broken line represents interval of 30 seconds between first and second part of figure. These explanations apply also to figures 3 to 5 and 8 to 15. Subject: Menninger (I). Note preliminary quickening, initial slowing, acceleration and subsequent slowing which has not passed off at the end of thirty-three seconds. the period of slowing comes so soon after the beginning of the effort, one would a prior? expect that it would coincide with the preliminary rise in systolic pressure. This expectation is borne out by the radial sphygmogram, for here the slowing of the pulse is often seen to ac- company the reduction of the dicrotic elevation (vide infra). Even when the effort is so severe as ultimately to suppress the pulse at the wrist during the fall of pressure, still during the part of the effort in which this slowing occurs, the pulse continues to be palpable. 3. Acceleration. Ten of the thirteen subjects showed a period of cardiac acceleration which varied much in amount, the maximum reduction in the length of the cycle being to 50 per cent. This phe- 490 PERCY M. DAWSON AND PAUL C. HODGES 0F 3 aa 150 os pas ee fi ae pass oa ice ie eS A N : ‘ 42 50 Fig. 3. ‘‘Typical” effect of Valsalva experiment upon heart rate. Subject: P. C. Hodges (II). For explanation of the symbols see legend of figure 2. Note preliminary quickening, enormous initial slowing, relatively small acceleration and moderate subsequent slowing, which has not disappeared at the end of forty- two seconds. 04 ——- 150 o8 75 A “A /2 50 Lb 37 Fig. 4. Less ‘‘typical”’ effect of the Valsalva experiment upon the heart rate. Subject: F. J. Hodges (I). For explanation of the symbols see legend of figure 2. Note preliminary quickening, absence of initial slowing, presence of accelera- tion and of very pronounced subsequent slowing, which has returned to within nominal limits by the end of seventy-five seconds. OB att Se LEY bo eb ee ee De i Bo A A Fig. 5. Less ‘‘typical” effect of the Valsalva experiment upon the heart rate. Subject: Dawson (1). For explanation of the symbols, see legend of figure 2. Note the absence of preliminary quickening and of initial slowing, presence of slight acceleration, and marked subsequent slowing, which has not disappeared by the end of thirty seconds. CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 491 nomenon usually begins during the effort and continues from one to ten beats after the effort has ceased. It occurs, moreover, at about the time that the fall in blood pressures might be looked for, and might therefore be thought to be the accompaniment (or result) of the low pressure. Here again the sphygmogram confirms the a priori expecta- tion in that the acceleration and the dicrotism are synchronous. 4. Subsequent slowing. In all but one experiment there occurred at from one to ten beats after the end of the effort, a very marked slow- ing of the pulse (figs. 2 to 5). Here the maximum lengthening of the cycle was to 185 per cent. The slowing was marked but usually not maximal at the outset. This final period was the longest of all, the slowing continuing often for forty or fifty beats and returning to normal sometimes with marked oscillations. Character of the changes in rate Duration of Q-T complex. If we assume that the duration of vens tricular systole corresponds to the time interval between the Q- and T- waves of the electrocardiogram, then it is possible to determine the degree in which each of these phases of the cardiac cycle shares in the variations in duration of the complete cycle. A study of the electro- cardiograms undertaken to elucidate this point shows that the varia- tions are chiefly diastolic in character. The measurements in three typical experiments are presented in table 4. Duration of the atrio-ventricular interval. Detailed study of the electrocardiogram was an after-thought. The string galvanometer was resorted to originally for the purpose of determining the heart rate after other methods had failed. The tracings were for the most part of little more value than was needed for the purpose of counting the heart beats, no efforts being made to cut out adventitious vibra- tions. In several experiments, however, the A-V interval could be made out. This interval was occasionally altered. With a quickened heart beat there might be shortening of the A-V interval while with a slowing, the interval might be prolonged. Usually, however, the interval remained unchanged. The results relating to these points are embodied in the accompanying table (table 5). At one point (24 seconds) in the tracing obtained from F. J. Hodges (I) the P-wave could not be found. The heart was at this time beating very slowly and the picture was that of a nodal rhythm. The tracing is unfortunately not suffi- 492 PERCY M. DAWSON AND PAUL C. HODGES ciently distinct to justify confidence, although suggestive enough to encourage further inquiry. In other tracings of hearts beating with equal slowness, the P-wave was clearly marked. TABLE 4 Effect of Valsalva experiment on duration of Q — T complex DURATION OF SUBJECT, ETC. SECOND ; : Ventricle | Ventricle Cycle systolic diastolic 49 1.00 0.24 0.76 65 0.76 0.26 0.50 Dawson, no. 1, 6-20 seconds........... 5 0.60 0.28 0.32 ae 0.58 0.26 0.32 i De 0.50 O23 0.27 4 0.82 0.26 0.56 Dawson, no. 2, 6-19 seconds........... 29 0.82 0.26 0.56 16 0.50 0.24 0.26 51 1.26 0.24 1.02 50 1.20 0.24 0.96 52 1.20 0.24 0.96 12 0.80 0.24 0.56 Menninger, II, 16-34 seconds. ......... 10 0.78 0.24 0.54 | 11 0.74 0.24 0.50 34 0.48 0.24 0.24 34 0.48 0.24 0.24 35 0.46 0.24 0.22 The first column shows the name of the subject to which is added a numeral indicating to which of the two or more experiments on this subject reference is made, also the numerical designation of the seconds at which the effort began and ended respectively, e.g., in Dawson no. 1., effort was begun during the 6th second and ended in the 20th after the beginning of the experiment. In the second column are the numerical designations of the seconds corresponding to the cycles analyzed in the succeeding columns. The third, fourth and fifth col- umns are the analyses of cardiac cycles of different lengths, taken from all parts of the electrocardiograms. Note that duration of the cycle is mainly dependent on variations in diastole; while variations in systole are small and inconstant. Variations in the Q-, R- and S-waves of the electrocardiogram. The electrocardiograms were also examined to ascertain whether during the changes in rate there were any accompanying changes in the Q-R-S- complex. Nine tracings were found appropriate for this study (table 6). The changes observed were of three sorts, a, a general reduction CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 493 TABLE 5 Effect of Valsalva experiment on atrio-ventricular interval A-V ETC. SUBJECT, Cc INTERVAL SECONDS CYCLE 7 0.86 0.12 8 0.86 0.12 30 0.48 0.12 48 1.20 0.12 Menninger I, 17-34 seconds, April 8, 1915........... 0.60 0.12 PG: Hodges I; 10-23 seconds.: .2.2.00....5. 4. ce eee 0.84 0.12 2 0.76 0.15 , 28 0.96 0.15 Menninger III, 23-69 seconds...............6...005 69 0.74 0.15 78 1.14 0.15 9 0.70 0.12 Kolls II, 16-48 seconds, April 20, 1915.............. 63 0.80 0.12 ami OF (SS oo ie) = 00 21 0.48 0.12 25 1.58 0.18 1 0.80 0.12 Menninger II, 5-21 seconds, April 8, 1915........... [ 0.72 0.13 —X for) Kolls III, 14-48 seconds, April 20, 1915............. 24 0.56 0.09 (va 07727 ||, Oxt2 w 0.80 0.12 j 8 0.88 0.12 F. J. Hodges I, 12-20 seconds, April 10, 1915........ 23 0.48 0.08 24 22 0.00 Wi ees Lazer O82 Here the first column gives the name of the subject, the date of the experi- ment and the period of effort. The last is indicated by two hyphenated num- bers, the first being the second during which the effort was begun, the second that during which the effort was ended. The second column gives the serial number of the second corresponding to the values which follow it on the same line; the third and fourth, the duration in seconds of the cardiac cycle and atrio- ventricular interval respectively. For example, in Menninger I thirty-five seconds after the beginning of the experiment (that is, during the second which followed the cessation of the effort) the duration of the cardiac cycle was 0.48 second, while that of the atrio-ventricular interval was 0.12. Note that occasion- ally atrio-ventricular interval varies in same direction as duration of cycle. TABLE 6 Changes in the Q-, R- and S-waves due to the Valsalva experiment R- AND S-WAVES NAME, ETC. Ligh: mee S-WAVE Q-WAVE hia ened Acceleration eee vessel cea aD 20 21 23? , 0.72 0.48 0.46 0.48 seconds..... Kolls I, 18-32 80 | 14 36 seconds..... 0.62 0.80 Kolls II, 16- \ 87 saa IG 26 43 48 seconds.. 0.56 0.60 | 0.60 0.68 ae ae oak 34 0. CoV eee erence 0.68 0.48 | 0.68 0.48 0.48 Onds!- eee eae MES ee, 100s S2ted (8 , 0.78 0.86 0.80 0.46 0.48 seconds..... Be or ae ey lic Mee. 2041 7 2 ; 0.62 0.60} 0.48 0.48 0.60 0.58 Ondshen se mae eS POs di ae ailuisheoe —|12 23 0.82 0.60 | 0.50 0.44 0.60 0.58 Onds ee eee P. C. Hodges | aa Sas 73 Nause Il. 2" 56 The first column shows name of subject, number of experiment on that sub- ject and duration of effort, i.e., serial number .of seconds during which effort began and ended; the second gives maximum reduction of duration of cardiac cycle, i.e., percentage of normal cycle to which the duration fell. The third and fourth columns are best explained by these examples: In Hodges I a decrease in the R- and S-waves began at 10th second (during preliminary quickening) and stopped at 20th second (during acceleration). Below figures 10 and 20 are figures 0.72, 0.48 indicating duration in seconds of corresponding cardiac cycles. A second example is that of Kolls II. Here there were two separate periods of reduction of waves in question, one during the preliminary quickening and one during the acceleration. The fifth column shows beginning of increase of S- wave, viz., 21, 21, 10, and disappearance of increase, viz., 23?, 36, 23. Where there are three consecutive figures as 21, 34, 36, then 21 denotes beginning of increase, 34 a sudden rise in the degree of increase and 36 disappearance of in- crease. Durations of corresponding cycles are placed below these figures. The siath shows beginning and end of increase of Q-wave. Note more constant pres- ence of reduction of R-S waves, and relative infrequency of increase of Q-wave. 494 CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 495 in the size of the R- and S-waves; b, an increase in the Q-wave; c, an increase in the S-wave sometimes accompanied by a simultaneous decrease in the R-wave. No changes were observed when the heart became slower than normal but only when it quickened its pace. The general reduction in size occurred in seven experiments. In three of these the decrease began with the preliminary quickening and con- prererarepapneramarae Fig. 6. Electrocardiogram taken during a Valsalva experiment. Subject: Menninger (II). Upper line is a chronographic record showing fifths of a sec- ond; middle line is the electrocardiogram; lower line, signal showing beginning and end of effort. Note preliminary quickening, initial slowing, acceleration with increase in S-wave, and subsequent slowing. tinued all through the effort and for a few seconds after the effort was over. In three other experiments the decrease accompanied the pre- liminary quickening, then disappeared again during the acceleration. This change seems therefore not to be the result of the Valsalva experiment as such, but merely the effect of decreased vagal tonus. Finally in one case the change appeared during the preliminary quick- ening only. 496 PERCY M. DAWSON AND PAUL C. HODGES The other two changes both imply an increase of the negativity of the apex as compared with the base of the ventricle. The increase im the Q-wave occurred in two experiments on the same subject. In one of them (Dawson no. 2) the phenomenon was very marked; in the other (Dawson no. 1) it was distinctly observable although less pro- nounced. The change began during the effort and persisted for a few seconds after the cessation of the latter. It should be noted that in neither of these experiments was there a preliminary quickening. The increase in the S-wave sometimes began at the time of cessation . of the effort (one experiment) and always lasted until the acceleration . gave place to the subsequent slowing. It might, however, begin much earlier in the effort (two cases). It never accompanied the preliminary quickening and may therefore be irfferred to result from the mechanical conditions of the Valsalva experiment itself. In the present unsatisfactory state of our knowledge regarding the cause of the Q-, R- and S-waves and the significance of their varia- tions, the writer feels that all he is justified in doing is to record the foregoing facts without attempting an interpretation. It is true that the tables might be turned and the facts in question might be utilized for the elucidation of the reverse problem, namely, the meaning of the Q-, R- and S-waves, but for dealing adequately with the latter further experimentation along these lines is desirable. The sphygmogram Owing to some curious statements in the literature in regard to the significance of the sphygmogram obtainable during Valsalva’s experi- ment (4), it seemed well to devote some attention to this matter. Of several sphygmograms obtained one is given as an example (fig. 7). Here it is seen that the normal pulse waves undergo a series of changes which may be described as follows: a, a fall of the base line; b, elevation of the base line, decrease in size of the beats and obscuration of dicrotic wave; ¢, fall of base line but not to normal, hyperdicrotism, and con- tinued small size of pulse wave; d, return to normal by decrease in dicrotic wave, fall in base line and increase in size of pulse. One finds no difficulty in interpreting these changes: a, The fall in the base line is the result of the strong inspiration with which the Valsalva effort begins and which assists in emptying the radial venae comites. |, The rise in the base line is due to distention of the radial artery result- ing from the rapid rise in arterial pressure and also to the filling of the CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 497 venae comites resulting from the rise in intrathoracic pressure which obstructs the venous outflow from the arm. c, As the arterial pressure falls rapidly, the base line sinks but does not reach normal since the veins are still engorged. d, With the return of normal pneumatic conditions within the thorax, the veins empty rapidly and the base line falls to normal. The hypodicrotism and the hyperdicrotism are evidences of high and low blood pressure respectively. These explanations, as far as they go, seem to the writers to accord well with known facts and to render the efforts of some investigators not a little superfluous. We2, Wor. 27, 19/F.. -- Fig. 7. Radial sphygmogram during Valsalva experiment. Chronograph records fifths of a second. 2 indicates approximately beginning and end of experiment including preliminary slow and deep inspiration. Note variations in base line (an index of engorgement of radial vessels), and in height of dicrotic elevation (an index of arterial pressure). LIFTING It has been the view usually accepted that lifting and the Valsalva experiment are essentially alike insofar as the disturbances of the circulation are concerned. This statement is in the main correct. The blood pressures and pulse rate usually do show a picture which corresponds to that seen in the Valsalva experiment although often these changes are less pronounced. As in the case of the Valsalva experiment, there are two series of variations to be considered, the blood pressure changes and the variations in the pulse rate. We shall consider these in turn. Blood pressure (systolic) The blood pressure changes will be rather summarily dismissed, otherwise, because of the essential resemblance between the effect of lifting and that of the Valsalva maneuver, we would only be engaging in vain repetitions. Three experiments were performed and showed the primary rise and the fall but no second rise until after the cessation of the effort. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 4 498 PERCY M. DAWSON AND PAUL C. HODGES It is, however, highly desirable that these studies should be carried further. Enough material has not been gathered to enable the writer to study individual variations. Pulse rate The pulse rate on the other hand received adequate attention. Its behavior also resembles in the main that observed in the Valsalva experiment. There are, however, a few exceptions as will be seen presently. The method employed was as follows: The subject stood upon a strong square piece of wood (about 80 x 80 em.) to the center of which was fastened by a hook a registering spring dynamometer. The sub- ject stood over the hook and passed around the body and over one shoulder a sling which was secured to the free pole of the dynamometer. The sling was then shortened until the subject stood with the knees shghtly bent. At the word of command the legs were extended and the subject made a vigorous attempt to lift the square piece of wood upon which he stood. The amount of effort was registered by the dynamometer. The subject ceased his effort with a loud gasp which was recorded by the operator who had also marked upon the electro- cardiographic tracing the moment at which the command to lift had been given. During the whole of this procedure electrocardiograms were being taken from the two forearms. The alterations of the pulse rate are presented best by means of the accompanying table (no. 7) and charts (8 to 15). The former contains most of the data embodied in the discussion while the latter are shown chiefly for the purpose of illustration. On perusing table 7 one becomes aware of the following facts: a, In general the changes in heart rate are similar to those produced by the Valsalva experiment, but are quantitatively less. b, The preliminary quickening is a very inconstant phenomenon. ‘This accords with the supposition already advanced that the preliminary quickening is due to an accentuation of inspiration for it often occurs that one does not precede an effort at lifting by taking a deep breath. c, The initial slowing occurred in numbers 1 and 4, but was absent from numbers 6 and 12. This may indicate that the phenomenon in question is at least partly dependent upon the severity of the lift for the dynamom- eter record was lower in numbers 6 and 12 than in numbers 1 and 4. d, The subsequent slowing is one of the most constant features; its CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 499 appearance may however be delayed for as much as one minute (nos. 5 and 10). It is not always the last sign to disappear when the effort is reduced for although this would seem to be the case sometimes (no. 12) it is not always so (no. 15). e, The acceleration is also a pretty constant feature. It may extend throughout most of the period of effort (nos. 5 and 7). On increasing the effort, it may linger after the secondary slowing has disappeared or on decreasing disappear before it (nos. 15,12). f, The erratic records (nos. 7,13 and 14) require special con- TABLE 7 Changes in heart rate due to lifting PARES INITIAL Sona} NUMBER SUBJECT, ETC. QUICEEN- SLOWING ACCELERATION Sees 1 131600 oe Sl ee eee ? 126 Absent 166 2 OlIiNetbe tac cierancsenoereck: 89 Absent Absent 139 3 HeMoubleniisee. yacc cee 88 Absent 82 123 4 Pressenitim Ls). a.) | Absent 120 89 120 5 Boadman62).6.08......8 720) AAbsent:| “Absent 80 116 (1) 6 Hodges! Di yascives). act! ... 85 Absent 66 134 7 Dawsonville... ............. |) Absent) | Absent 86 (2) 140 8 Schilaiiete aera nacre Absenitn |leAbsent 7D) 159 9 OISKGCY 2c As onieareteis are. «| Absent |) Absent 70 Absent 10 | Dawson II*...............| Absent | Absent | 79 152 (3) 113] INTORIS Hae Rik cere Pontiayte ox 87 Absent 78 177 12 Pressentin II..............] Absent | Absent Absent 113 fe pCO Gubler Uy. oacs 5.5. ? Absent | 97,114 (4) | 125 em elley i. 3. oes. chs dns cao ? Absent | 121, 146 (5) | 128 15 Dawson 18................| Absent | Absent 74. Absent 16 Dawson dO... s46.ee -slasee| Absenta| Absent 60 114 TE AOIEET a) eases trey ie Rea 87 123 76 136 WBA WETA GO cn Sob c cde asses hs adios 96 102 81 131 The first column gives reference numbers which are used in discussion in text For remaining columns see legend of table 3. No change of less than 10 per cent is regarded as exceeding the limits of normal variation and consequently such changes are designated ‘‘absent’’ in table. Interrogation point and averages as in table 3. (1) Signifies maximum change reached late (60 seconds); (2) inter- rupted by a slowing (see page 505); (3) coming on late; (4) omitted from both averages; here a slight acceleration occurred (97 per cent) but the last and long- est part of effort was accompanied by slowing (114 per cent maximum); (5) omitted from both averages; here usual acceleration was replaced by slowing, shortest cycle during effort was 121 per cent, the longest 146 per cent. In num- bers 1, 4and 15 effort was much more severe than in numbers 6, 12 and 16 respec- tively. *Shown in figures. Note that changes although essentially similar to those due to Valsalva experiment are less in amount. 500 PERCY M. DAWSON AND PAUL C. HODGES uv. 150 26. Aa ; come: 15 Figure 8. More ‘‘typical’’ effect upon the heart rate of lifting. Subject: Pressentin (I). For explanation of symbols see legend of figure 2. Note the initial slowing, acceleration and subsequent slowing. 0.4 Iso 08 aie 75 42 0 Fig. 9. Less ‘‘typical’’ effect of lifting upon the heart rate. Subject: Schlatter. For explanation of the symbols see legend of figure 2. Note the absence of the pre- liminary quickening, and the initial slowing. The acceleration and subsequent slowing are still pronounced. a a ee ee ee ee es Oe a ee ee Fig. 10. Less “‘typical’’ effect of lifting upon the heart rate. Subject: Daw- son (I). For explanation of the symbols see legend of figure 2. Note the absence of the preliminary quickening and the initial slowing; the presence of two periods of acceleration separated by a slight slowing, and of the subsequent slowing. 0.4 150 08 A ae 7 Fig. 11. “Atypical” effect of lifting upon the heart rate. Subject: Kelley. For explanation of symbols see legend of figure 2. It is exceptionally unfortu- nate that in this case the gap in the record (the half-minute pause) occurs where it does. It is possible that during this interval the usual secondary slowing may have taken place. Note the presence of the initial slowing, of a return to nor- mal following the cessation of the lift. ee CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 501 0.4 150 ‘ 7S Fig. 12. First of three figures showing the effect upon the heart rate of pro- longing the Valsalva experiment. Subject: Kolls (I). For explanation of the symbols see legend of figure 2. This short effort is presented for the sake of com- parison with figures 13 and 14. All three were obtained from the same subject. Note typical picture consisting of preliminary quickening, primary slowing, acceler- ation and subsequent slowing. ae F Fas Fig. 13. Second of three figures showing the effect upon the heart rate of pro- longing the Valsalva experiment. Subject: Kolls (II). For explanation of the symbols see legend of figure2. This effort is longer (32 seconds) than that of the preceding figures (14 seconds). Note preliminary acceleration, very slight pri- mary slowing (which amounts merely to a return to normal), moderate secondary acceleration, subsequent slowing. a4 150 A OK. A 42 50 Fig. 14. Third of three figures showing the effect upon the heart rate of pro- longing the Valsalva experiment. Subject: Kolls (III). For explanation of the symbols see legend of figure 2. This long effort lasted sixty seconds. Note the preliminary acceleration, the primary slowing, the acceleration, the subsequent slowing. There is also a period of slowing before the official end of the effort. 502 PERCY M. DAWSON AND PAUL C. HODGES 04 seen oTteeee | ; 150 a. ceeeee 55K A A 3 0. Bini Pe nino a PE 9 ee) SE oa ae Ca 0.4 Aa G Ses ise ery, Albina teas oa i A A 0.8 i—] So , : | 0.4 4 1 a , - . Coe ah Soe cane nes tate Seale 0 : Sree ; 47K A A 0.8 15 Fig. 15. The effect of lifting upon the heart rate as modified by the severity of the effort and by gasping. Subject: Dawson. For explanation of the sym- bols see legend of figure 2. The three curves are constructed on the same scale. The arrows pointing downward toward the third curve indicate the moments at which the subject gasped while making this record. The weights lifted which correspond to the three curves are, when arranged in order from above downward, 55, 35 and 47 kilos. Note that the last curve shows no greater deviation from the normal than the second curve excepting the fact that the acceleration took longer to pass away. This is in spite of the greater effort as compared with the second. TABLE 8 Effect of lifting on duration of Q-T complex DURATION OF SUBJECT, ETC. SECOND - 3 ; Ventricular | Ventricular Cycle systolic | diastolic 48 1.04 0.22 0.46 39 0.80 0.22 0.24 Dawson II, 15-36 seconds......:....:.. 13 0.68 0.22 0.46 0.48 0.24 0.24 ———* SS i) for) 22 0.68 0.24 0.44 Pressentin I, 3-10 seconds............. 1 0.60 | 0.24 0.36 12 0:52 | . 0.22 0.30 36 0.64 0.24 0.40 Pressentin II, 3-30 seconds............ 2 0.60 0.22 0.38 | 27 0.56 0.22 0.34 For explanation see legend of table 4. Note similarity of this table to corre- sponding table for Valsalva experiment (no. 4). CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 503 sideration: 1, The record no. 7 is easily interpreted by supposing that a fluctuation of the attention caused this subject to relax his efforts and then to reinforce them just before the cessation of the pulling. 2, The record no. 13 might be explained in the same way, but for the fact that we would then expect the slowing to continue after the cessation of the effort until the final return to normal. If this objection be valid we are left without an explanation of this record. 3, The record no. 14 is characterized by a great slowing during the effort. Here again one might urge that the subject ceased her efforts before she was aware of it, and that the actual termination of the effort does not coin- cide with the intentional termination. But the same objection that has just been advanced in the case of no. 13 applies here with redoubled force for the unusual slowing does not persist after the effort has come to its intentional conclusion but is actually replaced by a marked ac- celeration. The justifiable supposition is that the intentional termination of the effort is the actual termination of the effort, but if so the unusual slowing remains without an explanation (cf. pp. 505 and 485). Duration of the Q-T complex. Three electrocardiograms were found suitable for this study and the results were similar to those occurring in the case of the Valsalva experiment (table 8). Character of the changes in rate Duration of the atrio-ventricular interval. What has already been said in regard to the unsatisfactory character of the electrocardiograms in the case of the Valsalva experiment applies with still greater force in the case of the lifting experiments. We shall however consider such results as have been obtained. Of seventeen experiments only one showed a P-wave of adequate distinctness, the rest being spoiled by vibrations due to adventitious alternating currents. This single record (Dawson I) showed no change in the atrio-ventricular interval although the cycle was shortened from 0.64 to 0.44 seconds. These cycles occurred in the second and seventeenth seconds respectively, the period of effort being from the third to the twelfth seconds. Variations in the Q-, R- and S-waves of the electrocardiogram. As in the case of the Valsalva experiments the electrocardiograms were also examined to ascertain whether during the changes in rate there oc- curred any modifications of the Q-, R- and S-waves. Ten of the seven- teen experiments were found satisfactory for this study and the three changes observed in the case of the Valsalva experiment were noticed, and shall now be considered in order. 504 PERCY M. DAWSON AND PAUL C. HODGES 1. General decrease in the size of the R- and S-waves. Five of the experiments were negative while four gave results similar to those obtained in the Valsalva experiments (table 6). experiment (Kelley) differed entirely from the rest. The Here the maxi- remaining mum acceleration was to 94 per cent but the record showed that during TABLE 9 Changes in the R- and S-waves due to lifting NAME, ETC. PER CENT R- AND S-WAVES een Acceleration Kelley, 4-14 seconds.......... { nie ees) Se 5 7 10 16 Jolivetiouts lacs eee eee ee 0.56 0.54 | 0.60 0.66 Pressentinai, j2.40/..5 are. ab. stone 90 No change No change 84 3 22 Hy Boublew Mis, eelicei eels 0.60 0.60 Pressembunl Wee eects aieceraue orev ete | 80 No change No change RO GMAR s trsp eae ee ee ard eee 79 No change No change PAWSON S oe ete ce he eon oe 74 No change | -No change ; 70 6 19 Keiakey vised NAO 0.56 0.46 67 15 36 Dawson II, 15-36 seconds { 0.64 0.60 Dawson 9. 2.004 Are job adore: 65 No change No change For explanation see table 6. The figures in parentheses are the discordant results (Kelley) referred to in text. table for Valsalva experiment (no. 6). Note similarity of this table to corresponding the greater part of the lift the duration of the cycles was increased, attaining a maximum of 144 per cent. Along with this increase in duration of cycle there was a marked increase in the R-Q complex lasting from the third to the fourteenth second. No explanation is offered at the present time for this phenomenon. eae CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 505 2. Increase in the S-wave was not found at all, and 3. Increase in the Q-wave was found in only one instance at the period of maximum acceleration when the duration of the cycle had been reduced from 0.58 to 0.46 seconds. CERTAIN MODIFICATIONS OF LIFTING AND THE VALSALVA EXPERIMENT The effect upon the heart rate of prolonging the Valsalva experi- ment was studied in a single case. (Figs. 11, 12 and 13.) Here the usual four phases appeared in the first, the shortest, effort. On length ening the period of effort there was no conspicuous change in the char- acter of the results. But when the period was still further lengthened there appeared a marked slowing before the recorded end of the effort. On this record is marked the only intentional termination of the effort but if the attention or strength of the subject should flag during the experiment and the effort therefore be relaxed, a considerable amelio- ration of the pressure conditions within the thorax would result with- out the subject becoming aware of it. This state of things might give evidence of its existence through a premature slowing. The latter would disappear again when the effort was reinforced as it would be prior to the intentional abandonment of the effort which brings the experiment to its official termination. Support is given to this explanation by the somewhat similar results shown in figure 10. Here the writer (Dawson) was the subject and was watchful of his mental states during the course of the experiment (lifting). A fluctuation in the degree of effort was synchronous with a premature slowing which was succeeded by an acceleration followed in turn by a second slowing, the one which we have been designating the subsequent slowing. It has been seen that the pulse variations are of two sorts, the pre- liminary slowing which coincides with a respiratory movement, and the other three phases which are synchronous with changes in pressure. The initial slowing accompanies the primary rise in pressure; the acceleration, the fall in pressure and the subsequent slowing, and the return of the blood pressure to or above normal. It is a suggestive thought although perhaps not a justifiable assumption at the present time to regard the changes in heart rate as indices of blood pressures. In such a thought or assumption it is implied that in Kolls III there was a secondary rise in systolic pressure comparable to that described by Bruck (1). 506 PERCY M. DAWSON AND PAUL C. HODGES The effect wpon the systolic blood pressure and pulse rate of gasping during the Valsalva experiment, and during lifting was studied in several instances. This seemed desirable because of the fact that certain forms of physical effort are characterized by periods of strain inter- rupted by gasps. All very prolonged strains are so interrupted. Ob- servations upon the blood pressure changes were confined to the Val- salva experiments while those dealing with the heart rate were con- cerned with lifting only. In the former study the method employed was the following: The pressure in the cuff of Erlanger’s sphygymomanometer was raised to a high level and allowed to fall slowly during the performance of a Val- salva strain interrupted by gasps. At every 5 mm. of fall the observer marked with a key upon the drum. When the record was completed the pulse curves were examined to locate the first appearance of pulse tracings showing one of the systolic characters, namely, the separation of the catacrotic from the anacrotic limb. The pressure (shown by the manometer) which was synchronous with this form of pulse wave was taken to be the systolic pressure at that moment. In three experiments in which the manometer pressure fell from 180 mm. the systolic pressure was found to be 140, 130 and 130 mm. respectively, during gasps. The normal systolic pressure of the sub- ject observed was 118 mm. In studying the heart rate the usual electrocardiographic method was employed. The results obtained are shown in fig. 15. Three experiments were performed upon the same subject (Dawson). In the first 55 kilos were lifted, in the second 35, and in the third 47. The duration of the lift was progressively longer in the order given (10, 16 and 21 seconds respectively) and during the third effort the subject gasped repeatedly. As shown in the figure the changes in rate were greatest in the first experiment while those in the other two were about equal, the somewhat greater acceleration in the second being balanced by the greater subsequent slowing in the third. Thus the changes produced in the third experiment (with gasping) were no greater than those in the second in spite of the fact that in the third experiment the duration of the lift was 34 per cent longer and the weight lifted 37 per cent heavier. CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 507 LABOR PAINS As already stated in the introduction, parturition resembles con- siderably the Valsalva experiment in its mechanics. There is here the same closure of the glottis and forced expiratory effort, and we may infer that the cardio-vascular reaction is similar in the two phenomena. A chance observation upon a rabbit which was unexpectedly found to be pregnant during the course of an experiment seems worthy of being recorded. In one of a series of experiments on the blocking of nerve impulses through the application of heat, the sciatic of this rabbit which was under the influence of chloral, urethane and ether was stimulated with a moderate tetanic current. This stimulus was UO MAU MMS Fig. 16. Changes in respiration and mean blood pressure during labor. Sub- ject: rabbit under chloral, urethane and ether. Upper line records respirations by means of tambour connected with trachea cannula; middle line, mean blood pressure recorded by means of cannula in carotid artery; lower line, seconds and zero pressure. Mean pressure before spasm 114 to 122 mm., at height of rise 128 mm., at depth of fall110 mm. Hg. The record shows no change in rate of heart rate nor of respiratory movements. Note the decrease and cessation of respiration, a rise followed by a fall in mean pressure. sufficient to start a series of labor pains which resulted in the abortion of several nearly-mature fetuses. As the carotid of the rabbit was at the time connected with a mercury manometer, a record of the mean blood pressure was obtained during these labor pains. The respiration was also being recorded by means of a side tube from the trachea cannula. On examining this record (fig. 16) one observes a uniform movement of the respiratory recorder and of the mean blood pressure. The latter shows waves of the first (cardiac) and of the third order (vaso- motor) (5). The waves of the second order (respiratory) are absent. From the normal level the mean pressure rises slowly for the 508 PERCY M. DAWSON AND PAUL C. HODGES duration of about three of the large waves. It now rises sharply, next falls more slowly to a level which is subnormal and from which it re- turns to normal in the duration of about three of the large waves (sixty seconds). These pressure phenomena are nc*+ dissimilar to those obtained in the case of the Valsalva experiment bus since the closure of the glottis has been rendered ineffective by tracheotomy, the changes in question must result from a totally different set of causes. It is suggested that the cause in this case is the increase in peripheral resistance due to the contraction of the uterus. The writers have frequently observed a well marked and similar pressure change in the rabbit on stimulation of the peripheral end of the sciatic where the only area of constriction was the lower leg and foot and it seems not unnatural to suppose that these changes in pressure may readily result from the constriction and relaxation of so large and vascular an organ as the uterus at term. During the sharp rise in mean pressure the respirations show a short increase, a diminution for several movements, a cessation for a couple of seconds and as the pressure falls a rapid return. The subsequent movements are at first supernormal but normality is reached in about thirty-five seconds. The respiratory pause seems to have been in inspiration for the last weak movement was inspiratory (a down stroke) and the first of the returning movements expiratory (an upstroke) Under the conditions which existed in this animal, one would have expected the respirations to show first a deep inspiration followed at once by the strong although mechanically ineffective expiration. The phenomena observed, namely, gradual suppression and gradual return of the respiratory movements, are puzzling and the writers do not hazard a conjecture as to their significance. Suffice it then to em- phasize that in constructing a picture of the blood pressure changes in parturition one must include changes which are independent. of the closure of the glottis. SUMMARY OF RESULTS The following summary is submitted without further discussion. Credit to previous investigators together with the comment and criti- cism of one of us (D) will be reserved for the historical résumé already referred to (footnote 3, p. 482). I. In the Valsalva experiment the following phenomena were observed: CARDIO-VASCULAR REACTIONS IN VALSALVA EXPERIMENT 509 1. The systolic pressure rose rapidly, primary rise (McCurdy). This was followed by an extensive fall (Valsalva and Weber). 2. In place of the extensive secondary rise described by Bruck (to 180-200 mm.) a rise to not more than 140 mm. was observed, i.e., a height readily accounted for by the mild degree of asphyxia which occurred in these experiments. 3. When in the anesthetized dog the inflation of the lungs had reduced the mean blood pressure by 92 mm., neither violent abdominal compression nor severe dyspnea raised this pressure above normal again. On relieving the inflation the mean pressure rose somewhat above normal. 4. On cessation of the Valsalva effort there wag a moderate rise of the systolic pressure. 5. The pulse rate showed a preliminary quickening which accom- panied the inspiratory movement (cardiac cycle was sometimes reduced by 30 per cent), an initial slowing (cardiac cycle was sometimes in- creased by 33 per cent), an acceleration (cardiac cycle sometimes decreased by 50 per cent). The last usually began during the effort and continued for several beats after the effort had ceased, often not reaching its maximum until that time. 6. Subsequent to the effort there was a slowing (cardiac cycle some- times lengthened by 85 per cent) which might continue for 40 to 50 beats. 7. As the time which elapsed between the Q- and T-waves (electro- cardiographic) showed but little variation, the changes in the cardiac cycle may be regarded as chiefly diastolic. 8. The atrio-ventricular interval was usually unchanged but might be shortened with the shortening of the cardiac cycle. 9. The R- and S- waves might be reduced during the preliminary quickening or the acceleration or both. The acceleration might be accompanied by an accentuation of the Q-wave, or by an increase of the S- with a simultaneous decrease of the R-wave. 10. The changes in the sphygmogram were readily attributable to the changes in venous and arterial pressure. IJ. Lifting was characterized by the following phenomena: 11. The changes in systolic blood pressure were essentially similar to those which occurred in the Valsalva experiment. 12. The changes in pulse rate were often essentially similar to those occurring in the Valsalva experiment. They were, however, less regularly present and might be less in amount. In order of the decreas- 510 PERCY M. DAWSON AND PAUL C. HODGES ing frequency of their occurrence they were a, subsequent slowing and acceleration; b, initial slowing; c, preliminary quickening. 13. The shortening of the cardiac cycle was chiefly diastolic. The R- and S-waves might show a simultaneous decrease. The Q-wave was sometimes accentuated. No increase in the S-wave was noted (with the exception mentioned on page 504, viz., no. 14). III. When the Valsalva or lift was modified, the usual picture was somewhat changed as follows: 14. The effect upon the heart rate of greatly prolonging the Valsalva experiment was to decrease the intensity of the effort and with this the extent of the changes in the heart rate. 15. When the Valsalva experiment experienced a series of interrup- tions consisting of a single gasp each, the systolic blood pressure rose with each gasp 10 to 20 mm. above normal. 16. When a lifting experiment experienced similar interruptions, the changes of heart rate were less in extent than would otherwise occur even when the weight lifted was greater in amount. IV. Observations made during parturition: 17. During a labor pain in the anesthetized and tracheotomized rabbit, the mean blood pressure experienced changes (rise followed by fall) which are attributable to the changes in peripheral resistance due to uterine contraction and subsequent relaxation. 18. The respiration also changed decreasing in amplitude to a stand- still in the inspiratory phase and then gradually returning to normal. The writers have much pleasure in acknowledging the assistance of those students of medicine and of physical education who served as subjects in this research. To the names already mentioned in the text should be added those of Misses Glassow and McFadden. REFERENCES Bruck: Deutsch. Arch. f. Klin. Med., 1907, xci, 171. Dawson AnD Hopass: This Journal, 1916, xl, 139. Hopaes: Thesis for Bachelor of Science, Univ. Wisconsin, 1915. Lewis: Journ. Physiol., 1906, xxxiv, 391. Mayer: Wiener Sitzungsb., 1876, Ixxiv, 281. McCorpy: This Journal, 1901, v, 95. Weser: Konig. sichs. gesell. d. Wissensch. Math. Phys. Classe, 1850, ii, 29. Also Arch. gén. de méd., v. series, 1853, i, 399. Toor WN FH SOME ASPECTS OF THE NEUROMUSCULAR RESPIRATORY MECHANISM IN CHELONIANS HELEN C. COOMBS From the Department of Physiology, Columbia University Received for publication October 29, 1919 In 1795 Robert Towson, at G6ttingen, first showed that tortoises do not swallow air, as had previously been believed, but that by the contraction of certain muscles, the lung is compressed and expels air, “‘then, ceasing to contract, the other muscle contracts and draws the former, (e.g., the lungs), within; thus a vacuum is formed into which the air rushes, as in the respiration of animals with a thorax” (1). In 1863 Mitchell and Morehouse (2) described the normal respira- tion of the turtle as being due to the alternate contractions of certain antagonistic groups of muscles. ‘‘Inspiration is effected by the con- traction of the flank muscles, which in appearance strongly resemble the diaphragms of superior animals. Expiration is effected by muscles which he within the breast-box and consist of anterior and posterior bellies connected by a strong tendon continuous across the mid-line and common to both sides of the animal. These muscles act together and compress the viscera against the lungs.’”’ According to these authors the lung does not take an active part in the respiratory move- ments, but is compressed by the contraction of the expiratory muscles, while its passive expansion, due to the contraction of another set of muscles, causes more air to rush in and fill the lung cavity as in mammalian respiration. The same authors found that the neural apparatus of the respiratory mechanism consists essentially of the vagus supplying the larynx and lungs, of the spinal nerves distributed to the respiratory muscles of the trunk, and of the medulla oblongata through which the synchronous movements of the glottis and flanks are controlled. The functional activity of the vagus in connection with the lungs is not emphasized, however, 11 view of the passive réle played by the lungs in normal respiration. In 1878 Paul Bert made further observations, confirming the fact that respiration is carried on by contraction and dilatation of the oll 512 HELEN C. COOMBS thoracic-abdominal cage. He also made the important observatiom that the lungs themselves can be made to contract by direct electrical stimulation (3). Frangois-Frank, in 1906, made a comprehensive and detailed study of the muscular mechanism of respiration in the turtle, with an analysis of the normal respiratory curves. He studied the lungs and their innervation by the vagus. He also observed spontaneous contractions of the lung tissue so long as the medulla was intact (4). In 1908 Prévost and Saloz (5) called attention to the fact that stimu- lation of the vagus produces constriction of the bronchioles in the turtle, and in 1918 Jackson and Pelz demonstrated that while stimu- lation of the vagi produces constriction of the lungs, stimulation of the sympathetic chain with a very weak tetanizing current produces dilatation (6). In this paper the experimental work under discussion is twofold, dealing with the activity of the lungs themselves under certain con- ditions, and the effects of experimental lesions of the cerebrum, optic lobes and medulla upon the respiratory muscles which are effective in maintaining normal respiration. The activity of the lungs is first considered. Turtles were lightly etherized and the cranial cavity was opened. The cerebral hem- ispheres were removed, and the technique followed was that devised by Jackson in which, after removing the plastron, the fore and hind limbs were removed as well as all the viscera except the lungs and heart. The latter was connected with a recording lever. The vagi were freed for some distance in the neck, and ligatures were placed loosely around them. A small glass cannula was inserted in the trachea, by means of which, after being half inflated, the lungs were connected with a very sensitive recording tambour, which indicated very small changes in volume. " 3.0 0.19 0.24 10:50 | Injected 1 ce. of alcoholic extract of duodenum 11:50 15.8 0.35 0.40 12:50 0.8 0.31 0.38 12:50 | Injected 1 cc. of residue of duodenum after alcoholic extraction 1:50 B20 0.08 0.14 2:50 0.6 0.05 0.15 Each of these experiments was confirmed on two other animals, one having a very refractory Heidenhain stomach. IS PEPTIC DIGESTION A FACTOR IN LIBERATING THE GASTRIN ACTIVITY? Since appreciable peptic hydrolysis occurs in the preparation of the stomach extracts it was thought that possibly the extraction or libera- tion of the gastrin activity was facilitated in that case and that possibly peptic digestion of other tissues might yield more active preparations than simple extraction by 0.4 per.cent HCl. Accordingly various tis- sues were finely divided, well mixed and separated into two portions; one to be extracted in the usual way with 0.4 per cent HCI and the other digested for five days at 35° to 40°C. in 0.4 per cent HCl with the addi- tion of toluol and 5 grams of scale pepsin (1: 3000 U.S. P.) per 250 to 350 grams fresh tissue. The other steps in the preparation were as previously described. Both were finally treated with alcohol in the usual way, but no extraction with absolute alcohol was carried out. The final solutions represented as before 4 to 5 grams fresh tissue per cubic centimeter. As a control the amount of activity introduced by the pepsin had, of course, to be determined. Ten grams scale pepsin (1:3000 U.S. P. Armour) were dissolved in 0.4 per cent HCl, shaken with toluol, and digested at 35° to 40°C. for five days with frequent shaking. This was then heated to boiling, filtered and treated in the usual way. The final volumes were so concentrated that 1 cc. of the absolute alcohol soluble fraction was equivalent to 0.55 gram of pepsin and 1 ce. of absolute alcohol insoluble residue equal to 1 gram of the scale pepsin. A protocol attached shows that most of the activity goes into the absolute alcohol soluble fraction, but even this is not large. DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 531 Dog I. Pawlow stomach, August 1, 1916 QUANTITY OF JUICE IN CUBIC CENTIMETERS PER CENT FREE HCl TOTAL HCl TIME 10:40 | Dressed 11:40 1.0 0.00 0.03 11:55 | Injected absolute alcohol soluble fraction equivalent to 1 gram of scale pepsin (1 to 3000) 12:55 2.4 0.14 0.17 1:55 0.5 2:20 | Injected absolute alcohol insoluble fraction equivalent to 1 gram of scale pepsin (1 to 3000) 3:20 16 | 0.03 A more refractory Heidenhain stomach gave just a trace of activity, while dog V (gastric fistula with vagi crushed one year previously) se- creted from the entire stomach during the hour following the injection only 9 cc., free acidity 0.24; 0.32 total per cent HCl. Reference to the long list of inactive preparations in table 4, in which the extracts injected corresponded to 0.01 gram, 0.07 gram and 0.1 gram of scale pepsin respectively, demonstrates clearly that any activ- ity of the tissues extracted by the pepsin treatment is due to sub- stances other than that introduced with the enzyme. Duodenum. Reference to table 1 shows that when both absolute alcohol soluble and insoluble fractions are injected, the duodenal extract is quite as active as the stomach. The use of the scale pepsin does not appear to affect the activity of the preparation in any constant way; at any rate the comparative study suggests a loss or gain in activity, depending on which animal one considers. In five out of seven cases the stomach preparation was found more active than the duodenal extracts. 532 LUCKHARDT, KEETON, KOCH AND LA MER TABLE 1 Comparison of activity of gastric and duodenal mucosa ACIDITY EXTRACT EQUIVALENT TO 4 TO 5 GRAMS QUAN- Sea ee TISSUE tTity* | Free | Total per cent) percent HCl HCl [=| Stomach 4.9 | 0.29 | 0.34 Dog I, Pawlow stomach Duodenum and HCl 3.0 | 0.17 -|-0.25 [ Duodenum and pepsin + HCl 2.5 | OL255|0E35 : j Stomach 1.5 | 0.19 | 0.26 ge tne Duodenum and HCl 1.0 | 0.00 | 0.05 Duodenum and pepsin + HCl 1.4 | 0.09 | 0.17 Stomach 7.5 | 0.33) | 02389 Dog III, Gastric fistula Duodenum and HCl 7.3 | 0.19") 0234 Duodenum and pepsin + HCl 12.4 | 0.39 | 0.46 Dog IV, Gastric fistula. Stomach 9.0 | 0.37 | 0.41 Vagi crushed 1 year Duodenum and HCl 1.6 | 0.03 | 0.08 previously Duodenum and pepsin + HCl 6.5 | 0.22 | 0.27 Dog V, Gastric fistula. Stomach 15.0 | 0.46 | 0.50 Vagi crushed 1 year Duodenum and HCl 10.5 | 0.41 | 0.67 previously Duodenum and pepsin + HCl 8.2 | 0.44 | 0.50 e : | Stomach 11.7 | 0-15 f043 cate eat Estults 9)! Taodenum and HC! 4.5 | 0.23 | 0.42 Bee Duodenum and pepsin + HCl | 15.5 | 0.14 | 0.26 : Stomach 5.2 | 0.30 | 0.37 ara OTL. Duodenum and HCl 6.2 | 0.26 | 0.35 6 ory Duodenum and pepsin + HCl 5.0 | 0.05 | 0.13 *Quantity represents excess of secretion in experimental over control period. Liver. The first extract of liver was prepared in August, 1916, and gave activity far above any crude preparations which we have studied. The pepsin digests appeared to be somewht less potent than the simple acid extractions. These points are illustrated in the four protocols attached. een) en Pe DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 533 Dog I. Pawlow stomach, September 19, 1916 QUANTITY OF JUICE IN CUBIC TIME enemas PER CENT FREE HCl PER CENT TOTAL HCl 7:55 | Dressed 8:55 0.5 0.04 9:05 | Injected 1 ce. of liver extract (by HCl) 10:05 15040) 0.43 0.047 11:05 6.0 0.47 0.51 Dog I. Pawlow stomach, September 21, 1916 QUANTITY OF JUICE IN CUBIC TIME So ST PER CENT FREE HCl PER CENT TOTAL HCl 8:20} Dressed 8:50 0.6 0.09 8:50 | Injected 1 ce. of liver extract (by pepsin and HCl) 9:50 6.5 0.39 0.44 10:50 1.0 0.34 0.40 Dog VI. Vagotomized gastric fistula, September 19, 1916 QUANTITY OF JUICE IN CUBIC : i ; TIME GEST TTT PER CENT FREE HCl PER CENT TOTAL HCl 8:25 | Dressed 8:55 4.5 0.33 0.47 9:10 | Injected 1 ce. of liver extract (by HCl) 10:10 95.0 0.41 0.49 11:10 6.5 0.33 0.44 Dog VI. Vagotomized gastric fistula, September 21, 1916 QUANTITY OF JUICE IN CUBIC HCl TIME aera PER CENT FREE HCl PER CENT TOTAL HC 8:35 | Dressed 9:05 8.5 0.36 0.44 9:05 | Injected 1 cc. of liver extract (by pepsin and HCl) 10:05 35.0 | 0.38 0.46 11:05 4.0 0.29 0.41 These results were confirmed on five other animals in a series of twenty-three experiments. Their uniformity was quite striking. An attempt was made to confirm these results on liver extracts made on October 26 and November 11 of the same year. The first of these stood for two days in the ice-box and the second one was made immediately 534 LUCKHARDT, KEETON, KOCH AND LA MER upon receipt of fresh liver. On the two preparations of October 26, (one HCl extract, the other HCl and pepsin extract) four experiments were conducted. Reviewing the behavior of these animals we regarded one experiment as positive, two as showing a trace of activity and the fourth negative. The positive experiment is cited to show the low grade of activity as compared with experiment on same animal on September 19. Dog I. Pawlow stomach, November 9, 1916 TIME y eben I Ad PER CENT FREE HC] PER CENT TOTAL HCl Secretion 1 hr... 1h 0.08 Injected 1 ce. of liver extract (by HCl) Secretion Ist hr.. 2.9 0.09 0.16 Secretion 2nd hr. 1.0 0.14 0.21 On the preparations of November 11, five experiments were run using three Pawlow stomachs, two of these being quite sensitive. Dog IX had both splanchnic nerves cut as they emerge from the thorax beneath the pillars of the diaphragm. The two more sensitive animals re- sponded, but the other one (dog VIII) gave negative results. The activity was somewhat greater than in the extracts of October 26, but was in no way comparable to results on the August preparations. TABLE 2 Activity of pancreas preparations DATE AND METHOD OF EXTRACTION August | October 26 Novorper ANIMAL TYPE OF STOMACH HCl HCl HCl HCl | pep-| HCl | pep-| HCl} pep- sin sin sin Ti MB awl owe taken eo setae orisha lee saier utile cheaters =— | +) = |) aie Ts) Heidenliaine es eee ter rai astoenchhs srseebca a Poe TT | \Gasiniettehilann: sens ceriiert M.S acer ic ee eee | IV | Gastric fistula, vagi crushed 1 year pre- VLOUG lye aac elas wd ares On Oe see se ica al ha | V | Gastric fistula, vagi crushed 1 year pre- WLOUSI Vac eee eee tings oar ce ko ete Fan Ui ol (icc et WI. ("Gastrictfistulat 5-4 tears cleats. MEARS Stets =| ar VII | Gastric fistula, vagotomized................ — VEEL wRawlow:Paere tan oo eee cok cue aie ae = |= = TX). Pawlowssplanchnics cuit: s,...-. octys cst sake =| 2 ale DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 535 Pancreas. Reference to table 2 shows that the August preparation with pepsin extraction was uniformly active on all animals. In only one case (dog V) was there any trace of activity in the HCl extract. The other preparations of October 26 and November 16 gave two posi- tive and two questionable reactions out of seventeen experiments. An idea of the activity of extracts may be secured from the attached protocols. August preparations Dog I. Pawlow stomach, September 24, 1916 US OFS UICn EN CUBIC PER CENT FREE HCl PER CENT TOTAL HCl = CENTIMETERS 9:50 | Dressed 10:50 1.6 0.00 0.04 11:10 | Injected 1 ce. of pancreas extract (by pepsin and HCl) 12:10 5.7 0.33 0.37 1:10 0.5 0.46 0.55 Dog IV. Gastric fistula, vagi crushed one year previously, September 24, 1916 TIME eS Ea AES PER CENT FREE HCl PER CENT TOTAL HCl 9:50 | Dressed 10:50 0.2 0.06 11:10 | Injected 1 cc. of pancreas extract (by pepsin and HCl) 12:10 34.6 0.48 0.47 1:10 4.6 0.40 0.42 With these may be compared the positive experiment on dog I with the preparation of October 26. Dog I. Pawlow stomach, November 8, 1916 QUANTITY OF JUICE IN CUBIC TIME Saas PER CENT FREE HCl PER CENT TOTAL HCl 9:30 | Dressed : 10:30 16 0.09 10:30 | Injected 1 cc. pancreas extract (by pepsin and HC) 11:30 2.5 | 0.09 | 0.12 The other positive experiment was on the preparation of November 16, the animal used being an extremely sensitive Pawlow stomach, the splanchnic nerves of which had been sectioned. 536 LUCKHARDT, KEETON, KOCH AND LA MER Thyroid. Consideration of table 3 reveals the fact that thyroid ex- tracts give a rather definite activity by our method. These prepara- tions were made in October under the conditions which had resulted in negative extracts of liver and pancreas. It is further evident that the pepsin does not add to or detract from their activity since in three animals (I, III, VI) the HCl extract was more potent; in two (IV, V) the pepsin and HCl was the more active; in another (VIII) there was no difference; and lastly, dog VII gave results diametrically opposed at various times. The variability in results is due to the state of the stomach at the time of experiment. As every one knows, this is such a variable factor that the assay of activity frequently requires many repetitions. TABLE 3 Activity of thyroid extracts PER CENT ACID IN INCREASE! wWOUR FOLLOWING IN QUAN- ANIMAL PREPARATION TITY INJECTION ice* | __ Free Total cc. HCl 3.5 0.18 0.26 Dog I, Pawlow stomach......... ; HCland pepe, eos 0.12 0.23 : HCl 17.5 0.24 0.29 Dog III, Gastric fistula.......... HCland pepsin 16 0.02 0.15 Dog IV, Gastric fistula, vagi f HCl 6.5 0.29 0.36 EUS eae ey ee tert nies \| HCl and pepsin 20.0 0.36 0.42 : : HCl 14.0 0.38 0.42 Dog te ae fistula, vagl HC] 19.0 0.49 0.52 CTUGHEG Viet. Soe eee FIC] and pepsi 26.4 0.45 0.50 Dog VI, Gastric fistula, vagoto- {| HCl 7.6 | 0.10% | s0ua2 INIZEG eye oe ec eR eon HCl and pepsin |— 0.4 0.18 OF27 : . - : HCl 3.7 0.16 0.24 see a Vagotomized, -£a8ttte)))) Gi and pepsin ||=0) 5111 Sea ee StuUlaieeee eer aekoe reer Ee land senate 79 0.40 0.46 HCl 3.6 0.27 0.34 Doe iVillibs Paw lowes eerie mC rand’ pepsin 3 6 0.20 0.36 * This quantity was determined by subtracting the quantity secreted in 1 hour control period from the first hour experimental period. The (—) sign indicates control output was larger in control than experimental period. DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 537 Negative extracts. Table 4 contains a condensed summary of experi- ments on brain, muscle, spleen, thymus, peptone mixture and gastric juice. The table and the individual protocols show that these prepa- rations do not influence the secretory activity of the stomach. TABLE 4 Negative extracts and preparations NUMBER NUMBER OF EX- iz QUESTION- TISSUE ee orien NEGATIVE AB POSITIVE PT Sct ee ct Oe ts ee 13 8 10 3 0 MIELOr Ue ics . cried oh ae le: 15 8 15 0 0 SU EET ARG Geet 5 Aa ae Oe anne a 10 6 7 2 1 (DVT SS as oe 12 6 12 0 0 Berets DIAC I ULC Og. cara ict ior Ste, o's eso ubim te, 0184) 19 8 19 0 0 Fibrin pepsin HC] digest ........... 4 4 4 0 0 Filtrate from the lead acetate precip- itate of fibrin pepsin HCl digest... 4 4 4 0 0 Gastricjuice. The gastric juice was collected under varying conditions of stimulation. Thus samples were obtained from Pawlow stomachs under food and gastrin stimulation, from gastric fistulae under spon- taneous secretion and from human stomach (Mr. V. reported by Carl- son (5) ) following the chewing of food. The juice (1 ec.) was injected without concentration in some cases (3 experiments) ; in the other cases, the amount injected represented from 4.7 to 11.5 cc. original gastric juice. Animals with gastric fistulae and Pawlow stomachs were used to assay the preparations. Inno case was there a positive response. Fibrin peptone proteose mixture. Popielski (2) has called attention to Witte’s peptone as a source from which his vasodilatins could be pre- pared. It seemed important to investigate this as well as the tissues previously discussed. Instead of starting with a commercial prepara- tion whose history was unknown, a fibrin digest was selected. In this way we could be certain that any activity developing must be due to some of the hydrolytic products and not to substances extraneously introduced. The protocol of this preparation follows. Beef fibrin was washed free from blood in flowing water, pressed in cloth and the net weight (175 grams) was determined. This was next suspended in 7000 ec. of 0.2 per cent sodium hydroxide for six days, strained and filtered. To the opalescent filtrate was added one volume water and sufficient of a 0.5 per cent acetic acid solution to cause a good flaking out of the fibrin. This latter was allowed to settle, washed 538 LUCKHARDT, KEETON, KOCH AND LA MER by decantation with distilled water four times, using 8 liters for each washing. Finally, it was filtered by suction and moist weight (206 grams) determined. This constituted the purified fibrin, the substrate from which the peptone was made. i Of this substrate, 103 grams moist weight, (15 grams dry weight) was mixed with 300 ce. of 0.4 per cent HCl and 25 ce. of a 0.1 per cent solution U.S. P. pepsin (Armour’s) preserved with toluol and incubated at 35° to 40°C. After three days, 5 cc. more of a 1.66 per cent solution of same pepsin was added; the mixture was incubated another 24 hours, filtered, and the filtrate concentrated in vacuo to 50 ec. The filtrate was then precipitated with 300 cc. of redistilled (95 per cent) alcohol, the precipitate removed and the filtrate again concentrated in vacuo to dryness. Alcohol was removed, the residue dissolved in water and diluted to 25 ec. One cubic centimeter of this solution represented about 4 grams by weight of the fresh material and approximately 0.004 gram of scale pepsin. This constituted the test solution of proteose- peptone which was injected into the animals. Reference to table 4 shows that the experiments all gave negative results. This solution was further precipitated by basic lead acetate and after the removal of the lead from the filtrate tested against the same animals. This procedure,was followed because in the course of studies on purification of gastrin from stomach mucosa the active prin- ciple had been found in the filtrate from the basic lead acetate precipi- tate. These experiments also showed no activity. . We may conclude that an acid pepsin digest of purified fibrin gives no products which are capable of causing gastric secretion when injected intramuscularly. DISCUSSION Solubility of gastrin bodies in alcohol. Our experiments demonstrate that the gastrin preparations exhibit rather definite evidence of vasodi- latation as shown by the reddening of the nose and buccal mucosa of the experimental animals. We have not taken records of the blood pressure on intravenous injection. However, we published a tracing (1) showing that the blood pressure was not appreciably reduced when the injection was made into the muscles of the animal and we called attention to the fact that the maximum period of secretion came after the evidences of vasodilatation had subsided. In the data just pre- sented it is clear that absolute alcohol cannot be relied upon to separate DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 539 the two physiological activities, i.e., the vasodilatation observed after intravenous injection and the secretogogue action resulting from intra- muscular injection or, in other words, the vasodilatins of Popielski from the gastrin bodies, if they really be different substances. Formerly we concluded that the duodenal preparations presented a secretory activity definitely less than similar preparations from the stomach. When however we combined the alcohol soluble fraction, previously discarded as containing vasodilatins, the activity of the two tissues was the same. Maydell (6) states that if secretin is treated with alcohol and ether it is possible to separate the two component parts of gastric secretin, one of which evokes secretion and the other dilates vessels. Up to date we have been unable to secure his original report and for this reason cannot discuss the results further than to quote the words of the abstractor. Tomaszewski (7) extracted the residue of his preparation (stomach mucosa extract dried at 90°) with absolute alcohol. Practically all of the active substance was removed from the insoluble residue by this treat- ment, but only about one-seventh of the original activity could be found in the filtrate. A precipitation of proteins by six volumes absolute alcohol on the contrary appeared to increase the activity in the filtrate. This he attributed to a physical change brought about in the mixture, perhaps freeing the body in question more definitely from the proteose- peptone molecules. Again, he precipitated the acid preparation with six volumes of alcohol, the residue was discarded and then a second precipitation was made upon the filtrate after reconcentrating. After three such treatments the activity in the last filtrate was almost com- pletely lost. This experiment only means that the major portion of the active substance must have been carried over in the precipitates and thus disappeared. We conclude that these experiments furnish satisfactory confirmation that gastrin bodies are difficultly soluble in absolute alcohol. Relation of gastrin bodies to the vasodilatins of Popielski. It is our belief that it cannot be determined at present whether the power of stimulating the gastric mucosa and of causing vasodilatation are two properties of the same substance, or whether they are to be attributed to bodies of different chemical structure. However, Thomaszewski’s (7) experiments are of interest at this point. He confirms our former state- ment that intravenous injections of known active extracts gave little or no secretory response, but a maximum of toxic manifestations. Such a secretion as results he believes is purely a mechanical squeezing out of 540 LUCKHARDT, KEETON, KOCH AND LA MER residual juice that may have been present. Thus, in addition to his extracts he found that 20 cc. of 5 per cent Witte’s peptone solution also gave a small secretion when introduced intravenously, but no secretion at all when injected subcutaneously. Therefore he considers that there are two separate mechanisms involved. The one is elicited by intravenous injections and is perhaps due to mechanical factors, blood alterations and vascular changes. This is the secretion due to vasodilatation and the secretion which he claims will be given by any proteose-peptone solution as, for example, Witte’s peptone. The other is elicited on the subcutaneous injection of a body that is a real secretory excitant. Tomaszewski also repeated the experiments of Edkins on cats, but he was able to get only an insignificant increase in acidity of | the sodium chloride solution introduced into the stomach. This result he believes is to be explained as a reaction of the first type and that therefore Edkins has in no sense demonstrated the existence of a gas- tric hormone. Effect of pepsin on the extraction. It is possible that the pepsin and HCl digestion occurring in the process of preparing the stomach ex- tracts may be a factor in the production of the active substance, either liberating it through the more thorough breakdown of the tissues or through the actual hydrolysis of a particular complex. In other words, the cleavage of the molecule by pepsin may be significantly different from the simple HCI digestion, and to this difference the activity of the preparations may be attributed. However, our experience shows that pepsin does not influence the extraction in any constant fashion. Thus it appears to increase the activity in the cases of duodenum and pan- creas, to lessen it in the liver preparations and not to influence it in the case of the thyroid. However, Tomaszewski has published one experi- ment showing that incubation of an active extract with gastric juice for two days reduces its activity to one-third the value of the original. He concludes that pepsin may destroy this class of bodies, but trypsin does not. Distribution of gastrin activity. Our results indicate that the stomach, duodenum and thyroid have the same concentration of active substance per gram of fresh tissue. The liver and pancreas showed activity in one preparation far above any other tissue extracts, but the activity was practically absent in other samples prepared some months later. We are unable to explain this result at present. However, we will have occasion to show later that there are at least two classes of substances which produce secretory DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 541 activity. These we know behave differently in their precipitations. Gastrin is extremely stable in acid media, but the stability of the other class of substances has not been investigated. It is possible, therefore, that in the case of the pancreas and liver the activity may be due to this second class of substances. The readiness with which these two tissues autolyze and the ease with which they may be converted into excellent bacterial media may explain the formation of this second group of active substances. Finally, the liver is an extremely mobile metabolic organ and variations in the nutritional state of the animals might explain the discrepancy. Whatever may be the explanation, our method did not give uniformly active preparations from these tissues. Spleen, thymus, brain, muscle, gastric juice and fibrin peptone-pro- teose digest were uniformly negative. These results differ from Rogers and associates who found the spleen, pancreas and liver active, with thymus showing a smaller concentration of the active substances. It must be recalled that these investigators used an alkaline physiological salt solution for their extraction while we were using 0.4 per cent HCl. This of course does not mean that bodies possessing secretory activity may not occur in these tissues. It simply indicates that our method which has uniformly given active preparations from stomach and duo- denum gives such an active preparation from the thyroid, but not a uniformly active one from liver, pancreas, spleen, thymus, brain, stri- ated muscle, gastric juice and fibrin digest. It is an interesting question whether the active substance is preformed in the tissues and merely liberated by the acid digestion or whether it is a product resulting from the acid hydrolysis of protein. We have some evidence suggesting the latter to be true. Some of our earliest observations not reported up to present, are to the effect that water extracts of well dried gastric mucous membrane freed from fats are not active as secretogogues. The fact that gastrin is present uniformly in some tissue digests and not in others does not rule out the possibility of its being a building stone in certain proteins. Ifit be an extractive then its physiological possibilities assume greater interest. The absence of this substance or these substances from the gastric juice should be em- phasized. We will have occasion to return to this point in a later communication. In view of our other experiments attempting to purify the stomach preparation, it seemed inadvisable to pursue further the question of dis- tribution of gastrin bodies until we knew more of their chemical behavior and nature. 542 LUCKHARDT, KEETON, KOCH AND LA MER SUMMARY 1. Gastrin bodies are soluble in absolute alcohol. Therefore the method of absolute alcohol extraction cannot be used in separating the vasodilatins of Popielski from them. 2. When an extract corresponding to 1 gram of Armour’s scale pep- sin (1/3000) is injected, a slight secretory activity results. Secretion does not occur after an injection of 0.10 gram pepsin, the quantity used in the peptic digestion of the various tissues. 3. The stomach and duodenum contain approximately the same concentration of gastrin bodies. 4, A liver preparation made at one time gave an activity far in excess of the extracts from any other tissues. Similar extracts, prepared later, were practically inactive. 5. The pancreas extracted with pepsin and hydrochloric acid (on one occasion) gave more activity than the stomach and duodenum, but less than the liver. The hydrochloric acid extract at the same time was inactive. Similar preparations on another occasion (two months later) were inactive. The reason for this difference has not been determined. 6. The thyroid preparations presented about the same order of ac- tivity as the stomach and duodenum. 7. The spleen, thymus, brain, muscle, gastric juice and fibrin pep- tone-proteose preparations were uniformly inactive. 8. Pepsin and hydrochloric acid is not a better medium for extraction than hydrochloric acid alone. 9. The existence of two classes of bodies causing gastric secretion is suggested. Whether these bodies are extractives from special tissues or hydrolytic cleavage products has not been determined. The inves- tigation in the distribution of gastrin bodies has been temporarily aban- doned for the more promising studies into the chemical nature of the product derived from the gastric mucosa. (1) (2) DISTRIBUTION AND EXTRACTION OF GASTRIN BODIES 543 BIBLIOGRAPHY KEETON AND Kocu: This Journal, 1915, xxxvii, 481. PorretskI: Pfliiger’s Arch., 1909, exxvi, 483; Ibid., 1909, exviii, 191. PoPIELSKI AND PANEK: Ibid, 1909, exxvii, 222. PopirmusKr: Cen- tralbl. f. Physiol., 1910, xxiv, 635; Ibid., 1910, xxiv, 1102; Centralbl. f. Biochem. u. Biophys., 1910, xi, 724; Pfliiger’s Arch., 1912, exliv, 135; Pfliiger’s Arch., 1913, cl, 1. Fawcett, Rogers, RAHE AND BeEsBe: This Journal, 1915, xxxvii, 453. Rogers, Rave, Fawcerr anp Hackett: Ibid., 1915, xxxix, 345. Rogers, RAHE AND ABLAHADIAN: This Journal, 1919, xlvii, 79. Cartson: This Journal, 1912, xxxi, 151. MaypbELL: Dissertation, Kiev, 1917, cited from Physiol. Abstracts, 1917- 18, ii, 146. ToMASZEWSKI: Pfliiger’s Arch., 1918, clxx, 260; Pfliiger’s Arch., 1918, elxxi, 1. PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE I. ON THE LUCIFERINE AND LUCIFERASE OF CYPRIDINA HILGENDORFII! SAKYO KANDA From the Marine Biological Laboratory, Kyushu Imperial University, Tsuyazaki (Fukuoka), Japan Received for publication November 13, 1919 INTRODUCTORY Raphael Dubois (2) found in 1885 that two substances are concerned in light production of the West Indian cucullo, Pyrophorus noctilucus and also of a luminous mollusc, Pholas dactylus. He called one of these light-producing substances “‘luciférine,”’ which is not destroyed by boiling, and the other ‘‘luciférase,”’ which is destroyed by boiling and was assumed to be an oxidizing enzyme. K. Newton Harvey who investigated the mechanism of light pro- duction in luminous bacteria, fire-flies and others, “believed Dubois’ interpretation” of these experiments to be correct. But after he studied the light production of a Japanese ostracod crustacean, Cypridina hil- gendorfii, he was led “‘to wholly different conclusions regarding the ex- istence of luciférine and luciférase”’ (6, p. 322). He has proposed new views concerning the light-producing substances and has adopted new words, ‘“‘photophelein” and “photogenin” for Dubois’ luciférine and luciférase. The writer has attempted to test Harvey’s conclusions with Cypridina hilgendorfii and has found many experimental facts essentially contra- dictory to those of Harvey and rather in accordance with those of Dubois. MATERIAL AND COLLECTION The material used by the writer for all of the following experiments was the same species of ostracod crustacean, Cypridina hilgendorfii 1 A preliminary report of this work was published in the Japanese Journal of Zoology, 1918, xxx, 409, 445. 2 The writer’s thanks are due to Dr. Naohide Yatsu for identification of the animal. 544 ~ > uum ei A me PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 545 used by Harvey, which is abundant in Tsuyazaki gulf, Japan. It is caught the year round but is most abundant during April and November. The animal in general is strongly negatively heliotropic, so that it is best collected at night by means of a porcelain jar about 30 cm. high and 20 cm. in diameter. The head of a shark (on which the animal will feed) is placed in the jar, which is covered by a piece of cloth with a small hole in the center. Ten or more of such jars may be used in series connected with a long rope. These jars are submerged to the bottom of the sea about 15 feet deep. The animals go into the jars being attracted by the smell of the fish heads, and stay there feeding. This is a very striking case of chemotropism. In this way large quan- tities of the animals are readily collected. Harvey states that the animal is not readily caught on moonlight nights on account of its negative heliotropism. If, however, the method just mentioned above is adopted, no difficulty is encountered even on moonlight nights. The positive chemotropism of the animal is much stronger than its negative heliotropism. Harvey also states that ‘‘an- other non-luminous species (Cypridina x) is often obtained from the fish heads together with C. hilgendorfii. It is positively heliotropic to lamp-light’”’ (6, p. 319). ‘“‘Non-luminous Cypridina x” is, however, not caught at Tsuyazaki gulf, so far as the writer’s experience goes; but some individuals of Cypridina hilgendorfii, and sometimes many of them, are found to be positively heliotropic to strong daylight. THE MAXILLARY GLAND AND LUMINOUS SECRETION OF CYPRIDINA HILGENDORFIL A full account of the maxillary gland and its luminous secretion will be found in the papers of Harvey (6) and Yatsu (8). It is, however, not out of place to call special attention to the fact that Miller, the discoverer of this species, first pointed out the presence of ‘‘two groups of gland cells of different nature” with different secretion products. “Furthermore, he advanced the view that light is produced by the interaction of these two substances” (8, p. 438. Watanabe also states that the maxillary gland secretes a colorless transparent fluid and a yellow homogeneous substance (7, p. 87). Yatsu has especially em- phasized these points in his paper based on his histological studies (8), (though Harvey has entirely overlooked them) ; particularly “the pres- ence of two kinds of gland cells and the absence of a reservoir for the secretion granules common to all the gland cells” (8, p. 488). THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, No. 4 546 SAKYO KANDA Special attention is called to the fact, which has been made clear by the statements referred to above, that the animal produces no light in the maxillary gland cells but the light appears outside of the body of the animal when the light-producing substances secreted by the cells meet in the sea water. In other words, Cypridina hilgendorfii has in no sense any luminous organ or organs as have many other animals, fire- flies,forexample. . Inthe writer’s opinion, therefore, Harvey was entirely misled on this point when he uses such phrases as ‘‘the luminous parts” or “non-luminous parts of Cypridina hilgendorfi,” “‘the luminous or- gans of Cypridina,” “the luminous gland” or “luminous gland cells” and so forth all the way through in his paper (6). This point should be remembered when the distribution of the light-producing substances is discussed later on. THE PRESERVATION OF THE MATERIAL For various reasons, living Cypridinas are not good material to use for experimental work. The writer, therefore, always used dried ani- mals, except for some special purposes. Harvey states that he dried Cypridinas over CaCl, (6, p. 321). But this is an extremely slow process and the light-producing substances are not often in a satisfactory condition, most probably being injured by moisture in the course of the drying time. The writer adopted, there- fore, the following rapid method, which proved satisfactory, as table 2 will show. He made a wire ring to which white cloth was fastened. The animals taken out of the sea water were placed on the cloth. The water on the animals was then removed by absorbing it in blotting- paper as completely as possible. After this they were spread on sheets of dry blotting-paper and exposed to direct sunlight. They were stirred up from time to time in order to dry themevenly. At 30 to 35°C., they could be thoroughly dried in a few hours. If the animals thus dried are placed in a desiccator with CaCl. they may be kept over eight months without impairing their power to produce light when again moistened. Harvey states that removal of the animals ‘‘from sea water also in- hibits the ejection of the secretion” (6, p. 320). But this statement is not quite correct. In the course of drying, the animals may live for thirty minutes or longer even in direct sunlight (the time depending upon the temperature) and continue to eject the secretion which is readily observed by the naked eye, as judged by the color. The light- producing substances, therefore, may be lacking in some animals, so PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 547 that they may not be of much use when they are finally dried. The quicker the drying, therefore, the better the method, beyond doubt. The animals dried with the method just mentioned may also be preserved as a whole in pure ether. Or they may be crushed and sifted with a proper sized sieve in order to separate bodies and shells; and then the bodies, either as they are, or after extracting the fatty substances by several changes of ether during the course of a few days, may be preserved in pure ether as experimental material for as long as eight months or more. Of alcohol, ether and chloroform as preserva- tives, ether is the best owing to its low specific gravity. The material readily sinks into it. This is not the case with chloroform. Moreover, ether is most convenient to work with because of its quick evaporation when the material is taken out of it. THE SEPARATION OF THE MAXILLARY GLAND On account of its minute size it is next to impossible, if not abso- lutely impossible, to dissect out the maxillary gland by itself. At any rate, the writer confesses that he has not succeeded in this attempt. Nevertheless, he believes that the following method is an improve- ment upon that of Harvey (6, p. 325): Under a lens the anterior part of dried crushed Cypridina where the maxillary gland is located, is carefully dissected out with one pointed dissecting knife in each hand. But it-is hardly necessary to mention that other substances besides the light-producing substances are also cut out. The dissected gland cells examined under a low power of the microscope are very dark, though they are yellow while living. The dissected anterior parts, however, cannot be kept very long in con- tact with air, though they may be preserved a few days in a vacuum desiccator without impairing the power of light production. The light-producing substances are undoubtedly destroyed by moisture or oxygen or by both. THE EXISTENCE OF LUCIFERINE AND LUCIFERASE As already stated, the existence of two different substances neces- sary for the production of light by Cypridina was pointed out by Miiller, Watanabe and Yatsu on the basis of morphology. Harvey also demonstrated this fact experimentally in Cypridina, as Dubois did in a beetle and a mollusc. Unfortunately, however, the latter 548 SAKYO KANDA two investigators differ completely in their interpretation, although the results seem essentially similar in nature. At any rate, the fact that the existence of two different substances is concerned in the light production of Cypridina, is readily demonstrated in the following way. About twenty of the dried individuals are placed in 50 cc. of dis- tilled water. Immediately after this treatment, a brilliant ight ensues. After a few hours’ standing, the mixture is filtered through heavy filter-paper and the filtrate, which gives no light by itself any longer,’ is kept for testing. For convenience’s sake, this water extract may be called “A” solution. On the other hand, another set of twenty individuals are put into 50 cc. of boiling distilled water. No lght results in this case. After a few minutes of boiling, the cooled mixture is filtered and the filtrate (which produces no light by itself) is kept for testing. This may be called “B” solution. If now a small amount of ‘A”’ solution is added to a large amount of “B” solution, a brilliant light results at the moment and place of the contact of the two solutions; and the light spreads all over. Its brilliancy does not, however, last very long, but it remains dim for many hours to trained eyes in a dark room. This phenomenon may be explained on the hypothesis that two light-producing substances are secreted respectively by two different kinds of cells in the maxillary gland. At any rate, that the presence of two different substances is concerned in the production of light by the animal is obvious. That is to say, one of these two substances is destroyed by boiling while the other is not. This thermolabile sub- stance which may be temporarily called X substance, therefore, is left in the ‘‘A”’ solution and the thermostable one, which may be called Y substance, is in the ‘“B” solution intact. The mechanism is this: When the mixture of the material and water is allowed to stand, all the Y substance is used up and the X substance is left in the mixture. On the other hand, when the mixture is boiled the X substance is destroyed while the Y substance is left intact, because the former is destroyed by heat before it can use the latter for Hght production. On this view, the Y substance is “the source of the light” and the X ’ Light may be observed by trained eyes even after ten hours, though faint. Harvey might have mistaken it as light produced when this solution was mixed with many chemical substances. 4 The beaker for boiling water should be tall and the flame has to be lowered after the water has reached the boiling point. No material must stick on the wall of the beaker when the material is put in. PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 549 substance is something which causes the former to emit light, as already stated. This conclusion is opposed to that of Harvey (6, p. 323) and is similar theoretically to that of Dubois. According to Harvey, the substance destroyed by boiling ‘“ possesses certain properties . . . . characteristic of enzymes’ (6, p.- 323). But it can not be regarded as an enzyme because it is ‘‘slowly used up” in the reaction (6, p. 324). He, therefore, invented a new word, “photogenin,”’ to indicate the “light producer’ for Dubois’ luciférase. And this, claimed Harvey, is “the source of the light”’ (6, p. 323). In the writer’s opinion, however, it is hasty to conclude that the substance is “used up” when it slowly disappears in the reaction, because it may not be “used up” in the reaction, but its nature may be changed due to the instability of the enzyme itself. It is a well-known fact that a solution of any enzyme loses its activity if kept. “This is, in great part, due to the complex colloidal state of these substances” (1, p. 313). On this view, then, the disappearance of the thermolabile substance furnishes strong evidence of an enzyme, although Harvey declares to the contrary. The writer, therefore, could find no objection in using Dubois’ luciférase to indicate the nature of the action of the thermolabile or X substance found in the gland cells of Cypridina. For this reason he will hereafter use the word ‘‘luciférase”’ to indicate the X substance. As to the Y substance which is not destroyed by boiling, Harvey names it “photophelein’” for Dubois’ luciférine, that is the “light assistor”’ (6, p. 3824). This ‘‘is something which causes the luciférase to emit light” (6, p. 323). As already pointed out, however, the writer regards this substance as the source of the light, contrary to Harvey’s interpretation. The writer therefore defends Dubois’ lucif- érine, as he thinks it adequate to indicate the light-producing nature of the thermostabile substance. As far as the interpretation of the facts and the phraseology of the light-producing substances are concerned, Dubois and the writer are quite in agreement with each other. According to the former, however, the luciférase of Pholas is an oxidizing enzyme while the luciférine is capable of oxidation, with light production, by means of the luciférase. Since the writer has recently found that the phenom- enon of light production of dried crushed Cypridinas is by no means an oxidation,’ he can neither hold the luciférase of Cypridina as an oxidizing enzyme nor the luciférine as a substance capable of giving 5 The experimental evidence will be published in a separate paper. 550 SAKYO KANDA light by oxidation. As Harvey found (6, p. 328), the writer also found that Cypridina luciférine can not be oxidized with light production by such oxidizing agents as neutral H,O., PbOs, BaOs, KMnO, and K,Cr.0;, although Dubois found just the contrary in Pholas luciférin. It is possible, however, that the conditions in Pholas dactylus may be radically different from those found in Cypridina. The writer has, therefore, no intention to make any claim against Dubois’ conclusion. THE DISTRIBUTION OF LUCIFERINE AND LUCIFERASE According to Dubois, Pholas luciférine exists only in the luminous organs of the animal while the luciférase occurs throughout its body and also in many other non-luminous animals. On the contrary, Harvey finds that Cypridina photogenin (Dubois’ luciférase) is only in the luminous organs of the organism, while Cypridina photophelein (Dubois’ luciférine) “is found in many other non-luminous animals and in the non-luminous parts of Cypridina hilgendorfii” (6, p. 328). The writer removed the shells of ten dried Cypridinas and cut them into anterior parts with the maxillary gland cells and posterior parts with no gland cells. Special precaution was taken not to mix one set with the other. The water extract of these posterior parts was mixed with a water extract of dried crushed Cypridinas in which the luciférase was left. This operation was, of course, performed in a dark room. If the luciférine (Harvey’s photophelein) existed in these posterior parts, as Harvey claims, the mixture of these two extracts just mentioned should produce light. But this was not the case. The anterior halves with the maxillary gland, however, gave a brilliant light when moistened. On the other hand, the water extract of these posterior parts was mixed with the boiling water extract of dried crushed Cypridinas in which the luciférine was left. If the luciférase (Harvey’s photogenin) is present in the posterior parts of the animal, the rhixture of these two extracts should produce light. This was not, however, the case. Each of these two series of experiments was re- peated several times with no exception. The writer is forced, there- fore, to conclude that neither luciférine nor luciférase occurs in the posterior part of dried Cypridina. In other words, the luciférine and luciférase of Cypridina are found in the maxillary gland cells of the animal but in no other part. There is a possibility, however, that either luciférine or luciférase may exist in other parts of living Cypridina beside the maxillary gland, PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 551 even though they occur only in the maxillary gland of the dried animal. In other words, either of these substances might be destroyed by drying, although existing in the living. The fact that both the luciférine and luciférase in the maxillary gland cells of the animal are dried without impairment hardly favors this idea. But the writer repeated Harvey’s experiments. One living Cypridina at a time was removed from sea water by means of cloth fixed on wire and was placed on blotting-paper to remove the water adhering to the shell. Of course it was still alive after this treatment. It was then quickly cut by sharp scissors into the anterior and posterior parts. Five of these posterior parts with no maxillary gland cells were mixed with the water extract of these anterior parts with the gland cells or with the water extract of dried crushed Cypridinas. If the luciférine (Harvey’s photophelein) were present in the posterior part of the animal, as Harvey claims, the mixture of these two should give light, as the luciférase (Harvey’s photogenin) is left in the water extract of these anterior parts or dried Cypridinas allowed to stand. The results were very irregular. Sometimes no light was produced and sometimes one or two of the posterior parts showed dimly lighted points, though they lasted for only a short time. On the other hand, five posterior parts of living Cypridinas were mixed with the boiling water extract of the anterior parts of the living or of the dried crushed Cypridinas. If the luciférase exists in the posterior part of the animal, the mixture of these two should give light, since the luciférine was left in the boiling water extract. The results were again irregular just as those of the other series described above. If, therefore, one judged from the positive results of these two series of experiments, 1t might be said that both luciférine and luciférase occur in the posterior part of the animal where no maxillary gland cells are located. On the other hand, if judged from the negative results it is quite logical to say that neither luciférine nor luciférase is found in the posterior part of the animal. The latter alternative seems the truth, because the writer believes that the results ascribed to be “‘posi- tive” are nothing but false results derived from the rough method of preparing the material for experiment. The difficulty of manipulating the living Cypridina should be taken into serious’ consideration. Harvey’s contention expressed in such statements as, ‘By a careful quick scissors cut, the head end of Cypridina containing the luminous gland can be separated from the posterior half without any contamina- 552 SAKYO KANDA tion of the latter with luminous secretion,” (6, p. 325), is not beyond criticism. As already stated, Cypridinas give off the luminous secretion, if they are removed from sea water and are touched with any object, scissors, for instance. It is, therefore, quite possible that the secretion may adhere to the posterior part of the animal. If so, any “ positive’’ results would be deceptive. Unless this objection in preparation of the experimental material is completely wiped out, any positive results are of no value. Even Harvey’s statements, if subjected to a careful examination, suggest the possibility of false results, thus: ‘‘We must try the experiment immediately because this substance disappears if the extract stands in presence of oxygen. In absence of oxygen or if the extract is boiled immediately (but not too long a time) the sub- stance does not completely disappear even after one hour. There is, therefore, in the non-luminous parts, the substance photophelein which disappears even in the absence of photogenin (from luminous gland) unless the solution be boiled or oxygen excluded”’ (6, p. 325). This quick disappearance of photophelein in the absence of photo- genin seems to the writer to mean an adherence of a very small amount of the former to the posterior part of the animal. Furthermore the same “‘positive’’ results could sometimes be obtained even in the mixture of the posterior parts with no maxillary gland cells and distilled water. That is to say, the posterior parts of the animal carefully prepared sometimes produced light when mixed with water. Not only that, but a careful examination in a dark room revealed the fact that some of the posterior parts of living Cypridinas dissected out in the same manner as above were already producing light before they were brought in contact with water or any solution— the luciférase solution, for example. These facts undoubtedly prove the adherence of luminous secretion to the posterior parts of the animals, most probably to the shells while handled in cutting. With the same point in mind, the shells of dried Cypridinas care- fully removed for this special purpose were tested with distilled water. Numerous bright small points of light appeared. This fact of course could not be explained unless the light-producing substances secreted are assumed to have adhered to the shells while the animals were handled and dried in the sun. If Cypridina luciférine occurs in any other parts of the animal where the maxillary gland cells are not found, as Harvey claims, the existence PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE pene’ of two different kinds of gland cells, which is histologically proved, becomes totally meaningless. On the contrary, their existence con- stitutes very strong evidence of the separate and specific nature of their functions in the production of luminous secretion. Harvey also claims that he found the photophelein (Dubois’ lucif- érine) in many other non-luminous animals. He has given a list of many animals tested and he adds that “of these forms Lepas and Chiton gave the best light and of these two only Lepas gave light with dilute Cypridina photogenin”’ (6, p. 326). The writer, therefore, made cold and boiling water extracts of Lepas and Chiton and tested them with extracts of Cypridina luciférine and luciférase. In both cases the results were absolutely negative. From the variety of evidence, both experimental and _ histological, considered above, the writer is forced to conclude that Cypridina luciférine and luciférase occur only in the maxillary gland cells of the organism but not in any other part or in other non-luminous animals. In other words Cypridina luciférine and luciférase are specific in the strict sense of the word. In comparing the results regarding the distribution of light-pro- ducing substances in Pholas, Cypridina and other non-luminous animals obtained by Dubois, Harvey and the writer, the following table will assist in visualizing the contradictory points of each investi- gator’s claim discussed above. TABLE 1 Distribution of the luciférine and luciférase in luminous and non-luminous animals SPECIES OF DUBOIS HARVEY KANDA ANIMAL Luciférine Luciférase Photophelein| Photogenin | Luciférine Luciférase Non-luminous animals — — aa — _ _ Luminousani-| In | Out] In | Out| In | Out] In | Out| In | Out} In | Out mals gland|gland/gland/gland|gland/gland|gland|gland/gland|gland|gland/gland Pyrophorus noctilucus and pholas- dactylus = — + + Cypridina hil- gendorfii sil eae bla eh || A wel | Bee dle 554 SAKYO KANDA THE EFFECT OF PRESERVATIVES ON LUCIFERINE AND LUCIFERASE As already stated, the writer tested the effect of pure ether, alcohol and chloroform as preservatives of dried Cypridinas. They were pre- served in small bottles. Dried whole Cypridinas and dried crushed ones were separately preserved in each substance. From time to time they were tested with distilled water to see whether they would pro- duce light or not. ‘The strength of light produced by them was com- pared with that of freshly collected and dried Cypridinas as control. The following table will show the results tested after eight months. Of course the preserved animals with the luciférine and luciférase never emit light in these narcotics. The reason is that the luciférine and luciférase are insoluble in the chemicals. They are therefore of a non-lipoid nature. Solubility in water and light production go in union. TABLE 2 Effect of preservatives on the luciférine and luciférase of Cypridina hilgendorfi ETHER ALCOHOL CHLOROFORM Zz an S s a so 7) me) o a TIME & oa Co) > oO nS a F 2 z= 2 a 2 a Be 26 ) me i) EB S B eB & es o a is St be = rz q Ay = oO = (e) = 0 Hightanonthse: yecees. 4) Oe d 6 5 A slaies 2 1 * Figures indicate degree of lumination. The figures in the table indicate the degrees of brightness compared with that of control. The figure “1,” for example, means the lowest of all, although it is quite bright. It will be noticed that the bright- ness of the light produced by dried whole Cypridinas preserved in a desiccator with CaCl, for eight months is not very different from that of freshly dried crushed ones used for control. The question arises: if the luciférine and luciférase are insoluble in the preservatives mentioned above, what causes the gradual loss of light-producing power? This is hard to answer. But a few possible explanations may be mentioned. In the first place the preservatives used may not be pure, because they were not purified or redistilled for this purpose. As is well known, the so-called absolute alcohol on the market is about 99.5 per cent at best. Even a little water con- tained in alcohol might be a cause of slow destruction of the light-pro- ducing substances. The same may be said in the case of ether. In PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 555 the case of chloroform the material did not sink for a long time due to the former’s high specific gravity and perhaps the light-producing substances might be impaired by the moisture or oxygen of the air. The writer is therefore convinced that if extra-pure chemicals are used as preservatives and if the Cypridinas are thoroughly dried, the material may be kept quite long. At any rate the writer concludes that the reason why Cypridina luciférine and luciférase never gave light when mixed with ether, alcohol or chloroform, is their insolubility in these chemicals. As far as ether is concerned, this conclusion accords with that of Harvey. He says: ‘Dried crushed Cypridinas may be extracted with six changes of ether during the course of two days without impair- ing in the least their power to produce light when again moistened. The luminous substance is therefore of a non-lipoid, ether-insoluble nature, as might be expected from the fact that it is extruded from the animal as a clear water-soluble, non-fluorescent secretion” (6, p. 321). He therefore distinctly recognizes the fact of ether-insolubility of the luminous substances, and also the fact of indestructibility of the sub- stances in ether. In other words, he recognizes the fact that no light is produced when both the luminous substances, luciférine and lucif- érase, are present together in ether. Strangely enough, however, the same investigator states “It (Cypridina luciférase) will also give light if mixed with many pure substances as chloroform, ether, benzol, thymol, saponin, oleic acid, atropin, NaCl and others. Since most of the above substances could not possibly be oxidized by the luciférase, I conclude that they cause in some way the giving out of light in what Dubois terms luciférase”’ (6, p. 323). Judging from these statements Harvey seems to identify the action of many pure chemical substances so diverse in nature with that of luciférine. If so, the writer cannot quite understand why he uses the special word “‘photophelein” (or Dubois’ luciférine) to indicate one of two light-producing substances which acts in the same way toward the luciférase (Harvey’s photogenin) as chloroform, ether, benzol, thymol, saponin, atropin, pilocarpin, hydrochinon, pyrocatechin, chloral hydrate, butyl alcohol, oleic acid, ortol, aesculin, dextrine, NaCl, MgSO., (NHz) eSOz or KiFe(CN)6., or is omnipresent ‘through- out the body of Cypridina hilgendorfii” and in “many non-luminous animals.’”? Of course Harvey states, “It is hardly worth inquiring into the nature of the substances in each particular extract which may for convenience be collectively spoken of as photophelein, since 556 SAKYO KANDA I have found a great many simple bodies which, mixed with concen- trated photogenin in powder or crystal form, give rise to a bright light”’ (6, p. 327). If so, the substance ‘ photophelein,”’ which Harvey thinks he has found in the posterior part of Cypridina hilgendorfii and in many other marine animals, might be NaCl or MgSO; which are present in sea water in a large amount. The writer tested Cypridina luciférase solution mixed with pure ether, alcohol, chloroform, NaCl and MgSO: in various ways to see whether it gives light or not. The results obtained were absolutely negative. Because, as already stated, there is no possibility of light production in the mixture of the luciférase solution and pure ether, aleohol and chloroform, since no light is produced when the luciférine and luciférase are present together in these chemicals. For this reason Harvey’s conclusion that the ‘“luciférase is the source of the light and the luciférine. . . . . is something which causes the luciférase to emit light’’ is not tenable at all. Although his statements do not make clear how he conducted his tests, Harvey gives two tables in which ‘the effect of saturation of solutions (one Cypridina to 25 cc.) of photophelein and photogenin with the four substances,” that is, chloroform, ether, benzol and thymol, is summarized (6, p. 331). In the first place, however, we must con- sider from purely physico-chemical viewpoints the properties of these four substances. That is to say, to what extent these substances could be “saturated” with the luciférine and luciférase solutions. Ether is soluble to the extent of 1 part in 12 of water, chloroform to the extent of 0.712 part in 100 parts of water at 17.4°C., benzene to the extent of 0.082 part in 100 parts of water at 22°C. and thymol “is only very sparingly soluble in water.’’ It seemed to the writer to be extremely doubtful that any one could get real results which are to be considered as the effect of saturation of solutions of the luciférine and luciférase with these four chemicals, though Harvey claims that he did. The writer found that dried crushed Cypridinas produced a brilliant light when they were placed in the saturated water-solutions of ether, chloroform and benzene. No difference of brightness between the light of the animals in the solutions and that of the control in distilled water was perceptible. In other words, the solubility of the luminous substances was not affected by the small amount of ether, chloroform and benzene contained in the saturated solutions. Since neither luciférine nor luciférase when they are present together is destroyed PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 557 even by pure ether, chloroform and benzene, as has already been shown (except: benzene), it is no wonder that such a small amount of ether, chloroform and benzene as is in the saturated water-solutions of these substances is not detrimental to either luciférine or luciférase. Harvey claims, however, ‘‘ether is especially destructive to the photophelein”’ (6, p. 331). It was not ether, in the writer’s opinion, that was espe- cially destructive to the luciférine in Harvey’s experiments but it was water. As repeatedly pointed out, even a mere trace of water is destructive to both the luciférine and luciférase. On the other hand, the writer also tested ether and benzene which had dissolved water to the full extent, that is, to the extent of 2.25 ce. in 100 parts of ether and to the extent of 0.211 cc. in 100 parts of ben- zene. The results were all negative, that is to say, dried crushed Cypridinas produce no light in ether and benzene thus treated. It may be worth mentioning that the writer found that living. Cypridinas give light when put in pure ether, aleohol and chloroform This puzzle is readily explained by the following consideration: The living Cypridinas carry sea water inside their shells and secrete the luminous substances when they are placed in the chemical substances: and the secretion meets the sea water carried by themselves. The appearance of light is the result. Neither alcohol nor chloroform plays any role in this process. PROTEIN TESTS ON LUCIFERINE AND LUCIFERASE Harvey states that ‘“‘despite the fact that the light from the natural secretion of Cypridina is very bright, a sample of the secretion, col- lected by shaking many Cypridinas in a small volume of sea water and filtering, responds to none of the common biochemical tests. i that both Cypridina luciférine and luciférase solutions are positive for proteins. . . . .” (6,p.322). The writer, however, found that both Cypridina luciférine and luciférase solutions are positive to the biuret, xanthoproteic and other tests. The experiments are in progress and will be given in a separate paper. THE EFFECT OF TEMPERATURE ON LUCIFERINE AND LUCIFERASE Harvey found that “Cypridinas dried over CaCh, ground, and the powder suspended in sea water give a beautiful light which disappears when heated to 56°, but returns on cooling. If heated to 65° and cooled, the light also returns, but does not return if heated to 70° and 558 SAKYO KANDA then cooled”’ (6, p. 329). He also found that “the luminous mate- rial of Cypridina. .. . . . is unaffected by cold and will glow brilliantly at 0°C.’’ The writer tested the effect of temperature on light production of the animals in a different way from that of Harvey and found practically the same results as the latter. The method will be stated in brief. The writer heated 40 cc. of distilled water to various temperatures and put twenty dried crushed Cypridinas in it for about thirty seconds (or sometimes for one minute). The water with the animals was then poured into 300 ce. of distilled water which was at about 25°C. In this way the effect of different temperatures upon the light-producing substances was studied, as shown in the accompanying table (table 3). TABLE 3 Effect of temperature on the luciférine and luciférase of Cypridina hilgendorfiu HARVEY KANDA TEMPERATURE Heated Cooled Heated Cooled 0 - + SUMMARY AND CONCLUSIONS 1. The animal used for these experiments was Cypridina hilgendorfii. 2. The animal is generally negatively heliotropic, but some individ- uals of the species are positive to light. 3. The living animal is not good material for experimental work. It is better to use the animal thoroughly dried in sunlight. 4. The dried animal may be preserved in a desiccator with CaCl, for over eight months apparently without impairing in the least its power to produce light when again moistened. It may also be pre- served with some deterioration in pure ether, alcohol or chloroform for over seven months. 5. Ether is the best preservative. el ~ PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 509 6. The maxillary gland cells of the dried animal are dark red. The red substances produce light when dissolved in water. 7. Water is essential for the production of light. 8. Two substances are concerned in the production of light: the one is destroyed by boiling and the other not. As these two substances show practically the same reaction of luciférine-luciférase as those in Pholas dactylus studied by Dubois, the writer uses Dubois’ words, “luciférine”’ for the thermostable substance and “luciférase” for the thermolabile substance. 9. Harvey’s theory of photophelein-photogenin does not fit the author’s experimental results. 10. The luciférine and luciférase of Cypridina hilgenforfii are found only in the maxillary gland cells of the animal. Neither luciférine nor luciférase is found in the non-luminous animals. 11. The luciférine does not produce light when mixed with H2O2, KMn0Ox,, PbO, and so forth. 12. The dried animal does not produce light in pure ether, alcohol and chloroform. The luciférine and luciférase are not soluble in these chemical substances. They are therefore of non-lipoid nature. 13. The living animal may produce light in pure ether, alcohol and chloroform. These chemical substances, however, play no réle in such production of light. It is due to sea water adhering to the animal. 14. The dried crushed animal gives light in a saturated water solu- tion of ether. This is due to the water, not to the ether. 15. The luciférine and luciférase solutions give the color tests for proteins. 16. The effect of temperature is given in table 3. The writer concludes that Harvey’s new theory and new words are not tenable. The so-called photogenin (Dubois’ luciférase) is not the source of the light but may be an enzyme. The photophelein (Dubois’ luciférine) is the source of the light. This work was supported by the private contribution of Mr. Jihachi Hamano. 560 SAKYO KANDA BIBLIOGRAPHY (1) Bayuiss: Principles of general physiology,London, 1915. (2) Dusots: Compt. rend. Soc. Biol., 1885, xxxvii, 559. (3) Dusots: Ann. Univ. Lyon, 1892, ii, 167. (4) Harvey: This Journal, 1915, xxxvii, 230. (5) Harvey: This Journal, 1916, xli, 449. (6) Harvey: This Journal, 1917, xlii, 318. (7) WATANABE: Dobutsugaku Zasshi, 1896, ix, 86. (8) Yatsu: Journ. Morphol., 1917, xxix, 2. Sn a ako Gee PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE II. Tue Propuction or Ligut By CyPRIDINA HILGENDORFII IS NOT AN OXIDATION! SAKYO KANDA From the Marine Biological Laboratory, Kyushu Imperial University, Tsuyazaki (Fukuoka), Japan Received for publication November 13, 1919 INTRODUCTORY That the production of light and oxidation are intimately related to each other is a very general phenomenon judged from common sense and considered from the viewpoints of physics and chemistry. It is, therefore, no wonder that we find in the literature on the subject of the relation of oxygen to light production in living organisms, the statement that the production of light by bacteria, Noctiluca, Pholus, fire-flies and others, is an oxidation (4, p. 353-358). Harvey (2, p. 321) who has studied the action of oxygen on the pro- duction of ight by Cypridina hilgendorfii, concludes: Oxygen is necessary for light production as may be seen by placing the crushed animals in an hydrogen atmosphere, or by bubbling hydrogen through a glowing extract of the animals. The light never completely disappears even after a long time, but remains dim so that very little oxygen (as no special precautions were taken to remove the last traces of oxygen from the hydrogen, prepared in a Kipp generator) is sufficient to give light. Upon readmitting oxygen, however, a bril- liant glow results. Every other species of animal investigated likewise requires oxygen for phosphorescence. The writer also imagined that ‘‘oxygen and water are necessary for the ~ production of light by Cypridina hilgendorfii”’ and fancied that luciférase “is an oxidizing enzyme” (3, pp. 321 and 448). But the results of his recent work, which was carried out under more careful experimental conditions, have contradicted his expectation. That is to say, he has found that the light production in Cypridina is by no means a phe- 1 This paper was published in the Japanese Journal of Zodélogy, xxxi. 561 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 4 562 SAKYO KANDA nomenon of oxidation. Oxygen is not necessary for the production of light by this animal as will be shown in this paper. The following experiments were conducted in the Science Depart- ment of the Kyushu Imperial University, Fukuoka, Japan. The writer expresses his appreciation of the interest and suggestions of Dr. Tsuneya Marusawa, Professor of Physical Chemistry in the University, through- out the course of the work. To Prof. Ayao Kuwaki and all the mem- bers of the department, the writer acknowledges his gratitude for the privileges of the laboratory and their interest in the work. The work was supported by the private contribution of Mr. Jihachi Hamano. MATERIAL Cypridina hilgendorfii were thoroughly dried in the direct sunlight. The animals were then crushed and the shell and body separated by means of a sieve. The body material containing the maxillary glands was extracted with several changes of ether during the course of a few days “‘without impairing in the least their power to produce light when again moistened.’ For convenience’s sake, the bodies thus pre- pared are hereafter called the ‘experimental material’? or simply “material.” DESCRIPTION OF APPARATUS Believing in Harvey’s statements, the writer first tried “by bubbling hydrogen through a glowing extract” or a glowing mixture of distilled water and Cypridinas to see whether the light disappears or not. Hy- drogen gas was prepared in a Kipp generator. The method was, how- ever, so crude that he could not obtain decisive results. At the sugges- tion and also under the direction of Doctor Marusawa, therefore, the writer conducted all his experiments with the following apparatus. The arrangement of the apparatus as shown in figure 1 is typical al- though it was variously modified for the different gases employed. The apparatus consists of two wings, right and left. Each wing has an experiment bottle, H or H,, of a capacity about 60 c¢e. The bottle is fitted with a tight rubber stopper in which three glass tubes, A, B and C, or Ay, B, and Ci, with one stop-cock for each, are inserted with the arm bent at 90 degrees. In the case of the bottle H, the glass tube B with a stop-cock 6 is connected to a T-shaped glass tube, D, with a stop-cock, d, by means of rubber tube. The second arm of the glass tube D is connected to the flask W by means of glass and rubber tubes, PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE re =) Fe 563 564. SAKYO KANDA and the third leads to the Kipp generator, K, through the gas wash bottle. The glass tube C, which has a stop-cock c, and is T-shaped, is connected to the flask W by means of glass and rubber tubes, and a third arm has a stop-cock g by means of which water from the flask W is caused to flow. The glass tube A with a stop-cock a is connected to a T-shaped glass tube J, the second arm of which leads to a Gaede oil pump, P, through a manometer, M, and a safety bottle and the third to the glass tube A, of the experiment bottle EZ; at the left wing of the apparatus. The arrangement of the bottle 2, is exactly similar to that of the bottle HZ, though sometimes a gas holder is used, depending on the gas under investigation. So the Kipp’s generator is not in perma- nent use. The apparatus was placed in a dark room, to facilitate the observation of the production of light. THE METHOD OF EXPERIMENT As the chief purpose of the following experiments was to determine whether the production of light by Cypridina hilgendorfil is an oxida- tion or not, a hydrogen experiment was always conducted together with any other gas experiment as a control, besides a control for which air was used. Special care was, of course, exercised when an oxygen ex- periment was performed. The principle of the method is that the pro- duction of light by the material can be observed in the water free from any gases or in any pure gas. In the first place, therefore, the distilled water to be used should be thoroughly heated by boiling for a few hours. While the water is still boiling the flask W or W, is fitted with a tight rubber stopper which carries two glass tubes, long / and short s, tightly fitted. Each of these two tubes has a stop-cock. Steam comes out from the outer ends of these two tubes within a few minutes, since the water is still boiling. These tubes are then closed by the stop-cocks e and f. The glass tube s or s,, as the case may be, is now ready to be connected to one arm of the glass tube D or D, by means of glass and rubber tubes. One of the most important procedures in this method is to have any desired gas for experiment prepared in the Kipp’s generator or the gas- holder connected with the gas wash bottles and with the glass tube D or D, before the glass tube s or s, of the flask W or W, is to be connected to one arm of the glass tube D or D,;. In other words, any desired gas should be ready to pass through the gas washers and the glass tube D OF ODDITIES PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 565 or D; before the glass tube s or s; of the flask W or W; is connected to one arm of the latter, and the stop-cock of the glass tube s or s; is opened. Because the water and also the space in the flask W or W, should be filled up with the desired gas, while the water becomes cooler and cooler. Special care should, of course, be taken that any air in any glass and rubber tubes should be excluded by all possible means, whenever any connection is made. Before the glass tube s or s, of the flask W or W, is, therefore, to be connected to the tube D or D,, the washers should be evacuated and then filled with the gas to be used up to the stop-cock h or hy. All being thus prepared, the tube s or s; is connected to the tube D or D, without any unnecessary delay, while the water is still boiling and then the stop-cocks e or ¢;, h or hy and d or di, inorder, areopened. Hot steam and gas come out from the end of the arm, because the water is still boiling and the gas is slowly generating. All air is thus driven out from all spaces and the stop-cock d or d, is closed. At the same time the flask W or W, is removed from the flame. Everything thus ar- ranged, the water and space in the flask W or W, are gradually filled up by the gas until the temperature of the water becomes equal to that of the room. When the temperature equilibrium of the inside and out- side of the flask W or.W, has been established, the glass tube | or |; is pushed down into the water, and is now ready for its connection to one arm of the glass tube C or C, of the experiment bottle # or EF). Four bottles, two for the experiment and two for the control, of simi- lar capacity and shape, are cleaned and dried thoroughly. In each of these bottles 0.2 gram of the experimental material is placed. The ex- periment bottles are then fitted with tight rubber stoppers each of which carries three glass tubes as previously described. One arm of the glass tube C or C; which is T-shaped is now connected to the glass tube / or 1; of the flask W or W;. And then the gas-saturated water is introduced up to the stop-cock c or ¢, by the management of one arm of the tube C or C;. The bottle H or Eis fixed on aniron stand. Now the connections of the tube B or B; through the tube A or A; to the T-shaped tube J and the third arm of the last to the pump, P, through the manometer, M, and also a safety bottle, are made by means of heavy-walled rubber tubes. All arrangements thus made, the appa- ratus is now ready for experiment.’ In the first place, the stop-cocks a; and 6; are opened and at the same time the pump is started to evacuate all air in the bottle H, and the 2 It is convenient to begin any experiment from the left wing first. 566 SAKYO KANDA rest of the apparatus. A complete evacuation of the air, however, is impossible by one operation. In order to complete the evacuation of the air, therefore, it is necessary to fill up all the spaces again with the gas desired, which has been prepared for the experiment. The stop- cock d; is now opened, and the gas fills all the spaces. After this filling is complete, the stop-cock d; is closed and the pump is again started. This same procedure is repeated ten times, though five times are suf- ficient. After the last filling with the gas, the stop-cocks a; and 6; are closed. The last procedure is simply to pour a necessary amount of water from the flask W, into the bottle HZ; for an observation of the pro- duction of light by the material in the bottle. But this should not be done until the bottle E has been prepared in the same way. Todo this the pump Pis again started to evacuate the gas in all tubes and spaces from the pump and up to the points of the stop-cocks a and a. After this evacuation, the stop-cocks a, b and d are opened. In so doing, another gas is introduced to fill the bottle E and the rest of the apparatus. After this filling the stop-cock d is closed and the pump is again started. After this evacuation, the stop-cock d is opened and fresh gas is permitted to enter. This procedure is repeated ten times as before. After the gas has filled bottle H and the rest of the apparatus, all the stop-cocks, a, b and d, are closed. The bottle # is now also ready to receive water. Now the stop-cock ¢ is opened and about 20 cc. of the water are poured from the flask W into the bottle H. At the same time, about 20 cc. of the ordinary distilled water are also poured in one of the control bottles which is fitted with a tight stopper. The time should he re- corded because it is a very important factor in the production of light. Then the stop-cock c; is opened and about 20 ce. of the water are poured from the flask W,in the bottle H,. At thesame time the second control is prepared with water. These procedures take about 2 to 3 minutes. Then the bottles # and FE, with the tubes, A, B and C, and A, B; and C; are freed from all the connections of the rest of the apparatus. At the same time the room is to be darkened. The produetion of light by the material in the bottles E and EF, is thus to be observed together. As already stated, special care should be taken that the controls are to be set up at the same time when the water is poured in the bottles # and F), since time is a very important factor for the observation of the intensity of the light produced by the material. At the time of pouring water, therefore, at least three persons are to be ready for the work, to save time. Sy Ee PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 567 The writer is convinced that the method and apparatus described above are satisfactory to determine the effect of any particular gas on - the material. THE PREPARATION AND USE OF GASES The gases used for these experiments were hydrogen, oxygen, nitro- gen, carbon dioxide and carbon monoxide. Of these gases, hydrogen, nitrogen, carbon dioxide and carbon monoxide were prepared in the laboratory, and oxygen in bomb was used. » The methods of preparing these four gases were those in common use. The writer thinks, how- ever, that brief statements about the methods of preparing them may not be out of place, because all the procedures carried out by the writer need to be open for free discussion. 1. The preparation of hydrogen: Hydrogen was prepared with a Kipp’s apparatus in which zine and about 50 per cent H_SO, were placed. The liberated gas was washed by passing it through four wash bottles. Distilled water was placed in the first, and saturated KMNO, solu- tion in the second, a solution of 30 grams of KOH + 10 grams of C.H;(OH); in 100 ce. of distilled waterin the third, and concentrated H;SO, in the last. As hydrogen was always used as a control in order to compare the action of this gas with that of any other gas, this Kipp’s apparatus was always fixed in the right wing. The last wash bottle of this apparatus, therefore, was always connected to the tube, D. The solutions in the wash bottles were renewed from time to time. 2. The preparation of carbon dioxide: Carbon dioxode was also pre- pared with a Kipp’s generator in which pieces of marble and about 15 per cent HCl were placed. The gas when liberated was washed by passing it through three wash bottles. Distilled water was placed in the first and second and concentrated H.SO, in the third. This last wash bottle was connected to the tube, D;, when used for experiment. 3. The preparation of nitrogen: In the first place, 200 grams of NaNOz, 300 grams of (NH4).SO, and 200 grams of KeCrO, were barely dissolved in separate beakers. The solutions were placed together in a 5-liter flask and 1500 cc. of distilled water were added to it. A con- densor was connected to the stopper of the flask in order to cool the nitrogen gas. To this condenser, a heavy Erlenmeyer flask as a safety bottle and two wash bottles were connected. In each of these bottles 30 ec. of concentrated H2SO., 20 grams of K,Cr.0; and 100 ce. of dis- tilled water were placed. 568 SAKYO KANDA The mixture contained in the flask was slowly heated on a low flame. When enough Ns» gas was liberated, the last wash bottle was connected to a gas holder which was already prepared to receive thegas. Special care was taken to have no air in any space. When the gas in the gas . holder was used for experiment, it was again washed by passing through the same wash bottles as mentioned above. 4. The preparation of carbon monoxide: In a liter distillation flask, 500 ce. of 80 per cent H2SO, were placed. In the rubber stopper of this flask, one long thermometer of 200°C. and one glass tube perforated on its sealed end of about 5 ecm. with many small holes were inserted. The mercury part of the thermometer and the hole part of the tube were entirely dipped into the sulphuric acid of the flask. The outer end of the tube was bent about 90 degrees and a suitable rubber tube with a pinch-cock was connected to it. The other end of the rubber tube was then connected to a separating funnel in which some formic acid was placed. To the distillation flask, a heavy Erlenmeyer flask as a safety bottle, and two wash bottles, in each of which 200 cc. of 20 per cent NaOH were placed, were connected. The sulphuric acid of the flask was slowly heated on a low flame to about 110°C., and the formic acid was added little by little. After enough gas was liberated, the gas was received in a gas holder. When the carbon monoxide gas was used for experiment, it was again washed by passing through three gas wash bottles. Two hundred cubic centimeters of 20 per cent NaOH were placed in the first and second bottles, and 200 cc. of concentrated H2SOx in the third, which was connected to the tube, D,. EXPERIMENTAL With the methods and apparatus described in the previous section, the writer conducted his experiments with every possible care and re- peated each series of experiments four or five times, even though no exception was found in any trial. The results of these experiments are summarized in table 1. The figures of the table show a relative superiority and inferiority of the intensity’ of light produced by the material on which five independ- 3 The observation of the intensity of light was made at the moment when the two experimental and two control bottles held together in both hands were given an equal shaking. A question may arise about the intensity of light, because the writer has not determined it by a quantitative method. According to the calcu- PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 569 ent gases and air were allowed to act. The figure “6,” for example, means the highest degree of light intensity compared with all others. Asrepeatedly stated, the experiments on the action of hydrogen and of air were always made as controls of any other gases. But as the ex- periments on the other gases, carbon dioxide and carbon monoxide, for instance, were not carried out together, it was impossible to compare their light intensities’ at the same time. The time for observing the light intensity in each gas, however, was carefully recorded in com- parison with that of light produced in hydrogen and in air. In this roundabout way, therefore, the action of each gas may be compared. Furthermore, it was enough if the actions of hydrogen and oxygen were TABLE 1 Comparative intensity of light produced by the material in various gas atmospheres and water LIGHT INTENSITY OF THE MATERIAL OBSERVED AT A GIVEN TIME IN THE ATMOSPHERES AND WATER OF THE FOLLOWING TIME OBSERVED IN DARK ROOM Be OPE uo Ne co CO: Air Os 1m. 6 6 6 6 6 6 3-4 m. 6 6 6 5 ie 4 5-8 m. 6 6 5 5 3 3 12-15 m. 6 5 4 4 3 2 20 m. 6 5 4 4 2 1 9-10 h 5 4 3 3 1 0? 13h 4 3 2 3? 1 160 h 1 1 0 0 ity 200 h 0? 0 0 accurately compared, since the essential problem of these experiments was to determine whether the production of light by the animal was an oxidation or not. There was an unmistakable difference of the light intensity between the actions of oxygenand air. The difference of the intensity of light produced in the hydrogen and oxygen atmospheres, as well as in the air, was so astonishingly marked that no one could lation of V. Henri et Larguier des Bancels, however, the retina is very “‘sensitive to an amount of light energy as small as 5 times 10-“ ergs. This is about three thousand times as sensitive as the most rapid photographic plate” (1, p. 512). If so, although there is no quantitative means to decide the weak or strong intensity of light observed by the retina, there is no indicator superior to the qualitative judgment of the retina. It is believed, therefore, that the decision made by the retina is most accurate. 570 SAKYO KANDA question it. And the action of the nitrogen gas in comparison with the action of carbon dioxide or of carbon monoxide was also very distinct. Superiority or inferiority between the actions of CO and CO, may be questioned, though the writer felt that the former was a little superior to the latter. It will be worth while to describe some other facts than those shown in table 1. In the water saturated with air and other gases, the color of light produced by the material is bluish white, a few hours after the treatment, while in the water saturated with hydrogen, a, it is decidedly blue. This blue color in the latter lasts almost as long as the light con- tinues. In the cases of air and other gases the light after shaking takes about three or four seconds to return to the state before shaking, while b, in the case of hydrogen it takes about eight seconds. In a perfectly dark room it is observed that the strong light after shaking lasts quite long. As the time ratio of durability, 1: 2, however, is not altered, no deviation is involved in this observation. Generally speaking, the “steady homogeneous glow” of light is to be observed, as Harvey pointed out. But c, the light in the hydrogen-saturated water is ob- served to be heterogeneous by the flowing and precipitating of ex- tremely minute particles when shaken in a dark room. Harvey’s claim of ‘complete proof of the truly soluble character of the light-producing substances” (2, p. 321) may not be true. The writer will try to make this point clear by using an ultramicroscope in the near future. In the resting state, d, the light in the case of hydrogen glows as if the solu- tion glows itself but the glow of the individual points of the material is not so marked as it is in the cases of air and any other gases, while in the latter cases the solution is clear. And e,in the latter cases also the solution forms when shaken but not so much as in the former: In the cases of all gases, when the production of light is near to the end, the solution does not glow as a whole but only at the surface when shaken. This f, is markedly so in the case of hydrogen. But in the case of air, the solution glows as a whole till the end. These are six characteristics of the effect of the hy degen gas on the production of light by the material besides those shown in table 1. Whether these are of any significance regarding the light production or are simply to be overlooked as meaningless, can not be settled unless further facts are found. After an experiment was over, a test was made with a lighted match to see whether the gas used was present or not. The results were al- ways positive even after seventy-five days had passed. Each solution PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 571 was also tested with litmus paper to see whether it was alkaline or acid. In the cases of hydrogen, nitrogen, oxygen and air, the solution was found to be neutral, although the solution was slightly alkaline after about two hundred hours in the last case. .In the case of carbon dioxide, the solution was always acid and was distinctly milky with white precipitation. In the case of carbon monoxide, the solution was also acid, though very faint; and the acidity seemed to increase a little after about two hundred hours. In the cases of oxygen, air and carbon monoxide, the solution and material became very black, while in the ease of hydrogen and nitrogen the solution was clear and the material became reddish brown. CONCLUSION AND DISCUSSION That the less the quantity of oxygen contained in the experiment bottle the more intense is the light produced by the material after the first minute is quite obvious according to the results stated above. The light produced is more intense in the water saturated with airthan in the water saturated with oxygen and is markedly more intense and durable in the water saturated with hydrogen or nitrogen than in the water saturated with oxygen or air. In other words, oxygen is not necessary for the production of light by the material. If so, there is no ground for the assumption that the production of light by the animal is an oxidation. If the production of light by the animal is due to an oxidation, as Harvey claims, the more intense light should be produced by the greater concentration of oxygen. This is not the case, and the writer therefore concludes that the production of light by the animal is not an oxidation. A question arises whether the production of light by the material is a reduction or simply an hydrolysis. In expectation of getting some light on this question, the writer made experiments with carbon monox- ide as a reducing substance. But the results were not so marked as expected. That is to say, the results were not as good as those in the case of hydrogen. If so, is this not a reduction? This question is not settled as yet. Carbon monoxide is a “poison gas.’”’ It may, therefore, act as a poison just as it acts on the hemoglobin of the blood. At any rate, it may be necessary to take into such biological consideration some factors besides the action of carbon monoxide purely looked upon from the viewpoint of chemistry. If so, it is no wonder that the action of carbon monoxide on the production of light by the material is inferior 572 SAKYO KANDA to that of hydrogen or nitrogen. Whatever the exact interpretation of the facts may be, no decisive conclusion can be made unless further facts are found. As to the action of carbon dioxide, there is the same question as in the case of carbon monoxide. If the light production of the material is not an oxidation, the intensity of light produced should be the same in carbon dioxide as in hydrogen or nitrogen. But this is not the case. However, if the fact is considered that carbon dioxide is dissolved in water and carbonic acid is formed, the riddle may be readily solved. Because it is a well-known fact that acid is injurious to organisms while alkali is not. Furthermore, there are other points in table 1 which require expla- nation. In the first place, the light-producing substances in the water saturated with oxygen disappear fastest in spite of the production of the poorest light. On the other hand, the production of light in the water saturated with hydrogen is most intense and most enduring. Such phenomena may be explained by the following assumptions. - That the higher the oxygen tension is the faster the oxidation of any sub- stance is obvious. Although the production of light in question is by no means an oxidation, the light-producing substances, especially the luciférine, may always be subjected to an oxidation and thus the dis- appearance of the substance is apparent. If it be true that the higher the oxygen tension is the more the light-producing substance or sub- stances may be oxidized (though this oxidation of the substances has no bearing on the light production), it is no wonder that the light be- comes weaker and disappears faster if the substance or substances be- come less and less by an oxidation. Such a consideration explains the facts that the light is weakest and lasts shortest in the water saturated with oxygen and on the other hand, that the light is strongest and lasts longest in the water saturated with hydrogen with no oxygen. If so, the action of the water saturated with carbon monoxide, car- bon dioxide and nitrogen might be the same as the action of the water saturated with hydrogen. But this is not the case. As to the action of the first two gases there may be possibly some physiological or bio- logical factors involved, as already stated. So the mere fact that free oxygen isnot present should not be looked upon as a complete explana- tion. Taking this as granted, then, how is it in the case of nitrogen? The difference in action of hydrogen and nitrogen is not very marked, but it is still easily distinguishable. It may be said, therefore, that the superiority of the action of hydrogen over any other gas is its own speci- PHYSICO-CHEMICAL STUDIES ON BIOLUMINESCENCE 573 ficity. In brief, the writer should confess that he has not yet clear knowledge with which he can explain these various difficulties. He expects to make special efforts to gather all possible data based on experiments. These difficulties may be of such a nature that they may be explained when further facts are available. SUMMARY 1. The intensity of light produced by the material is strongest and lasts longest in the water saturated with hydrogen. 2. The intensity of light produced by the material is weakest and lasts shortest in the water saturated with oxygen. Therefore the pro- duction of light by the material is not an oxidation. BIBLIOGRAPHY (1) Bayuiss: Principles of general physiology, London, 1915. .(2) Harvey: This Journal, 1917, xli, 318. (3) Kanpa: Dobutsugaku Zasshi (Journ. Zodl.), 1918, xxx, 409, 445. (4) Mancoup: Winterstein’s Handb. Verl. Physiol., 1910, ii, pt. 2, 225. THE PREPARATION OF ADENINE NUCLEOTIDE BY HYDROLYSIS OF YEAST NUCLEIC ACID WITH AMMONIA! WALTER JONES anp A. F. ABT From the Laboratory of Physiological Chemistry, Johns Hopkins Medical School Received for publication December 8, 1919 When yeast nucleic acid is heated in an autoclave with ammonia for several hours at 145°-148°, it completely loses its phosphoric acid so that its four nucleotides are formed (1), (2). But when yeast nucleic acid is similarly treated with ammonia at a lower temperature (105°-120°) it does not lose any of its phosphoric * acid but decomposes into nucleotides (38). After hydrolysis at the lower temperature, if one will treat the product alkaline as it is, with an equal volume of alcohol, a very useful separation occurs. The guanine complex is precipitated while the adenine com- plex remains in solution (3), so that the material can be separated sharply into two fractions. This is the analytical key to the separa- tion of adenine nucleotide from guanine nucleotide and without it the results discussed below could not have been obtained. It is based upon the fact that alkaline salts of adenine nucleotide are quite soluble in 50 per cent alcohol while the alkaline salts of guanine nucleotide are practically insoluble (4). From the adenine fraction Jones and Germann (3) prepared the free nucleotide by means of its lead salt and found that it contained a uracil group in addition to the adenine group. From the fact that precisely half of its phosphoric acid is ‘‘easily split’ and half is “‘firmly bound” Jones and Read (5) concluded that the substance contains its adenine and uracil groups in equivalent quantities, and is therefore adenine- uracil di-nucleotide. This conclusion appeared to be much strengthened by the following experiment which Jones and Read did not publish. An attempt was made to obtain uracil nucleotide after destroying the 1While this article was in press, the contribution of Levene appeared. Journ. Biol. Chem., 1919, xl, 415. 574 ADENINE NUCLEOTIDE yael adenine nucleotide with boiling mineral acid. One hundred grams of the supposed di-nucleotide were used for this purpose but 98 grams were destroyed leaving only 2 grams of undecomposed nucleotide material. This was found to have all the properties of adenine-uracil di-nucleotide including the formation of a crystalline brucine salt having the required chemical composition. Thus it appeared that the adenine- and uracil-nucleotides were not only present in equal quanti- ties, but were destroyed in equal quantities. Jones and Read (6) prepared from the di-nucleotide a crystalline brucine salt which had sharply the composition required for the brucine salt of adenine-uracil di-nucleotide.’ Levene (7) prepared this nucleotide by the method of Jones and Germann and converted it into its brucine salt as described by Jones and Read. He states that the brucine salt which he obtained possessed the analytical values required for adenine-uracil di-nucleotide, but found that these values are disturbed when the substance is repeatedly recrystallized from 35 per cent alcohol. After nine successive crystal- lizations, the surviving brucine salt had a chemical composition very close to that required for the brucine salt of uracil nucleotide. From this brucine salt a crystalline barium salt was prepared and found identical with the barium salt of uracil nucleotide that Levene (8) had already prepared from another source. Levene’s results appeared very convincing to us in so far as they went to show that the supposed adenine-uracil di-nucleotide is in reality a mechanical mixture of its component mono-nucleotides, and his results also suggested to us the possibility of preparing adenine nucleotide from the mixture; for Jones and Kennedy (9) had prepared adenine nucleotide and had found its properties such as should make its isolation from almost any mixture a comparatively easy matter. By the appli- cation of their method to this problem, we have isolated pure crystalline adenine nucleotide identical with the substance described by Jones and Kennedy. Thirty-eight grams of nucleotide were dissolved in 152 ce. of hot water and treated with a solution of 95 grams of brucine in the smallest possible amount of alcohol. On cooling, the material stiffened to a paste of crystals which was filtered on a Buchner funnel and crystallized from 35 per cent alcohol. Sixty grams of this brucine salt were treated 2 This can be accurately determined because adenine nucleotide and uracil nucleotide differ from each other markedly in composition. The one contains five atoms of nitrogen; the other, only two. 576 WALTER JONES AND A. F. ABT with 6 liters of cold 35 per cent alcohol and after digestion with occa- sional agitation for 24 hours, the alcoholic extract was filtered from the undissolved residue on a Buchner funnel and allowed to evaporate for several days in open dishes at the room temperature. As the alcohol escaped, a crystalline brucine salt was deposited which was collected, allowed to dry in the air and used for the preparation of adenine nucle- otide as described below (25 grams). The undissolved residue from the alcoholic extraction was again extracted with 600 ce. of cold 35 per cent alcohol and the dissolved brucine salt was recovered after evaporation at the room temperature. This had very nearly the composition of the brucine salt of uracil nucleotide. 0.6136 required 12.04 ec. of standard H2SO, (1 ce. = 0.003642 N). REQUIRED FOR FOUND Adenine nucleotide |} Uracil nucleotide Nee rapte ee ee ee ood oe 10.00 6.79 7.15 Thus the original extraction with 6 liters of cold alcohol had almost completely removed the brucine salt of adenine nucleotide, but of course the extract contained also a considerable amount of the brucine salt of uracil nucleotide. Twenty-five grams of the brucine salt mixture (from the 6 L of 35 per cent alcohol) were suspended in 800 cc. of boiling water and treated with a little ammonia which takes it promptly into solution. As the solution cooled, it was made and kept faintly but distinctly alkaline with ammonia. In this way the brucine is thrown down in easily filterable crystalline needles. The material was kept overnight in the ice chest, filtered from the brucine with a pump and shaken in a sepa- rating funnel with several successive portions of chloroform to remove the last traces of brucine. The solution was then evaporated® at 45° under diminished pressure to one-third of its volume and treated in the warm with a slight excess of lead acetate. The heavy- granular lead compound was filtered on a Buchner and washed by grinding with hot water in a porcelain dish. Finally the material was made into a smooth suspension with warm water, decomposed in the warm with sulphuretted * The neutralization of the ammonia with acetic acid as done by Jones and Kennedy should be omitted since the lead salt of adenine nucleotide is quite soluble in ammonium acetate. ADENINE NUCLEOTIDE 5w7 hydrogen and the solution obtained after filtering off the lead sulphide was aspirated to remove the excess of sulphuretted hydrogen. On standing over night this solution deposited snow white adenine nucle- otide in characteristic needles. The mother liquor from these crystals, upon evaporation at 45° under diminished pressure gave a further yield of the same crystalline material. Altogether 3.12 grams of adenine nucleotide were obtained. In its chemical composition, crystalline form and solubilities, in its characteristic ability to form supersaturated solutions, in the curious behavior of its water of crystallization and in all other respects the substance is identical with the adenine nucleotide of Jones and Kennedy. HOW O=P-O- C;sHsO3 ; CsHiNs HO’ I. 0.3108 required 16.2 cc. of standard H2SO,. (1 ce. = 0.003642 N). II. 0.2374 required 12.32 cc. of standard H2S0Ox.. III. 0.4615 after complete burning gave 0.3105 Mg NH,PO;.6H.0. IV. 0.4335 after complete burning gave 0.2892 Mg NH,PO0,.6H.0. V. 0.3416 heated 2 hours with 5 per cent H.SO, gave 0.2292 MgNH,4PO..6H20 and 0.3428 adenine picrate. . VI. 0.3001 heated 2 hours with 5 per cent sulfuric acid gave 0.1982 MgNH, PO,;.6H.O and 0.2994 adenine picrate. VII. 0.5349 lost 0.0262 at 127°. VIII. 0.3162 lost 0.0149 at 127°. FOUND REQUIRED FOR Cio His NsPO7H20 II III IV V VI VII | VIII INURO GEN: c.-.< epee ee' 19.18 |19.00)19.02 Total phosphorus... 8.49 8.51| 8.44 Partial phosphorus. 8.49 8.48] 8.36 INGEMIIC sce seks 28 37.00 37.20/36 .92 Wiaitenerirsenc:. 2.50% is 4.93 4.89) 4.71 The mother liquor from adenine nucleotide of course contained uracil nucle- otide. It was allowed to evaporate at the room temperature to about one-fifth of its volume, filtered from a trace, heated and neutralized with brucine in sub- stance. On cooling the deposited crystalline brucine salt was filtered off and recrystallized from 35 per cent alcohol (6.46 grams) ; 0.6841 required 13.12 ce. of standard acid (1 ce=0.003642N). THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 4 578 WALTER JONES AND A. F. ABT NITROGEN REQUIRED FOR BRUCINE SALT OF FOUND Adenine nucleotide Uracil nucleotide Cio Hus Ns (C23 Hos N20) 27 H2O Cs His N2PO9(C23 H2s N204)27 H2O 9.76 6.79 6.99 It seems a little curious that crystalline insoluble adenine nucleotide never appears until one has prepared a brucine salt of this material. The identical procedure described in this paper for preparing the sub- stance from a mixture of brucine salts had been already executed in preparing the original nucleotide mixture, yet this nucleotide mixture could be evaporated to a syrup without depositing any adenine nucle- otide. It might be supposed that an impurity in the nucleotide mixture which prevents the crystallization of adenine nucleotide, is gotten rid of in the mother liquor from the brucine salts. But we were able to prepare crystalline adenine nucleotide from this mother liquor just as from the crystallized brucine salts. Perhaps it is necessary to start with a mixture of the nucleotides, in which adenine nucleotide is in considerable excess. BIBLIOGRAPHY (1) Levene anv Jacoss: Ber. d. deutsch. chem. Gesellsch., 1909, xlii, 2703. (2) Levene anp LaForce: Ber. d. deutsche. chem. Gesellsch., 1912, xlv, 608. (3) Jones aND GrRMANN: Journ. Biol. Chem., 1916, xxv, 93. (4) Jonzes anp Ricuarps: Journ. Biol. Chem., 1915, xx, 25. (5) Jones anpD Reap: Journ. Biol. Chem., 1917, xxix, 123. (6) Jones aND Reap: Journ. Biol. Chem., 1917, xxix, 111. 7) Levene: Journ. Biol. Chem., 1918, xxxiii, 425. (8) Levene: Journ. Biol. Chem., 1918, xxxiii, 229. (9) Jones aND Kennepy: Journ. Pharm. Exper. Therap., 1918, xii, 253; 1919, xiii, 45. CHANGES IN THE HYDROGEN ION CONCENTRATION OF THE URINE, AS RESULT OF WORK AND HEAT G. A. TALBERT From the Physiological Laboratory of the School of Hygiene and Public Health, Johns Hopkins University ° Received for publication December 1, 1919 INTRODUCTION In a previous article (1) I reported on changes in the hydrogen ion concentration of the sweat caused by muscular exercise and heat. At the same time that I was making these determinations I was collecting samples of urine from the subjects in order to ascertain what changes, if any, might be effected in its reaction under precisely like conditions. Study of the sweat meant working in.a quite unexplored field, and this, in part at least, might be said of the experiments that I have to report on the urine, for so far as I have been able to ascertain the effects of heat on the hydrogen ion concentration of the urine have not here- tofore appeared in the literature. As to the effects of muscular exercise upon changes in the acidity of the urine, several observers have pub- lished their results from time to time. But the fact that the methods that were employed in obtaining the total acidity of urine and simi- larly constituted fluids are now known to be obsolete, and furthermore that the results reported were somewhat conflicting, suggested a reéxam- ination of the subject. The principal observers who have investigated this phase of the problem are Hoffman (2), Ringstedt (3), Oddi and Tarulli (4), Vozarik (6) and Aducco (5). All except the last named have maintained that muscular exercise increased the acidity of the urine; Aducco on the other hand has claimed a decrease in the acidity. It is well to state at this juncture that in my experiments the subjects were not required to take extremely fatiguing exercise. The work for the time was hard and the heat endured was fairly intense. The subjects remained with me about one hour, which included the time of stripping and dressing. 579 580 G. A. TALBERT Exactly the same methods were employed as were used with the sweat, and like precautions were here observed. It seems only neces- sary to repeat that the hydrogen ion concentrations were obtained by Henderson and Palmer’s (7) colorimetric method controlled by occa- sional electro-metric measurements (Clarke’s (8) technique). EXPERIMENTS As in the case of sweat, my first observations were upon subjects from the Baltimore Central Y. M. C. A. The exercise taken was basket ball, volley ball and the usual work of a gymnasium, which was generally followed by running, these exercises lasting from one-half to one and one-half hours. It soon became apparent that volunteer subjects in a large city gymnasium were not altogether dependable; consequently it was found much more desirable to have paid subjects come to the laboratory where their part of the work would come under closer scrutiny. As these experiments were performed at least two hours, and usually five hours, after eating, and as the intervals between taking of the samples were so short, it did not seem necessary to lay much stress on the diet, particularly as it was the relative rather than the absolute values which were most desired. I took note on all morning experiments when I discovered that the urine had not been voided since breakfast, especially when alkali- producing fruits had been eaten. The main purpose I had in view was to ascertain whether the effects of such diets might be observed in the absolute values. Some of the high alkaline findings may be due to the retention of the morning urine after eating the above type of fruits. On the whole, in my eighty-six observations of normal urines I found the fluctuating values which have been emphasized by many others. When the experiments were taken up in the laboratory, a stationary bicycle was employed for the muscular exercise and the sweat-cabinet , for the heat. In this connection, however, I might say that heat was not introduced into the experiments with the thought of studying its effect on the urine, but only for the purpose of obtaining heat sweat. It was accidentally and subsequently that I conceived the idea of studying the reaction of the urine under the influence of heat. Conse- quently, in my first experiments I took only a control sample of the urine and another after the heat sweat had been produced and the muscular exercise had been completed. But after performing three HYDROGEN ION CONCENTRATION OF URINE 581 experiments of this character I decided to test the urine as well as the sweat secreted after heat. So thereafter four samples of urine were obtained at about fifteen-minute intervals. The first sample was taken just before the subject entered the sweat-cabinet, the second just as he came out and was ready to exercise on the bicycle, the third imme- CONTROL AFTER HEAT AFTER WORK AFTER DRESSING 1 2 3 4 Fig. 1. A chart representing four typical examples of the changes in the hydro- gen ion concentrations in urine where work followed immediately after the sub- jection to heat. Ordinate represents the hydrogen ions in tenths of pH values. Abscissa, the four periods in which urine was taken in 15-minute intervals. Under no. 1, control; under no. 2, after heat; under no. 3, after work; under no. 4, after dressing. 582 G. A. TALBERT diately after exercise, and the fourth after he had dressed. Inffthis connection it might be stated that the person was subjected to a heat of about 30°C. which rose rapidly to 40° or 45°C. There were in all eleven experiments in which four samples were taken in the above order. The most remarkable facts shown by these tests were that: a, there were different concentrations of the hydrogen ions for each of the four periods; 6, it seemed impossible to establish anything like a fixed relationship between the periods and the H-ion CONTROL AFTER WORK AFTER HEAT AFTER DRESSING 1 2 3 4 Fig. 2. A chart representing four typical examples of the changes in the hydrogen ion concentrations in urine where heat followed immediately after muscular exercise. Ordinate represents the hydrogen ions in tenth of pH values. Abscissa, the four periods in which urine was taken in 15-minute intervals. Under no. 1, control; under no. 2, after work; under no. 3, after heat; under no. 4, after dressing. concentration. In most cases the samples obtained in the second, third or fourth period were more acid than was the first sample, which was the control. On the other hand it was most difficult to predict whether the second, third or fourth sample would give the greatest HYDROGEN ION CONCENTRATION OF URINE 583 hydrogen ion concentration. For the sake of clearness I have plotted four typical examples in figure 1, the ordinates representing pH values and the abscissa representing the four samples of urine collected at about fifteen minute intervals. In the next series of experiments my subjects took the exercise first and then were subjected to heat, which meant that the samples taken at the second and third periods were now reversed. Here again I found just as variable results. In this order ten experiments were performed in which I have plotted four typical cases as shown in figure 2. These two sets of experiments brought results, as far as urine was concerned, that were difficult to analyze, but yielded some of my most important data on sweat. They seemed to obscure more or less the effects of either work or heat alone upon the urine reaction. Thereafter the effects of muscular exercise and heat were tested separately. The former will be discussed first. Three samples of urine were now taken: First, just before the exercise, as a control; second, immediately afterwards, and third, after dressing. The results, given in table 1, are listed under subjects G., M.and X. The first two named represent single individuals, while X. represents the report on several individuals. The interval of time between the taking of samples was about twenty minutes. It is also to be noted that in fourteen experiments with subject G. a greater hydrogen ion concentration wasfound in the second sample than in the first or control. Two of them show a slight loss while one is unaffected. In comparing the third sample with the control I dis- covered that twelve out of the fourteen samples had a greater acidity, one sample exhibiting a loss, the other remaining the same. It should also be pointed out that there is only one instance in which the second sample shows the smallest amount of acidity. In examining the seven experiments performed with subject M., it will be observed that in every instance the urine tested immediately after exercise showed a greater acidity than the control. In the third period four samples are more acid than in the second period, while three show a loss. One of these three is less acid than the control. In the averages of all the pH values from G., a fairly conspicuous increase in acidity is evident immediately after the exercise, persisting even to the third period, but not so perceptible as the- total pH values of 6.04, 5.76 and 5.61, respectively, indicate. 584 G. A. TALBERT In examining the total averages in the case of M., one observes a difference of 0.3 pH between the first and second periods, with a dif- ference only of 0.03 between the second and third. Under *X. are sixteen experiments in which there were taken only a control and a sample immediately after, corresponding to the second TABLE 1 = URINE o URINE scorer | ons, | vue loo Seal wcasecr | rmam, | wmne lo Sines pH pH pH pH pH pH G. 5.65 5.6 5. M. 7.0 6.7 5.8 G. 5.5 5.55 5.55 M. 7.4 7.35 7.6 G. 5.9 5.4 5.3 M. 7.5 7.0 7.5 G. 5.9 5.1 5.8 M. 6.9 6.85 6.8 G. 7.0 6.2 5.8 M. 8.5 7.9 ru G. 5.6 5.5 5.45 M. 6.8 6.4 6.3 G. 5.9 5.7 5.7 M. 6.7 6.1 6.4 G. 6.0 6.1 6.0 G. 8.7 ree! 7.0 || Av. pH 7.2 6.9 6.87 G. 5.8 5.7 5.6 G. 5.75 5.75 5.35 x. 6.0 5.3 G. 5.8 5.5 5.3 X. 5.2 5.1 G. 5.7 5.6 5.0 x. 5.6 5.1 G. 5.4 5.3 5.2 X. 7.4 6.0 ws Ke, 7.4 6.9 Av. pH...| 6.04 5.76 5.61 ie 6.2 5.9 xt 5.7 5.3 xi 8.0 4.7 Ke Wed 5.3 x. ae? 6.5 x 5.9 5.7 Da 6.1 6.15 aK 6.0 5.7 x 8.0 5.6 x 5.55 5.5 De 6.2 6.0 Av. pH..| 6.48 5.67 period in the other experiments. Of these sixteen determinations, fifteen showed a decided increase in acidity as a result of work, while one shows a decrease of 0.05 pH. The averages here show 6.48 pH for control against 5.67 pH immediately after the exercise. These figures are very significant, and they appear all the more so when one notes the summary of thirty-seven observations as shown in table 2. HYDROGEN ION CONCENTRATION OF URINE 585 In the same manner the effects of heat were studied in order to show how heat may modify the changes of the hydrogen ion of the urine. In table 3 G., M. and X. signify the same subjects as before. In the case of subject G., in fifteen out of eighteen tests the second sample yielded an increase of acidity over the control, in test 2 it showed a slight loss, and in one it remained unaffected. In five instances the second sample was more acid than the third, while in eight cases the third sample was of greater acidity than the second; and in five the acidity was unchanged. An examination of the results from subject M. show relatively about the same changes. As regards the miscellaneous subjects under X., comparable results were not obtained on account of the fact that the tendency to an increased acidity is delayed to the last period. TABLE 2 Averages of pH values as affected by work OBSERVATIONS SUBJECTS CONTROL WORK AFTER RH heh Pek? Tete Ae OO ee ba tee ae 14 G 6.04 5.76 5.61 a M C22 6.9 6.87 16 x 6.48 5.67 37 Averages....... 6.57 6.11 6.24 An examination, however, of all the individual tests coupled with the general averages as shown respectively in table 3 and table 4 forces one to the conclusion that heat has a tendency to increase the elimination of acids through the urine. Out of thirty-one observations my results show twenty-nine with an increased acid elimination manifested either in the second or the third sample. The two exceptions are found under X. These results, then, are quite as striking as those obtained from the effects of exercise and, like the latter, are more significant when one considers the short duration of the experiments and the fact that the subjects were not exposed to excessive heat. The exact changes that muscular exercise may bring about in the urine by causing an increase of total acidity constitute a question that only further research can settle. It was once thought that, at least as far as normal urines were concerned, the variations of acidity might be explained by the relative proportions of the mono- and di-acid phosphates. | 586 G. A. TALBERT TABLE 3 = “URINE = URINE sunsecr | conron | near [22 MiNUTes|| supecr | Coxon | Hear [20 MINUTES hen jf Set ieobin lo Gibco iE ater G. 5.8 5.9 5.6 M 6.4 5.8 6.0 G. SG 5.45 5uo M 59 15) 5.05 5.05 G. 5.6 Dao 5.6 M 5.95 5.9 5.8 G. Deo 5) (5) 5.85 M 6.05 D225 AS G. 75) dl 5.8 M 7.4 7.4 Coil G. 6.8 6.0 6.0 M 8.0 7.6 U4 G. 5.9 5.8 5.85 M (A 7/ 6.1 6.4 G. 6.2 6.0 5.6 (er 7.6 7 A 7 3 Ay. pls. 6.52 6.16 6.14 G. 6.0 5.9 5.9 G. fa) E5535) 5.45 G. 5.6 5.5 5.4 X. af aes, Soul G 5.7 5.6 5.65 X. 6.2 6.6 6.1 yeu 5.4 52 5.2 Xx. (ay 7.0 a2 ‘eh 5.8 BiG ENG Xe 6.2 Ket 5.6 eG 5.75 5.75 5.35 X. 71 8.7 8.2 G. 5.8 5.4 5.3 DG 5.8 6.0 6.0 : add : 5s “ 20 g Av. pH.. 6.42 6.5 6.03 Aya pHi 5.9 Ee 7al 5.66 TABLE 4 Averages of pH values as affected by heat OBSERVATIONS SUBJECTS CONTROL ¥ HEAT AFTER ana en Ce ee er ae 18 G 5.9 yal 5.66 a M 6.52 6.16 6.14 6 xX 6.42 6.5 6.03 31 Averages....... 6.28 6.12 5.94 Some years ago Folin (9) and also Henderson (10) called attention to the inadequacy of such an explanation. Folin states: The current attractive, and in a measure plausible, belief that the acidity of urine is regulated by variations in the relative proportion of the two forms of ‘acid phosphates’ is therefore erroneous. If urine does at no time contain com- paratively strong acids in the free form, the reason is in part the variability of the ammonia formation and in part the presence of salts of organic acid. HYDROGEN ION CONCENTRATION OF URINE ' 587 Henderson states: If phosphorie acid be the chief substance which is varying in amount of base associated with it as the reaction of the urines varies, it is none the less true that other substances are varying too. However, only two, so far as we now know, will at the ordinary physiological ranges show considerable change in combined base. These substances are carbonic acid and uric acid, both of which occur in such small amount as to be of minor importance. In the examination of Secaffidi’s (11) work, which bears directly on the problem, one is led to believe that while the uric acid is normally quite negligible it may become something of a factor as a result of muscular fatigue. While he says that during the exercise there is no increase in uric acid, at the same time he states: Immediatement aprés l’interuption du travail musculaire, il se manifeste une augmentation plus ou moins considérable, mais en tout cas trés rapidé et trés sensible, dans l’élimination de l’acide urique. Les bases puriniques, dans cette periode diminuent parfois. L’augmentation dans l’élimination de l’acide urique est beaucoup plus sensible quand la fatigue musculaire, et specialement le repos consécutif, ont lieu & des altitudes qui ne sont pas trés grandes. Kenneway (12) has obtained the same results, with this difference: the increase following exercise is only of importance when unaccus- tomed muscles are brought into use. He claims that after frequent use of these muscles the elimination of uric acid becomes less in amount. Furthermore, he holds that the increased excretion of uric acid imme- diately following work is not due to the ‘‘sweeping out of that which is already present and would otherwise be retained, but that exercise heightened the activity processes by which uric acid is produced,’ and maintains that the decrease of elimination during work is due to defective oxidation of purine compounds. Hill and Flack (13) have stated that during severe muscular exercise “the use of oxygen and production of CO», are so rapid in the muscles that the circulation cannot keep pace with the demand.” Ryffel (14) claimed that the presence of lactic acid in the urine during exercise might be regarded as indicating the inadequacy of the supply of oxygen. My data are not in harmony with Kenneway’s that the continued — use of the same set of muscles decreases the elimination of uric acid and that the increase is found only as a result of the use of an unaccustomed set. In my experiments the exercise was done almost entirely on the stationary bicycle, and no depreciation in the acidity resulting from con- stant use of the same set of muscles. 588 G. A. TALBERT As to the causes of increased acidity of the urine as a result of heat, we are even more in the dark. Finally, it should be mentioned that my results conflict with the statement of Vozarik (15) that changes in external temperature, as warm baths, decrease the acidity of the urine. SUMMARY 1. Intense exercise from fifteen to twenty minutes’ duration increases the hydrogen ion concentration of the urine. 2. A man subjected to the heat of a sweat-cabinet for fifteen or twenty minutes will, in a large majority of cases, void a urine of higher hydrogen ion concentration. 3. If muscular exercise follows immediately upon subjection to heat, or vice versa, it is impossible to interpret the results. BIBLIOGRAPHY (1) Tatsert: This Journal, 1919, 1, 433. (2) Horrman: Maly’s Jahresber., 1884, xiv, 214. (3) Rinestept: Maly’s Jahresber., 1890, xx. (4) Oppr AND TaruLur: Maly’s Jahresber., 1894, xxiv. (5) Apucco: Maly’s Jahresber., 1887, xvii. (6) Vozartik: Pfliiger’s Arch., exi, 497. (7) HeNDERSON AND Patmer: Journ. Biol. Chem., 1912-13, xiii, 393. (8) CuarxK: Journ. Biol. Chem., 1915, xxiii, 478." (9) Foun: This Journal, 1905, xiii, 104. (10) Henperson: Journ. Biol. Chem., 1911, ix, 417. (11) Scarripr: Arch. Ital. d. Biol., liv, 359. (12) Kenneway: Journ. Physiol., 1909, xxxviii, 1. (13) Hitz anp Fuack: Brit. Med. Journ., Oct. 17, 1908. (14) RyrreL: Comp. Physiol. Soc., Oct. 17, 1908. (15) Vozarik: Loe. cit. INDEX TO VOLUME L BT, A. F. See Jones and Ast, 574. Acidosis during starvation, 1. ——, vital staining and, 20. Adenine nucleotide, 574. Adrenal secretion, isolated heart as indicator of, 399. Adrenalin, effect of, on catalase pro- duction, 165. Alkaline reserve, plasma volume and, in shock, 104. ALVAREZ, W.C. and E. StaRKWEATHER. XVIII. Conduction in the small intestine, 252. Alveolar air and respiratory volume at low oxygen tensions, 280. Anaphylaxis and peptone poisoning, nitrogen of blood in, 357. Anesthesia, catalases of blood during, 54. —. effect of, on certain metabolites, 82. Asapa, H. tion, 1. Vital staining and acidosis, 20. Asphyxial blood, hyperglycemia-pro- voking ability of, 177. Acidosis during starva- BECKER, C. E. See Rermann and Becker, 54. Beure, E. H. and O. Rippie. The effect of quinine on the nitrogen con- tent of the egg albumen of ring- doves, 364. Bevier, G. and A. E. SHevxy. Urea excretion after suprarenalectomy, 191. Bioluminescence, studies on, 544, 561. Blood catalases during anesthesia, 54. —— changes from short exposures to low oxygen, 216. physico-chemical Blood diastase, regulation of, 208. —., nitrogen of, in anaphylaxis and peptone poisoning, 357. —— pressure, effect of training and practice on, 443. —— ——., low, treatment of, in shock, 86. —— volume in lung edema, effect of oxygen on, 157. —— in secondary traumatic shock, 3 Burag, W. E. The effect of adrenalin, desiccated thyroid and certain in- organic salts on catalase production, 165. CANNON, W. B. Studies on the conditions of activity in endocrine glands. V. The isolated heart as an indicator of adrenal secretion in- duced by pain, asphyxia and excite- ment, 399. Cardiac and respiratory centers, reac- tion of, to low oxygen, 327. Cardio-vascular reactions in Valsalva experiment, 481. Caruson, A. J. See LuckHarpt, PHIL- Lips and Carson, 57. Catalase, influence of organic sub- stances on, 520. —— production, 165. Catalases of blood during anesthesia, 54. Chelonians, neuromuscular respiratory mechanism in, 511. Circulatory responses to low oxygen tensions, 228. Compensatory reactions to low oxygen, 302. Conduction in small intestine, 252. 589 590 Coomss, H. C. Some aspects of the neuromuscular respiratory mechan- ism in chelonians, 511. Crystalloids in treatment of shock, 119. Cypridina hilgendorfii, luciférine and luciférase of, 544. —— ——,, production of light by, is not an oxidation, 561. PAWSON, P.M. Effect of physical training and practice on the pulse rate and blood pressures during ac- tivity and during rest, with a note on certain acute infections and on the distress resulting from exercise, 443. —and P. C. Hopcers. Cardiovas- cular reaction in the Valsalva exper- iment and in lifting with a note on parturition, 481. Diastase, blood, regulation of, 208. DEMA, lung, effect of oxygen on blood volume in, 157. Exus, M. M. Respiratory volumes of - men during short exposures to con- stant low oxygen tensions attained by rebreathing, 267. Endocrine glands, conditions of activ- ity in, 399. ERLANGER, J. and H. S. Gasser. Studies in secondary traumatic shock: VI. Statistical study of the treat- ment of measured trauma with solu- tions of gum acacia and erystalloids, 119. VII. Note on the action of hyper- tonic gum acacia and glucose after hemorrhage, 149. See GASSER and‘ERLANGER, 104. See Gasser, ERLANGER and MEEK, 31. Exercise, effect of, on blood pressure, 443. Eye, infra-red radiant energy and the, 383. FrUSIMOTO, B. Studies on the regu- lation of the blood diastase, 208. INDEX (GASSER, H. S. and J. ERLANGER. Studies in secondary traumatic shock. V. Restoration of the plasma volume and of the alkali reserve, 104. —, J. ErRuancer and W. J. MEEK. Studies in secondary traumatic shock. IV. The blood volume changes and the effect of gum acacia on their ‘development, 31. ——. See ERLANGER and Gasser, 119, 149. Gastrin bodies, distribution and ex- traction of, 527. GitHEeNs, T. S. See Mettrzer and GITHENS, 377. Glucose, gum acacia and, after hemor- rhage, 149. GoupscuMiIpT, S. See WILson AND Go.upscumipT, 157. Grece, H. W., B. R. Lutz and E. C. ScHNEIDER. Compensatory reac- tions to low oxygen, 302. 5; The changes in the content of hemoglobin and erythro- cytes of the blood in man during short exposures to low oxygen, 216. Gum acacia and glucose after hemor- rhage, 149. —— ——, effect of, on blood volume changes, 31. —— solutions in treatment of shock, 119. HARTMAN, F. L. See Remmann and Hartman, 82. Heart, isolated, as indicator of adrenal secretion, 399. Heat, work and, effect of, on pH of sweat, 433. : —_— ——, —— —,, on pH of bi urine, 579. Hemorrhage, gum acacia and glucose after, 149. Hisanosu, K. On the distribution of the non-protein nitrogen in cases of anaphylaxis and peptone poisoning, 307. INDEX Hopers, P. C. See Dawson and Hopaes, 481. Hydrogen ion concentration of urine, 579. Hyman, L. H. Physiological studies on Planaria. II. Oxygen consump- tion in relation to regeneration, 67. Hyperglycemia, asphyxial, 177. [NFRA-RED radiant energy and the eye, 383. Intestine, small, conduction in, 252. Intracranial pressure and body tem- perature, 102. —— ——., effects of increasing, 352. JONES, W. and A. F. Apr. The prep- aration of adenine nucleotide by hydrolysis of yeast nucleic acid with ammonia, 574. KAKINUMA, K. Studies on the extract of lung, 9. Kanna, 8S. Physico-chemical studies on bioluminescence: I. On the luciférine and luciférase of Cypridina hilgendorfii, 544. Il. The production of lght by Cypridina hilgendorfii is not an oxi- dation, 561. Keeton, R. W. See Luckwarpt, Keeton, Kocu and La Mer, 527. Kocu, F. C. See Lucknarpt, Krr- Ton, Kocu and La Mrmr, 527. LA MER, V. See LuckHarpt, KEE- ToN, Kocu and La Mer, 527. Larson, J. A. See May and Larson, 204. Light, production of, by Cypridina hilgendorfii, is not an oxidation, 561. Luciférine and luciférase of Cypridina hilgendorfii, 544. LuckHarpt, A. B., R. W. Keeton, F. C. Kocu and V. La Mer. Gastrin studies. II. Further studies on the distribution and extraction of gastrin bodies, 527. 591 Lucxuarpt, A. B., H. T. Pariuies and A. J. Caruson. Contributions to the physiology of the stomach. LI. The control of the pylorus, 57. LuckresH, M.. Infra-red radiant en- ergy and the eye, 383. Lung edema, effect of oxygen on blood volume in, 157. extract, studies on, 9. Lutz, B. R. and E. C. ScHnerper. Alveolar air and respiratory volume at low oxygen tensions, 280. Circulatory responses to low oxygen tensions, 228. The reactions of the cardiac and respiratory centers to changes in oxygen tension, 327. See Greaa, Lurz and ScHnetr- DER, 216, 302. MANN, F. C. Experimental surgi- cal shock. V. The treatment of the condition of low blood pressure which follows exposure of the ab- dominal viscera, 86. May, E.S. and J. A. Larson. Posture- sense conduction paths in the spinal cord, 204. Meek, W. J. See Gasser, ERLANGER and MEErkK, 31. Mettzer, S. J. and T. 8. GITHENs. The nose-licking reflex and its inhi- bition, 377. Metabolites, effect of surgical proced- ures on, 82. Mitus, C. A. A note on the question of the secretory function of the sym- pathetic innervation to the thyroid gland, 174. Moore, L. M. Experimental studies on the regulation of body tempera- ture. III. The effect of increased intracranial pressure on body tem- perature, 102. The effects of increasing the intracranial pressure in rabbits, 352. 592 ERVOUS, sympathetic control of thyroid function, 174. Nitrogen of blood in anaphylaxis and peptone poisoning, 357. —— of egg albumen after quinine feed- ing, 364. Nose-licking reflex and its inhibition, Sic QXIDATION, production of light by Cypridina hilgendorfil, is not an, 561. Oxygen consumption during regenera- tion in Planaria, 67. effect of, on blood volume in lung edema, 157. ——, low, alveolar air and respiratory volume at, 280. —— ——, blood changes from short exposures to, 216. ——, —-——, compensatory reactions to 302. ——, ——, effect of, on respiratory vol- ume, 267. ——, ——, reaction of medullary cen- ters to, 327. tensions, low, sponses to, 228. circulatory re- PARTURITION , cardio-vascular re- actions in, 481. Peptone poisoning, nitrogen of- blood in anaphylaxis and, 357. Puiturrs, H. T. See LuckHarpt, PuHILuies and Carson, 57. Planaria, studies on, 67. Plasma volume and alkaline reserve in shock, 104. Posture-sense conduction spinal cord, 204. Pulse rate, effect of training and prac- tice on, 443. Pylorus, control of, 57. paths in UININE feeding, effect of, on nitro- gen of egg albumen, 364. INDEX EGENERATION, oxygen consump- tion during, 67. ReEIMANN, S. P. and C. E. BrEcKER. The catalases of the blood during anesthesia, 54. — and F. L. Harrman. Effect of anesthesia and operation on certain metabolites, 82. Respiratory centers, cardiac and, re- action of, to low oxygen, 327. —— mechanism in chelonians, 511. —— votime, alveolar air and, at low oxygen, 280. SS SS , effect of low oxygen on, 267. Rippie, O. See BEHRE and RIDDLE, 364. ALTS, inorganic, effect of, on cata- lase production, 165. Sarant, Y. Experimental studies of the ureter. The cause of the ureteral contractions, 342. ScHNEIDER, E. C. See Greae, Lurz and ScHNEIDER, 216, 302. See Lutz and ScHNEIDER, 228, 280, 327. Suevky, A. SHEvEKY, 191. Shock, experimental surgical, 86. , plasma volume and alkaline re- serve in, 104. ——, secondary traumatic, 31, 104, 119, 149. ——, traumatic, treatment of, 119. Spinal cord, paths of posture-sense in, 204. STARKWEATHER, E. See ALVAREZ and STARKWEATHER, 2052. Starvation, acidosis during, 1. Stomach, physiology of, 57. Suprarenalectomy, urea excretion af- ter, 191. Sweat, effect of work and heat on pH of, 433. E. See Bevier and (TAKEDA, M. On the protective action of some organic substances on catalase in acid medium, 520. INDEX Tapert, G. A. Changes in the hy- drogen ion concentration of the urine, as a result of work and heat, 579. Effect of work and heat on the hydrogen ion concentration of the sweat, 433. Temperature, body, intracranial pres- sure and, 102. Thyroid, desiccated, effect of, on cata- lase production, 165. ——, function, sympathetic nervous control of, 174. Training, effect of, on blood pressure in man, 443. REA excretion after suprarenalec- tomy, 191. Ureteral contractions, cause of, 342. Urine, hydrogen ion concentration of, 579. 593 VALSALVA experiment, cardio-vas- cular reactions in, 481. Vital staining and acidosis, 20. WILSON, D. W. and S. Go.p- scHmMipT. The influence of oxygen administration on the con- centration of the blood which accom- panies the development of lung edema, 157. Work and heat, effect of, on pH of sweat, 433. —, on pH of urine, 579. YAMAKAMI, K. The hypergly- cemia-provoking ability of asphyx- ial blood, 177. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 50, NO. 4 0 SINDING SECT. MAR 4 1966 K QP Americen Journal of Physiology a A5 v.50 cOp.2 Biological & Medical] Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY STORAGE — a Le ee Pipe : rye EE a se ~— - Co ae ON maren nee