. fatat ag ads 5 Vets ae afte ws aa if st} Be 7 Rat te Bes fA ae ihe iia a y 7 ~ THE AMERICAN JOURNAL OF PHYSIOLOGY VOLUME XLIII : ie ae os BALTIMORE, MD. 1917 is CONTENTS No. 1. Apriz 1, 1917 MiecropissEcTIon Srupies. I. THe VISIBLE Structure OF CELL PrRoro- _ PLASM AND DeatH CuanGEs. Robert Chambers, Jr..................... Tue OxyGen Pressure Necessary FoR Tissue Activity. Montrose T. CM cack SE 2. | + |. CRM el cose aatice 13 Tue PHysIoLOGy OF THE ATRIO-VENTRICULAR CONNECTION IN THE TURTLE. Ill. THe INFLUENCE OF THE VAGI AND OF THE SYMPATHETIC NERVES ON ITs RHYTHM-FORMING PowER. C. C. Gault...............-...5..-.. 22 Tue ConpiTions DETERMINING THE RATE OF ENTRANCE OF WATER INTO FERTILIZED AND UNFERTILIZED ARBACIA EaGGs, AND THE GENERAL RELATION OF CHANGES OF PERMEABILITY TO ACTIVATION. Ralph S. Lillie. . <1: Gai a gO OU Frc OC a 43 THE eeetx 3 OF piemietinie’c ON THE + Cavanane Comment: OF THE TISSUES. SS OP Se) | Rr al 58 CHANGES IN THE AMOUNT OF SALIVARY SECRETION ASSOCIATED WITH CERE- rte. fe. oh, Daghley...:.... 22% ilgeweemeees oh G6 es oe 62 SrupIeESs IN THE PHYSIOLOGY OF THE RESPIRATION. I. THE CAPACITY OF THE AIR PASSAGES AND THE PERCENTAGE OF CARBON DIOXIDE IN THE ALVEOLAR AIR DurING Rest AND Exercise. R. G. Pearce............. 73 Tue GRADIENT IN SUSCEPTIBILITY TO CYANIDES IN THE MERIDIONAL CoNn- DUCTING PATH OF THE CTENOPHORE Mnemiopsis. C.M.Child......... 87 CARBON *D1oxIDE AcIDOSIS, THE CAUSE OF CARDIAC Dyspnea. John P. ER SS OR a = LY ea 113 Tue Reactions or Kirrens ArrER DECEREBRATION. Lewis H. Weed..... 131 No. 2...-May 1, 1917 Tue Excitation or Microscopic AREAS: A NON-POLARIZABLE CAPILLARY rmmarm-- I redertck a er rat... ... .. veer a se ce eee 159 Tue Accuracy oF MovEMENT IN THE ABSENCE OF EXCITATION FROM THE EGAN, ICS. 0MUMRMOM. ...|. . UMIN e k . ews sw niga 169 THe CarpIo-SKELETAL Quotient. W. L. Mendenhall..................... 195 CONTRIBUTIONS TO THE PHYSIOLOGY OF THE StomacnH. XLI. THe ALLEGED INFLUENCE OF THE REMOVAL OF THE SALIVARY GLANDS ON THE SECRE- mon or Gastric Juice. A.M. Swanson. oe... 2.2 e ee ce een eee 205 Tue ComposITION oF SALiva IN RELATION TO THE INCIDENCE OF DENTAL cama. Sahn albert Marsha... seer ok. oe wee 212 Tue Epinepnric Content OF THE BLoop In ConpiTiIons or Low Bioop PRESSURE AND SHock. Edgar Alden Bedford.......................... 235 iil lV CONTENTS Bite Pigment Mertasouism. VII. Brine Pigment Output INFLUENCED BY HEMOGLOBIN INJECTIONS, ANEMIA AND BLoop REGENERATION. G, H. Whipple and C. W... Hooper. . 00:28 pokes os os bok oes 2 Bite Pigment Meraspouism. VIII. Brine Pigment Output INFLUENCED BY HEMOGLOBIN INJECTION; SPLENECTOMY AND ANEMIA. C. W. Hooper and G. He Whiley. ois 05555 0. air dL yen 2 Bite Prament Merapouism. IX. Bite Pramenr Ourrut INFLUENCED BY HeMoOGLOBIN INJECTION IN THE COMBINED EcxK-BiILE Fistunta Doa. C. W. Hooper and G: H. Whipple, cee is see eee Tue Errects oF ADRENIN ON THE DISTRIBUTION OF THE BLoop. II. Vot- UME CHANGES AND VENOUS DISCHARGE IN THE SPLEEN. R. G. Hoskins and &. #6. "Lee Gunning, ... 0.4 SSRN. ese oa a THE Errects oF ADRENIN ON THE DISTRIBUTION OF THE BLoop. III. Vot- UME CHANGES AND VENOUS DISCHARGE IN THE KipNrEy. R&R. G. Hoskins and R. Ey Lee Gunning...... 25 ee ae Le ae eS ee FurtTHER OBSERVATIONS ON THE DIFFERENTIAL ACTION OF ADREWSEEM: Frank A. Hartman and Lois McPhedran:... 3.020. 024 AN Hyproxytic Stupy or Curtin. Sergius Morgulis...................... Tue Fare or Sucrose PARENTERALLY ADMINISTERED. Shigenobu Kuri- YON cB Se es oie Sia se © gb eae tee ERS esa ge ttctral was dc. SS hr THr PERFUSION OF THE MAMMALIAN MepvuLua: THe Errecr or CARBON DioxipE AND OTHER SUBSTANCES ON THE RESPIRATORY AND CARDIO- VASCULAR CentTERS. D. R. Hooker, D. W. Wilson and Helene Connett. . No. 3. JUNE 1, 1917 Tue Return oF UREA FROM THE KIDNEY TO THE BLoop. 7’. Addis and A. EB. Shevkey 806 80. GE RRR OA SEE AROSE. ee A CoMPARISON OF THE Errects oF BREAKFAST, OF NO BREAKFAST AND OF CAFFEINE ON WorRK IN AN ATHLETE AND A Non-Atuuete. J. H. Hyde C..B..ftoot and H. Curl... 2. RRR A Sa ee ee THe Errects or ADRENIN ON THE DISTRIBUTION OF THE BLoop. IV. Errectr oF Massive Dosgs ON THE OUTFLOW FROM Muscie. R. FE. Lee Gunning Errects oF ADRENIN ON THE DisTRIBUTION OF THE BLoop. V. VOLUME CHANGES AND VENOous DIscHARGE IN THE INTESTINE. R. G. Hoskins and R. EB. Lee Gunning... 22 pees oes es so THe VASCULARITY OF THE ADRENAL Boptzs. R. Burton-Opitz and D. J. Edwards... 50. 3 ee OO ne Tar INFLUENCE OF SECRETIN ON THE NUMBER OF ERYTHROCYTES IN THE _ CrrcuLatine BLoop. Ardrey W. Downs and Nathan B. Eddy........... THE AcTION or ULTRA-VIOLET RADIATION IN KILLING LivinG Cre.us Suc# as Bacteria. W. EF. Burge..:. Siem. s.. a ae) sc Tue Errecr or Tuyrorp FEEpING ON THE CATALASE CONTENT OF THE Tissues. W. EH. Burge, J. Kennedy and A. J. Neill.................... RuEorropic Responses oF EpINEPHELUS Srriatus Buiocu. Hovey Jordan FURTHER EvipENCE REGARDING THE ROLE OF THE VAGUS NERVES IN PNEU- MONIA. W. T. Porter:and L. H. Newburgh. i506) +-++)-+- >. sean 258 275 298 304 311 328 343 351 CONTENTS Vv __ OROKINASE AND Sativary Digestion Stupies in THE Horse. C. C. Palmer, _ A.L. Anderson, W. E. Peterson and A. W. Malcomson.................. 457 _ Aw Application or Boyte’s Law To Putse Waves IN CLINICAL MEASURE- bP MENT OF Boop Pressure. A. M. Bleile.........................2.-5. 475 Pe o ; No. 4. Jury 1, 1917 THE INFLUENCE OF THfROID FEEDING UPON CARBOHYDRATE METABOLISM. 5 Shigenobu Kuriyama Le, eS OP eee oe 481 VARIATIONS IN IRRITABILITY OF THE REFLEX Arc. III. Fuexton REFitex ‘ ARIATIONS, COMPARED WITH THOSE OF THE NeERVE-MuscCLE PREPA- Pig emnaeme Le Porter:............. cs: ouee eee ee 497 - Tue Benavior or Ho.oruurians 1N Batancep Intumination. W. J. Bg RR 510 4 - Tue Errect or Dextrose GIVEN INTRAVENOUSLY ON BLOOD CoMPoOsITION 3 AND URINARY SeGRETION. David M. Davis.....................:...... 514 ‘ Furtuer Srupies oN THE Errect or ADRENALIN UPON MuscuLarR FAarTiGue. EE eee 530 Tue Errect oF PHospHORUS POISONING ON THE CATALASE ConTentT. OF THE Ee: AAPOR le Lea ck ee ca es 545 Tue NATURE AND PRoPpERTIES OF METATHROMBIN. Arnold R. Rich......... 549 Tue CHANGES IN CLOTTING PoWER OF AN OXALATED PLASMA ON STANDING. eC tte Wa... ea ae CME call os wk wwe ace 571 Tue Diastatic AcTION OF SALIVA IN THE Horse. R. J. Seymour.......... 577 Tue RELATION BETWEEN THE THROMBOPLASTIC ACTION OF CEPHALIN AND its DEGREE or UNSATURATION. Jay MacLean......................... 586 - o THE AMERICAN | JOURNAL OF PHYSIOLOGY VOL. 43 APRIL 1, 1917 No. 1 MICRODISSECTION STUDIES. I. Tue VISIBLE STRUCTURE OF CELL PROTOPLASM AND DEATH CHANGES ROBERT CHAMBERS, JR. Cornell University Medical College, New York City Received for publication January 8, 1917 A study of the physical properties of protoplasm requires for its foun- dation a knowledge of the consistency of protoplasmic structures in the — living cell. Barber’s pipette holder, a mechanical device for manipu- lating microscopically fine pipettes and glass needles in a hanging drop has proved most valuable for this purpose as with it one may dissect living cells under direct microscopic observation. The technique used will be described in a paper which will shortly appear. One must be fully alive to the difficulty of discriminating between protoplasmic structures and artifacts frequently produced during dissection. Con- siderable experience was, therefore, found to be necessary for a proper interpretation of the results obtained. Marine invertebrate and plant ova lend themselves well to micro- vivisection as their post-operative development is easily followed. The bulk of this paper, therefore, deals with results obtained from studies on the protoplasm of Asterias and Arbacia ova at Woods Hole and that of Echinarachnius, Cerebratulus and Fucus ova at South Harpswell. It is remarkable how much tearing and pulling with a needle living protoplasm will undergo without showing injury. One may puncture a cell with a needle and drag the needle through the cytoplasm back 1 This paper is based on the work of two summers at Woods Hole Marine Bio- logical Laboratory, and one summer at the South Harpswell Laboratory. The writer wishes to express his indebtedness to Prof. F. R. Lillie for accommodation at Woods Hole and to Prof. H. V. Neal for facilities accorded at South Harpswell. An abstract was read before the American Physiological Society, December, 1916. 1 2 ROBERT CHAMBERS, JR. and forth cutting through the sides of the cell and if the procedure be slow and gradual the tear closes up behind as the needle proceeds and the process may be continued almost ad libitum without producing an ill effect. If, on the other hand, the needle be carried rapidly through the cytoplasm, a few thrusts only are necessary to induce rapid disor- ganization. The effects of injury are probably cumulative. If in- jurious effects be made to follow one another without giving the cell time for recovery, the additive effects of the injury soon manifest them- selves in disorganization of the protoplasm resulting in the death of the cell. The cytoplasm. In the marine eggs studied, the protoplasm consists of a hyaline fluid matrix in which are imbedded granules of various sizes. ‘The fluid offers no perceptible resistance to the needle and an indication of its very slight viscosity lies in the fact that when the néedle is moved through the fluid, the only granules displaced are those in the immediate vicinity of the needle. The fluid is water miscible. If, for example, the cell surface be torn, the interior cytoplasm pours out and mixes completely with the surrounding water. Also, if a drop of water be injected slowly and gradually into the egg by means of the mercury injection method, the water diffuses throughout the cytoplasm diluting it. A dilution can be produced in this way sufficient to pro- duce Brownian movement of the imbedded granules. The surface film of such an egg is so much weakened that a mere touch suffices to burst it open when everything except the smallest granules disappears in solution. This water soluble hyaline protoplasm coagulates with ease on mechanical injury. Mere compression will often cause an egg to coagulate into a solid mass. It can then be cut into pieces which hold their shape. This is apt to lead one to the erroneous conclusion that the substance of a cell is usually a solid protoplasmic gel. When a concentrated aqueous solution of neutral red or brilliant cresyl blue’ is injected into the interior of an egg, the dye spreads from the tip of the pipette diffusely staining the hyaline cytoplasm. Within a few seconds a granule here and a granule there begins to stain. Ina short time all the dye is taken up by the scattered granules leaving the cytoplasm colorless. At the tip of the needle where the concentration of the stain is at its greatest all the granules finally stain. Elsewhere colorless granules lie scattered among the colored ones. The stain is ? For the use of the brilliant cresyl blue and for noting the effect of acetic acid on the macrosomes, I am indebted to Mrs. M. R. Lewis. VISIBLE STRUCTURE OF CELL PROTOPLASM 3 never permanent. Within a few minutes to an hour, depending on the amount injected, the stain completely fades away. - A continuously stained condition can be maintained only by the presence of a superfluity of the dye. The granules in the cytoplasm of the eggs studied vary considerably within certain limits in size (fig. 1). They have been described by E. B. Wilson (2) and Kite (6). One may classify them into two groups. The smallest granules or microsomes, as they may be called, are minute specks considerably below 1 u in diameter but plainly visible with the high powers of the microscope, their refractive indices differing consid- erably from that of the surrounding cytoplasm. The larger granules or macrosomes (Wilson’s alveolar spheres), vary from 2 to 4 y in Fig. 1. Illustrating the protoplasm of a living Echinarachnius egg with microsomes and translucent macrosomes (the latter 2 to 4u in diameter) crowded together in a hyaline fluid. diameter and may be circular, oval or irregularly polygonal in shape, exhibiting simple outlines with no diffraction rings. They are closely crowded throughout the egg and among them are scattered the micro- somes. The macrosomes, constantly but in almost imperceptible de- grees, change in shape and position and it is probable that they are as constantly disappearing and reappearing in the hyaline cytoplasmic matrix. They are very translucent in Echinarachnius and so contrib- ute to the remarkable transparency of the egg. In the Cerebratulus egg they are comparatively opaque. The difference between the micro and macrosomes is brought out most prominently on injury of the cell by acids or by the needle. If a hanging drop of sea water containing eggs be subjected to the fumes of 4 ROBERT CHAMBERS, JR. acetic acid for a few seconds, the macrosomes quickly disappear leay-. ing the microsomes in a hyaline coagulated mass of cytoplasm. The dissolution of the macrosomes occurs also when disorganization of the protoplasm is induced by rapidly repeated tearing of the cytoplasm with the dissecting needle. The macrosomes swell and gradually but rapidly fade into the surrounding cytoplasm which now becomes very liquid and in which the microsomes exhibit a dancing Brownian move- ment. The disorganized area then spreads on all sides, and, if the initial injury be extensive and no protective membrane intervenes (see farther on), soon involves the entire cell. If the injurious tear be such as to destroy the surface film in one spot, the disorganized liquefied cytoplasm flows out and mixes completely with the sea water and, in a few seconds the entire cell, except for the dancing microsomes, disap- pears from view. The microsomes appear to be the most resistant structures of the cell. Wilson’s conclusions (2) that there exists “‘a complete gradation in size from the largest alveoli’ (macrosomes, as I call them) ‘down to the microsomes” is based, I believe, on a misconception derived from crushing protoplasm. On crushing protoplasm, the macrosomes go into solution and the disintegrating protoplasm as it flows out, tends to break up into spherules of varying sizes, by the formation of surface films. These spherules vary extremely in size and may fuse with one another on touching. They swell readily in water or may coagulate and are quite different from the macrosomes of the uninjured egg. Kite (3) describes the formation of more or less rounded masses in dy- ing Asterias eggs and rightly distinguishes them from structures nor- mally present in protoplasm. Frequently before all the disorganized cytoplasm has poured out and mixed with the surrounding water, the outflow ceases and the whole mass changes into a solid coagulum which may or may not show a network structure. The complexity in constitution of the protoplasm must account for the fact that a very slight difference in the manner of injury, or in the state of the protoplasm, at the time of injury may pro- duce such widely differing end results, viz., a coagulation into a solid jelly or a complete dissolution into the surrounding medium. The co- agulum consists of a hyaline gel with granules arranged in a delicate meshwork. Occasionally upon disintegration the cytoplasm begins to set before the macrosomes have quite gone into solution. Such a mass often resumes dissolution with a succession of peculiar spasmodic jerks a3 VISIBLE STRUCTURE OF CELL PROTOPLASM 5 apparently due to the macrosomes which, here and there, swell and -earry into solution regions surrounding them. In cells killed by fixing reagents, the death changes take place with- out destruction of the cell wall. If the fixing agent act rapidly, the mac- rosomes tend to swell simultaneously and coagulation sets in before the microsomes are irregularly dispersed. The precipitation network - incident to coagulation is produced regularly and the result is a fairly symmetrical alveolar structure with microsomes imbedded in the alveolar walls. Flowing may occur in the disintegrated cytoplasm if the fixing agent be slow in producing coagulation and a variety of dis- torted figures may result. The prolonged action of acetic or hydro- chloric acid tends to dissolve the microsomes and, with various fixing Fig. 2. Drawing of a sector of an Echinarachnius egg in the astral stage fixed in Bouin’s picroformol and stained with iron hematoxylin. The effect of fixation _is seen in the fibrous appearance of the astral rays and in the network structure of the coagulated cytoplasm. reagents, a variation in coagulation effects occurs so that one obtains granular and network precipitates which may simulate protoplasmic structure but in reality completely mask it and often lead to erroneous conclusions regarding cell inclusions. The cell aster is a structure whose nature is entirely masked on fixa- tion. In the living cell the rays are probably paths due to a centripetal flow of a hyaline fluid in an otherwise temporarily gelatinized cytoplasm. When a cell containing an aster is fixed, the reagent used disorganizes the cytoplasm in the way described above. The rays are then com- pressed between artificially produced alveoli, and their substance be- 6 ROBERT CHAMBERS, JR. ing precipitated, gives the appearance, familiar in fixed material, of fibers or rows of granules which seem to merge peripherally into an alveolar network (Fig. 2). An acid reaction of protoplasm on mechanical injury can be demon- strated in the Arbacia egg where many of the macrosomes are deep reddish brown chromatophores. On injury to the cell, the color diffuses into the liquid cytoplasm, changing from brown to an orange pink hue,? denoting a distinct increase in acid reaction. Also in cells whose granules are stained wth neutral red, injury causes the dye to diffuse out of the macrosomes as they dissolve and give to the liquid cytoplasm a rose color. Acid fuchsin also, which is avidly taken up by acid reg- ions, stains the injured disorganized cytoplasm while the normal eyto- plasm remains colorless. The peripheral layer of the cell. The surface layer of the egg cells stud- ied is very dense in consistency as compared with the cell interior into which it merges insensibly. In the unfertilized egg, the cell granules are imbedded in it up to the very line of division between the egg and surrounding medium. With the needle the surface may be pulled out into long strands without otherwise disturbing the contour of the cell. On being released the strands tend to curl and retract slowly till they disappear. If a more rapid tear be made, and if the cell be under com- pression, the spot torn bulges out as the internal cytoplasm presses on the weakened surface. The surface layer of the swelling protuberance is very easily broken, upon which the interior may pour out. The cytoplasm then either disintegrates entirely in the surrounding water or, if remaining normal, reéstablishes a film on its surface. When left undisturbed the new surface film gradually strengthens into a definite ectoplasmic layer and the protuberance slowly retracts until the origi- nal contour of the egg is reéstablished. If the point of attachment of the protuberance be small, the protuberance may be pinched off to form a spherule of cytoplasm which to all appearances is normal. If a tear of the surface layer or of any part of the cytoplasm be so injurious that disorganization sets in, a film may form around the disorganized area separating it from the sound cytoplasm. The recovery of an injured cell is only brought about by the formation of a membrane-like film which prevents extension of the injury. A sue- ’ This color reaction can be demonstrated in an aqueous suspension of the echinochrome made by drawing off a quantity of the colored body fluid into dis- tilled water. VISIBLE STRUCTURE OF CELL PROTOPLASM 7 cession of films may form, as-one after the other, they succumb to the steady advance of the destroying process and the film which finally holds out may enclose only a fraction of the original cell but what it encloses will be normal protoplasm. The retention of the forming sur- face film is aided by the presence of solid structures in the cytoplasm. These act as bases which shorten the span to be bridged over by the new film. In the mature fertilized egg, the developing sperm aster which is a gelatinized ball of cytoplasm frequently acts in this way. When the surface layer is injuriously torn and the disorganized in- terior cytoplasm flows out, the torn edges of the surface layer curl out -and rapidly dissolve, upon which the entire cell disappears in the sur- rounding water. If the disorganization of the interior has been pro- duced without destruction of the surface layer, the surface often becomes transformed into a rigid coagulum enclosing the fluid products of the disintegrated protoplasm. The readiness with which the surface film can be reformed is seen in the following experiment on the unfertilized mature Arbacia egg. If an egg be compressed in a hanging drop and then pushed along with a blunt needle, peculiar currents can be produced in the egg substance. The currents pass directly from the pushing object in a straight line through the egg to the anterior end where they curve outward and flow back along the surface to be caught again in the flow from the pushing object. The egg surface is thus being continually reformed by an outflow of its interior, much as the falling sides of a fountain of water are formed by the jet that is streaming up at the centre. The readiness, however, with which the internal cytoplasm may be trans- formed into the substance of the surface layer is limited. If the egg be pushed beyond this limit no surface layer forms and the cytoplasm at once disorganizes and dissolves in the surrounding water. If one stops before reaching this limit, the egg can be made to round up and will continue a normal existence. In protozoa, the surface layer of protoplasm, the ectoplasm, is very pronounced. If the ectoplasm of Paramoecium bursaria be torn, inter- nal pressure causes the endoplasm to flow out through the tear which is at first a gaping hole in the ectoplasm. The contractility of the ectoplasm, however, is such that the torn rim curves in. In this way the free edges of the tear tend to approach and if the tear be slight they meet and further outflow of the endoplasm is stopped. In time the concavity on the surface produced by such an injury fills out and the cell resumes its normal shape. If the gap in the tear be too wide for 8 ROBERT CHAMBERS, JR. this method of repair, a surface film bridging the gap forms which may break repeatedly until the outflow lessens the internal pressure suffi- ciently to allow the newly formed film to persist. Within a few hours, evidence of the tear is no longer visible, either the film stiffens into a definite ectoplasmic layer or the entire body of the cell contracts to bring the original edges of the gap together making the ectoplasm again a continuous layer. A paramoecium may be readily cut into pieces by squeezing it between the coverslip and a fine glass rod, or a knife which is not sharp enough to cut through the ectoplasm. The knife edge bears down upon the surface of the cell until the floor of the groove formed touches the ectoplasm on the other side. The surfaces of con- tact fuse and further pressure of the knife separates the cell into two portions possessing no gap through which the fluid endoplasm may escape. The cautions usually given in the technique of cutting up protozoa is significant when we bear the above in mind. The knife must be as sharp as possible (which will still be blunt from the point of view of the paramoecium), the cutting edges must be free from nicks and, in cutting, the experimenter must bear down upon the knife without giving it a drawing movement. Either of the two latter pre- cautions, if unheeded, prevents a continuous surface of contact for the ectoplasmic layer of opposite sides and when the knife is removed the endoplasm will flow out through the gap thus destroying the cell. In the marine ova studied, the ectoplasmic layer is very thin but the same condition holds true, viz., that cut pieces will persist only when the cut surface can be bridged over by a morphologically definite film. In summary, we may say that the surface layer is a highly extensile, contractile and viscous gel capable of constant repair. Its establish- ment and maintenance is a property essential to protoplasm. With the film intact the mass of protoplasm’ maintains itself and the life of the cell is assured. When the film is destroyed the cytoplasm flows out, the macrosomes swell and disappear, the whole mass completely disorganizes and disappears in solution in the surrounding water.* In regard to the difference in permeability of the surface layer of a cell and its interior, I have so far only tested the diffusion of the three 4 Kite describes the cytoplasm of the Asterias eggs as ‘‘a quiet translucent gel which can be cut into small pieces with comparative ease.’’ His paper is a pioneer one in microdissection research. The observations recorded were necessarily fragmentary and the differences between the surface layer of the egg and its in- terior as also the ease with which protoplasm forms gel membranes escaped his notice. VISIBLE STRUCTURE OF CELL PROTOPLASM 9 basic dyes, neutral red, cresyl blue and janus green (6). A droplet was injected into the cell simultaneously with the application of a similar droplet to its external surface. For all three dyes diffusion into the cytoplasm took place equally rapidly whether applied to the egg sur- face or injected into the interior. . Insect germ cells. The germ cells of certain insects, Periplaneta, Disosteira and Anasa, were studied both in modified Ringer’s fluid and in the serous fluid of the insects used (3,4). By pricking the walls of the cysts in the testes, spermatocytes in different stages of development flow out as isolated cells. In the serous fluid they may be kept alive two or three days, during which all stages of cell division may be ob- served. Except for the nucleus and the mitochondrial network which surrounds it, the resting cell consists of a hyaline fluid cytoplasm. By very slow action with the needle, strands of protoplasm may be pulled out which retract on being released. Injury, however, very * easily manifests itself upon which the cell outline fades and the cyto- plasm disappears in solution in the surrounding medium. Adult somatic cells. These vary greatly in consistency owing prob- ably to the amount of metaplasmic material into which their proto- plasm has been transformed. Of the various types, only a few possess the fluid consistency of embryonic and germ cell protoplasm. With the exception of the leucocytes, they are comparatively resistant to mechanical injury and are tough and fairly rigid bodies. The nerve cell is probably a rigid gel. The muscle fiber is also a gel. Its sub- stance can be easily pulled out into strands which retract completely when released. Continued injury causes the muscle substance to pass into a very rigid hyaline gel which may be cut into discrete non- glutinous pieces. Gland cells swell readily when punctured and torn with the needle. They seem to exist both in sol and gel states. Mucosa cells consist of a soft, very extensile gel. They tend to round up when isolated. The leucocyte possesses a protoplasm which is very like the undifferentiated protoplasm of the germ cells. If a leucocyte with pseudopodia be touched with the needle, the pseudopodia are retracted and the cell becomes spherical. If the tip of a very fine needle be inserted gradually into the leucocyte a puncture may be made without apparent injury. The diminutive size of the cell, however, renders necessary very little accummulation of injurious effects to disorganize the cell. This usually occurs with almost explosive rapidity, the entire cell going into solution. Dead leucocytes are coagulated bodies and may be cut into non-glutinous pieces. 10 ROBERT CHAMBERS, JR. The cell nucleus. Observations on the cell nucleus of all the ova studied agree with those of Kite (5) that the resting nucleus is hyaline and exists in the sol state. In the germinal vesicle of the immature egg a definite membrane seems to bound the nuclear substance. It is extensile but easily destroyed. Evidence that this membrane is a morphological structure is shown on withdrawing some of the nuclear contents with a micropipette. The nucleus then partially collapses throwing the nuclear surface into irregular folds. Suspended in the | nuclear fluid in which it may be pushed about with ease is the germinal spot. or nucleolus distinctly visible on account of the difference in its refractive index from that of the-nuclear fluid. It frequently contains one or more vacuoles and does not appear to be solid for it may be cut into two, each part rounding up like a droplet. A rapid. tearing of the germinal vesicle with the needle point pro- duces an injury which is accompanied by some remarkable changes. The nucleolus at once swells and fades from view, the nuclear membrane ° entirely dissolves and, as the nuclear fluid comes into contact with the surrounding cytoplasm, immediate disintegration takes place. The destructive action spreads and may involve the entire egg unless a protective film forms to enclose the area of destruction as in a vacuole. A vacuole of this kind gradually works to the surface of the egg, where it is eventually extrudéd and expelled. When this has occurred the egg resumes its normal appearance although smaller than before and minus its nucleus. The disintegrative action of the nuclear substance very quickly disappears on extraction from the cell. It is, however, possible, by: acting rapidly, to produce the destruction of one cell by injecting into it the substance of the germinal vesicle of another cell. If:the nuclear substance remains more than five or ten seconds in the micropipette it is found to be innocuous on injection. With the nor- mal breakdown of the germinal vesicle in the maturing egg, the cyto- plasm acquires an increased sensitiveness to injury by the needle. This sensitiveness gradually passes off during the polar body formation. In the mature egg both nucleus and cytoplasm are comparatively resistant to injury. The nucleus is a hyaline sphere which behaves like an immiscible fluid drop in the cytoplasm. One may divide it into two and each part rounds up into a droplet. On coming into con- tact the droplets run together. Such a process is no hindrance to normal nuclear activity, for in one case an egg so treated was sub- sequently fertilized and segmented normally. Rapid thrusts of the - needle into the nucleus cause injury, which manifests itself either in et ; — a ee VISIBLE STRUCTURE OF CELL PROTOPLASM ll a swelling followed by complete dissolution or a rapid coagulation with the production of a granular meshwork precipitate simulating the nuclear network of fixed cells. SUMMARY 1. Protoplasm is a hydrophilic colloid which, in early germ cells, egg cells and Protozoa, usually exists in the’sol state with a surface layer in the gel state. Adult somatic cells generally are gels in which one cannot demonstrate a cell membrane possessing a — different from that of the cytoplasm within. 2. The microscopically visible granules in the cytoplasm of the egg of Arbacia may be classified into two groups: (a) The microsomes, which are considerably less than one micron in diameter and constitute the most resistant parts of the cell maintaining themselves after com- plete disorganization of the cell; (b) the macrosomes, which range from 2-4 micra in diameter and are very sensitive to injury. 3. The external surface of the egg cell is a gel which passes gradually into the sol in the interior. The surface gel is very extensile and con- tractile and is readily regenerated on injury. Tearing of this surface, if unrepaired, results in the pouring out of the internal cytoplasm and dissolution. 4. A remarkable property of protoplasm is its ability to form a pro- tective gel film not only on its external surface but also around an in- jured area which is in the process of disorganization. The disorganized mass thus insulated is eventually expelled from the cell. 5. A continuous but gradual application of mechanical injury can be sustained by a cell for some time without evidence of harm done. ' A short but rapid application produces instant local destruction, the spread of which may involve the entire cell. 6. Disorganization of the cytoplasm of the egg cells studied takes place in the following way: First, the macrosomes swell and go into solution, and second, the liquid hyaline cytoplasm may flow out and disappear in the surrounding water or it may suddenly set forming a rigid coagulated mass. The coagulation structure gradually coarsens with the production of a network or granular precipitate. 7. Injury is accompanied by a swelling and an apparent increase in the acid reaction of the part involved. 8. The comparatively rigid ectoplasm and the fluid endoplasm of Protozoa are directly comparable with the surface layer and the internal cytoplasm respectively of the marine ova studied. 12 ROBERT CHAMBERS, JR. 9. The surface layer and the internal cytoplasm appear to be equally permeable to the basic vital dyes used. 10. The germinal vesicle of an immature egg consists of a hyaline liquid enclosed in a gel like membrane. The nucleolus is an immiscible droplet floating in the vesicle and is very sensitive to mechanical injury. 11. The contents of the germinal vesicle of an immature egg, if brought into contact with egg cytoplasm, either by mechanical rupture of the vesicle or by injection, produce instant destruction of the ey- toplasm. This is not true for the mature nucleus nor for the segmenta- tion nucleus. 12. The cytoplasm of an immature egg is comparatively impervious to mechanical tearing, that of a mature egg is very much more sensitive. 13. In the mature egg the nucleus behaves as a fluid droplet whose - substance is immiscible with the cytoplasm. It may be divided into two droplets which unite on touching. It coagulates with ease on mechanical injury. BIBLIOGRAPHY (1) Cuampers: Biol: Bull., 1917, xxxii. (2) Wiuson: Journ. Morph., 1899, Suppl. to xv, 1. (3) CuamBers: Science, N. S., 1914, xl, 824. : (4) CuamBERs: Science, N. S., 1915, xli, 290. , (5) Krre: This Journal, 1913, xxxii, 146. (6) Krre: Biol. Bull., 1913, xxv, 1. ADDENDUM é If the injection of an aqueous solution into an egg be not very carefully and gradually done (see page 2) the mechanical compression caused by the force of the injection will produce a coagulation film about the injected droplet to form a vacuole. This film exhibits diosmotic properties similar to that of the gelled surface of the egg for, as Kite observed (6), such a vacuole filled with hypertonic sea water increases in size while one containing distilled water loses its water and decreases in size. THE OXYGEN PRESSURE NECESSARY FOR TISSUE ACTIVITY MONTROSE T. BURROWS Pathological Department of the Johns Hopkins University, Baltimore, Maryland Received for publication January 9, 1917 The ordinary plasma culture is sealed into a very small chamber of a hollow ground slide. The cultures are made by placing small frag- ments of tissue into layers of liquid coagulable plasma on the surface of a cover glass. The cover is placed over the hollow ground slide so that the tissue and the plasma hang in the hollow chamber. It is sealed to the slide with vaseline and paraffin (1). From the air in the hollow chamber the cells obtain the oxygen necessary for their activi- ties and it has been of interest that in such a small chamber the cells may continue to grow actively for several days. This observation had already led to the belief that these cells may grow in an atmosphere of a low concentration of oxygen. This belief was further substan- tiated by the fact that the renewal of the air had little or no notice- able effect on the growth or activity of the cells. That the oxygen essential for activity in the culture is derived from the air chamber was determined by placing two small glass tubes under the cover and replacing the air in the chamber by hydrogen and sealing. No growth took place in these cultures. A considerable amount of work has recently been done to determine at ordinary atmospheric pressure the per cent of oxygen necessary to maintain a flame as well as to maintain the life of organisms. Clowes (2) and Loevenhart (3) noted that alcohol ceases to burn in an atmosphere containing 15 per cent of oxygen. The latter author noted that ether and Madison illuminating gas cease to burn at 13 per cent of oxygen, while hydrogen continues to burn until the atmosphere contains 6.6 per cent of oxygen. With the animal conditions are different. Loevenhart found rab- bits would live in an atmosphere containing 3.5 per cent of oxygen, but that they died when this was reduced to 3 per cent. In the present series of experiments an attempt has been made to 13 14 - MONTROSE T. BURROWS determine the effect of pure oxygen and various partial pressures of oxygen on the growth of the cells in vitro. Although at the present time the apparatus is still not perfect and a few of the results given here have a slight error, it seems that in their present form they warrant publication. MATERIALS AND METHODS The gas and mixture of gases tested have been kept at atmospheric pressure. The partial pressure of oxygen has been lowered by dilut- ing it with nitrogen. The culture chambers used have a large air space, forty to sixty times as large as that of the ordinary culture chamber. To make these determinations it was necessary to prepare pure oxygen and nitrogen and to devise a culture from which gases could not diffuse. The cultures, therefore, were made in sealed glass chambers. Fig. 1.. a, The complete culture chamber; b, the sealed culture chamber. The culture chamber. The culture chamber as shown in figure 1, a, is blown from soft glass tubing + inch in diameter. In place of the cover glass used in the ordinary cultures, one small part of one side of the tube is blown out, made very thin and flattened. Opposite this place the tube is again blown out and flattened. This latter flat sur- face forms a base for the culture and allows free transmission of light which is essential for examining the cultures under the microscope. On each side of the central portion the tube is constricted so that it can be easily sealed with a small flame. Beyond this point, it is left full size but corrugated to allow a rubber tube to be tightly fitted. - These culture chambers are easy of construction and are made in the laboratory as they are needed. _ The culture chambers thus prepared are boiled in soap and wiles left to soak several hours in sulphuric acid, rinsed in tap and distilled water. ‘They are then steamed for several minutes to remove soluble substances in the glass and sterilized by dry heat. a ee ee ee oe OXYGEN PRESSURE AND TISSUE ACTIVITY 15 _ Oxygen. The oxygen is prepared by dissociating a 1 per cent aqueous solution of sulphuric acid with an electric current. This method of preparing oxygen is well known. The particular apparatus 0 ae oe ee 7 Fig. 2. An automatic oxygen-hydrogen generator. used works. automatically and is shown in figure 2. The current used is D.C. 110 volts, which has been passed through a resistance of one to four 16-candle power carbon lights. The apparatus is made 16 MONTROSE T. BURROWS automatic by interrupting one of the wires with a platinum pointed | switch, which is operated by a float. Oxygen prepared in this manner is something over 99 per cent pure. It contains, however, a small amount of ozone, hydrogen and water vapor. Since the ozone attacks organic matter and may become in- jurious especially if rubber tubing is used at any point, the gas, before using, is bubbled through olive oil. To remove any traces of acid, it is further passed through a tube containing soda lime and bubbled through an aqueous solution of potassium hydrate 40 per cent. No rubber connections are used in any part of the apparatus. The ap- paratus is continuous except for one joint which is ground glass. Nitrogen. The nitrogen is commercial. It contains about 2 per cent of oxygen which is removed by bubbling the gas through several long cylinders containing pyrogallol and potassium hydrate. These are mixed according to the proportions given by Hempel (page 149) (4). One hundred and twenty grams of potassium hydrate are dissolved in 80 cc. of water. To this are added 5 grams of pyrogallol dissolved in 15 cc. of water. Hempel states that no carbon monoxide is formed when a solution of these proportions is used. To insure against this possibility the gas is bubbled through two cylinders containing Sand- meyer’s solution (Hempel, page 203) (4). It is then passed through a tube containing soda lime and bubbled through a solution of 40 per cent potassium hydrate. Preparation of cultures and tissues. For these experiments fragments of heart muscle, and skin of chick-embryos from five to sixteen days of age have been used. The medium is plasma prepared from the blood ‘of adult chickens. When thin layers of plasma are placed in these large culture chambers some slight evaporation may take place. To prevent this from causing an injurious hypertonicity, the plasma, previous to being used, is diluted 0.1 with sterile distilled water. Fragments of both heart muscle and skin are planted in each culture. It is impossible to prepare these cultures in the ordinary manner. The method used has been to mix together quickly outside four or five pieces of tissue with a drop of liquid plasma. The plasma and tissue are then sucked, before the plasma clots, into the end of a long slender pipette which may be passed into the culture chamber, and the drop of plasma containing the tissue fragments deposited on the thin, flat surface which has been prepared for it (fig. 1, a). The culture chamber is then gently shaken so that the drop of plasma spreads in a layer no greater than 0.5 mm. in thickness, and the tissue fragments OXYGEN PRESSURE AND TISSUE ACTIVITY 17 become scattered in this layer over the surface of the glass. The fragments of tissue are 1 mm. or less in diameter. As soon as the plasma has clotted, the chambers are turned over so that the culture hangs into it. The side walls of the culture chamber are moistened with sterile water so as to further prevent as far as _ possible any evaporation of the culture medium. In each experiment, one, two or three cultures are used. Besides these in most instances, three control cultures are also prepared. The air in one of the con- trols is replaced by nitrogen gas. In the other it is replaced with pure oxygen, the third is sealed at once and acts as an air control. This last culture, which contains air, controls the tissue and plasma as well as the activity observed in the cultures where pure oxygen gas is being tested, but it does not control any possible toxic substances in the ni- trogen. An attempt to partially control this was made by substitut- ing air for oxygen in several experiments. It was possible by this means to use for the very low dilutions of oxygen a quantity of nitro- gen, which has been proven not to be toxic in those cultures where a higher partial pressure of oxygen has been tested and an active growth of cells has been observed. The cultures prepared have shown no evidence of toxic substances in the nitrogen gas. Since it is essential for all these determinations that the gases used be saturated with water vapor at the temperature at which the cells are grown, the measuring of gases and the filling of the culture cham- bers with the gas to be tested are always made in the incubator. The arrangement of the apparatus within the incubator is illustrated in figure 3. The oxygen and nitrogen are passed through slender glass tubes, A and B respectively, across to the middle of the incu- bator where they are each bubbled through cylinders containing dis- tilled water, c-A and c-B respectively. From the water cylinders the gas passes into T-tubes, one arm of each of which is open and fitted with a stopcock. To this open arm the cultures in which the pure gases are to be tested are attached. The other arm passes to the measuring cylinder. The measuring cylinder is arranged so that not only these but other gases may be tested. The gases are measured over mercury. In order to keep them saturated a thin layer of freshly boiled distilled water is run over the mercury surface before they are admitted. The measurements are made at the temperature of the incubator. The measuring cylinder is fitted at the top with a 3-way cock. One arm of the cock is open and to this the onleam chambers (b) are attached, by means of a rubber tube. MONTROSE T. BURROWS 18 ‘sraqureyo oinyjno ‘q:‘aepurjéo Burysem ueSorytu ‘“g-o ‘ieputjAo Surysea uesf{xo ‘y-9 fueZo1yzru Suth11890 vf iia: = = " ANONNANNNNRARONNDNNANANTARNARALU TRA ALANA A a Z ZZ TTTTT TITEL LLL LL hhh ated i astlTTELDMLSDLLELSLDSDS SOP d V eorek Se ed == —_ > — FZ = p mT m m — eS ae ——ane MMOD EQ = 7 q g ’ . 1 a ae OXYGEN PRESSURE AND TISSUE ACTIVITY 19 The air in the culture chambers is replaced by pure gases or a mix- _ ture of gases in the measuring cylinder by allowing these gases to pass rapidly through the culture chambers at regular intervals during a period of one or two hours. The free open end of the culture chamber is fitted with a rubber tube and a piece of glass tubing. The free end TABLE 1. PARTIAL Lea * rs] n> a roam | or | 6 fabea| 6° c $2 | & Sa 33 #2 “— Reet}teot| @S2 | 8 fone | ae | @ | %s eubie | cubic | S£4 | £2 g z oa a 36 meurs|metes| #22 | $8] 88 | Bs | & | Bs of O | of N a." ps ae 4g PS a7 : Various times........ 100 O |100.0* | 50 |++++/++++/764.1*|712.1* Various times........ 80 | 20 8 I+4+44/4+4+4+4+ Various times........ 60 | 40 6 | +++ | +++ Various times........ 40 | 60 7 |4+4++/4+4++4+ Various times........ 20 | 80 15 |++4++4+|/4+4+4++4+ Various times........ 15 | 85 6 |++4++/++4++ Various times........ 12 | 88 4 |4+4++/4+4+4++4+ Various times........ 10 | 90 6 |++4+4/4+4+4++4+ June 2, 1916..........| 9 91 1 | +++ | +++ Various times........ 8 | 92 6 |++++/++++ Various times........ 7 93 1 +. dott June 21, 1916......... 6 94 1 2 ey 2 et See June 28, 1916......... 6 94 2 + +++ mec. 14, 1016:........ 6 94 1 0 Deo.-12, 1916. .:...... 6+| 94—| 6.6 1 + +++ |745.0 | 45.6 Dee. 19, 1016......... 5+| 94+) 5.63 1 0 ++-+ |760.0 | 40.66 Bree, ae, 1916.2... .... 5 95 5.0 1 0 |++++)|764.1 | 35.6 Various times........ 5 | 95 3 0 |} +++ Various times........ 4 96 4 0 8 2,4 Various times........ 0 | 100 0.40*| 50 eat ee on ee *These figures represent but one determination made in the series but since the other experimental conditions were constant they are representative of the series. of the latter is passed just underneath the surface of water contained in a small dish. After the air in the chamber has become entirely replaced by the gas to be tested and the medium of the culture may be assumed to have become saturated with this gas at atmospheric pressure, the culture chambers are sealed by fusing the small constricted portions of the tube, figure 1, b. The cultures are left in the incubator and 20 MONTROSE T. BURROWS examined periodically under the microscope, which is fitted with a warm box. The maximum growth attained is recorded. The measuring cylinder used in these experiments is a large one and the general method of measuring the gases is not very accurate. For this reason the measurements were in several instances checked by removing a sample of the gas and determining its oxygen content by the Haldane method. The pure oxygen and nitrogen gases were. also tested by the same method (see table).! It is also evident that in all these cultures another slight error must have resulted from sealing the tube and temporarily heating the air in the tube. This would make the actual pressure somewhat less than that recorded. DISCUSSION AND CONCLUSIONS Results of the experiments are given in the accompanying table. Tissues of chick embryos grow somewhat better in the spring and sum- mer than in the winter. In the first column the date of the experiments is recorded. ‘The growth is recorded in terms of the area of new cells which form about the fragment. Since there is no absolutely accurate method of measuring this, it is indicated by the relative term,+. In many cultures the cells grow in a single plane, while in others they _ grow in several planes and the density of the growth is not the same in one culture as in another. As indicated by the corrected readings of the oxygen percentage in the gaseous mixtures, column 4 of the table, many of the readings in — column 3 are probably slightly higher than the table indicates. Including, however, these possibilities of inaccuracy in the measure- ments it seems quite evident that certain definite conclusions may be drawn: 1. Cells may grow in an atmosphere of pure oxygen.’ 1The author is indebted to Mr. H. L. Higgins of the Department of Pedistries for making these determinations. * During the period of the development of the method the ecita were found not to grow in pure oxygen. The gas attacked the rubber tubing used in con- necting the cultures. The medium in the culture became orange red in color. This gas gave a definite test for ozone. It was not until the gas had been passed through olive oil that the results reported in the paper were obtained. Whether the ozone attacks the culture medium, liberating toxic substance, or whether it attacks the cells, was not determined. From the disintegrating rubber tubing SO. was liberated. The failure of growth may have been due to the presence of SOx. OXYGEN PRESSURE AND TISSUE ACTIVITY 21 2. The growth in an atmosphere of pure oxygen, although often slightly more rapid, is not_greater than the growth in a partial pres- sure of oxygen-no-more than 9 per cent or 10 per cent. 3. Although the growth becomes less when the partial pressure of oxygen is lower than 9 per cent or 10 per cent, very evident growth activity is seen in an atmosphere where the partial pressure of the oxygen is as low as 45.6 mm. Hg. 4. A method has been devised which in its full analysis will allow one to study in more detail many of those conditions which regulate oxidation in tissue cells and to study conditions which regulate oxygen pressure in the tissues. These results have become of interest in that they show that the activity of the cells within the cultures is little influenced by changes in the oxygen concentration or partial pressure when it remains above a certain amount. . ; Again, they are interesting in that the lowest partial pressure in which growth took place is closely related to the venous oxgygen ten- sion of mammals. The author does not know the venous oxygen tension of chickens. In the light of these facts an attempt is now being made to determine the venous oxygen tension of chickens, to study the effect of the addition of various substances to the plasma, and to measure with greater accuracy the lowest partial pressure of oxygen in which the cells may not only grow but show other forms of activity. BIBLIOGRAPHY (1) Burrows: Journ. of Exper. Zoél., 1911, x, 66. (2) Crows: Proc. Royal Soc., London, 1894, lvi, 2. (3) LorvenHarT: Harvey Lectures, 1914-15, 98. (4) Hempexi: Methods of gas analysis (trans. and enlarged by L. M. Dennis), New York and London, 1906. THE PHYSIOLOGY OF THE ATRIO-VENTRICULAR CONNECTION IN THE TURTLE III. Tue INFLUENCE OF THE VAGI AND OF THE SYMPATHETIC NERVES on ITs RuyTHM-FoRMING POWER C. C. GAULT From the Osborn Zoological Laboratory, Yale University Reeeived for publication January 18, 1917 INTRODUCTION In a recent paper Laurens (1) has described the results of experi- ments on the rhythm-forming power of the atrio-ventricular connection of the turtle, Malacoclemmys geographica, when the connection is stimu- - lated electrically. Haberlandt has recently published a series of arti- cles dealing with funnel rhythm in the frog (2, 3 and 4) and in the last of these he includes experiments on the influence of the vagus on the: production of funnel rhythm in the frog and turtle. He found that by the stimulation of the vago-sympathetic trunk the capacity of the A-V funnel to form automatic impulses, giving rise to fibrillation and to high frequent V contractions, is increased, so that long-lasting after effects are obtained which are not obtained by funnel stimulation alone. Atropin he found not to decrease this action of the vago-sympathetic trunk and concludes, therefore, that the effect must be due, in part at least, to the action of sympathetic fibers. Haberlandt draws a close comparison between his results and those that have been obtained on the production of ‘‘nodal rhythm” in the mammalian heart. Laurens (1) found that it was impossible to induce by electrical stimu- lation of the funnel the automatic formation of rhythmic impulses in the intact heart of Malacoclemmys. It was his intention to carry this study of the rhythm-forming power of the funnel further by investi- gating the effects of vagal and sympathetic stimulation on its produc- tion and to attempt to see whether some difference between the nerves of the right and left sides could be demonstrated. The carrying out of this problem entailed an examination, more or less thorough, of the general influences of the vagus and sympathetic nerves on the tur- 22 VAGI AND SYMPATHETICS ON A-V CONNECTION 23 tle heart, involving the repetition of much work that has been done al- ready. It was necessary in the first place to find out whether the right and left vagus nerves could be shown to have a control over different parts and different functions of cardiac muscle—owing to the differen- tial effects that have been described both for the turtle heart and the heart of the mammal—and whether any influence of the sympathetic nerves could be demonstrated. It seems best to consider briefly at this time the results that have been obtained by others. HISTORICAL Location of the pace-maker. Muskens (5) by suspending two parts of. the sinus showed that the sinus and large veins of the turtle (Pseu- demys rugosa and elegans) are a system of contractile units, the con- traction wave, in most cases, being seeri to start from the right vena cava. He also observed antiperistaltic contractions of the sinus and large veins in exposed hearts that were beating normally. Garrey (6) concluded from inspection of the turtle heart that the beat arises at the junction of the right pre- and post-caval veins, the sinus contracting after this region. He claims that the right caval veins possess a greater rhythmicity than the other parts of the heart and determine the rate of the whole heart. Meek and Eyster (7) have shown that the origin of the beat in the turtle heart is in the sinus and in a definite part of this, namely, the sino-auricular ring, somewhere to the right of the left venous valve (8), probably near the right venous valve, where the best connection be- tween the sinus and the right auricle is found (9). Schlomovitz and Chase (10) also find the pace-maker to be in the right sino-auricular junction. Action of vagus and sympathetic. The vagus has an inhibitory effect on the rate and strength of beat and on the conductivity and excita- bility of cardiac muscle. The effect of the sympathetic nerves is oppo- site to that of the vagus on rate and strength of beat and on conduc- tivity. There is some doubt as to the action of the sympathetics on excitability. Recent work has indicated that, due to a difference in distribution, the nerves of the right and left sides do not exert their effects in equal degree. Garrey has most recently investigated this matter in the turtle. That the right and left nerves were not equally effective in stopping or slowing the turtle heart had been repeatedly observed by previous investigators. According to Garrey (6 and 11) there is a preponderant inhibitory 24 Cc. C. GAULT action of each vagus upon the corresponding half of the heart, a homo- lateral distribution and function. The action of the right vagus is mainly negatively chronotropic and it has a preponderant action in stopping the heart, due to the fact that its effect is directly upon the right caval veins, where, as Garrey believes, rhythm is initiated. In many cases, in certain selected individuals, the left vagus was unable to affect the rhythm, but, by decreasing excitability, conductivity and contractility, blocked the impulse, thus producing or increasing S—A block. In cases of A-V block where the left vagus had no chronotropic influence, the effect was always to increase the degree of block, stop- ping or slowing the V, the A rate being unchanged. The right vagus, on the other hand, decreased the block, owing to its greater chrono- tropic effect. He obtained this effect also when the intracranial vagus was stimulated, thus showing that it was not due to sympathetic stimulation. ) . Greene and Peeler (12) agree with Garrey in so far as they found, in central cardiac inhibition, that when the left vagus was sectioned there was no change in the rhythm of the heart or in the state of inhibition, but that when the right vagus was cut, the heart resumed its normal rhythm. Garrey pointed out that the depressing effect of the vagi upon con- ductivity and strength of contraction is sometimes wholly obliterated by the coincident slowing of rate. Dale and Mines (13) and Mines (14) have recently shown that in the frog the action of the vagus is primarily to produce increased resistance to transmission from A to V, decreasing the rate, (and to shorten the duration of the electrical disturbance in © the V), while the action of the accelerators is to improve the rate of conduction between A and V (and to increase the duration of the elec- trical response in the V). These authors make no mention of a differ- ence in degree between the action of the right and left nerves. There is considerable evidence that in the mammalian heart there is a division of labor between the right and left nerves. Cohn (15 and 16) found a great qualitative difference in the action of the two vagi on the heart of the dog, the right controlling principally the Si node, and by its negative chronotropic influence the beat of all chambers, while the left has its greatest influence on the conduction of impulses over the A-V connection, producing either a delay, or incomplete block or complete cessation of V contractions. The left vagus may also have some influence over the A, and may rarely cause S—A block. Cohn and Lewis (17) showed that the left vagus has more effect on the A-V VAGI AND SYMPATHETICS ON A-V CONNECTION 25 junction than the right has, while the right effects the production of - impulses in the A more than the left does. Lewis (18) has investigated the matter further and his results show that it is not so simple as at first _ might seem to be the case. He admits that S-A rhythm is more read- ily inhibited by the right than by the left vagus, but questions the view that the A—V connection is influenced chiefly by the left vagus. He finds that, while the left vagus may have more influence than the right ‘in producing heart block, the right has a larger control than the left on “nodal rhythm” and that this control is even greater than the control of the right nerve over S-A rhythm. This latter in turn is about the same as the control of the left nerve over “nodal rhythm,” while the control of the left nerve over Si rhythm is weakest of all. In brief the right has a greater control over both nodes, and the control of both yagi is greater over the A—V node than over the S-A node. Rothberger and Winterberg (19 and 20) have reported that in the dog the right vagus and sympathetic influence particularly the S-A node, the left nerves particularly the A-V node. There are numerous exceptions due to the presence of fibers which go in each case to the other node, particularly in the left accelerator the presence of fibers which have a chronotropic influence on the S—A node, so that all parts of the heart are under the excitatory influence of the accelerators, the influence being strongest on the S—A node, the right increasing the rate more than the left (20). They were able to show, moreover, that the accelerators innervate their own side of the heart more or less sepa- rately, for in many cases of combined right stimulation, right extra- systoles appeared, and of combined left stimulation, left extrasystoles appeared, these breaking through the vagal standstill. In connection with other work on the production of ‘‘nodal rhythm” they found it - impossible to isolate the inhibitory chronotropic fibers in the vagi which go to both the Si and A-V node because they are mixed, and that it is only exceptionally possible to demonstrate an inhibitory effect of the vagus on the A—V node, moreover that it is only exceptionally that one finds the negative chronotropic fibers for the S—A node exclusively in the right vagus. Sometimes also the right vagus produced complete standstill, while the left produced merely slowing of the A and dropping out of V contractions, thus showing that it was distributed to both nodes. Ganter and Zahn (21) also report that the two vagi influence nodal tissue in unlike degree, the right being stronger than the left on stimu- lus formation in the Si, and the left being stronger than the right on 26 Cc. C. GAULT - the A—-V node, particularly in its effect on lowering conductivity. The distinction is by no means always a sharp one, the left vagus usually having a slight negative chronotropic effect as well. In a still later paper Rothberger and Winterberg (22) give further evidence regarding the distribution of the right and left nerves. In cases of A flutter they found that strong accelerator stimulation in- ~ ereased the strength and the rate of contraction, the right having the greater effect, although the duration of A flutter and fibrillation is in- creased. The rate of the V contractions is increased, owing to im- proved conduction. The vagi also, when strongly stimulated, increase the rate of the A contractions, the effect of the right being stronger, and just so much stronger, as the left vagus has less of an inhibitory effect. By decreasing conductivity the rate of the V contractions is decreased. Robinson (23 and 24) reports that the left vagus, just as it does not inhibit rhythm, does not inhibit A tachycardia and may only possibly inhibit A fibrillation. The right vagus has no influence on A fibrilla- tion, but inhibits the A tachycardia which is coexistent with the A fibrillation, and changes the flutter into a fibrillation. The right vagus is the more effective in increasing the susceptibility of the auricles to faradisation, and in holding them in the abnormal activity. In some hearts vagal stimulation alone can initiate the same abnormal A activ- ity cause! by A faradisation. The left vagus is more effective than the right in replacing the abnormal auricular activity by the normal sequen- tial beat. ae The influence of the cardiac nerves on the production of nodal, or A—V. rhythm. ‘There are several references in the literature to this matter. - These have to do particularly with the vagus nerve and Haberlandt in his papers has pretty thoroughly reviewed them. Rothberger and Winterberg also give a brief review and the last mentioned authors have most carefully and recently investigated this matter. Winterberg (25) found that the accelerators have no influence onthe origin or duration of V fibrillation and may shorten A fibrillation, while the vagi assist in the production of A and V fibrillation. Rothberger and Winterberg (19 and 20) have shown that stimulation of the right accelerator cannot produce “nodal rhythm,” while the left accelerator, which has only a slight chronotropic influence on the S-A node, did so in 30 per cent of their experiments, stimulation of the right accelerator causing this to disappear. The failure of the left accel- erator to cause A—-V rhythm is referred to the admixture, already men- VAGI AND SYMPATHETICS ON A-V CONNECTION 27 tioned above, of fibers which go to the S-A node. The stimulation of the vagus alone never produced V fibrillation; combined with the accel- erator it did-so-in-10 per cent of their experiments. Owing to the preponderant distribution, in certain cases, of the right and left vagi and sympathetics to the S-A node and A-V node respec- tively, they thought that combined stimulation of the left accelerator and the right vagus should give ‘‘nodal rhythm.” Their early experi- ments (19) were negative. Later (20) by combining such stimulation, in those cases where the inhibitory fibers for the Si are practically only in the right and those for the A—V node in the left, they obtained results in two experiments. In a third (p. 361), combined stimulation of the left accelerator and either vagus produced “nodal rhythm,” thus indi- cating an equally intense inhibitory effect of the two vagi. They also found in another case (p. 362) that the stimulation not only of the left accelerator, but of the right as well, produced in combination with stimu- lation of the right vagus a ‘‘nodal rhythm.” These various effects of the cardiac nerves bring up for consideration the question as to whether they really have any direct influence on the ventricle. Hering (26 and 27) and his collaborators hold that they do; Erlanger (28) does not believe that the vagi have any significant chro- notropic influence on the ventricles and that the A-V bundle contains no inhibitory fibers for the ventricle. Haberlandt (3) has recently described both positive and negative chronotropic and inotropic influences of the vagi on the slow automatic V contractions of the frog. He compares his results on the inotropic effect with those which Gaskell (29, p. 85) obtained, in one case, on inhibiting the independent auriculo-venticular rhythm through the | coronary nerve by stimulating the right vagus. The positive influences which Haberlandt obtained would seem to be due to the excitation of sympathetic fibers in the vago-sympathetic trunk. Tonus. It is conceivable that the rhythmical variations in tone of the turtle auricle may be influenced by the possible differentiation in function of the right and left nerves. The influence of the nerves on these tonic oscillations has been investigated by a number of workers, among them being Fano-and Fayod (30), Bottazzi (31 and 32) and Oinurha (33), with the general result that the vagus increases the tone, while the sympathetic depresses or inhibits it. Gaskell (34) was of the opinion that the vagus nerve did not decrease the state of tonic contraction of the A, although he obtained a positive variation when 28 Cc. Cc. GAULT the vagus nerve was stimulated, as has recently been substantiated by Meek and Eyster (85) and Samojloff (86). The supposed cause of the rhythmic variations in A tone is worthy of brief notice, particularly so since this matter has recently been given extensive consideration by Gesell (37 and 38). Rosenzweig (39), who found that vagal stimulation did not increase the tonus, came to the conclusion that the rhythmical variations in tone were due to the grad- ual death of the preparation. He suggested that they might be caused by the contraction of the smooth muscle cells lining the auricular cavity, an assumption which Bottazzi (32) confirmed, thus giving up his sarcoplasm contraction theory. Oinuma (33) agrees with this. The anatomy of the sympathetics and vagi. The course of the sym- pathetic nerve and its innervation of the heart has been described for various species of turtles by various authors. According to Gaskell and Gadow (40) the rami cardiaci come chiefly from the first thoracic ganglion. In Testudo graeca they are sent off from the middle cervical ganglion and enter the heart, together with the vagus fibers, between the pulmonary vessels and the vena cava. In Chelone imbricata the ramus cardiacus from the middle ganglion enters the heart along the aorta, running close with the vagus of its side. In Emys europaea rami cardiaci likewise arise from the median ganglion and’ perhaps anastomose with the cardiac branches of the vagus. It has been shown by Bottazzi (31) that cardiac branches of the sympathetic in E. europaea arise from the middle and inferior ganglia, but chiefly- from the latter. Oinuma (33) agrees with this. In Emys caspica Kazem- Beck (41) describes the main ramus cardiacus as arising in the first . thoracic ganglion and running along the superior vena cava to the heart, although cardiac branches are also sometimes given off from the inferior and middle ganglia. According to Wesley Mills (42) the rami cardiaci in Pseudemys rugosa arise from the first thoracic and the middle ganglia, the former branches appearing to be the more important. Ranson (43) has recently studied the vagus in Chelydra serpentina. In this animal the sympathetic and the thoraco-abdom- inal branch of the vagus run together into the thoraco-abdominal cavity. The sympathetic leaves the vagus opposite one of the lowest cervical vertebrae and turns toward the inferior cervical ganglion of the sympathetic. At the point at which the two nerves separate there is developed the middle cervical sympathetic ganglion and the gan- glion thoraco-abdominale vagi. a er VAGI AND SYMPATHETICS ON A-V CONNECTION 29 "ANATOMICAL A description of the sympathetic in Malacoclemmys geographica will not be out-of place since it differs in some details from the conditions which have been described in other Chelonians. cervicale superior situated just under the base of the skull (fig. 1) a small branch passes anteriorly through the jugular foramen into the skull along with the vagus and accessory nerves. The cer- vical sympathetic arises from the posterior side of the ganglion and passes caudalwards with the vagus, just under the omohyoid muscle and median to the carotid artery. The two nerves run separately down the neck, and do not unite to form a common vago-sympathetic trunk nor have connections between them been observed. The sympathetic receives no rami communi- cantes from the first five spinal nerves. At the level of the fifth to sixth cervical vertebrae the sympathetic enters the ganglion ‘cervicale me- dium. A small nerve may sometimes be seen connecting this ganglion with the vagus, but its presence is not typical. From the ganglion cer- vicale medium the sympathetic runs towards - the median line and at the level of the seventh to eighth cervical vertebrae enters the ganglion cervicale inferior, which receives a ramus com- municans from the eighth spinal nerve. It is continued downward to the first thoracic gang- lion (Th. G) giving off a small branch to the plexus brachialis as it passes. Gaskell and Gadow have found considerable variation and fusion of two or more of the ganglia in this region, corresponding to the stellate ganglion in ‘mammals. That the main rami cardiaci arise in or pass through the first thoracic ganglion is corrobo- rated by experiments. Gaskell and Gadow (40) have shown that stimulation of the sympathetic chain between the first thoracic and the middle cervical ganglion, or of the ramus cardiacus, From the ganglion a Fig. 1. Ventral view of neck of Malacoclem- mys showing the course of the vagus and of the sympathetic nerve of the right side. G.C.S., Ganglion cervicale su- perior;G.C.M., Gangli- on cervicale medium; G.C.I., Ganglion cer- vicale inferior; T.G., Ganglion thoracico-ab- dominale vagi; Th.G., first thoracic ganglion; C.R.S., cardiac ramus of the sympathetic; C.R.V., cardiac ramus of the vagus. 30 Cc. C. GAULT in Testudo graeca “causes a most marked acceleration of the heart and augmentation of the auricular contractions.” Kazem-Beck (41) pointed out that in Emys caspica stimulation of the ramus cardiacus from the first thoracic ganglion produces an acceleration of six to eight beats per minute following the stimulation. Dogiel and Arch- angelsky (44) found that in Emys europaea stimulation of the inferior ganglion caused an acceleration of about 4 beats, while stimulation of the first thoracic ganglion with a current of the same strength pro- duced an acceleration of about 8 beats per minute. No mention is made by the last two authors of an augmentation of the A contrac- tions. Bottazzi (31) has shown in E. europaea that stimulation of the sympathetic chain between the middle ganglion and the first tho- racic ganglion produces an augmentation of the height of contraction as well as an acceleration. According to Greene and Peeler (12) the augmentatory apparatus of the turtle is poorly developed, if not in fact absent. Garrey (11) demonstrated what he considers an in- direct augmentatory effect in cases of A—-V block in the turtle after vagal stimulation, in that the block disappears for a while. Also that intraventricular block was improved after stimulation of the cervical vagus, but not of the intracranial vagus. EXPERIMENTAL Since Laurens (1) found that in the intact heart of Malacoclemmys a funnel rhythm could not be initiated by electrical stimulation of the A-V connection, it seemed interesting enough to see what the influence of stimulating the cardiac nerves at the same time would be. As has already been mentioned above, Haberlandt found that stim- ulation of the vago-sympathetic trunk in the frog and turtle had a beneficial influence on the formation of funnel rhythm. Since the work of Rothberger and Winterberg has shown that the vagus when stimulated has no ability to produce ‘‘nodal rhythm,”—whatever effect these nerves may have on such rhythm after it has been produced,— and that the stimulation of the left stellate ganglion could start “nodal rhythm,” particularly when the vagus was also stimulated, it seemed that this matter might not be so simple as Haberlandt implies. The investigation, the results of which are now to be reported, was under- taken at the suggestion of Dr. Henry Laurens. My sincere thanks are due him for his assistance and helpful criticism at all times. A preliminary report of the results has already appeared (45). VAGI AND SYMPATHETICS ON A—V CONNECTION 31 METHODS The turtle Malacoclemmys geographica was used in all of the ex- periments. The animal was rendered insensible by decerebration, anaesthetization, or by destroying the brain and spinal cord, decere- bration proving the most satisfactory and therefore most generally employed. The plastron was removed and the fore limbs amputated after ligating the subclavian arteries. In all of the experiments care was taken to reduce the loss of blood to a minimum and to keep the circulation intact as much as possible. After the pericardium had been slit up the heart was fastened to a slender strip of cork by two small porcupine quills and suspended for mechanical registration from the right auricle and the apex of the ventricle. The cardiac nerves were exposed and shield electrodes placed on the thoracico-abdominal gan- glion of the vagus and the first thoracic ganglion of the sympathetic, or between this and the inferior ganglion.' The conclusions reached in this paper are based on results obtained during the first two to three hours of experimentation, and before the strength of Si impulse had markedly decreased. Effect of sympathetic stimulation on the normal heart beat _ Stimulation of the sympathetic by relatively weak interrupted cur- rents produces no effect on the rate of beat. With stronger currents the action of the sympathetic becomes more apparent. Of 32 experi- ments, sympathetic stimulation produced a visible effect on the heart in 18. Of these, augmentation of the A contractions was produced in 10, augmentation and acceleration in 5, and acceleration alone in 3. The amount of augmentation, as calculated from the height of the registered contractions, was usually small (from 1 to 3 mm.), although an increase of as much as 6 mm. was not uncommon. The amount of acceleration varied from between 2-3 to 6 beats per minute. No difference between the effects of the right and left sympathetics could be noticed. The maximum effects became apparent fifteen to sixty seconds after the commencement of stimulation, though some- times not until after it had been discontinued, and lasted for longer or shorter periods of time. 1]n many cases, as observed by Gaskell and Gadow, the first thoracic and the inferior ganglion are fused. In such cases the electrodes were placed either on the ‘‘fusiform ganglion’”’ or between this and the middle ganglion. oe Cc. C. GAULT Effect of sympathetic and vagal stimulation on variations in A tone The action of the sympathetic in abnormal conditions, such as in- crease in auricular tone, A-V block, V fibrillation and V-A rhythm is more striking than the augmentatory and acceleratory influences on the normal beat. Bottazzi (82) observed in one case that the right sympathetic was more effective in reducing A tonus than the left was, but that in all other cases the effect of both was the same. In the intact heart of Malacoclemmys, variations in A tone rarely appear. In only one turtle used in this investigation did tonus oscil- lations spontaneously appear and then only in the left auricle. This animal had been in the laboratory aquarium for some time, and with the onset of cold weather had eaten little or no food and was in a very weakened condition, the heart beat being slow and weak. During the preparation for registration considerable blood was also lost. Stimulation of either of the vagi with interrupted currents too weak to stop the heart caused an increase in the tone of the left auricle, the left vagus apparently being the more effective. Stimulation of the right or left sympathetic during a period of increased tonus, whether automatic or set up by vagal stimulation, caused the tone to decrease, the left sympathetic apparently again being the more effective. Rosenzweig’s view regarding the appearance of variations in tone seems to be substantiated by the fact that in only one turtle did mark- ed variations in tone occur spontaneously. As is well known, and as Laurens has observed on the auricles of Malacoclemmys, marked varia- tions in tone can be produced by mechanical or electrical stimulation, or by flooding the auricles with cold Ringer, etc. Laurens (1 and unpublished observations) found that in experiments on excised hearts nearly every one showed marked oscillations in tone, both in the beat- ing and still (sinusless) heart. — Effect of sympathetic stimulation on A-V block Block was produced by the method used by Laurens (46) of cutting through parts of the A-V connection. If the part cut is small, either no block is caused or a very transitory one of 3:1 or 2:1 rhythm. By cutting through a sufficiently large portion of the funnel severe and long-lasting partial or total block can be produced. | Strong faradisation of the sympathetic nerves has been found to . improve the block. To describe a typieal case: After complete block VAGI AND SYMPATHETICS ON A—-V CONNECTION 33 had been produced (fig: 2) the right sympathetic was strongly stimu- lated. Soon after the stimulation was discontinued the V gave a single contraction, and then began to beat slowly as shown in the figure. By Phage He Hil | | | if Te HTT ventricle. Fig. 2. Improvement of A-V block by sympathetic stimulation. Time in this and succeeding figures in seconds. RPT ventricle . Fig. 3. Continuation of figure 2, showing continued improvement resulting from further sympathetic stimulation. repeated stimulations the block was further improved, the A and V finally beating in a 2:1 ratio (fig. 3), although the block could never be brought to disappear entirely. 34 Cc. C. GAULT The action of the vagus and sympathetic on funnel rhythm - To produce funnel rhythm essentially the same method as employed by Laurens (1) was used. Two fine needle electrodes connected with the secondary of a Harvard inductorium were inserted into the- V jugt below the A—V boundary so that they penetrated the A-V funnel. If ‘strong interrupted currents are sent in fibrillation of the V is pro- duced which in most cases spreads to the auricles. The. fibrillation, however, does not persist after the current is broken, and is often in- terrupted by the normal sinus rhythm during the stimulation. Stim- ulation of the funnel sometimes produces a rapid V rhythm which also does not persist after the stimulation (fig. 4, B), Sirhythm reappearing almost immediately after the stimulation is discontinued. If the vagus is stimulated with relatively strong interrupted cur- | rents at the same time, that the funnel is; stimulated, there is produced, in a few cases, V fibriftxdion, which, sometimes changing into a regular but rapid V—A rhythm, lasts over for varying lengths of time after the stimulation has. been discontinued (figs. 5, 6, 7 and 8). In order to attain this result it has~ been found that the vagal stimulation must be of sufficient strength to completely inhibit the’ Si impulses, valthough, of course, even when this is so, automatic funnel impulses are not always set up. All cases of the rhythmic production of auto- matic funnel impulses were obtained'by stimulation of the funnel and the right vagus. Conjoint stimulation of the funnel and the left vagus causes ventricular fibrillation during the stimulation as with right stimulation, but apparently the block produced by the left vagus is not Binion strong to prevent the reappearance of the normal sinus rhythm as soon as the stimulation ceases (see fig. 4, A). This may be due to the fact that the left vagus principally affects conduction between the A and the V and that the Si impulses, not having been weakened to any great extent, break through to the V. If the vagus is stimulated during this funnel rhythm with a strength of current that would ordinarily stop the heart, it produces merely ‘a decrease in the amplitude of the A contractions with no effect on the rate. Stimulation of either of the sympathetics, on the other hand stops the funnel rhythm, after which the normal Si rhythm reappears, the right and left sympathetics apparently being equally effective (figs. 5 and 6). Conjoint stimulation of the funnel and of either sympathetic never produced funnel rhythm (see fig. 4, C).. Nor did conjoint stimulation CONNECTION Y. pcod'E SYMPATHETICS ON AND VAGI UoIzZeINWUIYyG ‘_ ‘orjoyzedurds YYSIA puw jouuN; Jo uoryenunys yutofuog ‘9 ‘SNA 4joy puv f ‘ jeuuny jo UOT} RINUITYS qurofuog V ‘ur Aqa ‘9uo[’ JoUUNy Jo jouunj oonpoid 04 sone “Pp ‘Sy a jauun} RU A way splwa oa @[O144UdA 36 Cc. C. GAULT of either of the vagi and one or the other of the sympathetics result in the production of funnel rhythm. This latter subject is, however, worthy of a more detailed investigation than has. been made. é hin Vina Nhnn Nan fla ventricle #f {vm Wy Wl, An WV i vey funnel Fig. 5. Conjoint stimulation of funnel and right vagus resulting in the pro- duction of V fibrillation followed by V-A rhythm with partial block. Stimu- lation of the right sympathetic stops the funnel rhythm. Fig. 6. Conjoint stimulation of funnel and right vagus resulting in V fibril- lation followed by V-A rhythm showing partial block. The V shows pulsus alternans. Stimulation of the right sympathetic stops the funnel rhythm. In a number of cases the funnel rhythm develops partial V—A block (see figs. 7 and 8, also 5 and 6). This block is characterised by the dropping out of A systoles after a gradually shortening V—A interval, evidenced most clearly by the gradually decreasing length of the suc- _ VAGI AND SYMPATHETICS ON A-V CONNECTION 37 cessive A-A intervals. The block, with the decrease in rapidity of production of the funnel impulses, eventually disappears. Laurens (1) has described such a case of-block in detail. In the cases which I have observed, either the impulses are arising in groups characterized by a 2) ventricle 5S bmn right vagus® cAsacAL EUs AMCaAMMMMRARASAAAANGS GAARA AMAR LAA Fig. 7. Funnel rhythm (rapid V-A) showing V-A block following conjoint stimulation of funnel and right vagus. ANa A i | | | fl) } || | Ht} i Hj | | TV HHI HHH aan | WA UY j PLJIWNUEL UYU YUN LY UUs lean fh VN NI ) Nn ry) Afi a yy hay Pe a Ee Fig. 8. Funnel rhythm, showing V—A block, following conjoint stimulation of funnel and right vagus. gradual quickening of impulse formation, or the conduction towards the A improves from beat to beat until the impulse falls into the re- fractory period and an A systole is missed. Often there is a slight pause before the last As, which is then more vigorous (fig. 5). 38 ¥ Cc. C. GAULT The impulses for this V-A rhythm originate in the A-V funnel, probably below the level of the A-V boundary. They therefore spread in both directions, going to the V and upward to the A. There is no measurable variation in the lengths of the V—V intervals, and we are therefore obliged to admit that the impulses arise at regular intervals, but that they are conducted upward with increasing ease, until an impulse falls into the refractory period. The fact that in many cases there is a pause before the last As, which is then more vigorous, would seem to indicate that, in some cases at least, the dropping out of the As was due to a refractoriness of conduction. THEORETICAL CONSIDERATIONS. The results of my experiments have given no reason for believing that the distribution of the two sympathetic nerves is in any wise different, and have given only indirect evidence for the vagi. A series of observations on comparing the effects of right and left vagal stimu- lation on the right and left auricles respectively seeméd to indicate that it is only extremely seldom that there is a differencé between the influence of the two nerves... In cases of A—V block, caused by cutting through parts of the A~V connection, stimulation of the sympathetic results in a partial removal of the block, decreasing it by improving the conductivity of the A~V connection as well as increasing the strength of the Si impulse. | In the. case illustrated (figs. 2 and 3), which is t¥pical, the contractions. of the V are at first at long intervals, and with each succeeding stimulation become more and more numerous, indicating a decrease in the degree of block which is dependent for the most part upon an improvement in conducting power of the remaining fibers of the A-V connection. It seems certain when the A-V connection is cut that the irritability of the V is not reduced, but that the A impulse which is able to reach it over the intact but injured funnel fibers is decreased, so that it is subliminal and the V therefore fails to respond. The irritability of the V, owing to its inactivity, increases until the weakened A impulse getting through proves strong enough to elicit a contraction, after which the irritability of the V again falls to zero, and one or more V systoles are missed (partial block), or the impulse, although the ir- ritability of the V is maximal, is always subliminal and the block is. complete. It has been shown above that sympathetic stimulation in the turtle causes an increase in strength of the sinus impulse, which coupled with eS VAGI AND SYMPATHETICS ON A-—V CONNECTION 39 the increase in-contractile power and the improvement in conduction causes an increase in strength of the A contractions. Owing to all of these effects of the sympathetic, the impulse reaching the V is increased in strength to such an extent that it elicits a contraction. We have not sufficient evidence to warrant an assertion regarding the effect on: excitability. Owing to the increase in strength and the improvement -in conduction the irritability of the V does not have to rise:so high as it did before, in order to be excited to contract and it therefore con- tracts more often. In other words the strength of the impulse able to reach the V over the A~V connection is the important matter, and the strength of this impulse depends most on the conductivity of the connection fibers. Laurens (46)- has recently discussed block with particular reference to cases produced by cutting the A—V connection, and therefore injuring the fibers remaining so that their conductivity is reduced. The argument that most depends on the improvement in conductivity gains support here from two factors of sympathetic im- provement, namely, acceleration and+the fact that the augmentation of the Si impulses is not very great in amount or in duration. If the conductivity of the A—-V connection were not improved, an acceleration _ of the rate could have only the result of increasing the block. And if the conductivity were not improved permanently then the original degree of block would return soon after the stimulation ceased owing to the gradual decrease in the strength of the Si impulses. . The action of the vagus on funnel rhythm is due to its inhibiting effect on the normal pace-maker, or on the impulses originating there. Stimulation of the funnel in certain cases starts automatic impulse formation. If these are not interfered with they will continue for a longer or shorter time, arising automatically and rhythmically and initiating a beat of the heart which is characterised by a V—A sequence, owing to the fact that the V is nearer the point of origin of the impulses than the A is. The left vagus has been claimed to have, in certain cases, less of an influence on the normal pace-maker, in so far as the frequency of impulse formation there is concerned, and to have a much greater effect in reducing conductivity. It might be concluded from the failure of the left vagus, when stimulated conjointly with the funnel, to produce conditions favorable to the formation of a funnel rhythm that this was due to the fact that the Si impulses, weakened only slightly, were able to break through and reach the funnel and thus put a stop to the formation of impulses that might have started automatically to arise there. 40 Cc. C. GAULT Haberlandt, in certain cases, obtained funnel rhythm by stimula- tion of the vagus or of the funnel alone. This seems to have been in preparations which had previously been long and severely stimulated. It has also been observed in the present investigation that in old prep- arations the ability of the funnel to initiate rhythm is relatively in- creased, due to the weakened condition of the Si impulses. Haber- landt’s results of obtaining funnel rhythm when the vagus alone is stimulated offers substantiating evidence of the relative increase in the ability of the funnel to form automatic and rhythmic impulses when the Si impulses are weakened. It speaks for the function of the funnel rhythm in cases where the Si impulses are not able to in- itiate the rhythm of the heart. It has also been observed during the course of the present investigation that as has often been observed before, the funnel rhythm may arise automatically in hearts which have been exposed for some time. Furthermore Laurens (1) has shown that in the still (sinusless) heart funnel rhythm is more frequently and easily accomplished than in the excisedand beating heart. The stopping of funnel rhythm by sympathetic stimulation (see figs. 5 and 6), falls in line with the explanation given above of the action of the sympathetic in improving A-V block, and the reverse of that just given for the influence of the vagus on producing funnel rhythm. As mentioned above the sympathetics stimulated conjointly with the funnel, or conjointly with the vagi, never led to the production of fun- nel rhythm. So far as our evidence goes, it points to an exclusive dis- tribution of sympathetic fibers, so far as impulse formation is concerned, to the normal pacemaker, or sinus. By their stimulation the strength of the normal Si impulse is increased. Conductivity being also in- creased, the normal Si impulses come to the V with increased vigor, thus blotting out any attempts on the part of the funnel to automati- cally form rhythmic impulses, or when these impulses have been formed stopping them, and assuming again the initiation of the heart beat. In this investigation sympathetic stimulation failed but once to re- verse the sequential beat to normal. It seems therefore that, in the turtle, vagal stimulation aids in the formation of an automatic funnel rhythm by inhibiting the Si impulses, either by decreasing them or decreasing the conductivity of the muscle. Sympathetic stimulation, by increasing the strength of the Si impulses and by improving conduction, prevents the formation of an automatic funnel rhythm by swamping with Si impulses the impulses which might begin to arise automatically in the funnel. Vagal stimulation does not VAGI AND SYMPATHETICS ON A-V CONNECTION 4l affect funnel rhythm after it has gotten started, but, by decreasing the conductivity and the contractility of the A muscle, decreases the strength of the A contractions. Sympathetic stimulation causes a re- versal of funnel rhythm to a normal Si rhythm by its influence on the strength of the Si impulse and on the conductivity of cardiac muscle. No very close comparison regarding the differential influence of the right and left nerves on ‘nodal rhythm”’ can, therefore, be drawn be- tween the results that have been obtained on the mammalian heart and those obtained on the turtle heart. SUMMARY 1. In Malacoclemmys geographica, the vagus and sympathetic nerves run separately in the neck and are not united to pee a com- mon vago-sympathetic trunk. 2. The sympathetic rami cardiaci come principally from the first thoracic ganglion. 3. Stimulation of the sympathetic results generally in an increase in strength of Si impulses, and an acceleration of rate of production, the former being more often obtained. 4. Stimulation of the vagus causes an increase in auricular tone. Stimulation of the sympathetic decreases and abolishes it. (Variations in tone obtained in but one case.) 5. Stimulation of the sympathetic, by improving conduction, de- creases heart block produced by cutting the A—V connection. _ 6. Stimulation, in the intact heart, of the A~V funnel by strong in- terrupted currents causes V fibrillation, or rapid V rhythm, which stops when the stimulation is stopped. 7. Conjoint stimulation of the right vagus and of the A—-V funnel with strong interrupted currents sometimes produces a funnel rhythm which lasts over, in different experiments, for varying lengths of time, after the stimulation has been discontinued. 8. Vagal stimulation during funnel rhythm with a strength of cur- rent sufficient to normally still the heart decreases the height of the A contractions but does not affect the rate. Stimulation of the sympa- thetic, by increasing the strength of the Si impulses and improving conduction, stops funnel rhythm, after which the normal sequential beat returns. 9. Conjoint stimulation of the funnel and sympathetic does not pro- duce funnel rhythm, nor does conjoint stimulation of either sympa- thetic and one of the vagi do so. 42 Cc. Cc. GAULT BIBLIOGRAPHY (1) Laurens: This Journal, 1917, xlii, 513. (2) HaBeruanpT: Zeitschr. f. Biol., 1913, lxi, 1 (3) HaBerRLANpT: Zeitschr. f. Biol., 1914, Lxiii, 305. (4) HaBeruanpT: Zeitschr. f. Biol., 1915, Ixv, 225. (5) Musxens: This Journal, 1898, i, 486. (6) Garrey: This Journal, 1911, xxviii, 330. (7) Merk anv Eyster: This Journal, 1912, xxxi, 31. (8) Merk anv Eyster: This Journal, 1916, xxxix, 291. (9) Laurens: Anat. Rec., 1915, ix, 427. (10) ScHLOMOVITZ AND CHASE: This Journal, 1916, oe 112. (11) Garrey: This Journal, 1912, xxx, 451. (12) GREENE AND PEELER: Journ. Pharm. a Exper. Therap, 1916, vii, 591. (13) Date anp Mines: Journ. Physiol., 1918, xlvi, 319. (14) ‘Minus: Journ. Physiol., 1914, xlvii, 419. (15) Conn: Journ. Exper. Med., 1912, xv, 49. (16). Conn: Journ. Exper. Med., 1912, xvi, 732. (17) Conn AND Lewis: Journ. Exper Med., 1913, xviii, 739. (18) Lewis: Heart, 1914, v, 247. (19) se sligiewnataes uNnp WinTERBERG: Arch. f. d. gesammt. Physiol., 1910, CaRSY, 559. (20) Rowinwneea UND WINTERBERG: Arch. f. d. gesammt. Physiol., 1911, exit; 343, oe (21) GANTER UND ZAHN: ‘Arch; fa gesammt. Physiol., 1913, cliv, 492. (22) RoruperGEeR unD WinTERBERG: Arch. f. d. gesammt. Physiol., 1914, elx, 42 (23) Roptnson: Journ. Exper. Med., 1913, xvii, 429. . (24) Rostnson: Journ: Exper. Med., 1913, xviii, 704. (25) WInTeRBERG: Arch. f. d. gesammt. Physiol., 1907, cxvii, 293... (26) Herina: Arch. f. d. gesammt. Physiol., 1905, evii, 125. (27) Herinea: Arch. f. d. gesammt. Physiol., 1905, eviii, 281. ee (28) Ertaneur: Arch. f. d. gesammt. Physiol., 1909, exxvii, 77. (29) GasKELL: Journ. Physiol., 1883, iv, 43. : (30) Fano er Fayop: Arch. Ital. de Biol., 1888, ix, 143. (31) Borrazzi: Arch. Ital. de Biol., 1900, xxxiv, 17. (32) Borrazzi: Zeitschr. f. Allg. Physiol., 1907, vi, 140. (33) Ornuma: Arch. f. d. gesammt. Physiol., 1910, exxxiii, 500. (34) GasKELL: Journ. Physiol., 1887, viii, 404. (35) Mrex anp Eysrer: This Journal, 1912, xxx, 271. (36) Samostorr: Zentralbl. f. Physiol., 1913, xxvii, 575. (37) GesExL: This Journal, 1915, xxxviii, 404. ‘ (38) GrseLu: This Journal, 1916, xxxix, 239. (39) Rosrnzwete: Arch. f. Physiol., 1903, 192. (40) GasKEeLL anp Gapow: Journ. Physiol., 1884, v, 362. (41) Kazem-Bucx: Arch. f. Anat., 1888, 325. (42) Miuus: Journ. Physiol., 1885, vi, 246. (43) Ranson: Journ. Comp. Neur., 1915, xxv, 301. (44) DoatrL uND ARCHANGELSKY: Aryeh: f. d. gesammt. Physiol., 1906, exiii, 1. (45) Laurens AND Gautt: Proc. Soc. Exper. Biol. and Med., 1916, xiii, 182. (46) Laurens: This Journal, 1916, xlii, 89. ” a ee THE CONDITIONS DETERMINING THE RATE OF EN- TRANCE OF WATER INTO FERTILIZED AND UNFERTIL- IZED ARBACIA EGGS, AND THE GENERAL RELATION OF - CHAN = OF PERMEABILITY TO ACTIVATION RALPH S. LILLIE ‘From the Marine Biological Laboratory, Woods Hole, and the Department of Biology, Clark University Received for publication January 19, 1917° ‘In his recent book, “The Organism as a Whole,”’ Professor Loeb raises an apparently serious objection! to my interpretation of the observation (recently made by me) that in dilute sea-water fertilized Arbacia eggs take up water by osmosis several times more rapidly than unfertilized eggs.2 This phenomenon had seemed to me to in- dicate that the permeability of the egg-surface to water is decidedly increased by fertilization, and hence to confirm the view that a general increase of surface-permeability is an essential factor in the activation- _ process, a view which Professor Loeb appears to share; yet he dismisses the above observation in a footnote as unimportant, and suggests that the increased entrance of water after fertilization is due to the removal of the layer of viscous material or “jelly”? which normally invests the unfertilized egg and disappears at fertilization. Apparent- ly in his opinion it is this jelly—and not the relatively impermeable state of the egg-surface—which retards the entrance of water into the unfertilized egg; and,he concludes that no inference can be drawn from this phenomenon with regard to any change in the condition of the egg itself. An objection of this kind is best removed by experi- mental evidence, which is presented below. I may remark, however, that the observation. does not stand alone, but is supported by the results of a number of other investigators cited in my paper, all in- dicating that an increase of permeability is associated with fertiliza- tion. It seems therefore to me that this additional evidence ought not to be thus put aside on purely conjectural grounds; experiment 1 The organism as a whole from a physico-chemical viewpoint. Putnams, 1916, 122. ? This Journal, 1916, xl, 249. 43 44 RALPH 8S. LILLIE alone can decide in such cases; and the further observations which I have made during the past summer have shown definitely that the above objection has no basis; in point of fact the jelly does not inter- fere appreciably with the entrance of water into the egg. To avoid misunderstanding regarding the essential nature of the question at issue, it seems necessary, before proceeding with the dis- cussion, to distinguish clearly between two separate possibilities with reference to the relation which changes-of permeability may bear to the activation-process. The first possibility is that after fertilization, and as one of the secondary consequences of this process, the general conditions of permeability in the egg are permanently modified, and that the protoplasmic surface-layer or plasma-membrane from that time on remains more permeable and more subject to changes of per- meability than before; there is, in fact, considerable evidence that this is the case.2 The second possibility—which is the one suggested by the resemblance between the conditions of activation of the resting egg and those of stimulation in general—is that the primary event in the activation-process, as well as in the stimulation-process, is a tem- porary increase of permeability; upon this initial change follow the other changes expressive of the general response or activation of the egg-cell. A temporary or initiatory increase of permeability is thus to be distinguished from a permanent alteration in the general proper- ties of the plasma-membrane involving increased permeability. There is good reason to believe that both of these processes are concerned in the activation of the resting egg. This distinction, however, does not appear to have been made hitherto, and confusion may be avoided by recognizing it. In Professor Loeb’s earlier paper on the action of salt solutions upon fertilized and unfertilized eggs of Strongylocentrotus* he de- * My own observations show that the rate of entrance of water from dilute sea-water remains several times greater in fertilized than in unfertilized Arbacia eggs, at least until the second or third cleavage, and probably later. Similar con- ditions hold for the electrical conductivity of the fertilized eggs of Echinus, ac- cording to the recent observations of Gray. After an initial increase of conduc- tivity immediately following fertilization, the conductivity returns (within ten or fifteen minutes after fertilization) toward that of the unfertilized egg (Journ. Mar. Biol. Assoc., 1913, x, 50). Further investigation has shown, however, that this second change of conductivity is temporary and small compared with the original increase, and that the conductivity of the fertilized eggs always re- mains decidedly greater than that of the same eggs before fertilization (unpub- lished observations kindly communicated to me by Mr. Gray). ‘Biochem. Zeitschr., 1906, ii, 81. PERMEABILITY AND ACTIVATION IN ARBACIA EGGS 45 scribes experiments showing that pure solutions of NaCl are more toxic to fertilized than to unfertilized eggs; and he suggests, as a possible explanation of this fact, that fertilization may increase the permea- bility of the egg to NaCl or its ions; this explanation, however, he rejects in favor of one based upon the difference in the rate of oxida- tions in the two kinds of egg; the more rapidly metabolizing fertilized egg is more readily injured by any condition, such as lack of oxygen or presence of cyanide, which interferes with oxidations; and its greater susceptibility to injury by salt solutions probably has the same basis, since alkali is found to promote and cyanide to retard the toxic action. It is to be noted that these two explanations are in no sense incon- sistent with each other; an increased rate of oxidations may well be associated with, or even dependent upon, an increased permeability of the egg-surface, e.g., to oxygen. And in fact in a later publication Professor Loeb suggests that an increased permeability to-oxygen or _ OH-ions, resulting from the surface-alteration, may be a factor in the increase of oxidations following fertilization.6 This hypothesis is dif- ficult to test experimentally and remains conjecture, although, for reasons to be given below, it seems probable that this condition ac- tually does exist in. the sea-urchin egg. My own first experiments on the relation of permeability-change to activation were suggested by the idea of .a general parallelism between the conditions of activation in resting eggs: and of stimulation in ir- ritable tissues.: In stimulation there appears to be a temporary in- crease in the permeability of the plasma-membrane; it seemed therefore probable that a:change of the same kind might be the critical event in activation. If this be so, a parallelism should exist between the per- _ meability-increasing action of a substance and its. general effectiveness as an activating agent. This view is therefore supported by the fact, previously shown by Loeb, that cytolytic substances are in general good parthenogenetic agents; obviously such substances increase’ sur- face-permeability. My own earlier. experiments with Arenicola larvae had shown that pure isotonic solutions of Na and K salts increase the permeability of the pigment-containing body-cells and at the same time powerfully stimulate the musculature of these organisms; both effects are simultaneously prevented by the addition of CaClk.* It - seemed therefore probable, if an increase of permeability is the initial 5Chemische Entwicklungserregung des tierischen Eies, Berlin, 1909, 16, preface. 6 This Journal, 1909, xxiv, 14. 46 RALPH 8. LILLIE or critical factor in the activation of the resting egg, (1) that pure ~ salt solutions would cause activation, and (2) that this effect would be prevented by the addition of CaCl, to the salt solution. In ex- periments with starfish and sea-urchin eggs both of these expectations were realized. As index of permeability-increase in Arbacia eggs the exit of pigment was employed; and it was found that those salts which freed pigment most rapidly were in general the most effective as activating agents; the order of relative effectiveness for both per- meability-increase and activation, with pure isotonic solutions of Na and K salts, was (for anions): Cl<. Br | is a retardation of the rhythm which begins within a few moments. After fifteen to thirty minutes impulses are still evidently originating at the aboral (apical) end of the plate-row, but the amplitude of vi- bration is least at the aboral end and increases in the oral direction along the row, until in the oral half more or less of the row it is indis- tinguishable from the normal. After one-half to one hour movement in the most aboral plates of the row is almost imperceptible, but increases orally and may still be normal in amplitude over a longer or shorter dis- tance at the oral end. After one hour in KCN all movement of the plates in about the aboral fourth or third of the row has ceased, but the single plates of this region still respond to direct mechanical stimulation by one or a few beats of almost or quite normal amplitude. Such response may be limited to the plate stimulated, or the stimulus may be trans- mitted to the two or three plates adjoining on the oral side of the plate stimulated. Where this response is rhythmical, the rhythm is inde- pendent of and usually more rapid than that in the more oral regions of the row where rhythmic movement still persists. At about this time (one hour) or a little later another effect makes - its appearance, particularly in the four longer plate-rows. This is 92 C. M. CHILD the appearance, of an: independent rhythm over a longer or shorter distance at the oral end of the plate row, sometimes involving the oral third or even more of the row. The manner in which this rhythm makes its appearance is of interest. In some cases, for example, in KCN m 25 x 10" movement ceases in the most aboral plates, while in the middle half of the row the plates are still beating with a rhythm much below the normal. In this case this rhythm does not usually extend into the oral fourth more or less of the row, but this region shows an independent and more rapid . rhythm alternating irregularly with periods of complete quiescence. Such periods of quiescence are interrupted from time to time by the passage of an impulse from the more aboral region into the oral fourth. It is evident that at this stage the impulses from the middle region. do not ordinarily pass into the oral fourth and this is developing an independent rhythm more rapid than that of the more aboral regions. This rhythm, however, is still intermittent and in the periods of quies- cence summation of the impulses from the middle region may occur at the boundary between the two and a single impulse or sometimes two or three may pass all the way to the oral end. After this the oral portion may again become quiescent or may resume its independent rhythm. In some cases four regions of different behavior are distinguishable along a single row: the most aboral where there is no movement; a second or middle region in which rhythmic movement is proceeding with a slow rhythm; a third region which sometimes beats with the rhythm of the second and sometimes independently with a more rapid rhythm; a fourth region at the oral end of the row in which the rhythm is completely independent of all more aboral regions and most rapid of all. In such cases the most oral region has become completely independent of other parts and has developed its own rhythm, while the region next to it is at times independent of and at times subordi- nated to the region next aboral to it. In still another case five distinct regions appeared along a plate-tom: ; an aboral region in which movement had ceased; a second showing a slow rhythm; a third showing mostly an independent rhythm more rapid than the second but occasionally becoming subordinate to it and showing the same rhythm for a short time; a fourth region showing a still more rapid independent rhythm; a fifth region, the most oral portion of the row with a still more rapid independent rhythm. In this case the fifth region became independent first, then the fourth and then the third. — eee en i, EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 93 In general the effect of the inhibition at this stage—after one hour— is to make a longer or shorter portion at the oral end of the plate-row independent of the-impulses origitiating at the aboral (apical) end or at the most apical level which is still active. For some reason the im- pulse coming from aboral regions is no longer effective in controlling the oral region and as time goes on other regions in succession from the oral end become independent. The appearance of independent, more rapid rhythms in the oral regions of the row shows further that these regions, when they have attained a certain degree of independence of more apical regions are capable of initiating and maintaining a rhythm _ of their own or even several different rhythms at different levels, that of the most oral region being the most rapid. In some cases the oral portion of the plate-row not only becomes in- dependent of impulses from aboral regions but the direction of trans- mission of the rhythmic impulse actually undergoes reversal from aboral-oral to oral-aboral. This reversal has been observed in three cases in different animals. In such cases the reversed impulse begins at the extreme oral end of the row and travels a longer or shorter dis- tance in the aboral direction to a point where it meets an aboral-oral rhythmic impulse with a different rhythm, and there it ceases to be effective. The boundary between these two different rhythms pro- ceeding in opposite directions is at times perfectly distinct, one of two adjoining plates beating with one, the other with the other rhythm. At other times there may be an intermediate region of a few plates where impulses sometimes travel in one direction, sometimes in the other. As will be pointed out below, these cases of reversal of the direction of conduction are of particular interest. If the action of the cyanide is continued beyond this stage of com- plete aboral inhibition and oral independence with or without reversal, complete quiescence of the plates gradually progresses from the aboral end of the row in the oral direction and the rhythms in these parts of the row which are still active become slower and slower, the retarda- tion at any time being greatest in the most aboral and least in the most oral region. After one and a half to three hours according to con- centration and age, the older animals being less susceptible, movement has ceased over the aboral three-fourths to seven-eighths of the row and only the oral one-fourth to one-eighth still shows rhythmic move- ment, the direction of conduction being either normal or reversed, and the rhythm being much slower than when this region first became independent. Finally, movement ceases in this region. Here as else- > 94 C. M. CHILD where the single plates remain capable of response by one or a few beats to direct mechanical stimulation for some time after rhythmic movement has ceased, but such stimulation is usually not transmit- ted or at most affects only two or three plates. This condition may persist for an hour or more after cessation of movement but finally the plates become whitish and opaque and are certainly dead. This final change usually begins at the aboral end of the row and proceeds orally, but the difference in time between death of the plates at the two ends of the row is very much less than the dif- ference in time of cessation of movement and frequently the change extends over the whole roav almost at once. Not infrequently single plates suddenly begin rapid vibration just before they turn white. In no case has transmission of this vibration to another plate been observed and it usually continues only a few seconds, but occasionally for a minute or two. It is evidently the result of a stimulation con- nected with the changes immediately preceding death. If the animals are returned to sea-water after one to one and a half hours in KCN, or if the dish is left open so that the KCN gradually escapes, more or less complete recovery may occur. The changes along the plate-row in recovery are essentially the reverse of those in KCN. The rhythms become more rapid, this change being much greater aborally than orally. Plates near the aboral end of the row, which had ceased to move gradually resume movement with increas- ing amplitude of vibration until finally the whole row is again active. The most interesting phase of recovery is the subordination of the independent oral portions of the row to the aboral impulse. Where no reversal of direction of conduction has occurred in the oral region it can be observed that the aboral rhythm gradually impresses itself on the oral region, at first only intermittently but with increasing approach to continuity, until finally the oral region is again under control. In cases where two or three independent regions with dif- ferent rhythms arise in the oral region of the row, the subordination or control of these regions in recovery progresses in the oral direction until a single rhythm again extends over the whole length of the row. In cases where the direction of conduction has undergone reversal in the oral region, this region retains its independence for a longer time than otherwise, sometimes until the death of the animals, which usually occurs after a day or two in the laboratory. The reversal in direction of the impulse, or conditions associated with the reversal, have somehow made this region more independent of the other impulse. EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 95 A few other incidental experimental data and observations are brief- ly mentioned. The effect-of cutting across a row of plates is like that observed by Parker (4). Recovery on both sides of the cut occurs rapidly and on the aboral side of the cut the usual rhythm synchro- nous with that of the other row of the same quadrant is maintained, while oral to the cut an independent rhythm arises. Direct transmis- sion across a cut as recorded by Eimer (30) and Verworn (3), was not observed in Mnemiopsis. Several cases were found in which a plate-row had been separated into two parts by some injury, and the wound had healed, leaving a distance of one to several millimeters between the two parts. In all such cases observed, the part oral to the injury showed a rhythm independent of parts aboral to it, even though the separation of the parts by the injury was not more than one or two millimeters. If the plate-row is divided into several independent regions by ¢uts across it at different levels and the animal then placed in KCN the retardation of rhythm, decrease in amplitude of vibration and cessa- tion of movement occurs in general most rapidly in the most aboral portion and less rapidly in each succeeding portion in the oral direc- tion. In such cases, before inhibition has proceeded too far, several different rhythms are present, the slowest rhythm in the most aboral portion and successively more rapid rhythms in each successive por- tion in the oral direction. In this respect these cases where the plate- row is separated by cuts are like those described above, in which one or several different rhythms appear in the more oral regions of the row without any mechanical interruption of continuity. In the one case a physical, in the other a physiological isolation has occurred.. In general the susceptibility of the plate-rows as well as of the whole body-surface of Mnemiopsis is greatest in the youngest animals and de- creases with advancing age as in other animals (22). In the very young animals, where the plate-rows consist of only six to ten plates the gradient in susceptibility along the rows is slight and independence of the oral portion has not been observed. Apparently in those stages the length of the row is so short that the impulse undergoes but little decrement. In the large old animals independent rhythms ap- pear in the four long rows more frequently than in the four short rows, another fact indicating the existence of a spatial decremental factor in the transmission of the impulse. 96 - Cc. M. CHILD DISCUSSION The nature of transmission in the plate row. Considering first the question of the nature of transmission, the experimental data do not support the theory of direct mechanical transmission. All the phe- nomena of inhibition by cyanide are essentially similar to those ob- served in metabolic gradients along the main axes of simple organisms (23, Chaps. IV, V). The ability of the plates to respond to direct mechanical stimulation long after rhythmic metachronic movement has ceased and the fact that this response is either not transmitted at all or only to two or three plates make it probable that mechanical trans- mission does not play any very important réle. Before the plates cease to move the amplitude of their vibrations gradually decreases, while transmission still occurs and is effective to a certain limit of distance. But the appearance in KCN of independent rhythms in the oral regions of the plate-row before transmission and movement of plates have ceased in more aboral regions presents the greatest difficul- ties to the mechanical hypothesis. I# transmission is mechanical it is impossible to understand how one plate can beat with a certain re- tarded rhythm while the plate next to it orally develops an indepen- dent more rapid rhythm, or how, in the absence of the impulse from the apical region, any level of the plate-row may become the point of initiation of a new rhythmic impulse. The only conclusion possible in view of all the facts is that reached by Parker (4) that, while direct mechanical transmission may occur to some extent or under certain conditions, it is not the fundamental or chief method of transmission. The hypothesis of neuroid transmission is the only one which will ac- count for the facts. The susceptibility gradient. Assuming then that transmission is neuroid in character it is necessary to interpret the various phenom- ena .of inhibition by cyanide on this basis. The first and perhaps the most conspicuous feature is the gradient in susceptibility to KCN along the plate-row. Cessation of movement begins in all cases at the aboral end of the row and progresses in general in the oral direc- tion. Cessation of vibration of the plates, however, does not mean loss of the ability to vibrate, for plates that have ceased to vibrate in the regular progress of inhibition may still be induced to vibrate by direct mechanical stimulation. Evidently cessation of movement in KCN is the result of cessation or decrease below the threshold of the transmitted impulse. EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 97 Decrease in amplitude of vibration of the plates precedes complete cessation of movement, but the plates when stimulated mechanically after complete cessation of movement may show the full normal am- - plitude of vibration. This decrease in amplitude of vibration may re- sult in part from decrease in the irritability of the plate itself in KCN, but the fact that vibrations of full amplitude may follow mechanical stimulation even after the cessation of natural movement, proves that another factor must also be concerned. There seems at present to be no escape from the conclusion that a decrease in the intensity or physiological effectiveness of the transmitted impulse must be at least in part responsible for the decrease in amplitude of vibration. _ If this conclusion is correct amplitude of vibration must be a function of intensity of impulse, at least up to a certain limit and the decrease in amplitude and final cessation of the rhythmic movement in KCN must mean that the intensity of the impulse gradually approaches the threshold and finally falls below it. In the decrease in amplitude of vibration the same gradient along the -plate-row as in cessation of movement appears. In this gradient several factors may conceivably be concerned: there may be a gra- dient in rapidity of decrease in intensity of the transmitted impulse with a decrease in rapidity in the oral direction; or a gradient in the same direction in the rapidity of decrease of irritability of the plates, or the threshold of stimulation may be highest in the most aboral plates of the row and may decrease in the oral direction or possibly all these factors may play some part in the result. Another effect of KCN is the progressive retardation of the rhythm, the decrease in frequency of impulse. This effect also appears in the form of a gradient, the retardation being most rapid aborally and decreasing in the oral direction. This gradient, however, becomes visible only when the plate-row is separated by section at several levels into several independent portions, or when one or more oral regions become physiologically independent in KCN. In all such cases at the proper stage the rhythm is slowest in the most aboral portion and increases in each successive portion in the oral direction. In later stages of course complete inhibition occurs in the aboral region and retardation progresses in the oral direction. All these facts indicate the existence of a gradation of some sort: in the path of conduction along the plate-row. As regards the am- plitude of vibration, the retardation of rhythm and the cessation of movement, the action of KCN is most rapid at the aboral end of the 98 Cc. M. CHILD plate-row and decreases progressively to the oral end. Moreover, the aboral end of the plate-row, the region of highest susceptibility, is the region where the normal impulse originates and the transmis- sion of this impulse is from levels of higher to levels of lower suscepti- - bility. On the basis of the general relation between susceptibility to — KCN and many other agents and metabolic activity or physiological - condition mentioned on p. 89 above and discussed in various earlier publications (see especially 17, 18, 22, Chap. III, 28, Chap. III, 25), we are forced to conclude that the susceptibility gradient along the plate-row of the ctenophore is an indicator of a gradient in general metabolic activity, or irritability, i.e., the potentiality of metabolic activity as expressed in the condition of the protoplasmic system. According to this conception the aboral or apical region is the region of highest metabolic rate and the rate decreases in the oral direction along the plate-row. The normal impulse then is transmissible down the gradient as in the case of the nerve fiber according to Tashiro (10, 11, 12), in other words the region of highest metabolic rate dominates or sets the pace, so to speak, for other levels, and in isolated portions of the plate row the level of highest rate becomes the dominant or controlling region. The fact-that essentially similar relations between a metabolic gra- dient and physiological dominance and subordination are characteristic features of general body-axes in both animals and plants (23) must at least suggest the possibility that a certain fundamental similarity exists between physiological axiation and order in the development of the organism and in the transmission of impulses along a conducting path, whether protoplasmic, ‘‘neuroid” or of the highly specialized nervous type. In fact, if we conceive the general organic axis as a dynamic or metabolic gradient established by the general protoplas- mic transmission, with a decrement in intensity, of dynamic changes from a region of high metabolic activity, which is itself determined in the final analysis by the differential action of external factors upon the protoplasm concerned—if we accept this conception of the organic axis, we can trace a genetic relation between the general physiological axis in its simplest form and the highly specialized nerve-axis. The one represents in fact the most generalized, the other the most spe- cialized condition of the same thing. Physiological Isolation. Under the usual conditions the impulse is transmitted over the whole length of the plate-row, and conclusive evidence for a decrement in intensity or effectiveness in the normal EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 99 animal is lacking. It has seemed to me, however, that a distinct grad- ual decrease in amplitude of vibration of the plates toward the oral end of the plate-row could sometimes be seen when the animal was not strongly stimulated, but I cannot state this as a positive fact. What- ever the condition in the normal animal, it is evident that sooner or later in KCN the impulse from the aboral end loses its effectiveness at some point near the oral end, especially of the long plate-rows, and the oral region thus set free from the control of the impulse transmit- ted from more aboral levels soon initiates a rhythmic impulse of its own. Later, a second and in some cases even a third region may be thus set free from aboral control and develop its own independent rhythm. Stated in slightly different terms, one or more regions at the oral end of the plate may be successively physiologically isolated as the effectiveness of the original impulse decreases in KCN. Conditions on both sides of the point where the aboral impulse ceases to be effective probably play a part in determining the position of this point. The action of the cyanide is most rapid in the aboral portion of the row and the retardation of rhythm is greatest there. Since the oral region is less affected by KCN it is capable, if isolated from the aboral impulse, of initiating a much more rapid rhythm. It seems probable that when the aboral rhythm has been retarded to a certain degree an independent more rapid rhythm may arise at some point in the oral region in the intervals between the aboral impulses. If, however, the impulse from the aboral region retains its original intensity or effectiveness and is merely retarded in rhythm we should expect the independent rhythm of the oral region to be interrupted or modified by the passage of the less frequent aboral impulses over the region. This does occur in some cases in the early stages of independence but later the oral region becomes entirely inde- pendent and the aboral impulse does not produce any effect upon it. The only possible conclusion is that the aboral impulse has lost in intensity or been weakened in some way, so that its effective range, i.e., the length of path over which it is effective, is decreased. The fact that the effective range of the impulse is limited and under- goes a progressive decrease in KCN indicates very clearly that a de- crement in intensity or effectiveness occurs in transmission, at least under these conditions, and that consequently a spatial range of effec- tiveness exists, which decreases as the inhibitory action of KCN pro- ceeds. In short, the impulse behaves like a wave in a physical medium in that at a certain distance from its point of origin it dies out, so far 100 C. M. CHILD as the characteristic physiological effect is concerned, and this distance decreases progressively with the action of KCN. This decrease may conceivably be due either to a decrease in intensity of the impulse at the point of origin or to a decrease in conductivity of the path or more probably to both factors. As noted above, the facts indicate that a decrease in intensity does occur in KCN and it is probable also that a decrease in conductivity of the path occurs, in fact a decrease in rate of conduction is clearly visible. In any case it is evident that the aboral - end of the plate row is most susceptible to the inhibiting action and that even while the oral portions of the plate-row still retain their irri- tability and conductivity the impulse transmitted from the aboral end becomes ineffective at a greater or less distance from the oral end. Using the physical analogy we may say that KCN decreases the height of the wave at its point of origin and the conductivity of the medium, and so decreases the distance it travels before becoming ineffective. These cases of the escape of one or of successive regions at the oral end of the plate-row from the control of the impulse transmitted from the aboral end are cases of physiological isolation similar in character to cases of physiological isolation observed and experimentally pro- duced in the axes of the simpler organisms (16; 22 p., 228; 23, Chap. V). Physiological isolation of parts or regions in a metabolic gradient may be brought about in four ways: first, by growth of the protoplasmic or cell mass so that the length of the mass in a given axis is greater than the effective range of control of the dominant region; second, without altering the actual size of the mass, by decreasing the meta- bolic activity in the dominant region and so decreasing the effective range of control so that those portions of the mass most distant from the dominant region are no longer affected by it; third, by decreasing the conductivity of the path along the gradient, i.e., by decreasing its excitability, and in this way decreasing the effective range of the transmitted change; fourth, by excitation of a subordinate region to such a degree that it becomes independent of the excitation transmit- ted from other regions. Physiological isolation occurs in nature and can be induced experimentally in these four ways. In the physiological isolation of the oral region of the ctenophore plate-row the second factor, decrease in the activity of the dominant region, and the third, decried in conductivity of the path, are undoubt- edly concerned and in addition there is in consequence of the differ- ential susceptibility to KCN a relative increase in the metabolic ac- tivity of the oral, as compared with the aboral region which is essen- * EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 101 tially similar to the fourth factor, excitation of the originally subor- dinate region. But whatever the réle of these three factors in any particular case, the fact of physiological isolation of the oral region or of two or three regions successively is sufficiently evident. In the plate-rows of small young individuals physiological isola- tion of the oral region in KCN has not been observed, and in the four shorter plate-rows of large individuals it is much less frequent than in the four longer rows. Apparently, as might be expected if the impulse undergoes a decrement in effectiveness in the course of transmission, the longer the plate-row the more frequent is physiological isolation at its oral end. The effect of physiological isolation. In the conducting path of the etenophore plate-row, as in the axis of the simpler animals and plants, the effect of physiological isolation is essentially similar to that of physical isolation by section, viz., the reproduction of a new individual order, the development of a new individual. Physiological isolation in the chief axis of the lower organisms is the necessary condition for many if not all the processes of agamic reproduction, fission, budding, ete., and in the minor axes, of reduplicative reproduction of parts. The reproductive process in the physiologically isolated part of the cteno- phore plate-row consists in the initiation of an independent rhythmic impulse which begins at one end of the isolated portion and is trans- _mitted over its length. This physiologically isolated region then be- comes a new individual essentially similar to the previously existing individual—the whole plate-row—of which it was originally a part. The result is the same as the result of physical isolation of a part of the plate-row by cutting across the conducting path. Not only one but two or three such individuals may arise successively, beginning at the oral end of the row, as the original impulse becomes progressively weaker and its effective range decreases. In some of the simpler ani- mals, e.g., Planaria (16) series of new individuals: arise at the poste- rior end of thé body in essentially the same way, either as the result of increase in the length of the body or depression or removal of the an- terior end. These physiologically isolated regions of the plate-row, particularly in the earlier stages of their isolation, are sometimes temporarily sub- ordinated again to the original impulse, either in consequence of sum- mation of excitations at the boundary between the two rhythms, or perhaps by unusually intense impulses which have a greater effective range and so are able to pass this boundary. The same relations ap- 102 , C. M. CHILD pear between anterior and posterior zodids in Planaria. Summation of impulses or strong stimulation of the anterior body-region may bring the posterior zodids under complete control of the anterior region for a time, but as the animal returns to the usual condition they soon become physiologically isolated again to a greater or less degree. This initiation of a new and independent rhythm in the physiolog- ically isolated, oral portion of the plate-row is then in the broad sense a case of reproduction, differing from various agamic reproductive processes in the simpler organisms chiefly in its somewhat specialized character. The occurrence of this reproductive process depends upon the fact that these portions of the plate-row, while they do not ordinarily initiate a rhythmic impulse but are subordinated to the rhythmic im- pulse transmitted from the aboral end, still retain the capacity to initiate such an impulse where the original impulse is prevented by any means from reaching them or when its intensity falls below a certain level. Similarly the agamic formation of new individuals from parts of the body of Planaria and other forms depends upon the fact that these regions, while physiologically and morphologically parts of an individual still retain the capacity to become new individuals when physiologically or physically isolated from the control of the domi- nant region of the original individual. The effect of recovery from KCN. Where recovery is permitted to occur, the physiologically isolated region where a new individual has developed may be again subordinated to the dominance of the aboral: end of the plate-row and so may lose its independent rhythm, i.e., its in- — dividuality, and again become what it was originally, a part of a larger individual. This is a process of reintegration, the reverse of reproduction. This reintegration evidently results from an increase in the intensity and so of the effective range of transmission of the aboral impulse during recovery, until it dominates and obliterates the independent rhythm ‘in the part which was before physiologically isolated. Similar reversal of the reproductive process and reinte- gration may be brought about by fundamentally similar methods after agamic reproduction has begun in the lower animals. For example, in Planaria and other forms new ‘‘zoéids,”’ i.e., new develop- ing individuals in the posterior body-region, may be made to disap- pear by removing the anterior half or more of the original or parent individual and permitting a new head to regenerate from the cut end, which is much nearer the new zooid than was the original head. Since the distance between the regenerated head and the new zodid is much EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 103 less than that between the original head and the new zodid, the zodid is no longer physiologically isolated and disappears as an individual, becoming again-a-part. In some forms this reintegration may be brought about even after a considerable degree of morphological devel- opment of the new individual has occurred. This case differs from reintegration in the ctenophore plate-row merely in that in the one the intensity and so the effective range of the transmitted impulse is in- creased by recovery from KCN, while in the other the dominant region is actually brought nearer to the physiologically isolated region which then falls within the effective range of the transmitted dynamic changes. The rhythmic period in physiologically isolated regions. When a region at the oral end of the plate-row is physiologically isolated and develops an independent rhythmic impulse, the rhythmic period, i.e., the interval between impulses is shorter than the period existing at the same time in the more aboral region and when two or three regions are successively isolated physiologically and develop independent rhythms, the rhythmic periods in all these regions are shorter than in the aboral regions, but that of the most oral region is the shortest of all, that of the second region longer and that of the third region still longer. These differences in the independent rhythmic periods at different levels of the plate-row undoubtedly depend at least in part upon the - general metabolic gradient which appears in KCN as a susceptibility gradient. The aboral end of the row is most susceptible to KCN and the rhythm is most retarded there, while the oral end is least suscep- tible and the rhythm is therefore least retarded in the stages of KCN- action under consideration. Between these two extremes are interme- diate degrees of susceptibility, and when the aboral impulse becomes so weak that one or more oral regions are physiologically isolated the rhythmic period in each such region must depend in part upon its level in the general gradient and so upon the degree to which KCN has already affected it. But the rhythmic period in these physiolog- ically isolated oral regions of the row are often very short, even shorter than those in normal animals and the rhythmic activity may be irreg- ularly intermittent. The behavior of these regions when physiolog- ically isolated suggests the possibility that some factor which regu- lates and orders the rhythmic period in normal animals is not fully developed in these regions which are suddenly made independent. _ In fact, to state the case in unscientific terms, they behave as if they ‘were not accustomed to independence. Until we know more of the dynamic conditions which determine rhythmic activity, interpreta- tion of its changes under experimental conditions can not go very far. 104 Cc. M. CHILD The direction of transmission in physiologically isolated regions. In the physiologically isolated regions of the plate-rows the direction of transmission is, in the majority of cases, aboral-oral like that of the original impulse, but sometimes a reversal of direction occurs sooner or later in the extreme oral region, and the impulses run in the oral- aboral direction. If the direction of transmission is connected in any way with the metabolic gradient, as both Tashiro’s and my experimen- tal data indicate, a reversal in direction of transmission must be asso- ciated with a reversal of the gradient, and I believe that such reversal of the gradient has occurred in these cases. Since the levels of higher metabolic rate in a metabolic gradient are more susceptible to KCN in sufficiently high concentration than levels of lower rate, the gen- eral effect of KCN on such a gradient must be first a levelling down, a decrease in the metabolic differences at different levels, which may lead to complete obliteration and even to reversal of the gradient. This is true not only for KCN but for many other inhibiting agents and such reversals have been experimentally induced in the polar axes of organisms through differential inhibition, as experimental data soon to be published will show. ; If a wave of increased metabolic activity is an essential feature of the transmission of excitation in protoplasm, it is probable that the effective range of such a wave is greatest in the downward direction along a metabolic gradient, less along a metabolic level and still less in the upward direction along a gradient. If this be true then the levelling down and reversal in KCN of a metabolic gradient along a conducting path must be an important factor in decreasing the effec- tive range of an impulse transmitted along that path. A metabolic wave probably can not be transmitted up a gradient beyond the point where the metabolic rate or the protoplasmic condition before excita- tion is the same as that in the wave of excitation. Moreover, in a metabolic gradient in which rhythmic impulses originate, as in the conducting path of the ctenophore plate-row, it is evident that normally the impulses originate in the region of highest metabolic rate in the gradient and are transmitted down the gradient. There is every rea- son to believe that the same relation between gradient and rhythmic impulse exists in the physiologically isolated oral regions. Where the direction of transmission remains aboral-oral the original gradient still persists, and where transmission is in the opposite direction the . gradient has been reversed by the differential action of KCN on dif- ferent levels of the original gradient. The limitation of reversal to the EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 105 oral region of the plate-row in my.experiments is probably due at least in part to the fact that metabolic activity in the more susceptible abo- ral regions of the plate-row is inhibited so rapidly and to such an extent that these regions become incapable of initiating or transmitting im- pulses by the time a well marked reversal of the gradient has occurred in them. With less toxic agents or perhaps with lower concentrations of KCN it may be possible to reverse the direction of transmission throughout the whole length of the plate-row. -. In some species of ctenophores reversal in the direction of trans- mission occurs or can be experimentally induced in normal animals. Parker (4, p.411) states that in Pleurobrachea a rapid wave in the aboral- oral direction is sometimes ‘‘reflected”’ at the oral end of the plate-row and is transmitted in the oral-aboral direction, but rarely over more than one third the length of the plate-row. Much earlier Eimer (29, p. 226) observed reversal in the direction of transmission in Beroe and Chun (7, p. 182) records similar observations on Beroeand other species, while Verworn (3, p. 167; 30, p. 440) observed that such reversal .can often be induced by stimulating mechanically the oral end of a plate- row. I believe that all such cases of reversal in the direction of transmission are dependent upon temporary reversal of the metabolic gradient along the path, or that part of it where reversal occurs. In the case of ‘‘re- flection”’ recorded by Parker the excitation reaching the oral end of the row increases the metabolic rate there so rapidly and so far above the level of adjoining parts that an impulse is initiated there and is trans- mitted backward to a greater or less distance. Reversal after mechani- cal stimulation of the oral end of the row as observed by Verworn is evidently due to the increase in metabolic rate at that end in conse- quence of the stimulation and so the initiation of an impulse sufficiently intense to travel some distance in the oral-aboral direction. All such cases of temporary reversal in the direction of transmission are in reality temporary reversals of the physiological polarity of the conducting path and the readiness with which they occur or can be induced undoubtedly varies with the slope of the metabolic gradient and with the degree of permanency or irreversibility of the record in the protoplasmic condition of the dynamic gradient in different species. Physiological polarity in the lower organisms shows similar possi- bilities of reversal and alteration by very similar methods (23, pp. 96, 117, 132, 142) and, as in the ctenophore plate-rows, the readiness with which reversal occurs or can be induced in different species depends 106 Cc. M. CHILD upon the degree of permanency or irreversibility of the record in proto- plasmic condition and differentiation of the preexisting dynamic gradient. THE GENERAL SIGNIFICANCE OF METABOLIC GRADIENTS The conception of the physiological axis as consisting in its simplest form of a metabolic gradient together with the gradient in protoplasmic condition in the broadest sense which must be associated with a dynamic gradient has proved a fruitful working hypothesis in its application to normal processes and experimental modifications of development in both animals and plants (16-28). Many different lines of evidence indicate the existence of such axial gradients, and it is possible to control and modify development to a high degree through the differential action of external agents upon such gradients and to interpret in terms of gradients modifications produced in nature and experiment. If the conception is correct it means that the first step in the physiological integration which constitutes what we call the axiate individual or organism consists in the establishment, or the inheritance from a pre- existing individual of one or more such gradients.. The primary gradient ° represents the primary or chief axis and the region of highest metabolic rate in that gradient becomes the apical or head region, and in other axes the morphological and physiological order or pattern shows a definite relation to the metabolic gradient in those axes. In any such gradient the region of highest metabolic rate dominates or controls regions of lower rate, and is the primary factor in establishing the gradient, because in consequence of its activity dynamic excitatory changes of some sort are transmitted through or over the limiting sur- faces of the protoplasm to other less active regions and are more effec- tive in determining their metabolic activity than excitatory changes transmitted from regions of lower rate. In short the highest level of the gradient is to a greater or less degree physiologically dominant because the excitatory changes initiated in it are greater or more in- tense than those initiated at other metabolic levels. Since in general protoplasmic transmission a decrement in intensity or effectiveness occurs, such a transmitted change has a limited ef- fective range which varies in general with metabolic activity and proto- plasmic condition, and this effective range determines the spatial limit - of such physiological dominance. According to this conception the unity and order, the physiological integration of the organism is primarily dependent upon the trans- mission of dynamic changes rather than upon the transportation of ~ EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 107 chemical substances. In other words, physiological integration in the axiate organism is primarily “neuroid” in character rather than a matter of so-called chemical correlation, though I prefer “transmissive” and “transportative”’ to “neuroid’”’: and “chemical” as denoting the character of the fundamental condition in such an integration. _ As I have shown elsewhere (23) localization and differentiation arise in relation to different levels of the axial metabolic gradients, and in - a system so complex as even the simplest living protoplasm, there is no difficulty in accounting for the origin of qualitative from quantitative _ differences. As soon as differentiation begins, transportative or chemi- cal correlation begins to play an essential réle and is of course of great importance in further development. It is evident, however, that trans- portative or chemical correlation cannot be the starting point of physio- logical integration in the individual because a definite unity and order, a definite organization, in short an integration, must be present before such correlation is possible. It is this primary, fundamental organi- zation that the conception of metabolic gradients attempts to account for. This conception is in no sense a substitute for the conception of chemical correlation which plays so important a rdéle in present-day physiology. It is merely an attempt to establish the basis upon which chemical correlation becomes possible. The gradient is merely the starting point and as soon as the production of different substances at different levels of the gradient begins, which must be very early, trans- portative correlation becomes an essential factor in determining the further course of events. The gradient merely determines the primary pattern, and chemical factors may play the chief, or at least a very large rdéle in determining the character of the different parts of the pattern. lf we conceive the organism merely as a complex of specific chemical correlations we cannot account for the origin-and development of the nervous system. No adequate reason can be given for the transfor- mation of a system in which correlation is primarily transportative or chemical into a system with transmissive correlation. The transmis- sive factor must be, as we know it is, a fundamental property of living protoplasm, and if this is true, this factor must play a fundamental part in physiological integration. If the organism is primarily a trans- missive integration the origin, development and functional dominance of the nervous system become at once intelligible. Moreover, the central nervous system develops in the regions of highest metabolic rate in each of the primary axial gradients of the organism (23, p. 175). 108 c..M. CHILD This fact is highly significant as indicating that the nervous system is merely the final morphological and physiological expression of the re- lations which in their simplest terms are represented by metabolic ~ gradients. If this conception is correct we must expect to find that the nerve fiber is primarily a metabolic gradient and Tashiro’s observations (10, 11, 12) indicate that this is actually the case. The existence of a gradient in the “neuroid’”’ conducting path of the ctenophore plate-row is also © to be expected, and the similarity between its behavior and that of the chief axis of the simpler organisms follows as a matter of course. We must expect, moreover, to find that other organs in which rhyth- mic impulses are transmitted in a definite direction are likewise pri- marily metabolic gradients. The vertebrate heart, for example, is such an organ, and all the experimental data which we possess con- cerning its rhythmic activity indicate the presence of a metabolic gradient. The sinus-region which normally initiates the beat and so is physiologically dominant must be primarily the region of highest rate in this gradient. It is a familiar fact that when this region is in- hibited by cooling or otherwise, the beat may begin-in the uninhibited region nearest to the sinus and reversal of the direction of the beat by means of inhibition of the sinus end and stimulation of the bulbus end has even been induced. This is essentially similar to what occurs in the ctenophore plate-row and also in the axes of the simpler organisms. In the ascidian heart reversal of the direction of beat occurs periodi- cally under natural conditions probably in consequence of differential fatigue, the region of high rate in the gradient at any given time, which is the dominant region, the initiator of the beat at that time, becoming fatigued more rapidly and to a greater degree than regions of lower rate. In consequence of this differential fatigue the metabolic activity of the dominant initiating region decreases more rapidly than that of other levels and this leads sooner or later to cessation of the beat in _ the original direction. The existing gradient is also levelled down or perhaps reversed by differential fatigue and the region which was formerly the low end being least fatigued recovers more rapidly and so contributes further to reversal of the former gradient. In this way the low end of the gradient of one series of beats becomes the high end and the initiator of the next period, and in this manner periodic reversal of the direction of beat continues. According to this interpretation the ascidian heart is simply a reversible gradient, while in the heart of the higher vertebrate the gradient is much more stable, and reversal therefore less readily induced. —- ~ Si EFFECT OF CYANIDES ON Sema CTENOPHORE 109 It is, of course, not necessary to assume that transmitted impulses or changes along a metabolic gradient are always rhythmical, they may be tonic, irregular-or rhythmical. In the chief physiological axes of the organism they may be largely tonic with irregular, or more or less rhythmic changes as metabolic. changes in the dominant region occur. This conception seems at first glance to disagree with what we know concerning transmission in the medullated nerves of vertebrates. It has been stated repeatedly that in these nerves transmission under normal conditions shows no decrement in intensity or effectiveness and the further assertion has been made that the ‘all or none law’ applies ‘not only to the primary excitation but also to transmission in such nerves. If there is actually no transmission-decrement in such nerves then transmission to an infinite, or better an indefinite distance, would occur in a nerve fiber of infinite or indefinite length. It seems highly improbable that any physico-chemical medium is capable of such transmission; moreover, as regards the medullated nerve, it is a familiar fact that under various experimental inhibitory conditions such as partial anesthesia, cooling, etc., a decrement in effectiveness and a limit of effective range appear. It is difficult to believe that anesthesia or low temperature or other inhibitory conditions alter so fundamentally as this the nature of transmission in the medullated nerve. Thatthey decrease the conductivity or the intensity of the impulse or both and so increase the decrement and decrease the effective range can readily be understood, but that they determine a decrement and a limit of effective range where none is present normally, it is difficult to believe. There is much evidence for the normal existence of a decrement and a limit of effective range in the more primitive protoplasmic and neuroid conducting paths, and in view of this fact, the only logical conclusion seems to be that the medullated nerve is simply a so much better con- ductor of impulses than these primitive paths that within the lengths of nerve fiber available or ordinarily used for experiment, the decre- ment is slight or inappreciable. The evidence in the case taken as a whole seems to point very clearly to this conclusion as the only one possible, and if we accept this conclusion, the medullated nerve pre- sents no difficulties to the general conception of metabolic gradients. The attempt has been made in this paper to show, on the basis of experiment upon a relatively primitive ‘‘neuroid” conducting path that there are adequate reasons for believing that the body axes of organisms in their simplest terms, and the most highly specialized axes in the organism, the nerves, as well as other physiological axes inter- . mediate between these extremes are fundamentally similar in certain 110 ; Cc. M. CHILD respects. Considered in the light of its value as a basis for experimen- tal investigation, analysis and synthesis, the hypothesis justifies itself, and while it will undoubtedly undergo modification as time goes on, I believe it will serve to throw light on many physiological and morpho- logical problems. Objection may be made to the term “metabolic” as applied to the axial gradient. This term means no more than that differences in the degree of metabolic activity are associated with and serve as an indicator of the gradient. Since function and structure are indissociable, it goes without saying that a metabolic gradient can- not persist or even exist without associated differences in protoplasmic condition corresponding to different levels of the gradient and those who prefer to emphasize the structural or physical rather than the dynamic or chemical aspects of the gradient may prefer to call it some- thing else than a metabolic gradient. But whatever we call it, the facts indicate that a gradient in activity or in reactive capacity, deter- mined in the final analysis by the differential action of factors external to the protoplasm concerned, represents the first step in not only the physiological integration of the axiate individual or organism, but in that of axiate organs and parts. It must be noted, however, that such a gradient represents only one possible type of integration or individuation. Even in the organism many other kinds of integration undoubtedly occur, such as, for example, molecules, molecular complexes, colloid particles, crystals, ete. The conception of the gradient is concerned only with that sort of integra- tion which expresses itself as a definite, controlled, progressive order of events in space and time, occurring in living protoplasm, whether cell or cell mass, with a definite relation to certain directions or axes in the protoplasm. The specific protoplasm or even the cell, with all the possible kinds of integration or individuality which may be present in it is regarded as given, and the conception of the gradient is merely an attempt to answer the question, what is the nature of a definite physiological axis in a specific protoplasm, whether cell or cell mass? SUMMARY 1. In the conducting path along the row of swimming plates of the ctenophore, Mnemiopsis leidyi a gradient in susceptibility to KCN exists. This gradient is indicated by the fact that decrease in ampli- tude of vibration, increase in rhythmic period and cessation of rhyth- mic movement occur first at the aboral end of the plate-row and show "a regular progression toward the oral end. Since the plates remain EFFECT OF CYANIDES ON CONDUCTION IN CTENOPHORE 111 capable of responding to direct mechanical stimulation by beats of full amplitude after the rhythmic beat has ceased, decrease in amplitude and cessation-of rhythmic movement must be due primarily to changes in the transmitted impulse rather than in the plates themselves. 2. This susceptibility gradient is an indicator of a gradient in general metabolic rate and in protoplasmic conduction associated with it. According to the relation between susceptibility and general metabolic condition the aboral end of the plate-row is the region of highest meta- bolic rate in this gradient and from this the rate decreases in the. oral direction. 3. The effective range of the rhythmic impulse decreases in KCN until it may be less than the length of the plate-row. Under such conditions a longer or shorter region at the oral end, or two or three regions suc- cessively, became physiologically isolated and develop independent rhythms which have a shorter period than the more or less inhibited impulses from more aboral regions. This difference in rhythmic period is another feature of the gradient and results from the fact that the less susceptible oral regions are less inhibited and their rhythmic period less retarded than the aboral region. 4. Occasionally a physiologically isolated oral region shows reversal in the direction of transmission at a certain stage of KCN action. Such reversal is undoubtedly associated with reversal of the metabolic grad- ient through the differential susceptibility to KCN. 5. In recovery the effective range of the aboral impulse increases, and regions previously physiologically isolated are brought again under control. Amplitude of vibration also increases and rhythmic period decreases during recovery. 6. The behavior of this physiological axis under the conditions of experiment is fundamentally similar to that of the main body axes of organisms, which are also in their simplest terms metabolic gradients, and the experimental data serve as a basis for consideration of the general significance of metabolic or dynamic gradients as physiological axes, both in the organism as a whole and in its parts, even in axes so highly specialized as nerve fibers. BIBLIOGRAPHY (1) ENGELMANN: Jen. Zeitschr., 1868, iv, 321. (2) EneeLMANN: Hermann, Handb. der Physiol., 1879, Bd. I, Teil 1, 343. (3) Verworn: Arch. gesammt. Physiol., 1890, xlviii, 149. (4) Parker: Journ. Exper. Zoél., 1905, ii, 407. - 112 (5) Baauion1: Winterstein, Handb. der vergleich. Physiol., (6) Bauer: Zeitschr. f. allg. Physiol., C. M. CHILD 1910, Bd. IV, 102. 1910, x, 236. (7) Cuun: Fauna und flora des Golfes von Neapel, I Mee 1880. (8) TASHTRO: (9) TAsHTRO: This Journal, 1913, xxxii, 107. Biol. Bull., 1913, xxv, 282. (10) Tasutro: This Jounned: 1914, xxxiii, 37. (11) Tasuiro: (12) TasHtRo: (13) TasHIRo (14) TasHtRo (15) TasHtro (16) Carib: (17) Carib: (18) Curiip: (19) Curip: (20) Curiip: (21) Curb: (22) Curip: (23) Crip: (24) Curb: (25) Curip: (26) Curip: (27) Carib: (28) Curnp: (29) Ermer: (30) Verworn, M: Arch. gesammt. Physiol., This Journal, 1915, xxxvi, 368. Proc. Nat. Acad. Sci., 1915, i, 110. . AND ApAms: This Journal, 1914, xxxiv, 405. AND Apams: Internat. Zeitschr. f. .phys.-chem. Biol., 1914, i, 450. AND Apams: Journ. Biol. Chem., 1914, xviii, 329. Journ. Exper. Zo6l., 1911, xi, 221. Journ. Exper. Zoél., 1913, xiv, 153. Arch. f. Entwickelungsmech., 1913, xxxvii, Biol. Bull., 1914, xxvi, 36. Proc. Nat. Acad. Sci., 1915, i, 164. This Journal, 1915, xxxvii, 203. Senescence and rejuvenescence, Chicago, 1915, Chaps. III and IX. Individuality in organisms, Chicago, 1915. Biol. Bull., 1916, xxx, 391. Bot. Gaz., 1916, lxii, 89. Journ. Morph., 1916, xxvii, 65. Biol. Bull., 1916, xxxi, 419. Sci., 1914, xxxix, 73. Arch. f. mikr. Anat., 1880, xvii, 213. 1891, 1, 423. 108. CARBON DIOXIDE ACIDOSIS, THE CAUSE OF CARDIAC DYSPN KA fi JOHN P. PETERS, JR. From the Medical Clinic, Presbyterian Hospital, and the Coolidge Fellowship for Medical Research, Columbia University, New York Received for publication February 15, 1917 This paper deals with the discrepancies observed in a series of cases in a comparison of the carbon dioxide of the alveolar air and that of the plasma. Comparisons of a similar nature have been made by Van Slyke (1), Peabody, and Walker and Frothingham (2). METHODS _ For the plasma CO, the Van Slyke method (1) was used. In this method plasma is saturated with an atmosphere containing about. 6 per cent COs, the tension usually obtaining in alveolar air. The plasma is then introduced into the Van Slyke pipette and rendered strongly acid to release CO, from the carbonates. This carbon dioxide is pumped out by means of a Toricellian vacuum and measured. volumetrically. Corrections are made for temperature and atmospheric pressure and the volumetric reading reduced to the mgm. CO2 chemically bound, which it represents. Finally, multiplication by an empirical constant, 35, converts it to terms of the alveolar carbon dioxide tension that should correspond with the determined concentration of carbonates in the plasma. The blood for the carbonate determinations was iii avie within fifteen minutes of the time when the last alveolar specimen was ob- tained and always after the alveolar work had been completed. The latter precaution was taken to eliminate the possible action on the respiratory center of the excitement caused by the venous puncture. The blood was drawn directly into a centrifuge tube containing a small amount of neutral, recrystallized potassium oxalate, removed at once to the laboratory and centrifugated. In most cases the blood was drawn and centrifugated under a layer of albolene. The plasma was 113 114 JOHN P. PETERS, JR. aerated in a 250 cc. separating funnel with the author’s alveolar air’ and the carbon dioxide content determined immediately. All de- terminations were made in duplicate. All studies were made just before meals, either between 11 and 12 a.m. or 3.30 and 4.30. p.m. For the determination of the alveolar carbon dioxide the Fridericia (3) method was employed. It seemed best to use some modification of the Haldane method and to make the comparison with arterial car- bon dioxide, both because the original work of Van Slyke was done in this way and also because arterial readings might be expected to bring out discrepancies more clearly. (The work done by Peabody and by Walker and Frothingham shows that, for clinical purposes, this is an unnecessary precaution). Of the arterial methods the Fridericia is the simplest. It is impossible to use it unless the patient is intelligent enough to codperate; it is also uncertain in the presence of marked respiratory irregularities, such as Cheyne-Stokes breathing; and it obviously demands a certain minimum respiratory capacity to clear out the dead space of the machine. In view of these possible errors the cases here reported have been chosen with the greatest care and much interesting material has been omitted. Repeated determina- tions were made in all cases and none have been accepted in which the readings varied by more than 0.2 to 0.3 per cent. In all the tables Column I shows the observed alveolar carbon di- oxide, Column II the alveolar carbon dioxide calculated from the car- bonate determination and Column III the ratio of the actual to the calculated value. OBSERVATIONS Of course the deductions made from such a study are largely de- pendent for their value upon the accuracy of Van Slyke’s constant. In Walker and Frothingham’s paper (2) 116 observations are reported on 100 cases. When these cases are compared on the basis of the ratio of alveolar to plasma COs, it is found that 92 out of the 116 ratios fall between 0.90 and 1.10; 102 between 0.85 and 1.15. If 1 mm. Hg. is subtracted from each alveolar reading in an attempt to reduce venous to arterial figures, 105 out of 116, or 90 per cent lie between 0.85 and 1.15. Table 1 gives the results in five normal cases (the author, three mem- 1 Determinations of my own alveolar carbon dioxide repeated over a period of fifteen months have shown a maximum variation of 3 mm. Hg., between 44.40 and 47.60, under normal conditions. CARBON DIOXIDE ACIDOSIS AND CARDIAC DYSPNEA 115 TABLE 1 Part 1 Zz - CARBON DIOXIDE No. DIAGNOSIS DATE REMARKS 3 i 3 ea -7i2--4--+ 18 trials ae Beech aes Saee 11 trials tages ; Fiskioll. ccs taserececen dss 8 trials” Peo: op on Bogs aaeae yes ee (rade MP re ee 12 trials Total, direction correct 39 trials Total, direction incorrect 35 trials movement permitted by the position of the subject’s foot). The va- riations in two series of tests are shown in figure 3, other tests giving similar results. There is no correlation either in direction or extent between the active movements and the passive pattern which they were intended to reproduce. The subject’s failure to give any precise or constant reaction to the position or movement of his knee seems to prove that the anes- thesia of the leg is sufficiently extensive to exclude any reflex. control of the accuracy of movement based upon cortical excitations arising from the moving limb. The one remaining possibility of excitation from the limb calls for the postulation of receptors in the muscles which are stimulated by active contraction and not by passive tension. To 178 K. S. LASHLEY test this the relative lengths of the voluntary movements which the subject estimated as equal when different amounts of resistance were opposed to the movement were measured. ; AFFERENT EXCITATION FROM THE CONTRACTING MUSCLES Rough tests were first carried out with different amounts of resist- ance opposed to the subject’s active movements. He was asked to flex his knee so that his foot was drawn back 3 inches from a given po- sition (120 degrees exten- - ~ sion) in which his leg was | held by the experimenter seaport and to indicate verbally em when he had moved his foot through this distance. ERE , The latter precaution was ct [ ; taken to test the relation — between the duration of the motor innervation and —o any excitation of the lan- | guage mechanism which —o might exist. rae In the first trials the foot cous was allowed to move at a rate of only about 1 cm. Fig. 3. Record of attempts to duplicate by ac- per second. The average tive movements a pattern set by passive move- distance moved and re- ment of the leg. The direction and extent of the ported as 3 inches under passive movements are shown by the solid rec- this condition semaine tangles; the active movements of the subject are : shown by those in outline. The pattern was re- with a range from 0.9 to peated before each active movement. 2.5 em, (ten trials). In the next ten trials the pull against the subject’s movements was increased to such an extent that the knee was extended slightly during his attempts to flex it. In these tests an average forward movement of 1.04 em., with a range of 0.5 to 1.5 cm. was reported as a backward movement of 3 inches. During these tests the subject reported that he felt resistance but had increased the force of his movements until he had compensated for it completely. The mechanism by which the resistance was detected was not deter- mined. It may have been a residual joint or tendon sensitivity which ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 179 was too slight to be stimulated by unresisted movements, or to the _ deep muscle sensitivity to pressure, or to certain stimulation from strain on the muscles of the trunk which could not be altogether eliminated. The subject_indieated his knee as the source of stimulation, but was very uncertain. Whatever the locus of excitation, the stimulus was not specific enough to give a clue to the extent of movement of the foot. - Ina third series of tests the active movement of the subject was accel- _ erated by the experimenter, the subject being asked, as before, to stop _ the movement and indicate when his foot had moved 3 inches. The ~ _ average distance moved in ten trials was 28.94 cm. with a range from 26.0 to 30.5 cm. This was practically the maximum extent of move- ment allowed by the position of the subject. TABLE 3 The average length of voluntary movements (flexion of knee) stated by the subject to be equal in extent, when the amount of resistance opposed to the movement was varied RESISTANCE AVERAGE DISTANCE MOVED. (TEN TRIALS) grams cm. “1111 Accelerating 26.0 Limit of movement 0 17.4 266 Retarding 18.4 O44 Retarding 12.4 866 Retarding 9.6 1133 Retarding 8.7 1380 5 Retarding 8.3 SS CS oan 0.0 Knee fortibly extended .... ..-......-. 2 to 10 Extension The same tests were repeated with a better control of the amount of resistance offered to the movement, and with virtually the same The subject’s foot was supported at 120 degrees extension by a spring exerting 100 grams for each 4.5 cm. extension and he results. was asked to draw his foot back three inches. on the spring was varied from 266 to 1380 grams. The initial tension In other tests the ‘ foot was held motionless, in others the knee was extended during the attempt at flexion, and finally the spring was set to flex the knee. As far as possible the subject was kept in ignorance of the procedure. He was asked to indicate verbally when he had carried out the instruc- tions, The results of the tests are shown in table 3. The distances 180 K. 8. LASHLEY moved vary inversely with the resistance encountered but not in direct ratio. The foregoing tests seem to show that there are no excitations from the actively moving limb which are specific enough to give a clue to the extent of the movement. 3 A certain amount of adjustment to the resistance is indicated by the fact that the extent of movement is not inversely proportional to the resistance. The work done in extending the heavier springs (computed as distance times weight lifted) is greater than that performed with the lighter springs. Such a method of considering the data is misleading, however, for it considers that no work is done unless external resistance is opposed to the movement, whereas a certain amount of work is done in the contraction of free muscle, in the stretching of the antagonis- tic muscles, and in overcoming the resistance at the joint. It is impos- © sible to estimate the amount of force expended in this way, but if we as- sume that the work done in moving the leg without external resistance is the equivalent of lifting 50 grams for the distance moved, the total amount of work done in moving against each of the resistances record- - ed in the table is practically the same. There is no justificatidn for assigning this particular value to unresisted movement, but there is also no certain evidence that there was any ee for the dif- ferent resistances encountered. There remains the subject’s statement that he felt resistance and made adequate allowance for it by giving a harder pull. How much of the apparent increase in work done was due to this cannot be deter- mined without more data than are available at the present te upon the internal resistance to movement of muscle and joint. Such imperfect compensation for resistance as the subject may have made is irrelevant to the problem in hand since the subject failed to distinguish the extent of movement. No difference between a flexion of 26 cm. and an extension of as much as 10 em. was detected except in the amount of resistance encountered. The recognition of resistance was evidently not based upon any excitation which could — give evidence upon the direction and extent of movement. . We may conclude that we are dealing with an anesthesia to passive and active movements of the knee which is practically complete for a rate of movement of less than 20 cm. per second within an arc of 45 degrees in each direction from the right angle. With this established it is possible to test the accuracy of voluntary movements within these limits with the certainty that the intensity and duration of the inner- - vation involved in them are not reflexly controlled by afferent excita- tions from the moving limb. ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 181 THE ACCURACY OF ACTIVE MOVEMENTS Direction of movement. In the preceding tests where active move- ments were requested the subject made no errors in the direction of his movements. Only twenty additional trials were given in a formal test of accuracy in the direction of active movements, in all of which the movement was made correctly. In many hundreds of voluntary movements, however, I have never seen the subject make a mistake in direction, except when he misunderstood instructions. It seems certain that the voluntary excitation of a specific group of muscles is possible in the absence of afferent excitation from it. Extent of movement. Two somewhat different methods were used for testing the accuracy in control of extent of voluntary movements. In the first experiments the subject was asked to move his foot through a given distance (an estimated inch, 2 inches, etc.) while the experimenter gave a slight nearly constant support to the backward moving foot in order to control the inability of the relaxed quadriceps to support it against the pull of gravity. The inaccuracy of this method has been considered (page 176). The resistance was applied only to the back- ward movements, probably accounting for the fact that they are slight- ly shorter than the corresponding forward ones (tables 4 and 6), and does not influence the extent of the forward movements, which are equally accurate. When a movement was made its extent was re- corded and the foot was brought back passively to its initial position, usually about 110 degrees extension. None of the subject’s active movements exceeded the rate of 20 cm. per second so that the controls for slow rates of passive movement apply to all the active movements studied. In the first experiment the subject was asked to make ten attempts to move his foot through distances which he judged to be 3, 1, 2 and 3 inches. The averages for the different distances are given in table 4. The movements were all longer than the distance asked for but there was practically no overlapping between the movements estimated as different. The pattern set by the first voluntary movement was duplicated rather accurately in later movements. In these tests the ten trials for each distance and direction were given successively and it seemed possible that this might contribute something to the accuracy of the movement through the establishment of a rhythm of motor excitation. A series of tests was made therefore in each of which movements through distances of from $ to 6 inches were 182 K. 8. LASHLEY TABLE 4 Average distances, each based on ten trials, through which the subject moved his foot when asked to move through a distance which he judged to be that given at the left. Inches were used because he was not familiar with the metric scale ATTEMPT TO MOVE. FOOT AVERAGE DISTANCE MOVED inches cm. 4 Forward 2.88 + 0.37 Forward 1 Forward 3.86 + 0.14 Forward 2 Forward 5.16 + 0.50 Forward 3 Forward 13.42 + 0.87 Forward 2 Forward (later test) 7.00 + 0.27 Forward B Backward 1.46 + 0.25 Backward 1 Backward 3.40 + 0.24 Backward | Backward 7.34 + 0.68 Backward 3 Backward 11537 = 0:57": Backward made successively. The results of five such tests are given in table 5. In only one of the five tests, which began with the shortest and pro- gressed to the longest movement, was a movement shorter than the one preceding it. This case is marked in italics in the table. The subject was not told that he had made an error yet an apparent compensation appeared in the next movement, which is the longest made for that distance. A third series of the same general character was carried out in which all the movements of a given estimated distance were made successive- ly but the different distances to be estimated were taken in irregular order so that an estimate of the absolute distance moved rather than a TABLE 5 Same as table 4 except that the subject attempted to estimate distances from 4 to 6 inches successively DISTANCE MOVED SUBJECT ASKED TO MOVE |. AVERAGE Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 inches cm. cm. cm. cm. cm. cm. 4 Forward 3.0 5.0 3.6) 2.5 5.0 3.82 1 Forward 4.2 10.2 9.0 8.2 7.1 7.74 2 Forward 7.0 13.0 12.2 12.3 11.0 11.10 3 Forward 12:1 17.0 14.5 15.8 13.2 14.52 4 Forward 13.2 13.0 16.5 18.0 17.2 15.58 5 Forward 16.8 22.0 20.4 21.9 18.6 19.94 6 Forward 22.9 24.0 23kSUS PaTeBLO 22.0 23.04 . ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 183 comparison of the length of successive movements was required. The order in which they were taken up was 3, 6, 1,3 and 2 inches. The averages for the different distances with their standard deviations are given-in table 6. TABLE 6 Same as table 4 except that the different distances were estimated in irregular order DISTANCE TO BE ESTIMATED AVERAGE DISTANCE MOVED STANDARD DEVIATION 4 Forward 2.5 = 0.11 0.499 1 Forward 6.5 = 0.44 2.271 2 Forward 11.2 = 0.71 4.357* 3 Forward 16.3 + 1.32 6.187* 7a8 Forward 22.1 + 0.56 2.633 4 Backward 2.9 = 0.13 0.589 1 Backward 4.8 = 0.20 1.016 2 Backward 9.3 + 0.38 1.823 3 Backward 15.9 = 0.41 1.929 6 Backward 20.6 = .0.36 1.782 * The highstandard deviations here are the result of one trial in each group in which the subject reported that the movement had been made when his foot did not move. It is possible that the foot caught against the floor in these cases but as this could not be verified they were included. In all these tests there is a surprising accuracy in the extent of the voluntary movements. Too few trials were given at each distance to lend much significance to the coefficients of variation for the attempts to estimate given distances, but in every series of tests the average distances moved are roughly proportional to the distances which the - subject was asked to estimate. The pattern set by voluntary move- ment could be duplicated with a fair degree of accuracy and the inten- sity of innervation could be graduated in a series of distinctly different steps. Comparison of accuracy of movement with that of a normal subject. For the determination of the variability of movements when the sub- ject was asked to copy a pattern set by his own active movements and for a comparison of this with the normal variability, it was necessary to eliminate any influence which the experimenter might exert in sup- porting the subject’s foot. This. was done by supporting the foot by a light spring so that the knee was partially extended. The subject was-then asked to draw his foot back through a given distance and then 184 K. 8S. LASHLEY allow it to swing forward freely. The carriage of the recording appara- tus was arranged to stop at the limit of the backward movement. Series of twenty or more trials were obtained for estimated distances of 1,2 and 3 inches. The average extent of movement for these three distances with the standard coefficient of variations are given in table 7, and the distribution of variations is shown in figure 4. For compari- son, the results of a similar experiment on an apparently normal indi- vidual are included in the table and figure. This subject, a physician, TABLE 7 Variation in the extent of movements estimated as equal by the anesthetic subject and by a normal individual Bipot AVERAGE EXTENT cuaeemenon ygrererpion NUMBER OF wate | 7 NOTRE || eine | Serna inches cm. cm. Aeainthees b 1 | 4.62 = 0:08 0.254 + 2.12 84 pets me Gea 2 | 14.40 + 0.26 0.122 + 9.40 21 bie eae ee 3 | 23.98 + 0.44 0.167 +16.48 38 it 2.04 = 0.09 0.476 — 0.46 50 Normal subject.. 2 | 11.24 + 0.21 0.193 + 6.00 50 3 22.30 = 0.16 |- 0.076 +14.80 50 44 40 36,32 sa FIE é as tit i a is 2of fae 7 at ! 4 ae rity 16 "4 F yy ; . | 12. 7 : I ; * 8 ! : 5 . Le ; Af \ ; 4 5 ‘. / ¥ : \ , \ ae PS \ 0 2 4 6 8. 10 12) "R16: Ss a ae ee ee oe ae ee 34 CM Fig. 4. Distribution of variations in the length of voluntary movements in the anesthetic and in a normal subject. The ordinates represent the percentage of the movements in each series which were of the lengths given on the abscissae. Anesthetic subject.- - - - Normal subject. ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 185 was selected at random, the instructions and method of supporting the foot and recording the extent of movement were the same as those employed with the anesthetic subject so that the results with the two subjects may be taken as closely comparable. There is a surprising similarity in the results obtained with the anesthetic and with the | normal subject. The movements of both showed wide errors from the distances which they were asked to estimate and the normal subject was not greatly superior to the anesthetic in this respect. In the va- riability of the movements estimated as equal there was no constant superiority of either subject. The normal individual gave more uniform movements for the longer distance but varied more in estimation of - the shorter ones. We may conclude that the anesthetic subject’s control of his move- ments is not significantly less accurate than that of the normal indi- vidual, and it is not clear that for the simple movement studied the afferent impulses from the moving limb contributed anything to the accuracy of movement in the normal subject. The chief mechanism for the control of movement is located in some other body segment than that of the moving organ. THE RELATION OF RATE TO ACCURACY OF MOVEMENT Earlier studies of movement, particularly those of Loeb and Della- barre have indicated that the duration of movement may serve as a clue to its extent, in place of the changing pattern of stimulation from the moving limb. It seemed possible that the subject of the present experiments was depending upon the duration of movement by main- taining a constant motor discharge during time intervals corresponding to the distances through which he was asked to move. In the follow- ing tests the rate of movement was recorded with the distance. The subject’s foot was suspended with the knee extended to 110 degrees by a spring having a coefficient of 100 grams for each 4.5 cm. extension and he was asked to draw his foot back through distances and at a rate suggested by the experimenter. The results of the tests are summa- rized in table 8. From this it will be seen first, that’ the duration of movement is not proportional to the distance when the subject is al- lowed to choose his own rates but that the rate of long movements is less than of the short ones (tests A, B and C); second, that the move- ments may be made of equal extent, although the rate is quite different (tests D and F), third, that, except in test A, the variability in the 186 K. 8. LASHLEY time of movement is considerably greater than that of its extent; and fourth, that the variation in both extent and time of movement de- creases with increasing rate. The experiments thus show a degree of: independence in the rate and extent of movement which precludes the possibility that the extent of movement is determined merely by the control of the duration of the excitation of motor pathways. They indicate, on the contrary, that there is a control of the intensity of motor discharge which is independent both of the duration of excitation and of the effects of the discharge upon the effectors. The increase in accuracy with increased speed is in accord with the results obtained by Woodworth (6) in his study of the accuracy of automatic move- TABLE 8 Variation of the rate of movement compared with variation in the extent of movement. In tests A, B and C the subject was allowed to select his own rate of movement; in test D he was asked to move quickly, in E, still more rapidly, and in F, to jerk his foot back as quickly as possible. The averages are each based on ten trials STANDARD STANDARD AVERAGE DIS- | COEFFICIENT AVERAGE COEFFICIENT DISTANCE oped TANCE OF TIME Or REQUIRED VARIATION VARIATION cm. seconds inches A ee aus 4.27 0.334 0.69 0.272 1 AS: Seo ee pees 8.05 0.156 2.19 0.216 2 OR erica ts 11.52 0.153 3.97 0.398 3 | Be ap maerysenet oe 12.59 0.121 1.44 0.324 3 Bho otto pee 9.87 0.119 0.80 0.264 3 Gye aanipoeroaees 11.41 0.115 0.68 0.235 3 AVOTRDC ate iene Satis tant 0.166 0.285 ments and confirms his assumption that rapidity of normal move- ment interferes with its accuracy only by reducing the influence of the “current control,” of the excitations aroused by the moving organs. The time records showed further that the slow movements were not the result of a single muscular contraction but consisted, in prac- tically every case, of a series of from two to five successive contractions resulting in alternate acceleration and retardation of the movement. This furnishes additional evidence against a temporal control of move- ment and also raises the question whether an initial set is adequate to account for an accurate movement which is excited by a series of inner- vations, without some controlling mechanism which is active -contin- uously during the course of the movements. ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 187 REACTIONS TO ERROR OF MOVEMENT In occasional instances the subject stated that a given movement was longer than he had intended. Records of only eight such movements have been obtained but in every case the recorded movement was considerably greater than other movements of the series in which it occurred. The recognition of such movements is ascribed by Wood- worth (6) to sensory elements arising from the movement. If, as seems established by the tests recorded, the subject ‘of the present experi- ments was anesthetic to movements of the knee, the detection of error must be ascribed to some mechanism other than the receptor system of the moving organ. This demands a distinction not only between the initial set or intention of movement and the final adjustment due to sensory stimulation, but also the recognition of a third factor in the control of movement, the capacity for reaction to the intensity of inner- vation which is independent of both the initial set and the excitations from the moving organ. This suggests the old doctrine of the feeling of innervation, although an alternative hypothesis must be considered. This is outlined on page 193. EFFECTS OF FATIGUE In some of the earlier tests, after many repetitions of a given move- ment, the subject complained of feeling resistance to his movements and at the same time increased their length. It seemed that this might be the result of fatigue and a number of series of movements was there- fore made to test this more thoroughly. The subject was required to repeat a movement of a given length from 20 to 85 times. Resistance to the movements was offered by a spring which drew the foot back to the starting point after each movement. Table 9 shows the results of this test. In each case repetition of the movement led to a considerable increase in its extent. The progression in the length of movement in two series is shown in figure 5 which is based upon the average of successive groups of five trials. During the later trials of the long series the subject stated that he felt tired and that it seemed to require a greater effort to move his foot than had been necessary at first. We can scarcely interpret such data at present. The progressive increase in the length of movements estimated as equal seems almost certainly the result of the frequent repetition of the movement.. From the subject’s statement it seems probable also that the increase resulted 188 K. S. LASHLEY from some feeling of resistance or of increased effort necessary for the movement, which led to an over compensation. stimulation leading to this compensation offers an interesting problem. It does not seem probable that with the extensive anesthesia to all other forms of stimulation there should still persist a normal sensitivity to chemical changes in the muscles which give rise to the feeling of fatigue. The alternative seems to be some cortical mechanism by which the 28 The source of the 2 VA i IN é | V 22 A 2U 18 16 Pa = , 14 B Ly 12 10 3 6 | ee 4 — ee] | 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Tnals Fig.5. Theeffects of fatigue upon the length of movements estimated as equal. The ordinates represent the average length in centimeters of successive move- ments taken in groups of five. cm. extension TABLE 9 The effects of fatigue upon the length of movements intended to be of equal extent. The subject pulled against a spring which exerted a resistance of 100 grams per 4.5 LENGTH AVERAGE EXTENT OF MOVEMENT OF MOVEMENT INCREASE TOTAL NUMBER ATTEMPTED ; ; ? OF TRIALS First five trials . Last five trials inches cm. cm. per cent : 1 2.9 6.4 120 85 2 | 13.0 15.1 16 20 3 19.6 27.4 39 40 ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 189 increase in the threshold of excitability of the motor cells resulting from fatigue directly modifies the behavior of other action systems besides the one which is immediately involved in the movement. THE INTERACTION OF DIFFERENT MUSCLES IN THE CONTROL OF MOVEMENT Owing to the lack of adequate means for determining the degree of tension of the muscles of the subject’s thigh, it was not possible to determine the relative functions of the flexors and extensors in con- trolling movemént, but a few crude observations indicate that much of the normal complex interplay of the muscles is retained. In quick flexing movements a preliminary contraction of the quadriceps exten- sor is detectable although the inertia of the subject’s foot prevents the appearance of a form of reaction movement similar to that first described by Smith (5) for finger movements. There is also, seemingly, an increase in the tension of the quadriceps as the limit of movement is approached. . When the subject is asked to contract both flexors and extensors, to “make his leg tense,” an apparent fluctuation in the intensity of innervation results in oscillations of the foot, yet a given degree of extension is maintained much longer than when he is merely asked to hold his leg extended as in the experiment described on page 174. In the latter case only the extensors are in active contraction so it seems that the simultaneous excitation of both flexors and extensors permits of a steadier and longer motor discharge than is possible when only one set of muscles is innervated. The complete loss of the tendon reflexes and the great reduction in the tone of the muscles makes a reciprocal innervation of the antag- onistic muscles improbable and suggests that the interaction of antag- onistic muscles in the control of movement is regulated by some part of the nervous mechanism cephalad to the spinal segment from which the muscles are innervated. THE INFLUENCE OF TRAINING The final question arises as to whether the condition of control of movement in this subject is comparable to that of a normal individual or whether some mechanism of control has been developed by practice which was not functional at the time of the lesion. The present ob- servations were made five years after the spinal injury but the subject’s 190 K. S. LASHLEY history scarcely supports the view that his control of movement has been reacquired by practice during this time. For the first year after injury he practiced daily walking with crutches for two hours but he made so little progress that he became discouraged and gave up all attempts to recover the lost functions, spending his time either in bed or in a wheel-chair. Except for this relatively brief practice, which aimed only at a visual control of movement, there is no history of any activity which could develop an accurate control of short, slow movements. DISCUSSION ba Every adaptive movement seems to involve three physiologically distinct processes when we attempt to analyze its neurological mech- anism. These are, 1, the initiation of motor excitation resulting in muscular contraction; 2, its continuation by a series of disturbances propagated either in the central nervous system, or reflexly as a result of the motor discharge; 3, the cessation of excitation of the protago- nists and excitation of the antagonists. The first of these has much in common with the simple reflex twitch and is not of moment in the problem of the control of accuracy of movement unless in some way the extent of movement may be determined by the initial excitation of the motor pathway. The continuation of the movement implies the production of a series of tetanic contractions arising from succes- sive nerve impulses which have been excited by a single momentary stimulus. The duration of the tetanus varies with the extent of the contraction. Curiously enough, no more than vague suggestions have been advanced. to account for this change from brief to long-con- tinued excitation. The demonstration by Forbes and Gregg (2) that a single strong stimulus may induce the propagation of two or more waves of disturbance in the nerve may furnish the clue to the contin- uation of movement, but the increase in duration of excitation which they have shown is very slight and seems inadequate to account for move- ments continued for several seconds. Upon this the present work gives no data except the probable elimination of circular reflexes by which a contracting muscle might stimulate its own contraction; a pos- sibility perhaps adequately disproved already by operative experiments. The cessation of movement is no more explicable today than its con- tinuation. To be useful, a movement must end with the attainment of a result which is specific for a given stimulus; it must reach a deter- mined distance, exert a determined force, ete. In many cases the stim- a ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 191 ulus to cessation or inhibition of movement evidently comes from exteroceptors and does not directly involve the receptors of the moving organs, but in the experimental duplication of a pattern set by active movement no extero-stimulation is present to determine the cessation of movement and its duration and extent must be determined wholly within the organism, either by excitation of proprioceptors or by proc- esses carried out wholly within the central nervous system. Several alternative hypotheses to account for the cessation of movement in such cases have been formulated in such a way as to make experimental test possible. The first of these appeals to the local sign, assuming that the extent of movement is determined by a change from one pattern of stimula- tion in the moving organ to another. It seems to demand a vast if not infinite series of specific reactions to different patterns of stimula- tion. (Ladd and Woodworth, (3) 408.) A second assumes that the extent of movement is determined by the amount of excitation coming from the moving organ, the amount varying with the extent of movement. The hypothesis seems to de- mand the assumption of a priming or preliminary integration of efferent neurones before the initiation of movement. Third, there may be purely intracortical control by some spreading of excitation, of whose nature we can form no concept at present. Some evidence bearing upon the réle of excitations from the moving organ in the control of accuracy has been obtained by other experi- menters. Lack of space prevents any extensive summary of the lit- erature and the thorough review of Woodworth (7) makes this un- necessary. Indications of the relative independence of motor discharge from direct control by circular reflexes come chiefly from three sources. (1) The discovery by Bowditch and Southard (1) that the reproduc- tion by movement of a pattern distance set by visual stimulation is more accurate than that of one set by kinesthetic gives evidence of the importance of the preliminary set for accurate movement, although it throws no light upon the mechanism of the preparation for movement. (2) The observations of Loeb (4) upon the inequality of simultaneous movements of the two arms indicate that the movements of the two have a common source of control, the action of which is relatively independent of the extent of movement. (3) In his study of automatic movements Woodworth (6) found that the accuracy of movement increased in direct ratio to the speed. From this he concluded that such movements are controlled by the initial set. 192 K. S. LASHLEY The significance of the results of the present study depends upon the validity of the evidence obtained for anesthesia to slow movements. The various tests recorded have shown that: (1) The subject is unable to determine the position of his leg (except occasionally when the knee is hyperextended); (2) he can not detect the extent, duration, direc- tion, or even the presence or absence of passive movements of the knee, if such movements are made at a rate of less than 20 cm. per second; (3) he is unable to detect the movements and changes in degree of active contraction of the muscles of the thigh occurring during his attempts to hold the knee in a given state of contraction against the pull of gravity; (4) he is unable to determine whether or not his at- tempts to bend his knee have resulted in movement when various amounts of resistance are opposed to the movement. In the face of this evidence we can scarcely hold that he retains a sufficient sensi- tivity to movement to make possible a reflex control of the extent of movement or a distinction between the extent of actual movement and the intended extent, based upon any stimulation of the receptors of the leg. We are rather forced to the conclusion that the phenomena observed are independent of afferent excitation from the moving organ. The experiments have shown that the subject is able to control the extent of his movements with almost normal accuracy, to vary the speed and extent of movement independently, and to make rhythmic alterations of flexion and extension. The evidence for anesthesia makes it necessary to assume that all these activities may be carried out in the absence of excitation from the moving organ. The mech- anism of control must be sought either in the central nervous system or in some other body segment. Data on the accuracy of movement at different rates show that in the present case its extent is not deter- mined solely by its duration. This makes it necessary to assume some regulation of the intensity of motor discharge which is independ- ent of its duration. Is this determined immediately by the incoming stimulus to movement resulting in a ‘‘set”? by which a given intensity of motor excitation is aroused explosively without further possibility of control, or is there such a spreading of the motor impulse that some control of its intensity is possible during the discharge? A certain amount of evidence bearing upon this question has been obtained. It * It would be profitless to discuss here the nature or existence of attitude or set. Whether it be a priming of reflex pathways, the assumption of an altered muscular tonus, or what not, it seems distinguishable from the condition in which control of movement occurs after the initiation of the motor impulse. ACCURACY OF MOVEMENT IN ANESTHETIC LIMB 193 comes from three sources. First, very slow movements were not made by the subject as a steady contraction of the muscles but by a series of impulses following each other at intervals of one-tenth second or more. Second, the subject was able to detect the excessively long movements of a series and to state that they were longer than he had intended to make them Third, he complained of fatigue at the same time that he showed objective signs of some disturbance in the normal conditions of movement, while chemical sensitivity to fatigue products in the muscles of the anesthetic leg seems improbable. These points are not at all firmly established by the data at hand but all indicate that there is a spreading of the motor excitation which plays a part in the control of movement and may perhaps lead to some phenomenon such as that described as the feeling of innervation. The hypothetical explanation of such a condition which is most open to experimental test is that assuming a spread of the motor impulse to other action systems with reflex control from them. It may, perhaps, be tested by a study of the possibility of controlled movements within intervals less than the minimum cortical reaction time. Experiments to this end are now in progress. As evidence for the importance of the initial set, on the contrary, there is the fact, emphasized earlier by Woodworth (6) for automatic movements, that the accuracy of movement increases directly with the rate. The evidence at hand is not adequate to rule out either alternative. SUMMARY Active movements of the left knee were studied in a subject having “a complete anesthesia to movements of this joint. Evidence bearing upon the nature of the control of movement in such a condition suggests the following conclusions. 1. Accurate movement of a single joint is Debio in the absence of all excitation from the moving organs. The interaction of the various muscles concerned in the movement is not obviously different from that found in normal subjects. 2.‘In contrast to reflexly controlled movements, the accuracy of movements under such conditions is in direct ratio to their rate, that is, within the limit tested, the quicker the movement the more accu- rately it is made. 3. The control of accuracy of the movements is Sasiiely independ- ent of their duration; movements of different length do not result from a uniform excitation continued for varying time intervals but from variations in the intensity of motor discharge. 194. K. 8S. LASHLEY 4. It is probable that a control of the intensity of motor discharge after its initiation is possible in the absence of excitation from the organs activated. 5. The normal phenomena of fatigue occur when it is highly probable that the chemical sensitivity of the fatigued muscles is reduced. BIBLIOGRAPHY (1) Bowprrca anp SoutTHarD: Journ. Physiol., 1881, iii, 232-245. (2) Forsns AnD Greaa: This Journal, 1915, xxxix, 172-235. (3) Lapp anp Woopworts: Elements of physiological psychology, New York, 1911. (4) Logs: Arch. f. d. gesammt. Physiol., 1897, xli, 107-127. (5) Smrra: Mind, 1903, xii, 47-58. (6) Woopworta: Psych. Rev., Monograph Supplement, 1899, iii, 13, 114. (7) Woopworts: Le mouvement, Paris, 1903. THE CARDIO-SKELETAL QUOTIENT W. L. MENDENHALL From the Laboratory of Pharmacology in the Dartmouth Medical School Received for publication February 25, 1917 _ It has long been known that the strength of faradic stimulus neces- sary to provoke response in heart tissue is greater than that required to elicit activity in skeletalmuscle. The exact relationship existing between these two tissues with respect to strength of stimulus necessary to arouse contraction has not hitherto been definitely established. Doubt- less this has been due to unreliable means of measuring faradic stimuli. In the investigation of another problem in this laboratory it became necessary to make a preliminary study of the relationship existing be- tween the threshold of the frog’s ventricle and the threshold of its gas- trocnemius muscle. In this investigation the thresholds were determined by means of Martin’s method of quantitative faradic stimulation (1). 8 units were determined in all cases. Particular care was used in selecting healthy frogs for the experiments. Each frog was pithed and weighed. The heart was exposed by a wide free incision into the anterior body wall. The pericardium was slit throughout its length and the frenum was cut. A looped silk ligature was next thrown around the heart. Posteriorly it was in contact with the white crescent, anteriorly it was tied so that the aortae were included in it. In some preliminary experiments it was found that the ordinary Stannius ligature usually stopped the heart but that it would not remain stopped long. enough to determine the threshold, or would start beating the moment the threshold had been reached. If in tying the ligature the aortae were included, the heart remained quiescent for a time sufficient for the determination of the threshold. The electrodes used were the ordinary platinum electrodes of the Harvard Apparatus Company. The points were sharpened and adjusted 2mm. apart. After the heart ligature was tied the frog was _ placed under the recording apparatus. The tip of the ventricle was caught in a clip which was connected with a light heart lever writing on a slowly revolving smoked drum. Then the electrodes were thrust 195 196 W. L. MENDENHALL directly into the ventricle about 2 mm. from the auriculo-ventricular groove. The stimulation caused by placing the electrodes was usually followed by several rhythmical contractions of the ventricle. It was noted that in the first few minutes after placing the electrodes the threshold varied considerably, but that after about fifteen minutes the response became quite uniform and remained so for an hour and fre- quently longer. As soon as the heart showed uniformity of response the threshold was quickly determined. The usual routine was followed of finding first the threshold with no added resistance in the secondary — circuit and then finding the thresholds with 10,000, 20,000 and 30,000 ohms in the secondary circuit. The resistance of the tissue was then measured by the Kolrausch method. The heart was not moistened during the experiment since it was noted that the irritability appar- ently shifted rapidly with each application of solution. When neces- sary to prevent rapid drying, a moistened filter paper was suspended so . as to surround the heart without touching it and thereby serve as a moist chamber. Usually the experiment was completed within three- fourths of an hour and during this time the response was quite con- stant. One heart which was not moistened after it was connected with the recording apparatus maintained a uniform irritability over a period of two hours. The experiments were all performed at a tempera- ture of 20°C. After the heart thresholds were determined, a slit was made in the skin over the gastrocnemius muscle and the same electrodes that were used in the heart were thrust directly into the middle of the be’ly of the muscle. A line joining the two points of the electrodes would be at right angles to the long axis of the muscle. By this method it was found that the muscle became uniform in its response in about the same time that the heart required to become uniform in response. The least visible response of the muscle was taken as the indication of the threshold. In some experiments in which the thresholds were re- corded on a smoked drum it was found to correspond closely with the visible contraction observed in the intact muscle. Following the de- termination of thresholds with varying resistances in the secondary cir- cuit, the tissue resistance was measured by the method previously mentioned. Table 1 shows the results obtained. The average for the threshold of the ventricle is seen to be 191.0 Z units. The average of the 6 units is 98.3. In each instance the ratio of 6 to Z was determined. The . average is 0.50. The average threshold for the gastrocnemius muscle is 33.3 Z units, 16.1 B units. It is of interest to note that the threshold a ee ee ee ee es CC CARDIO-SKELETAL QUOTIENT 197 of the gastrocnemius in this series is greater than that reported by Martin in his series of eighteen experiments (2). This is due to the fact that the experiments-were not performed under comparable condi- tions. Martin has already called particular attention (3) to the im- TABLE 1 -Showing thresholds of heart and gastrocnemius muscle and their relationship as expressed by the cardio-skeletal quotient HEART GASTROCNEMIUS Cee ide uaa z 8 é z 8 e Z 8 4 110.0} 50.6] 0.46 10 126.8} 63.6] 0.50 | 30.0 | 12.4 | 0.41 | 0.23 | 0.19 27 134.5| 54.6] 0.40 | 12.1 4.9 | 0.40 | 0.08 | 0.08 12 138.2| 46.5] 0.33 | 40.2 | 15.2 | 0.37- | 0.29 | 0.32 24 138.2|} 81.4] 0.58 | 29.2 | 17.1 | 0.58 | 0.21 | 0.21 26 142.1 90.5} 0.63 26.1 16.3 0.62 0.18 | 0.18 25 147.1} 70.3] 0.47 | 26.4 | 13.0 | 0.49 | 0.18 | 0.18 28 147.1| . 8.5} 0.58 | 30.0 | 17.1 | 0.57 | 0.20 | 0.20 1B 151.8| 57.0} 0.37 | 40.2 | 12.4 | 0.30 | 0.26 | -0.21 8 151.8| 72.3] 0.47 | 45.0 | 20.9 | 0.46 | 0.29 | 0.28 2 160.1| 85.8] 0.53 6 161.1| 69.0] 0.42 | 26.1 | 10.1 | 0.38 | 0.16 | 0.14 30 166.5 | 99.0! 0.59 5 187.2| 112.2] 0.59 | 32.2 | 14.9 | 0.46 | 0.17 -| 0.13 71 1880!) 98.0] 0.49 | 13.5 7.0 | 0.52 | 0.07 | 0.07 17 195.5| 69.0] 0.35 | 32.4 | 11.4 | 0.35 | 0.16 | 0.16 1 198.7} 98.0| 0.49 15 212.5 | 111.9] 0.52 | 22.4 | 12.4 | 0.55 | 0.10 | 0.11 22 220.5 | 102.7} 0.46 3 222.0 | 125.2] 0.55 | 23.0 | 12.2 | 0.58 | 0.10 | 0.10 9 246.8] 110.9| 0.44 | 67.0 | 74.1 | 0.50 | 0.27 | 0.30 18 265.9| 139.8] 0.52 / 20° | 276.5| 198.5| 0.71 | 61.8 } 38.1 | 0.61 | 0.22 | 0.19 29 332.8 | 172.5| 0.51 | 30.4 | 16.4 | 0:53 | 0.09 | 0.09 14 355.5] 197.9| 0.55 | 45.8 | 21.5 |. 0.46 |.0.12 | 0.10 Average | 191.0] 98.3| 0.50 | 33.3 | 16.1 | 0.48 | 0.17 | 0.17 * portance of using similar conditions if there is a desire to duplicate the experiments of another observer. In this series the electrodes were 2 mm. apart and placed in the middle of the muscle. This was done in order that they might be used in the heart and thus provide uni- 198 , W. L. MENDENHALL formity in the application of the stimulus to both tissues. In Mar- tin’s series an electrode was thrust into each end of the muscle, the kathode being nearest the origin. Martin used isolated muscles, in this series the muscle was left in situ. The average 6 unit in this series is 16.1, in Martin’s series it is 8.1. Some experiments which will be discussed later were arranged in which the technique used by Martin was employed. The average 6 unit was then found to be 9.0. Attention is called to the ratio of the gastrocnemius 6 to Z. It is 0.48. This is practically the same as Martin found in his series of eighteen experiments upon the same muscle. Martin found the ratio of B to Z to be 0.49, and the average variation from this ratio was 15 per cent. In this series the average variation is 15.6 per cent. In the heart series the ratio of 8 to Z is 0.50. The average variation from this is 14 per cent. The cardiac ratio Fand the skeletal muscle ratio : are evidently of the same order of magnitude and from the evidence may be considered identical. In order to determine the relationship existing between the heart Z units (HZ), and the gastrocnemius Z units (GZ), also the heart 8 units (Hg) and the gastrocnemius 8 units (G8), the ratio of GZ to HZ and the ratio of GB to HB were determined in each instance. The average of the ratio on was found to be the same as the ratio ig Itis0.17. In the individual experiments it is noted that the ratios were usually equal or nearly so. This equality in the ratios GZ: HZ and G®: HB was so constant that it was thought to be indicative of accuracy in the deter-- mination of the thresholds. It seemed desirable to use some expres- sion to signify the relationship existing between the threshold of the heart and the threshold of the gastrocnemius so the term Cardio-skele- tal Quotient is suggested for that purpose. It will be used throughout this discussion to signify the ratio of the threshold of the gastrocnemius ° muscle to the threshold of the ventricle. Thus there is a cardio-skele- tal quotient for the Z units, indicated by sa and a cardio-skeletal quotient for 6 units, indicated by iy Theoretically es should equal Gp He and in the present series this was found to be true. The average quotient in each case is 0.17. CARDIO-SKELETAL QUOTIENT 199 The cardio-skeletal quotient may serve as a means of making correc- tions where obviously only one determination is in error out of the four usually made. Thus if the cardio-skeletal Z quotient is 0.21, then in accordance with the evidence obtained in this series the cardio-skeletal B quotient should also equal 0.21. If in making the determinations it is found that the cardio-skeletal 8 quotient is widely divergent from the cardio-skeletal Z quotient, it is possible by referring to the averages of the series to determine whether the error is in the Z units or in the 8 units. If the error is in the 8 units it may be ascertained, by reference to the averages in the series, whether the error is in the heart 8 units or the gastrocnemius 8 units. Finding that the error is in the heart 8 GZ _ Gp units the equation —~ HZ ~ He of Hg. An example will serve to make the point clear. In experiment 24 in table 1, the H8 was determined by the above equation. As originally determined the units and ratios were as follows: HZ 138.2, 8633.0, 8 0.23; GZ 29.2, Ge 17.1, 2.0.58; 2 0.21, 22 0.51. Referring Z ? ’ ? 7, ’ HZ Z ? Hs B in the table to the average cardio-skeletal Z quotient and the average cardio-skeletal 8 quotient it. will be seen that in this experiment the cardio-skeletal Z quotient is approximately that magnitude which is normally present. The cardio-skeletal 8 quotient however is plainly an error. Apparently there was a mistake in calculating the 6 units. ‘The G8 is 17.1. Reference again to the averages of the series reveals a may be used to determine the real value close correspondence in the figures. The gastrocnemius ratio Fis seen to be 0.58. This while differing from the average by about 20 per cent may be considered within the range of error. Inspection of the heart 6 units and the heart ratio f reveals the source of the error. The heart 8 is 33.0. This is a variation from the average as shown by the series of 66 per cent. The heart ratio f is 0.23. This is 54 per cent divergent from the average as shown by the series. Obviously therefore the error is in the 8 units of the heart. By use of the equation a = ae in which X is the H@ sought, it is found that the Hf is 81.4. This correction necessarily implies that the heart ratio of 8 to Z is always the same as the gastrocnemius ratio of 8 to Z. That such is the case is indicated by the table. This method of correction applies necessarily to 8 units. 200 W. u. MENDENHALL The evidence presented here indicates that the cardio-skeletal quo- tient is a fairly constant quantity, the average variation being a little under 30 per cent. This apparent constancy of the cardio-skeletal quotient suggested the possibility of its utilization as a means of study- ing the effects of various agents upon the heart, using as an index of the effect an alteration in the average cardio-skeletal quotient. In order to test the validity of the quotient, it was decided to make some practical applications of it. Use was made of the gastrocnemius ex- periments in table 1. Taking each experiment the gastrocnemius Z units and 6 units were divided by the cardio-skeletal quotient 0.17. The results were taken to represent respectively the heart Z units and the heart 6 units. The following averages are the results: Z units 196.1, B B units 95.1, 7 0.48. Reference to the actual determinations in table 1 shows the very close agreement of the averages. Individual experi- ments revealed in many instances wide variations from the actual de- terminations. This method clearly would not apply to one experiment but to a series. ; Another test of the validity of the quotient was made. In table 1 it will be noted that there are six experiments in which the units for the gastrocnemius were not determined or were discarded because of faulty technique. In this case the corresponding heart units were multiplied by the cardio-skeletal quotient 0.17 to determine the theoretical values of the gastrocnemius muscles. Table 2 shows the results obtained. Comparison with table 1 shows again a very close agreement between the TABLE 2 Theoretical determinations of gastrocnemius thresholds (omitted from table 1) from ’ the corresponding heart thresholds by means of the cardio-skeletal quotient 0.17 HEART GASTROCNEMIUS EXPERIMENT NO. 2 B g z “B p 4 110.0 50.6 | 0.46 18.7 8.6 0.46 2 160.1 85.8 | 0.53 27.2 14.5- 0.53 30 166.5 99.0 | 0.59 3.3 | 168 0.59, 1 198.7 98.0 0.49 33.7 16.6 0.46 22 220.5 | 102.7 | 0.46 37.4 17.4 0.46 ald 265.9 | 139.8 | 0.52 45.2 23.7 0.52 -Average......| 186.9 95.9 | 0.50 31.8 16.2 0.50 vay fio A Te a a CARDIO-SKELETAL QUOTIENT 201 average theoretical determinations in these six experiments and the average actual determinations in nineteen experiments of the series. Evidently the cardio-skeletal quotient is a reliable guide for the de- termination of threshold units where the conditions of experimental procedure are comparable. In case the methods are not uniform they may be reduced. to uniformity in some instances and the cardio-skeletal quotient still be applied. Thus in the series of eighteen experiments upon the frog’s gastrocnemius reported by Martin the average Z unit and 6 unit were 50 per cent lower than the same units reported in the present series. But attention has already been called to the different technique used by Martin. In his series the average Z units was 17.1, the average 8 units was 8.1, the average 4 0.49. In the technique em- ployed in this series it required approximately six times the strength of _ stimulus to make the heart contract that it required to make the gas- trocnemius contract. Using the technique employed by Martin it would require twelve times the strength of stimulus to make the heart contract that it required to make the gastrocnemius contract. Since by the latter method results are obtained that are 50 per cent lower than the same units obtained in the present series there would also be a lowering of the cardio-skeletal quotient by 50 per cent. Difference in the manner of application of the stimulus led to a corresponding dimi- nution in the strength of stimulus necessary to arouse the activity of the tissue. Accordingly it was assumed that the cardio-skeletal quotient in Martin’s series was 50 per cent less than the same quotient in the pres- ent series or 0.085. By use of this quotient theoretical determinations of the heart units in Martin’s series of gastrocnemius thresholds were made. Thus each gastrocnemius Z unit and 8 unit was divided by- the assumed cardio-skeletal quotient 0.085 and the results taken to repre- sent respectively the heart Z unit and 6 unit. The results were as fol- lows: average Z unit 195.0, average 6 unit 95.3, average £0.49. Com- parison with the general averages in table 1 shows a striking closeness of the results. In order to determine if the method of obtaining thresholds in the present series accounted for the difference in the same values as re- vealed by Martin’s series, a few experiments were arranged in which the technique was the same as Martin’s. The results are shown in table 3. That the assumption in regard to lowering of the cardio-skeletal quo- tient because of difference in technique is tenable, is clearly shown by 202 W. L. MENDENHALL the results. The heart units are of the same magnitude as those of the present series, but the gastrocnemius units are of the same order of magnitude that Martin found. The ratio e is somewhat higher than those of the series. This could not be accounted for except that the frogs had just begun dying rapidly and some unusual disturbance may have been present which slightly raised the ratio of 8B toZ. Themost interesting fact is the close agreement of the determined cardio-skeletal quotient with the assumed one for Martin’s experiments: 0.085 and 0.09. In table 4 is presented a summary of the results obtained where actual determinations were made, also where theoretical determina- tions were computed. TABLE 3 Showing thresholds of heart and gastrocnemius muscle using Martin’s technique for the gastrocnemius determinations. Relationship expressed by the cardio-skeletal quotient HEART GASTROCNEMIUS oe een EXPERIMENT ee Se Shik an es 6 4 A | 147.0] 71.6] 0.48 | 14.9 | 8.8 | 0.56 | 0.10 | 0.11 C | 152.7] 83.9] 0.58 | 16.7 | 9.5 | 0.57 | 0.10 | 0.11 B | 164.5] 80.5] 0.48 | 15.7 | 8.6 | 0.54 | 0.09 | 0.10 E | 191.3] 99.1] 0.51 | 17.5 | 9.4 | 0.53 | 0.09 | 0.09 G | 197.8] 113.7] 0.51 | 18.0 | 9.6 | 0.53 | 0.09 | 0.08 D | 204.2] 116.4] 0.56 | 16.9 | 9.0 | 0.53 | 0.08 | 0.08 F | 206.4] 117.7| 0.58 | 17.7 | 9.2 | 0.51 | 0.08 | 0.07. Average| 180.5] 97.5| 0.53 | 16.7 | 9.0 | 0.53 | 0.09 | 0.09 The curious effect of removing the heart entirely from the body was shown in one experiment. It was found impossible to stop the heart by means of the ligature so it was isolated by cutting through the auri- cles. It stopped promptly. Upon stimulating it was found to require 2496 Z units to obtain a response. The resistance of the ventricular tissue included between the stimulating electrodes and thegastrocne- mius tissue included in the same area was the same. ‘The average of each was 1500 ohms. The evidence presented in this series of experiments seems to justify the conclusion that the cardio-skeletal quotient is a body constant. Also this: quotient may be utilized in studying the effects of various CARDIO-SKELETAL QUOTIENT 203 TABLE 4 Summary of average results obtained by actual and theoretical determinations CARDIO-SKELETAL HEART GASTROCNEMIUS f QUOTIENT PROCEDURE ey Pe N 0. 0. 0. aver- | Z B e aver- | Z B ed aver- | Z B aged aged aged |} sa determina- } 25 1191.0/98.310.50/ 19 |33.3l16.110.48| 19 0.17 |o.17 Theoretical deter- || minations of heart units in ta- ble 1, by means |\ 19 |196.1/95.1/0.48 of the gastrocne- mius experiments and the cardio- skeletal quotient. |) Theoretical deter-. minations of gas- trocnemius thresholds (omit- Syeee table 1) 6 |31.8|16.2|0.50 by means of cor- ‘ paw fs responding heart experiments and the cardio-skele- tal quotient. Theoretical deter-— minations of heart thresholds from Martin’s series of gastroc- ee Pen” |) 18 |195.0195.3/0.49| 18 |17,1) 8.1]0.49] 18 |0.085/0.085 ments, by means of cardio-skeletal quotient 0.085. Figures in italics are Martin’s. i Actual determina- tions using Mar- tin’s technique |? 7 |180.5|97.5)0.53 7 |16.7| 9.010.53| 7 {0.09 (0.09 with the gastroc- nemius muscle. || 204 W. L. MENDENHALL agents upon the heart, and conversely it may be used to study the ef- fects of various agents on the gastrocnemius muscle. It may also be used as a means of correcting 6 units when there is a series to obtain averages from. SUMMARY 1. The threshold stimulus of the frog’s ventricle as shown by this series is 191.0 Z units, 98.3 8 units, and the ratio is 0.50. 2. The threshold of the frog’s gastrocnemius by the method used in this series is 33.3 Z units, 16.1 6 units; and the ratio e is 0.48. 3. The cardio-skeletal quotient is defined as the ratio of the gastroc- nemius threshold to the threshold of the ventricle. By the method employed in this series it is 0.17. 4. Utilization of thé cardio-skeletal quotient as a means of studying conditions that affect the heart is indicated. 5. By the method described in this series it is shown that it requires six times as strong a stimulus to make the frog’s heart contract as it does to make the frog’s gastrocnemius contract. BIBLIOGRAPHY (1) Martin: The measurement of induction shocks, New York, 1912. (2) Martin: Loe. cit., 85. (3) Martin: Loe. cit., 87. - CONTRIBUTIONS TO THE PHYSIOLOGY OF THE STOMACH - XLI. Tue ALLEGED INFLUENCE OF THE REMOVAL OF THE SALIVARY GLANDS ON THE SECRETION oF GASTRIC JUICE A. M. SWANSON From the Hull Physiological Laboratory, University of Chicago Received for publication February 25, 1917 There are many factors involved in| the secretion of gastric juice. Besides the reflex or psychic factor, we ‘have the saliva itself, (Pavlov (1) ), the secretagogues of the food (Schiff, Pavlov, Bayliss and Starling), and of the pyloric mucosa (Edkins). Tarulli and Pascucci (2), report gastric secretagogues in the spleen. In addition to these Keeton and Koch (3) tested various organs for the presence of secretagogues (“gas- trin’’) and found some positive, while others, such as the submaxillary glands, gave negative results. The present work was carried out to determine whether or not such a hormone exists in the salivary glands (and consequently affects the secretion of gastric juice by way of the blood), as has been affirmed by some but denied by others. Since the secretion of saliva is the initial process of the digestive act, one might expect some results on digestion, on health and on the secre- tion of gastric juice from extirpation of the salivary glands; but in an animal like the dog, which bolts its food, and in which the ptyalin is absent, (Carlson (4) ), with the secretion of the sublingual glands strongly alkaline, and with no evidence of an adaptation of the charac- ter of the saliva to diet (Garrey (5) ), the effect of the. removal of the salivary glands on gastric secretion is brought into question. Hemmeter, working on dogs, reached the conclusion that a hormone is present in the salivary glands, and that the absence of the salivary glands leads to diminished gastric secretion and peptic digestion. He did not discuss the effect of their removal on the acid secretion of the gastric juice. This depressed gastric secretion, he states, can be brought back almost to normal by feeding extracts of the salivary glands. In 205 206 A. M. SWANSON his work published in Science (6), he states that he used the Pavlov pouch, while in the conclusions from his work as published in the Trans- actions of the American Gastro-Enterol. Assoc. (7) and Biochem. Zeitschr. (8), he states that the work was done by use of the simple gastric fistula. He adds that ‘in some cases there is a secretion after removal of the glands” and concludes that this is due to one of the three following factors: (a) That the lobules of the parotid are not completely removed; (b) that the psychic secretion was not completely eliminated; (c) a possible abnormal secre- tion. In the dog we found that the removal of the parotid is not, after all, such a difficult task, because it is smaller than the submaxillary and because its relation to the latter and its location over the ear always give one a good clue to its location, while the space in which it lies can be increased in size by extending the head. To eliminate the psychic secretion would appear to indicate a diminished and possibly prolonged gastric secretion, and therefore lead to an incorrect interpretation of the results. The possible presence of an abnormal secretion would somehow have to be controlled by the salivary glands. Hemmeter’s conclusions are as follows: (1) In dogs with simple gastric fistula the extirpation of all of the salivary glands produces a marked diminution in the gastric secretion. This is also evi- dent in the analysis of test meals drawn by the test tube from animals with intact stomachs. It is necessary to prevent psychic secretion in order to bring about the phenomenon described; (2) even in animals with intact vagi it may sometimes happen that the removal of all the salivary glands causes a decided impairment of gastric secretion, so that a causative relation between the loss of the salivary glands and the reduced proteolytic and milk coagulating power of the gastric juice appears certain, even in these cases; (3) in nine salivary dogs in whom the gastric secretion has been decidedly diminished, it is not restored to the normal by the feeding of food.that has been well masticated and insalivated by other normal dogs; (4) when the gastric secretion is diminished a temporary restoration may be brought about by intravenous or peritoneal injection of extracts made from the salivary glands of normal dogs; (5) this temporary restoration of gastric secretion takes place even when the stomach is isolated from the central nervous system; (6) the presence of an exciting gastric secretion hormone formed in the salivary glands. Salivary gland extract fed directly with the food or placed into the stomach directly is not capable of exciting gastric secretion. Ground up fresh salivary glands cause approximately the same gastric secretion as an equiva- lent amount of ground beef in these animals. Loevenhart and Hooker (9), working on dogs with simple gastric fis- tulae, tried to determine the presence or absence of a salivary hormone REMOVAL OF SALIVARY GLANDS AND GASTRIC SECRETION 207 by feeding extracts of the salivary glands. They assumed that the presence of such a hormone would cause an increased secretion of gas- trie juice in normal dogs, and since they did not obtain evidence of such increased secretion they concluded that a gastric secretory hor- mone is not present in the salivary glands. Hemmeter (10) later took exception to their method and conclusions, emphasizing his findings that extracts of the salivary glands raise a depressed gastric secretion almost to normal, and that their experi- ments on dogs with normal gastric secretion could not be used to dis- prove his conclusions. EXPERIMENTAL PROCEDURE The method employed is based on the use of the Pavlov pouch, the secretion of which may be considered a true index of the course of the secretion in the main stomach. After allowing seven to ten days for complete recovery from the Pavlov pouch operation, the gastric juice was collected from the pouch for six to eight hours each day. The rate of gastric secretion was determined by measuring the juice secret- ed at intervals of one hour, beginning one hour before feeding a stand- ard meal of lean meat. In this manner a normal secretion curve was obtained over a period of seven to ten days. For each hour the rate, peptic digestion and acidity were determined and curves plotted. After the determination of the normal secretion curve the three pairs of sal- ivary glands were removed in one operation. This was preferred to two separate operations because we observed that a second anesthesia was prone to induce infection of the respiratory tract (distemper). After considerable experimentation, the most satisfactory modus operandi, and the one finally adopted, consisted in making two inci- sions, one on each side, extending from the ear to the angle of the man- dible, each being about 23 inches long. In the dog the three glands on each side approximate each other very closely; the parotid lying over the ear, its medial portion in close relation to the submaxillary; the sublingual consisting of two parts, the aboral portion lying directly on the submaxillary gland and the oral portion further up along the duct of the submaxillary in the sublingual triangle. The manner of collecting the gastric juice and determining the rate consisted in placing a perforated rubber tube in the pouch and collect- ing the secretion in a container somewhat similar to the one sketched by Keeton (11) for use on cats. 208 A. M. SWANSON TABLE 1 Summary of observations on the gastric juice of two dogs before and after extirpation of the salivary glands. The averages are made up from 20 observations of each dog, 10 before and 10 after removal of the salivary glands BEFORE AFTER DOG High Low | Average} High Low | Average 1 |15.75 | 7.25 |10.33 |18.50 | 7.00 {10.00 2 |26.25 |11.50 |17.39 |23.50 |12.25 |19.33 Rate of secretion in cu- bic centimeters....... 1 | 0.0931] 0.0639] 0.0761! 0.2263] 0.0712] 0.1204 2 | 0.2327| 0.0626] 0.1325 0.2934! 0.1384] 0.2381 Total acidity in per cent 1 none | none | none | 0.1733} none | 0.0536 0.1900} none | 0.0718] 0.2489) 0.0791) 0.1819 Pepsin concentration in millimeters (Mett). 1 18.25 |14.75 |18.99 |20.75 {11.50 |13.96 2 18.75 | 9.75 |18.61 |25.50 (14.75 |17.21 Free acidity in per cent { [ \ cr. L 2. nN rn 1 n ° 1 2 3 4 5 6 7 BHOURS Fig. 1. Represents the rate of gastric secretion on a standard meal of meat as determined by 20 observations on a dog with a Pavlov stomach, the continu- ous line indicating the average of 10 observations before removal of the salivary glands, while the broken line indicates the average of 10 observations after re- moval of the salivary glands. REMOVAL OF SALIVARY GLANDS AND GASTRIC SECRETION 209 _ The peptic digestion for a period of twenty-four hours was deter- - mined according to Mett as modified by Cobb (12). The free acidity was calculated by titrating 1 cc. of the gastric juice diluted with 20 cc. of distilled water with N/40th NaOH using dimethyl- amidoazobenzol as an indicator for the free acidity and phenolphthalein as an indicator for the total acidity. ___ Two vigorous dogs survived all the operative procedures, their wounds : healed perfectly and their health did not seem to be at all impaired. : Areewarn ° t 2 3 ry 3 6 7 “B [ouns} _ Fig. 2. Dog. 2. Represents the total acidity of the gastric juice as obtained by the average of 20 experiments. The continuous line indicates the total acidity as determined by 10 observations before the removal of the salivary glands. The broken line indicates the total acidity for a series of 10 observations after the removal of the salivary glands. Their mouths did not appear as dry as would be expected after loss of all the salivary glands, which is probably due to the numerous mucous glands present in the oral cavity. The dogs soon learned to swallow their food and their taste did not appear to be altered. RESULTS The rate or quantity of secretion of gastric juice in both dogs was not altered by complete removal of the salivary glands (fig. 1). 210 A. M. SWANSON The acidity of the gastric juice shows a greater variation, there being © a decided increase after the removal of the salivary glands (figs. 2 and 3 3). In dog 1 before removal of the glands there was at no time any free acidity, while out of thirteen days following there was only one day in which free acid was absent. In dog 2 free acid was present both before and after removal of the glands, but the free acidity after removal of the salivary glands showed a marked increase. The maximum total acidity in both dogs occurred, on the average, half an hour to one hour o2F Graenorn) ° t 2 3 4 5 6 7 8 [Hours] : Fig. 3. Dog. 2. Represents the free acidity of the gastric juice as obtained by the average of 20 experiments. The continuous line indicates the free acidity as determined by 10 observations before the removal of the salivary glands. The broken line indicates the free acidity for 10 observations after the removal of the salivary glands. later than in the control periods, and the subsequent fall in acidity was more gradual. The peptic digestion was about the same both before and after the removal of the glands in dog 1, but in dog 2 there was a slight increase. CONCLUSIONS 1. Our results contradict the theory of a hormone in the salivary glands stimulating the secretion of gastric juice. Extirpation of the x REMOVAL OF SALIVARY GLANDS AND GASTRIC SECRETION 211 salivary glands in the dog does not decrease the gastric juice secretion (appetite and secretagogue juice). 2. Extirpation-of the salivary glands causes a distinct rise in the acidity of the gastric juice. This increase in acidity is greater than: can be accounted for by the slight increase in the rate of secretion. The slight increase in quantity may be due to the absence of the alka- line saliva. 3. After extirpation of the salivary glands, the maximum secretion rate after a meal appears slightly retarded. This may be due to the absence of the water of the saliva, and to decreased psychic secretion, owing to the dryness of the mouth and consequent impaired taste. BIBLIOGRAPHY (1) Paviov: The work of the digestive glands (transl. by W. H. Thompson)» 1910. (2) Taruii anv Pascuccr: Cited from Luciani’s Human physiology (English transl. by F. A. Welby), 1913, ii, 175. (3) Kerron AND Kocu: This Journal, 1915, xxxvi, 353. (4) Carson AND CrITTENDEN: Proc. Soc. Exper. Biol. Med., 1910, vii, 52. (5) Garrey: Journ. Biol. Chem., 1907, iii, 40 and 41. (6) HemmeteR: Sci., 1907, xxvi, 473. (7) Hemmeter: Trans. Amer. Gastro-Enterol. Assoc., 1908, 5. (8) Hemmerter: Biochem. Zeitschr., 1908, xi, 238. (9) LonvennART AND Hooker: Proc. Soc. Exper. Biol. Med., 1908, v, 114. (10) Hemmeter: Proc. Soc. Exper. Biol. Med., 1909, vi, 33. (11) Kerron: This Journal, 1914, xxxiii, 25. (12) Coss: This Journal, 1905, xiii, 448. THE COMPOSITION OF SALIVA IN RELATION TO THE INCIDENCE OF DENTAL CARIES? JOHN ALBERT MARSHALL From the Department of Biochemistry and Pharmacology and the Laboratories of the Department of Dentistry, University of California Received for publication March 5, 1917 INTRODUCTION It has been previously reported by the writer (4) that the ratio of the neutralizing powers, or the power to maintain neutrality, of normal resting saliva and of the activated saliva, obtained by the chewing of paraffine, bears a definite relationship to the incidence of dental caries. In persons with carious teeth this ratio, expressed as a percentage, exceeds 80 while in persons whose teeth are temporarily immune from (or, more correctly, resistant to) caries the ratio falls below 80. In other words, as the difference between normal resting saliva and acti- vated diminishes so does the liability to the incidence of caries increase. Shepard and Gies, in discussing this relationship or “salivary factor” have maintained (5) that it is inconstant. They based their conclu- sions, however, upon data which included all the different types of stim- uli indiscriminately without regard to the nervous impulses and re- flexes produced by the sense of taste. In their experiments they used paraffine, sucrose, sodium chlorid, alcohol and certain combinations of these. It was subsequently stated by the writer (8) that comparisons of saliva, the samples of which have been collected under different con- ditions, are inadmissible since such procedure ignores entirely psychic influences. Recalculation of their figures confirms the findings origi- nally reported (4). In a second communication (6) the writer presented data which both substantiated and developed the above thesis. Reports were made of investigations conducted in some of the state institutions for the in- 1 Submitted in partial satisfaction of the requirement for the Degree of Doc- tor of Philosophy in the University of California and accomplished in part under the auspices of the Research Institute of the National Dental Association. 212 COMPOSITION OF SALIVA AND DENTAL CARIES 213 sane, and consisted of analyses of saliva from certain cases of dementia praecox and epilepsy. This work was likewise criticised by Gies (7) and answered in turn by the writer (9). No data have been presented which disprove any of the conclusions drawn in either paper (4 and 6) and it is in the further development of the consequences arising out of these conclusions that the following experiments were undertaken. There are two main questions which fall under consideration in this connection, namely, first, the origin of the change of the neutralizing power of saliva which occurs in response to certain stimuli, and second, the significance of this change in relation to the incidence of dental caries. In other words, whether the alteration of the salivary factor is a contributing cause of dental caries, or conversely, an effect pro- duced by dental caries, and lastly, whether there is a cause common to both altered factor and dental caries in which case the factor would become merely an incidental symptom. In connection with the first of these problems, I have sought to throw light on the origin of the differences between the neutralizing powers of different samples of saliva: (a) By dialysis experiments in which the attempt has been made to determine the relative magnitudes of the parts played by the inorganic, or at least the diffusible substances, and the non-diffusible, and presumably organic, constituents of saliva. (b) By the determination of the amino nitrogen content of various samples of saliva after hydrolysis with a view to estimating more ex- actly the part played by protein in contributing to the difference in properties and composition between normal resting and activated sa- livas. In connection with the second problem I have sought to extend the observations of Pickerill upon the relationship of diet and habit to the incidence of dental caries and furthermore to determine the influ- ence of the locality of the stimulus upon the neutralizing power of the secretion which is evoked. Part 1 ‘THE ORIGIN OF THE DIFFERENCES IN NEUTRALIZING POWERS DISPLAYED BY NORMAL RESTING AND PARAFFINE ACTIVATED SALIVA a. The relative magnitude of the parts played by diffusible and non- diffusible substances in determining the neutralizing power of saliva The experiments were carried out as follows: Saliva was titrated and a second set of samples was dialyzed for a period of time; then the liq- uids both outside and inside the membrane were separately titrated. 214 JOHN ALBERT MARSHALL The dialysis of the samples was made in these cases where a sufficient quantity of saliva could be obtained without conscious exertion on the part of the patient. Following the method of C!ausen (10) and Porter (11) a solution of gun cotton in an ether alcohol mixture was made. From this solution collodion thimbles were fabricated and then placed in recently boiled distilled water until ready for use. From 2 to 5 ce. of the sample were measured directly into the thimble and then placed in a small prescription vial. Boiled distilled water was pipetted into the bottle until the level of the sample exactly coincided with that of the water. Decomposition of the sample was prevented by adding TABLE 1 Comparison of undialyzed and dialyzed saliva RESTING SALIVA ACTIVATED SALIVA fe \23|,| 212/231.) 2 1/88/88) ,1 Fee 2 # (32/28) 2/ 8183/81 2 lyeslse|2| 298 gene) a we) 7g) 5) 3) 31812 | 5 |e.e “S|. | Sen 2 ea] $2) ) 2) | 28] a | & |S55| 22] 9) 3 | s |e 3/2 x |e 18 1 ee ae a eae ee < 4) eon ae a 5 Cubic centimeters Cubic centimeters Cubic centimeters Cubie ‘eontimeters of HCl of NaOH of HCl of NaOH E 4] 6.60} 3.00/2.00) 5.00/13.75) 9.20/4.20]13 .40/24.15)16.00/2.65)18.65/4.75!2.85)1.65/4.50 E 1} 9.00} 6.20/5.40}11.60) 9.50} 8.25/0.20) 8.45|17.75)14.00/0.95/14.95/5 .50/4. 40/0. 69/5 .09 E 6 |19.50)12.30]1 . 28/13 .58) 6.50] 6.00 6.00/48 . 30/46. 50/3 . 40/49 .90}1.20 E 7| 8.75} 6.50/0.00| 6.50) 9.80/10.70 10.70/25. 20/21 .40)2 . 40/23 .80|4.55/3.40)1.00/4.40 © E 9 /10.70} 8.10|1.00) 9.10) 7.30} 4.85/0.85] 5.70|15.80/14.45/1. 15/15. 60/3. 10/3 .00 3.00 E10 |14.00| 9.25/3.15]12.40) 5.00) 4.90/0.75] 5.65/16.35/13.70/2.00/15.70/4.80/2.10/2.25/4.35 E11 |17.45/10.25|7 .00/17.25) 7.40} 3.30/0.90] 4.20/19. 10}18.30/3.00/21 .30/2. 90/1 .75}1 .00/2.75 E12 | 7.25} 4.00/3.15) 7.15|11.75| 8.10)2.70|10.80 19.90/16 .70|2. 2018. 90/4. 85/3. 10/0.95)/4.05 E13 | 8.50} 4.85/3.20) 8.05/12.40} 9.60|3.00/12.60/22.70|20. 10/1 .50/21.60/5.40/4.00|0.90/4.90 E14 |16.40) 9.70/5.80/15.50} 8.90] 4.80|/3.90| 8.70|35.40/23 . 10/3 .40/26.50/1.70|1.00 1.00 one drop of chloroform and one drop of xylol. The bottle was tightly corked and placed in an air tight cabinet for twenty-four hours, at a temperature between 20° and 24°C. At the end of this time the liquid _ outside the membrane was titrated separately from that inside the membrane. ple, the sum of the two titration figures for either alkalinity or acidity should equal the original titration value. But a slight precipitation of phosphates, which is always evident, demonstrates that some loss of Theoretically, if there were no loss of COz, from the sam- CO2z has occurred. Data based on twelve, thirty-six and forty-eight hour dialyses showed a wider variation than those based on the twenty- . COMPOSITION OF SALIVA AND DENTAL CARIES 215 _ four hour limit and this later time, consequently, was chosen as a standard. _ The results of this work are reported in tables 1 and 2. The analy- ses, although of questionable quantitative value demonstrate, quali- tatively, that the greater percentage of alkalinity and acidity is found in that portion of the sample which has dialyzed through the membrane and is due, therefore, to inorganic constituents. With subject No. E4 the alkalinity of 10 cc. of the resting saliva was 6.60 ec. N/200 HCI. After dialysis the alkalinity outside of the membrane was 3.00 cc. and inside the membrane, 2.00 cc. The activated sample exhibited a TABLE 2 Dialysable proportion of neutralizing power = NORMAL RESTING SALIVA PARAFFINE ACTIVATED SALIVA 5 5 £bd 5 5 Lad = z sas 2 Ed S:35 i CONDITION 22 2 oe BZ 2 sé 3 si NUMBER OF we? wo ao 2 8 Ags 8 : MOUTH as ag 28s ‘S a¢ o@3 3 s> a> tos5e| 8° > | osbe ye 28 23 3°83 ae 43 g°32 P Rep st me | S28 su >s we e 33 as |@388] 83 a8 | a3.88 = Zz Z a) Z , A ZB E4 Immune) 20.35 18.40 | 90.01 28.90 23.15-| 80.10 | 70.41 E1 Immune| 18.50 20.05 | 108.38 23.25 20.04 86.19 79.57 E 6 Immune} 26.00 | 18.58 | 75.31 47.10 | 49.90 | 105.94 | 55.20 E7 Immune} 18.45 17.20 93 .22 29.75 28.20 94.79 62.02 E9 Carious| 18.00 | 14.80 | 82.03} 18.90} 18.60 | 98.41 | 95.23 E10 Carious| 19.00 18.05 95.00 20.15 | 20.05 99 .50 94.29 Ell Carious| 24.85 21.45 86.11 22.00 24.05 | 109.32 | 113.00 E12 Immune} 19.00 17.95 94.47 24.75 22.95 | 92.77 76.76 E13 Immune! 20.90 20.65 98.80 28.10 26.50 94.31 74.37 E14 Immune} 25.30 23.20 | 91.70 | 37.10 27.50 | 74.12 | 68.20 marked difference for the titration figure of the undialyzed sample was 24.15 and for the dialyzed, 16.00 outside the membrane and only 2.65 inside the membrane. . The acidities likewise show the same phenomena, the undialyzed normal resting saliva having a reading of 13.75 ec. N/200 NaOH and the dialyzed 9.20 ouside the thimble and 4.20 inside. b. The amino-nitrogen yielded by hydrolysis of norma! resting and activated saliva In the utilization of the Van Slyke apparatus for the determination of the amino-nitrogen in the saliva the author has employed a method 216 JOHN ALBERT MARSHALL which combines accuracy with simplicity. The wide application which this apparatus has found in blood analysis recommends it favorably to the problem at hand. The procedure outlined in Hawk (14) was fol- lowed with a few modifications. The technique of the analytical work was performed by Mr. S. A. Waksman and I take pleasure in acknowl- edging his service. It was at first thought best to analyze the samples as soon as they were obtained from the patient but this procedure is open to objection on account of the fact that there is so small an amount of gas evolved in the reaction that accurate readings of the gas volume are exceedingly difficult. Since it has been the custom in the salivary work to secure the material between eight and eleven in the morning it was found in- convenient to make the determinations at the same time. To over- come these objections all the samples were hydrolyzed. ‘Ten cubic cen- timeters of well mixed saliva were measured into a special digestion tube and 4 ec. of concentrated HCl were added. The tube employed was of soft glass about nine inches long and one inch wide. One end was sealed off and the other drawn out until the diameter was reduced to nearly + inch. After the introduction of the sample and the addition of the acid the small end was sealed. These tubes closed at both ends were placed in a water bath and digested at 100°C. for four hours. — Attempts to digest the samples by boiling over a flame and using a re- flux condenser were unsatisfactory as bumping and loss of the material could not be controlled. Having prepared the Van Slyke micro-apparatus in the usual manner ' 2c. of the well mixed hydrolyzed sample were carefully transferred to the measuring tube and run into the decomposing bulb. The bulb was shaken for ten minutes and the NO absorbed by the permanganate solution. The volume of nitrogen was read and the room temperature and barometer noted. Duplicate determinations were made and, in many instances, triplicates. The results of the analyses are reported in tables 3 and 4 and repre- sent the cubic centimeters of amino nitrogen at standard pressure and temperature yielded by 10 cc. of sample. This calculation was made so that the data would be comparable to those of the titration experi- ments. Inspection of the table shows that, in the greater percentage of those cases which were classed as-immune from dental caries, the ni- trogen content of the normal resting saliva is appreciably higher than that of the paraffine activated saliva. Those samples, however, taken COMPOSITION OF SALIVA AND DENTAL CARIES 217 from the mouths in which caries existed did not exhibit the same dif- ‘ference. These facts substantiate the work done on the dialysis experi- ments. For it was shown that the increase of total neutralizing power in paraffine saliva from immune cases was due to a larger amount of inorganic constituents. These later tests demonstrate that a smaller percentage of organic bodies, represented by the amino nitrogen values, is contained in this same type of saliva and that therefore the difference in titration equivalents must be due to inorganic material. TABLE 3 Caries am ORMAL | actIvVaTED | STIMULATED. CENTIMETERS SALIVA CUBIC SALIVA CUBIC DATE SERIAL NUMBER OF AMINO N CENTIMETERS CENTIMETERS pur 10 ce. oF AMINO N | or amino N aAMPEe PER 10 cc. PER 10 cc. SAMPLE . SAMPLE September 18, 1916..........| G2 2.20 3.00 September 21, 1916.......... G 2 2.50 3.00 September 25, 1916.......... G8 3.70 3.80 October 27, 1916> G28 3.70 3.60 October 27, 1916. G29 3.70 3.60 October 27, 1916. G3l 4.30 4.30. November 2, 1916........ G36 4.10 3.60 November 2, 1916.. G37 4.20 3.80 September 25, 1916.......... G9 3.10 3.30 October eee ss 565-01 G9 3.00 3.00 October 30, 1916.. G9 3.40 2.80 September 29, 1916,......... G10 3.30 3.40 3.80 November §8, 1916.......... G10 4.15 4.27 November 10, 1916.......... G10 3.77 3.63 November 10, 1916.......... G18 2.80 2.95 November 7, 1916.......... G18 3.10 3.20 November 9, 1916.......... G18 4.10 4.00 November 4, 1916.......... G38 3.90 3.80 November 5, 1916.......... G42 4,20 4.10 November §8, 1916.......... G44 4.40 4.50 November 15, 1916.......... ‘G9 4.27 3.70 .November 20, 1916.......... G9 5.80 3.57 November 23, 1916..........|. G10 4.02 4.02 November 2, 1916.......... G35 4.10 4.00 November 4, 1916.......... G43 5.90 4.87 3.90 © November 21, 1916.......... G52 4.48 4.84 November 21, 1916.......... G53 3.80 4.10 November 28, 1916.......... G58 3.12 3.39 218 JOHN ALBERT MARSHALL TABLE 4 Immunity NORMAL REST- PARAFFINE ELECTRICALLY SERIAL ING SALIVA, ACTIVATED STIMULATED Dar semzan | | °GOF INO | sucrtayon/ 0a, erat SAMPLE cc. SAMPLE cC. SAMPLE September 16, 1916.......... Gil 2.40 2.30 2.20 September 19, 1916.......... Gl 3.20 2.90 September 19, 1916.......... G 3 4.60 3.80 3.50 September 26, 1916.......... G3 5.70 3.30 September 21, 1916.......... G 4 3.00 2.60 September 28, 1916.......... G4 3.50 4.00 3.20 October 9, 1916.......... G4 3.80 3.20 3.40 September 21, 1916.......... G5 2.90 2.70 September 21, 1916.......... G5 3.20 3.00 September 28, 1916.......... G5 3.70 3.90 4.60 October 9, 1908. nee G5 3.60 3.00 2.90 . October 3; 191635 ss. seen G 6 3.40 2.60 October By 1016. 3. POs Gil 3.40 3.10 October O.- 1916; oo oe Gl1l 4.90 4.10 _ October $5 (1916. e307 ee G12 5.00 4.10 October DB 1016: os ve G12 4.20 3.80 October OB, F916) ecu G14 4.00 3.40 October 0, LORG: sta G15 4.20 3.20 October 13, 1916.......... G16 3.90 3.70 October 13, 1916.. G17 4.60 4.00 October 20, 1916.......... G19 2.90 2.30 October 20, 1916.......... G20 3.60 3.40 October - 72), 1916. .... gag G21 3.50 2.70 October 20, 1916.......... G22 4.20 3.40 October 20, 1916.......... G23 4.50 3.40 October . 28, 1916.......... G24 4.10 3.70 October 25, 1916.......... G24 . 4.20 3.90 October 27, 1916.......... G24 4.10 4.00 October 25, 1916. G25 5.60 2.90 October: 25, 1916:......... G26 5.60 3.00 October 28, 1916.......... G26 4.30 3.60 October 28, 1916.......... G32 , §.10 4.30 October 28, 1916. G33 5.60 4.50 October 28, 1916. G34 4.30 3.80 November 2, 1916. G34 4.20+ 3.90 November 18, 1916. G26 5.80 4.10 November 11, 1916.......... G26 5.75 4.00 November 4, 1916.......... G39 4.80 4.10 November 4, 1916.......... G40 4.60 4.15 November 4, 1916.......... G41 5.70 4.05 November 11, 1916.......... G45 4.87 4.63 COMPOSITION OF SALIVA AND DENTAL CARIES TABLE 4—Continued 219 a Ea NORMAL REST-| PARAFFINE ELECTRICALLY SERIAL ING SALIVA, ACTIVATED STIMULATED DATE NUMBER cc. OF AMINO SALIVA, CC. OF SALIVA, CC. OF N rw 10 cc. AMINO N In 10| amino N rn 10 SAMPLE cc. SAMPLE cc. SAMPLE November 11, 1916.......... G46 4.80 4.00 November 11, 1916.......... G47 6.24 4.37 November 11, 1916.......... G48 4.97 3.97 November 13, 1916.......... G48 4.63 4.05 November 14, 1916.......... G48 4.70 4.13 L November 15, 1916.......... G48 4.90 4.06 _ November 16, 1916.......... G48 5.04 4.20 : November 17, 1916.......... G48 3.51 3.30 November 23, 1916.......... G25 5.25 3.10 ; November 24, 1916.......... G25 5.44 3.19 4 November 24, 1916.......... G26 3.80 3.20 4 ‘November 13, 1916.......... G26 5.80 4.10 ; November 15, 1916.......... G26 5.75 4.00 November 6, 1916.......... G49 4.97 4.14 November 13, 1916.......... G49 4.87 4.14 November 14, 1916.......... G50 5.50 4.06 November 17, 1917.......... G50 5.21 3.99 December 4, 1916.......... G55 5.93 3.67 December 11, 1916.......... G55 5.70 3.67 On averaging these tables, the following figures are obtained: Immunity normal resting saliva............. Immunity paraffine activated saliva......... 4.47 cc. amino N 3.63 cc. amino N 0.84 20.7 per cent of the mean 3.89 cc. amino N 3.72 ec. amino N 0.17 4.5 per cent of the mean ' Caries normal resting saliva................. Caries paraffine activated saliva............. The difference in nitrogen content between the normal resting saliva in cases of immunity and that in caries is strikingly brought out by the above tabulation. In the determination of averages however there is always an error which must be measured before the data may be considered conclusive. The determination of the standard deviation of the mean demon- strates, however, the reliability of the results presented. These calcu- lations were evaluated from the formula given by Davenport (15) as follows: 220 JOHN ALBERT MARSHALL sum of the squares of the deviation from the mean number of measurements, Standard deviation =1 standard deviation number of measurements. Probable error of the mean = 0,6745 X af The appended table thus summarizes the data: CARIES IMMUNITY Standard deviation normal resting saliva.....| 0.645 0.913 Probable error: .....:: Wa.eoesie.. . Saha 0.08 -| 0.07 ; Bauals, i. ccdcs os... open ee eee see 2.1 per cent of | 1.8 per cent the mean Standard deviation paraffine saliva........... 0.5564 0.569 & Probablecerror. isos. Joes ene A2 ia 0.0735 0.0496 Btls: cs 5 5 sacks ss pect aes © oe 2.0 per cent of | 1.4 per cent the mean It is evident from these figures that the difference between the yields of amino nitrogen by normal resting and activated saliva in persons afflicted with dental caries is only twice the probable error of the mean; in other words, that there is either no difference, or at the most only a slight one between the average protein content of normal resting saliva and activated saliva secreted by such persons. In normal individuals, on the contrary, the difference between the amino nitrogen yields is no less than ten times the probable error of the determination and it is evident that in such persons there is a very definite divergence of com- position between normal resting and activated saliva. The normal rest- ing saliva of a person with caries approaches, in protein content, the activated saliva of a normal person, and stimulation by chewing paraf- fine results in little change either in protein content or in neutralizing power of the saliva secreted. These relations are illustrated in the appended table showing the average neutralizing powers (that is the number of cubic centimeters of N/200 acid and alkali required to change 10 cc. of saliva from neutra ity to phenolphthalein to neutrality to paranitrophenol) of normal resting and activated saliva in normal persons and in persons afflicted with caries. These averages are compiled from salivary analyses com- pleted within the last two years and comprise data obtained from over one hundred individuals. It is evident that saliva from subjects with COMPOSITION OF SALIVA AND DENTAL CARIES 221 carious teeth presents two distinct differences from saliva of immune subjects, namely, that the neutralizing power of the resting saliva is supernormal while that of the paraffine activated saliva is subnormal. Average total neutralizing power 2 PARAFFINE AVERAGE . 3 5p Aamcspa 2 cp ag ion See eae eee s 23.693 | 38.324 61.82 OO ee 2 oe 30.096 30.952 97.24 The normal resting saliva of persons with carious teeth is, therefore, characterized by (1) a relatively high neutralizing power and therefore, presumably, (2) a high content of diffusible substances; (3) a low con- tent of proteins. TABLE 5 Comparison of values of amino nitrogen with the total neutralizing power NORMAL RESTING SALIVA PARAFFINE ACTIVATED SALIVA NUMBER Total | Amino ‘ Total | Amino ; Hoi | Noon | "ee | Neer | uct | Nao | ngute | Nper | Salivary power | sample power | sample ce. ce. ce. ce. ce. ce. G24 17.10 | 5.00 | 22.10 | 4.10 | 38.50 1.00} 39.50 | 3.70 | 55.95 G26 27.40 | 4.70 | 32.10 | 5.60 | 58.45 | —1.50| 56.95 | 3.15 | 56.40 G26 20.00 | 4.50 | 24.50 | 4.30 | 49.00 0.50} 49.50 | 3.60 | 49.50 G27 15.00 | 12:90 | 27.90 | 8.00 | 41.10 1.75} 42.85 | 4.90 | 65.10 G9 27.75 | 2.55 | 30.30 | 3.40 | 36.55 1.50} 38.05 | 2.80 | 79.60 G45 28.50 |—1.60°'| 26.90 | 4.87 | 57.60 | —4.15) 53.45 | 4.63 | 50.30 G48 34.75 | 2.15 | 36.90 | 4.63 | 47.55 2.50) 45.05 | 4.05 | 81.90 G48 24.00 | 3.30 | 37.30 | 4.90 | 35.00 0.60} 35.60 | 4.06 | 76.70 G49 16.10 | 5.70 | 21.80 | 4.97 | 35.50 1.25) 36.75 | 4.14 | 59.30 G40 18.40 | 3.70 | 22.10 | 5.50 | 31.60 2.30) 33.90 | 4.06 | 65.20 G52 33.00 | 8.00 | 41.00 | 4.48 | 40.65 1.00} 41.65 | 4.84 | 98.45 G53 24.15 | 3.95 | 28.10 | 3.80 | 30.80 2.80} 33.60 | 4.10 | 83.60 G58 7.70 | 8.00 | 15.70 | 3.12 | 32.45 1.30) 33.75 | 3.39 | 46.50 G58 11.90 | 6.00 | 17.90 | 4.10 | 33.45 1.25) 34.70 | 3.70 | 51.60 G59 24.55 | 2.75 | 27.30 | 3.19 | 26.70 0.90} 27.60 | 3.00 | 98.90 The inter-relation between the amino nitrogen and the salivary fac- tor is shown in table 5. Subjects 52 and 59 were classed as carious, the others as immune. This connection of the one to the other is ren- dered all the more striking when the results of the dialysis experiments 222 JOHN ALBERT MARSHALL are kept in mind, for the data evaluated in table 1 pointed to the fact that the increase of the total neutralizing power is due primarily to an increased amount of inorganic substances. Conversely, the lowered amount of organic bodies in activated salivas coupled with their greater neutralizing power brings further evidence to substantiate this conclusion. From these results it is evident that immune persons secrete in re- sponse to stimulation by chewing tasteless substances a saliva which has a greater neutralizing power than normal resting saliva and is furthermore differentiated from normal resting saliva by a considerably lower content of protein and higher content of inorganic salts. The alteration in the character of the saliva is not merely due to dilution, consequent upon more rapid secretion, but involves a marked change in the relative proportion of the constituents. Persons with carious teeth differ from normal persons in that their normal resting saliva ap- ° proximates in composition and neutralizing power to the composition of the activated saliva, in other words the salivary glands of such per- sons behave as though they were constantly receiving stimulus analo- gous to that constituted by the act of chewing a tasteless substance. Such a stimulus might conceivably be provided by carious teeth them- selves, or on the other hand, both conditions may be attributable to a common underlying cause. Part 2 RELATIONSHIP OF DENTAL CARIES AND THE COMPOSITION OF SALIVA TO DIETARY CONDITIONS AND THE LOCALITY AND NATURE OF STIMULI PROMOTING SECRETION a. The relation of diet to the incidence of caries The alteration in the salivary factor may be due to either a direct or an indirect cause. If the former, then the presence of dental caries in an otherwise healthy mouth would initiate the change. If the latter, then the change is either incidental or comprises a portion of a vicious circle. Among the indirect factors which may initiate an acute disturbance may be mentioned diet, a lesion or an infection, a chronic peripheral nervous affection, a central nervous affection or lastly a defect in the processes of repair and growth correlated with a defect in salivary function. Diet comprises per se two factors, namely composition and taste. COMPOSITION OF SALIVA AND DENTAL CARIES 223 The former has been the subject of much discussion and research and the conclusions reached by the many authorities appear to be rather negative-in-character. Data hereinafter reported will deal more par- ticularly with this phase. It has been shown (6) that the salivary factor is constant in certain types of insanity to the same extent as in the normal individual and from this fact the deduction may be made that differences in neutral- izing power are not related to the central nervous system. In reviewing the literature concerned with the problems of dental caries and its possible relation to habits of cleanliness, climatie condi- tions, diet and general health of an individual or of a race, one notes a lack of uniformity in the recorded observations. This may be attrib- uted to the fact that the many different writers were influenced in their judgment by different standards of observation, so that teeth which superficial examination would designate as non-carious might disclose, upon a more thorough examination, exactly the opposite condition. . Pickerill (1) states that certain food investigations point to the fact | that the modern dietary of the civilized world differs from the diet of the uncivilized world in that the former is less hard but more tough and requires, therefore, more triturating but less crushing. From this conclusion the idea is advanced that, since different sets of muscles are used in triturating than in crushing, the over-development of these tri- turating muscles (buccinator and pterygoids) is responsible for both the abnormally shaped as well as the undeveloped dental arch. This, in turn, accounts for the increasing number of malposed teeth which accom- pany the under-developed arch. Malposed teeth are very susceptible to dental caries and therefore the increase of this disease among the mod- ern civilized nations is correlated to our changed habits in masticating. Pickerill notes further that of the races of the world, the meat eaters, or at least those whose food is largely protein in character, were quite as susceptible to caries as those whose diet was mainly vegetarian. This _is contrary to the views expressed by both Mummery and Patrick. The argument advanced by Pickerill is that the immune races, which include according to some authorities the Asiatics, Africans, Polynesians, Australians, et cetera, owe their comparative freedom from caries to the fact that their diets were both varied and sapid. Their universal use of masticatories resulted in the prevention of stagnation within the oral cavity (p. 314) a fact which other investigations appears to support. 224 JOHN ALBERT MARSHALL Dr. R. Thurnwald, in speaking to the author of his anthropological researches in New Guinea; mentioned that the inhabitants in that see- tion of the world seem to be comparatively free from dental caries. Their diet is mainly vegetable consisting of yams, sago, rice and sugar cane, etc.; meat is rather an accessory and, with the exception of the rather scarce mango, there are no acid fruits. The custom of chewing the betel nut plays, unintentionally, an important part, no doubt, in their oral prophylaxis. For at the age of puberty this custom, often connected with one of the initiatory ceremonies, is commenced and continued throughout the life of the individual. Contrary to the usual belief, this betel nut habit does not blacken the teeth. For that purpose there is employed a mixture of cocoanut oil and soot which is vigorously rubbed on the teeth. The betel nut is not used alone but is combined with seeds or leaves of a peppery nature together with pulverized lime from calcined shells. These three substances are taken into the mouth one after the other and are masticated between meals. The old men are edentulous and pre- pare a paste by mixing the material before chewing it. The natives expectorate profusely after the use of this mixture and the saliva is colored blood red. This coloration might be ascribed to the bleeding or to a compound formed by the action of the lime on the betel nut. The teeth are lost between the ages of forty and fifty and are exfoliated comparatively rapidly once the process has commenced. This exfolia- tion is accompanied by swollen and bleeding gums. Calculus is deposited upon the teeth in almost unbelievable quantity, and it is often the case with the adults to see the size of the lower in- cisors increased by these concretions to 300 or 400 per cent. : Underwood (2) after examinations of skulls from different collec- tions makes the following comment: In the hot belt of the earth including India, Africa and Southern China, bathing and washing are natural habits because of the heat; rinsing the mouth after meals and the use of sticks, toothpower, ashes and salt for cleansing the mouth is almost universal among the natives; while the foodis largely rice and _ no alcohol is used. In all of them caries isso rare that to all intents and purposes the natives may be regarded as immune. He states further that the people of the arctic regions whose personal habits, at least in regard to the care of the mouth, leave much to be desired, and whose diet is quite different, likewise enjoy immunity. He considers the Australian native equally immune. These observa- tions lead him to conclude that the use of artificial foods and the re- VY. Se COMPOSITION OF SALIVA AND DENTAL CARIES 225 placing of breast feeding exerts a direct influence in the “weakening of the tooth defences.” Just what constitutes “tooth defence” is not mentioned. = The effect of certain drugs upon the teeth has been dealt with by Austen (3). The systemic conditions which are supposed to favor the development of caries are anemia, dyspepsia, pregnancy, acute rheumatism, enteric and other continued fevers. Various salts and compounds of mercury, lead, bismuth, silver and copper were used in _ the experiments. Although it was found that the drug was partly excreted into the oral mucosa, yet it is rather an open question whether this excretion at one time may be so long continued as to accelerate or even cause any deleterious effect upon the erupted teeth. Another point however which may well be considered, is that the frequent drug- ging of growing children promotes a disturbance in the nutrition of the ameloblast and of the odontoblast thereby bringing about structural changes in the enamel and dentin respectively. Histological examina- tions conducted along this line of experimentation would undoubtedly throw light upon certain phases of present therapeutic methods. In table 6 are presented certain abstracts and notations upon the ~ teeth and diet of a few races from different parts of the world. Defi- nite information on the subject appears to be rather scattered for in many instances an author may detail the foods at great length, the manner of cooking and habits of eating but will overlook entirely the conditions of the masticatory apparatus. In so far as the relative ’ amounts of protein to carbohydrate in the diet are concerned, the data appear to confirm Pickerill’s (1) conclusion in this regard, namely, that the protein eating races are as susceptible to dental caries as those whose food is mainly carbohydrate. On the other hand they appear to nega- tive the popular impression that the teeth of primitive races are relatively immune to caries. b. Relation of type and locality of stimulus to the neutralizing power of the secretion of saliva evoked by the stimulus If the chronic disturbance of the neutralizing power be due to a chronic peripheral nervous affection or to a lesion or an infection remote from the salivary glands or teeth, such a disturbance would probably act through nervous reflexes. If the locality of the lesion is important, then by applying a definite stimulus to a circumscribed area in the oral mucosa such nervous irritation so produced might be expected to JOHN ALBERT MARSHALL 226 uop -uoT “OD Y JUeq ‘“uoJINgG “J "Y “BoLsy yseq ut sdoqys}oo,7 4s.11y uopuo7y ‘soig A9[suty, ‘uoJINg “J “Y ‘[Izeig Jo spuv[ysiy qysIOX MON ‘FI6GT “OD FY URTV ‘roouedg ‘eijerysny Jo A1041II97, WIOYWON 94} JO Sequy, oOANeN Ig ‘d‘ATWVq “SUING (OW ‘uoljyIpedxy uloF{ 94} Jo qa0doxy 2% OFZ ‘d v4 qaed ‘yT “[OA “BITBIJsNV “Og “20g ‘AoY suey, “SUIS “0 “A pues woyerg “5 “M “BIIejued “IBY JO FINE) OY} JO Oquy, [eUIst -10QV oy} jo ‘040 ‘Sony ‘sulogsny uopuoy “og » 1epunqiiy, “LIND “J “W ‘S0BY UBITVIISNY oy], uopuoy 44nN preg “NoIyyB ‘f ‘MOID pus yueg opseq soqep ‘peorq qoyprua ‘sor ‘1e99nq “yy ‘mOZ4yNy AJPAISSOOXO IOY}BI 09810{00dxe I[®? pus poxyjoq st pooq ‘4yvour pep uns ‘“eurlrey ‘osveyo ‘AIT ss -Ipo St 4evyy SuryyAue yevo soATYeN BUN} UVI[VUIUTVUL OY} [[B VA : - poo} [eulrue jo edvys oy} Url SuryyAue ysou[e ye Iepue} Wey} SoyeuUr pue 300} UMOp sIvOM oUTM} jo Suryeul wo1 ieqy Sulmeyo suvA ‘s1001 uley ‘SsBIs ,NPIBN,, ‘91qQvy -os0A AJUQ “MBI U9}BO SOUT} -oUI0S YSe,J “orp ployoid oB1eT 090Bq0} BINS WOIJ pouTeys pue | 100d oie “ooford you op yy], Avoop 04 a[qQvI] ATyensnuy Iepus} AI9A 4499 J, , (dour -W09 UOT{SOBIPUI puB syoVyAOOL,,, [eulog jo youviq ulopeg 10 vesoq uvlizeig suvI] -BIjsny UOy}ION uBlerysny uBI[erysny UBI[vIYSNY uBlerysny ALIMOHLAV SLIGVH GNV Laid HLGaL GoOva 9 HIAVL 227 COMPOSITION OF SALIVA AND DENTAL CARIES ‘qguoo ted gTp'g Aj UO Zuloq poururexe y}00} 90e*g UI selmvo Jo ORI OY ‘SedvI OUNTTUTT om jo ouo sv suviydd3y oy} 80410 [JoroyoIg Aq pojonb youryeg wy} OJON , uopuoyT Yueq ‘uoJINg “WJ ‘Yy ‘Bolsy “op ut sdoqysjooy 4suy "Ld ‘uopuoT ‘ouosuvyy a ouojsnay, ‘yoy "T “H ‘oou10og. Y WON Ystjlig puwe Yeaviwg JO SsoatyeN ‘shUd “Wf Poy ‘uopuoyT “4uoqy = ‘*AIaA00 ‘yoo ureydey “Pa -SIq jo sodvAo, €6Z ‘6061 ‘QT ‘yoouw'T 4j0¥49 uopuoy Qudq ‘uoyINg “yf “Y ‘BOUIFY | 4SBGT UT sdoqs}0o,q 4sa1q ‘shy “OH Pov -py ‘uopuoT “ueq ‘A19A00 “SI jo soBvAoA ‘yoo ureydeg £9% “606T ‘LT “oourrT 470790 uopuo'y — uyor ‘ouey "M ‘oT ‘SuBIydASsy uropopy jo SUIOJSNY PUT SAoUUBPY JO YUNODDW (BsonT 999) 4sy pus ‘s]Moj ‘saoquinono ‘sured ‘oor Awp 8 sourry x18/Posuvoyo y}09], ‘190 BIBYO UT OyBIpAYoqavo ATUTE A] posn you SI pUBw [NJoISVISIP ST avsng ‘gueo aed QT ‘S330 pue yvour ‘YSiT “Yerp Jo yuoo sod Vg ‘oY0 ‘OZIBUL ‘SaTqBIOBOA 19YIQ “4oIp jo yueo ted Og ‘eon, = “yorp Jo yued ied og ‘oyejod yooms & ‘sojouIED Yous 4300} oy osn Ayenqiyeyy «‘spaeyunap = pengiq -ByT ‘“elqejeZo0a AyUO oy} spanory (9}104B4IQO JO puvIST JO SOATFLU 90g) e3ns jo soryyuenb 4veid sopnypout 4orcy ‘OMI SNOISI[OL B YYOO} JO But -uvolo oy], ‘surdny, ‘suveq ‘suo -joul ‘s#3o ‘y]IUI fozreur 10 YorpTUT jo pwoiq ‘sorqvyesoa ATureUr Yor, quourviedure} WAeM jo UZIS B PoOpIsSUOD 4490} pa19}9BIg (uMouyUN ysourye 018 Y490} poABoop puv 9YOVY4OO F,,, IB[NFII pus oy M pouruexoe S9SvVo JO LEqUINU 9B1B] B JO 4uUeD Jod ¢o'z Ajuo ul yuoseid sowed (jBuLog 999) SuOI}eU JeY}O JO osoyy UBYY 19100d ynq ‘449003 poos [eioues uy poururexe sosvo jo ques zed (6'0z UI yueserd soTAeD WOUITIOD oTHeT}OO J, [eutog BM “BIG JO SOATIBN ay 10YBIO §}O.103] GILL, SIOTRH So[S] A]PUeLyy sourd My uo -poyy puv 4uoro “uy, ‘suvydAaq JOHN ALBERT MARSHALL 228 “PHI ‘PIAI “PMI cggt ‘uopuoT ‘Avrinyy uyor “euoyssuIATT sopreyy pues preg “ojo ‘Isequiez oy} 0} UoT}Ipedxy Uv jo SeATZeIIeN yormuog ‘SUBIUBUISeT, 94} JO ajITyT Alreq 18 pus g ‘dd ‘puvsuq ‘xeyyey ‘Uog »p Bury “yoy ‘a ‘H ‘stuvwsey, jo soursmoqy uopuoy “usq ‘uoJINgG “WJ ‘Y ‘Bolsy yseq ul sdoqsyoo,y yi q9IP PoxIy] ‘090 ‘s007R40d qooms ‘vavsseo ‘suryduind ‘oor ‘suvA ‘synu ‘suveq ‘ozreur “TTA U10D INJey ‘ezreu ‘adr ‘sa0yvuU0y ‘sa04%40d qooms ‘ouvo avsns ‘s83o ‘s;Mo yeour yueyd -9]9 “may ppImM ‘1aeq ‘fou0y ‘u10d ‘10943nq ‘y]IUI ‘eBprtssi0d —SuIpnjour “yerp pextur [e1ousr [Bl1oyvul YSno04 puv prvy SULMOY 0} oNp 44004 Jo ssaupunog pooj [vultue uo Aperyo oary ‘UIT? OUO 4B UO9yBe oIB soTIT) -uenb 48013 puv ssourpee13 4vo13 YyIM u9zve o1B sSTVay *S94Ty -odde snoulioue oARyT ‘spury SnolvA Jo sejqeyes0A Auvur ‘sq81 980 {pooy uvedoinq osnjoy 190q F9TTTANT “891989 -o30A Moj ‘snojoy ‘Yse_y “yors 43004 Aq poeuBe[o o1B Y4904 OUI, , S89]04}19A0u UOUTUIOOUN II® 38 you syoBVy}009 Jo saseg ‘ABoop ynoyjim suossed pyo UI SUNS 944 0} UMOP IvIeM U94JO puv pljos e10ul are s10y Y499I,,, 4390} [NJTINVog uoul -wooun ABdep fuioM YonuI 44997, 4490} poos Sut -ABY SB UOSMOPIAA PUB dIOIPIP] -TIq BT ‘yoog Aq payrodar ory «<490} poos AI0A,, pBy []® pout -UIBX9 SOAT}VU OMJ-AYIO] Zuoury 0008q0} Aq pouTeys puv 100d o18 “yoofoid you op 4499], eyoquinye yy eluesueyy eseqviqoy OOO SUBIUBUIS® J, SUBIUVUIS® J, SUBIUBUISY J, [eutog ALIMOHLOAV SLIGVH GNV LaIa HLaagL Gove ponuyu0pD—9 WIEVL | COMPOSITION OF SALIVA AND DENTAL CARIES 229 influence the neutralizing power. One of the easiest methods of stimu- lation is the use of an electric current which has been passed through an inductorium, and experiments along these lines were projected. In this series of experiments it was desired to determine what com- parative differences would result in titratable acidity and alkalinity by the use of the electric current at different parts of the oral mucosa. It has been demonstrated that the mechanical stimulus obtained by the chewing of paraffine excites a flow of saliva which is markedly different from that found normally in the mouth. With the employment of the electric current, obtained from an inductorium, a third sample was se- _ eured which differed in titration value from either the normal resting or paraffine saliva. The amount of current used and the locality at which it was applied did not appear to produce any marked deviation from the general result. Although the strength of current varied with different individuals, only that strength was used which at the end of two minutes produced a tingling sensation at the point of contact. Whenever this amount was appreciably increased it was found to be prejudicial to salivary activity, as an unnecessary nervous tension was thus produced. The apparatus consisted of two Edison Lelande cells, type Z con- nected in series with an inductorium and a key. The electrodes con- sisted of two platinum points mounted on a vulcanite handle. The electrode was applied to the mucous membrane at some predetermined point and the saliva thus obtained titrated in the usual manner. The different localities at which the electrode was applied were, first, the opening of Stenson’s duct opposite the upper second molar in the buccal mucosa; second, the openings of Wharton’s ducts and the ducts of Bartholin on either side of the frenum linguae; third, on the dorsum of the tongue at the juncture of the posterior with the middle third near the apex of the V formed by the convergence of the two lines of the circumvallate papillae; fourth, at the gingivae. In applying the cur- rent at the bilateral structures one side was stimulated for two minutes and then the opposite side. No inflammation of the mucosa at the point of contact was developed at any time. The results of these ex- periments are reported in table 7. In the first column is noted the _ serial number of the patient, in each column “A” the alkalinity of 10 ec. of sample expressed in cubic centimeters of N/200 HCl; in “B” the acidity of 10 cc. sample expressed in cubic centimeter of N/200 NaOH., and in “‘C” the total neutralizing power. The salivary factor appears whenever the paraffine saliva was taken; the distance of the JOHN ALBERT MARSHALL 230 Giva 109 woLova 9161 ‘9T ynsny | 00°92 OL'%L | 0288] O8'IT | OF 9% | GO'Zh OLS | SoH | O8 Fe | G22 | S097] 8D “QI6I ‘ZZ Ajoe | 08°2z IL 9F G¢'Se | OO IT | G¢°22 | GS OF OS IT | S8'tF | GO'Ie | OS 9 | STST 8-4) QI6T ‘ce Ateniqeg | 0¢'6Z 09°22 | OL'% | 06°6T : Gest | Of S | G87} 4 YO QI6I ‘ST Aavnuse | 006% 0L°SS | 02'8T | S6°S | SZ FI | 06°68 jOS'S | OF LE | 02°22 | OS F | OLLI Ga QT6I ‘81 Aaenuve | OC 6z 02°86 | 06°91 | 0S°% | OF FI | S8'E% le8'T 00°22 | S¢°%% | S%@'9 | OL'9T a QI6I ‘21 Aaenuee | 0¢'6Z | 02°98 | OG 4T | O€% | OO'ST | 06°92 |G2°z clr | 02 $2 | O'S | O02} & A QI6T ‘Qt Aarenuve | ¢2°6z% 06°29 OF 22 | OLS | OF'6I | OT 9E |06'T OCF | 09 FZ] OFS | OL'IZ| 2 A QI6I ‘¢T Arenuee 00°08 60 ZL 08°Sz | S29 | SO'6T | O22 |OL'Z | OS 6% | 02 E% | S69 | SZ°9T Ae | 9I6I ‘g Axenuee | 0¢°6Z 00°18 00°€Z | O12 | 06°ST | G9°2z |Gz'8 OF 6T | OF ZZ | O09 F | OST y a 9161 ‘ZI Aaenuve | 00°ZE 06°8G | €9°8T | OS I—| E1°0% | S6'0E |GZ°E— | OL FE | SZ ST | OF T | G6°9T Aa | gI6t ¢ Axrenuve | 000s 09°62 | SO'9T | O&'e | €2°OT | z'S% jogs GL LT | 0G ST | 0¢°6 | 00'°6 tg SJONP 8 UYOYWDG PUD Ss UOJLDY QI6I ‘I azoquieydeg | 00°2z 08 8¢ OF’ se | OS IT | 01°22 | OF OF |O¢'0 | O8'6E | OL: Ez | OF & | OF'OZ| 8 OD QI6I ‘Zt Aavnuve | 00° 2z 0°29 «=| $9 ZT | 0O'L | SOOT | S262 [G9 F | 02°Sz | Gest | 086 | ¢2°8 Ne QT6T ‘g Aaenuve | 00'2z 02°¢¢ | S6°2z | OS % | Sh'Sz | OT ZF [Oc T— | 08'SF | 00°9Z | OG'9 | 09'6T 9 4 QI6T ‘OT Aavnuve | 00° 2z 09°21T | 08°9Z | OS'S | O8'ZT | OT 9E [SE's C128 | Sh Gh | GL 4S} OL ZT | 3 QI6I ‘2, Aavnuve | 000 OF OL 0g°e2 | GL°21 | SG'2 | 06°82 |¢2°F | STZ | GE'0% | SLT | 09'°9 ancl SI6I ‘8z Joquieseq =| 00°22 0S 69 Gh'0Z | OO'TL | SF'6 | 06°22 [OFS | OS FZ | OF ET | O€'ST | OT'F Aes | QI6I ‘¢ Axrvnuse | 0g GZ os9¢ | 02°€% | 09°¢ | OZ'ST | G9°2e |Ga'e OF FE | OT'IZ | 08's | OS 'ST eee: | SI6I ‘8z goquteseq 0G 22 08 69 08°'Iz | 0S 9T | 08 F | S8°2z% |se°% o¢'sz | Of 6I | Sz'¢ | SOFT I a "mo *09 “09 5 °99 "92 *02 *09 pyonp s uosuary 9) a Vv fe) a Vv re) a Vv ‘d‘'N‘L| HO#8N IOH ‘d'N’L| HO®N IOH ‘d'N’L| HOGN IOH VAITVS VAITVS GaLVIAWNILS ATIVOINIOGRIaS ‘daLV IOWILS ANIGAVUVG VAIIVS DNILSTY TYNYON ugd@aWan e SNINUYS 10921999177 ,, 24 ATaVL q O16T ‘Tr Aaenuve Q16I ‘@ oquioaoyy ‘eT sd ol. ee _ 232 JOHN ALBERT MARSHALL secondary coil from the primary shows the comparative strength of the current and is expressed in centimeters. The data submitted demonstrate first, that the alkalinity of the electrically excited saliva is lower than that of the paraffine sample; sec- ond, that the acidity of the former is relatively higher than that of the latter; and third, that the neutralizing power of the saliva obtained by electric stimulus is lower in every instance than that collected hy the paraffine method. ‘This fact indicates that the use of the inductorium does not promote salivary activity to the same extent as the’ paraffine. It was hoped that differential areas of irritation in the mouth could be demonstrated by means of this electrical stimulation for it was de- sired to determine whether one part of the oral mucosa is more sensi- tive to extraneous influences than another, so that a local irritation in one circumscribed locality would tend to produce a greater change in the salivary neutralizing power than in another locality. The ex- periments, however, so far reported, have failed to supply any conclusive evidence. The repeated use of the inductorium on one individual produces a marked effect upon the relationship of the normal resting saliva to the activated paraffine saliva. This electric irritation alters the factor in a few days from one which is comparatively low to one with a much higher valuation. As examples of this condition may be cited subjects E1, E4 and G8. In the first instance the experiment was commenced on December 28, 1915. The salivary factor at this time was 69.3. - Electrical stimulation was applied and after eight days a second test was made. The normal resting saliva and the paraffine saliva were first collected and the application of the current repeated. It was found that the factor rose steadily from 69.3 to 79.6 and finally to 93.2. Similarly with E4 the factor at the start was 69.5 but rose in eleven days to 81.0. For subject G8 the factor evaluated on July 23, 1916 was 46.7; on August 1 it had risen to 74.7. One month later it had returned to nearly the same ratio as at the start and duplicate samples on suc- cessive days gave a factor of 58.8. The results reported are based on duplicate analyses and on duplicate samples obtained on successive days. The determination of the reaction of the taste impulses upon salivary secretion was attempted by comparing the results obtained with a nor- mal resting sample with those secured by the use of different substances of marked taste. Howell (12) and other investigators describe the taste sense as being composed of four fundamental sensations, namely, bitter, sweet, acid and salty. Tastes other than these are combinations of any two or more of the primary sensations and produce, therefore, a COMPOSITION OF SALIVA AND DENTAL CARIES 233 mixed stimulation of the sense organ. There was used for this ex- periment quinine on one day and sucrose on the second or third day — following. In securing the samples there was collected first a resting saliva and then the saliva secreted from the use of the taste stimulant. The re- sults are reported in table 8. In the first column appears the subject number. Figures in columns A, B and C represent respectively the alkalinity, acidity and total neutralizing power of each sample. It may be stated in general that the action of sucrose, the sweet stimulant, tends to lower the total neutralizing power of the saliva. Quinine, the bitter stimulant, appears to produce the opposite condition yielding saliva resembling that secreted in response to the stimulus of chewing . TABLE 8 Comparison of effects of different types of stimulation upon salivary secretion SALIVA SALIVA FROM QUININE SALIVA FROM SUCROSE NUMBER NORMAL RESTING STIMULUS STIMULUS ; A B Cc A B Cc ei.” B Cc Fl 16.25 | 6.95 | 23.20 | 20.65 | 4.55 | 25.20 | 13.89 | 10.70 | 24.50 F2 19.55 | 4.20 | 28.75 | 33.95 | 2.00 | 35.95 F2 25.50 | 5.00 |.30.50 | 25.80 | 3.75 | 29:55 | 17.15 | 8.00 | 19.15 F2 21.15 | 7.50 | 28.65 ‘| 18.45 | 5.50 | 23.95 F3 20.70 | 3.70 | 24.40 13.25 | 8.00 | 21.25 F3 18.40 | 5.40 | 23.80 9.95 | 8.25 | 18.20 F4 19.25 | 4.75 | 24.00 15.00 | 6.20 | 21.20 F5 19.10 | 5.75 | 24.85 10.65 | 8.00 | 18.65 G7 22.85 | 2.50.| 25.35 12.60 | 3.60 | 16.20 Fi 18.05 | 4.70 | 22.75 | 19.60 | 13.40 | 33.00 | 13.85 | 7.00 | 20.85 paraffine. The well known work of Miller (13) demonstrates that fer- menting carbohydrates such as would be found on the teeth tend to promote caries. From the above experiments it would be logical to infer that saliva favors the condition as well, since the neutralizing power of the sucrose-stimulated saliva approaches that of the normal resting saliva, and a factor evaluated on this basis would be of a magnitude cor- responding to that which under the paraffine stimulus would indicate the presence of caries. SUMMARY Dialysis of saliva shows that the total neutralizing power is chiefly due to inorganic constituents. The use of the Van Slyke apparatus in the determination of amino- nitrogen in saliva is a new application of this method. The results so 234 JOHN ALBERT MARSHALL obtained show that there is a definite correlation between the concen- tration of inorganic constituents, the amino-nitrogen content, and the neutralizing power of saliva; namely, that a high neutralizing power is associated with a correspondingly high percentage of inorganic constit- — uents and with a low percentage of protein. Data are presented which confirm Pickerill’s observations concern- ing the effect of different constituents of the diet upon dental caries. The use of the electric current as a salivary stimulant excites a se- cretion markedly low in alkalinity and correspondingly high in acidity when compared to the saliva resulting from the paraffine stimulus. The neutralizing power of saliva secreted in response to electrical stimuli is less than that secreted in response to the chewing of paraffine. Differential areas of stimulation in the oral mucosa cannot be demon- strated. ' The comparison of analyses of saliva obtained by a sweet stimulus (sucrose) with that obtained by a bitter stimulus (quinine) proves that the former yields a saliva low in neutralizing power and that the latter produces the opposite result. 2 In conclusion I wish to express my deep appreciation to Dr. T. Brails- ford Robertson for his interest evinced throughout the work as. well as — for the valuable suggestions offered. My thanks are also due to those students and clinic patients who have provided the necessary material for the experimental purposes. BIBLIOGRAPHY (1) PickeriLu: The prevention of dental caries and oral sepsis, 2nd ed., Phila- delphia, 1914. (2) UNDERWoop: Brit. Med. Journ., 1910, ii, 621. (3) Austen: Brit. Med. Journ.,; 1910, ii, 772. (4) Marsuauu: This Journal, 1915, xxxvi, 260. (5) Samparp anp Grins: Journ. Allied Dent. Soc., xi, 1916, 273. (6) Marsuatu: This Journal, 1916, xl, 1. (7) Gres: Journ. Allied Dent. Soc., xi, 1916, 488. (8) Marsuatu: Dent. Cosmos, 1916, viii, 1225. (9) Marsnauy: Dent. Cosmos, 1917, lix, 33. (10) Cuausen: Journ. Biochem., 1914, xvii, 417. (11) Porter: Quart. Journ. Exper. Physiol., 1910, iii, 375. (12) Howe: Textbook of physiology, Philadelphia, 1915. (13) Mitizr: Dent. Cosmos, 1902, xliv, 437. (14) Havx: Practical experimental physiology, 5th ed., Philadelphia, 1916. (15) Davenport: Statistical methods, with special reference to biological varia- tion, New York, 1904. So PP o> S\eE)5)3/ 2) 2/2)2) 3/28] 28 | & Simitigie| 2 1} 2) 2/2) 2 8") ek 1916 lbs. May 15 /16 |17|17 5O/11.1 |13.9)15.2 40.2 trace |36.0) Hemoglobin 113 per cent. Bled 122 ce. May 16 |17 |23/18) | 58| 7.6 |10.4| 6.8 24.8 0 |35.8| R. B. C. 6,- 444,000. He- moglobin 114 ; per cent May 17 |16*|17/20|15) 52) §.2*|13.7|19.8|17.0'50.5/14.7| + (35.8 May 18 /19 |17|15 51/12.0 |13.0)14.2 39.2 + |36.0)Stools contain no__sterco- bilin May 19 | 9*/25/20|15) 60) 5.8*|11.3)18.6/17.0/46.9/11.1) + (35.5 May 20 44 39.2 + {35.0 *5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 35.8 mgm. per six hours. Average of twelve control days before hemoglobin injection experiments is 35.9 mgm. or twenty-four hour collections are extremely trying for the dog and if repeated at short intervals will change the dog from a normal to an abnormal condition. As stated in our first paper, we are convinced that these short period collections give a more nearly normal picture than do the long or continuous observations where the dog can not even approximate a normal condition. 266 G. H. WHIPPLE AND C. W. HOOPER Tables 44 and 45 continue the observations on dogs 16-138 and 16-60 (tables 42 and 43). These tables show the effect of a slight bleeding before the injection of hemoglobin. The results of the hemoglobin injection are .striking—there is only a small output of bile pigment — above the mean in one dog and. no bile pigment increase in the other dog. Here is a very suggestive observation which may have something TABLE 45 ; Bile fistula. Bleeding. Hemoglobin injection Dog 16-60 BILE 3 REMARKS & : . >} / agape rch Bile pigments in milligrams 3 g DATE a 25 . . s . . gi ra Sree Mixed diet A heed om ae: ‘ F a 9 F a & zlgjgleis| £/£)2] 2/4/28] 28 | €| TINS ol ba os oe es eee 1916 lbs. May 15 |22 |22/21) | 65) 8.9 | 9.9] 8.9 27.7 trace |36.8| Hemoglobin 97 per cent. Bled 130 ce. May 16 /28 |19/20] | 67| 6.3 | 6.0) 5.0 17.3 trace |37.3} R. B.C. 5,728, - 000. Hem- oglobin 95 per cent May 17 |20*/15/13/12} 40) 3.8*| 4.1] 2.3) 2.4] 8.8] 0 |trace |37.8 ; May 18 |87 |17/13 67) 8.9 | 2.3) 2.3 13.5 trace |37.5| Hemoglobin 96 per cent May 19 |27*/16/22/15| 53] 8.5*| 6.4) 5.4) 3.4|15.2| 0 trace 37.0 May 20], 60 19.6 trace |36.8 * 5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 19.5 per six hours. Average of twenty control days before this experiment is 22.4 mgm. of importance back of it. May we assume that the greater part of this injected hemoglobin pigment radicle may have been deviated in the body to supply some pigment demand which was precipitated by this removal of blood? Before we answer yes to this question, as we might wish to do, let us consult the following tables 46 and 47 where the anemia is marked but the pigment retention not definite nor uni- form—in other words, the anemic dogs react like the normal controls. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 267 Tables 46 and 47 continue the history of these two simple bile fistula dogs. Anemia of a definite grade is produced by aspiration of blood from the jugular vein in sufficient amounts to cause a drop in the dogs’ hemoglobin to 50 per cent or less. This acute demand for hemoglobin pigment to make into new red cells might react on the general pigment metabolism (compare tables 44 and 45) as has been suspected. There is no direct evidence in these tables 46 and 47. We note the same wide variation in bile pigment excretion following a unit injection of hemo- globin. In fact, the variations are even more marked than in the normal periods of these same dogs (tables 42 and 43). The escape of bile pigments in the urine is more noticeable. The mean daily bile pigment output during this anemia period (dog 16-138, table 46) is subnormal, 25.6 mgm. as compared to the normal 35.9 mgm. per six hours. This is much more evident during the period of regeneration following the acute anemia (table 48). Dog 16-60 has a subnormal output to begin with, and does not show this drop in bile pigment secretion (table 47). Tables 48 and 49 show one important fact. During the period of acute regeneration following anemia produced by bleeding bile fistula dogs, there is a very low mean bile pigment output. Dog 16-138 has a normal mean daily output of 35.9 mgm. bile pigment per six hours (table 42). During the acute anemia period this mean output of bile pigment falls to 25.6 mgm. per six hours (table 46); but during the regeneration period following the anemia this mean bile pigment output per six hours falls to 11.2 mgm. or even to 7.2 mgm.—a drop to one-third normal or even less. Dog 16-60 shows a less striking reaction but one which is quite similar. It is to be recalled that this dog’s initial mean bile pigment output per six hours was subnormal (table 43). Dogs during the periods of blood regeneration following hemorrhage will put out less bile pigments than normal. There are several possible _ explanations and probably several factors concerned. The circulating red cells are somewhat below normal, and the number destroyed pre- sumably smaller than normal, consequently the amount of pigment furnished in this manner is subnormal. It is more than possible that there is a conservation of pigment substances under these conditions of blood regeneration, but this term pigment conservation must be limited to the body pigments at present, however much the temptation to extend it to pigments introduced from without (e.g., hemoglobin). We have no evidence here that any more injected hemoglobin may 268 G. H. WHIPPLE AND C. W. HOOPER TABLE 46 Bile fistula. Anemia. Hemoglobin injection Dog 16-138 BILE & REMARKS i) : . <>) Ane Bile pigments in milligrams 5 Zz DATE Cg ae BS g a § Se Mixed diet BO es ea A, 4 : . = 2 f ee E EVE SlEla|2/}2£|4/5| 3/85] #2 | & SIS oitie) & bee ted et eet Bo oe 1916 lbs. May 22 |31 |25/30} | 86/12.5)11.9]16.4 40.8 + |34.5| Hemoglobin 115 per cent. Bled 122 cc. May 23 |17 {15/20} | 52/9.6 |11.8/14.9 36.3 trace |34.3) Hemoglobin 102 per cent. Bled 282 cc. May 24 |18 /15/17} | 50/9.8 |11.2/11.8 32.8 trace |34.5) Hemoglobin 94 per cent. Bled 200 ce. May 25 |16*/22/20/21) 63/7.9*|19.6/24. 2/18 .6/62.4/36.8] +--+ |34.3} Hemoglobin 66 per cent May 26 |20*/26/21/18) 65/9.1*/23.6/19.8]15.4/58.8/33.2} ++ |36.0 May = 27 |29 |20/18 67/7.9 | 5.4] 6.0 19.3 trace |36.3 May 29 /|26 |16|22) | 64/8.2 | 8.6] 9.9 26.7 + |34.8) Hemoglobin 88 per cent. Bled 350 ce. May 30 |16 /15/12| | 43/7.2 | 8.1! 9.4 24.7 trace |34.8) Hemoglobin 63 per : cent. Bled 200 cc. May 31 |19*|18/22|22) 62|5.6*|14.5/33.6/15.8/63.9/38.3) ++ |34.5| Hemoglobin 52 per cent June 1 |20*/25/25/23) 73/7.2*/13.5|15.8/14.0/43.3]17.7| ++ |34.8 June 2 {16 |16|23 55|7.9 | 9.0] 9.9 26.8 trace |34.3 June 3 60 18.8 trace |34.5| Hemoglobin 71 per cent. Bled 350 cc. June 65 {18 /18)19 55/8.9 | 9.4! 8.5 26.8 trace |33.5| Hemoglobin 60 per cent. Bled 350 cc. June 6 |14*/19}18)14) 51/4.7*|15.3/17.8|14.6/47.7/22.1| ‘+ 133.5 Hemoglobin 38 _ per cent June 7 |24 |21/21 66/7.6 | 4.7] 7.6 19.9 trace |34.3 June 8 [25*/25/24/32 81/5.1*|10.8/14.0}12.2/37.0]11.4 + 134.5 Hemoglobin 51 per cent June 9 (22 |22/23 67/4.9 | 4.4! 5.2 14.5 + |34.5 June 10 52 20.0 + |34.0} Stools contain no stercobilin * 5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 25.6 mgm. per six hours. Average for twelve control days with no anemia is 35.9 mgm. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 269 TABLE 47 Bile fistula. Anemia. Hemoglobin injection — Dog 16-60 BILE o a apna er Pay Bile pigments in milligrams rs - DATE z| 2% a a oS} BR : é A) ¢A8} #8 é}siéis|° ; ; a | o | Bo > BiSE\El3| |) 5)2)5)3188) B Pemeitial 2 12) 2 |) 2 \s"| 83 1916 May 22 {31 |25)30) | 86/12.5 |11.9|16.4 40.8 trace May 23 /23 |18/24) | 65) 9.9 | 9.4/11.9 31.2 trace May 24 |17 /17\34| | 68) 6.8 | 6.8/11.5 25.1 trace May 25 |16*|13)18/20) 51) 6.4*| 7.3)13.8|17.3|38.4|17.3\trace May 26 /|26*/28/34/27| 89) 8.8*|15.8/26.4/20.1/62.3/41 .2|trace May 27 |28 |19/20) | 67| 7.6 | 6.0) 8.1 21.7 trace May 29 |18 |19|19 56} 8.1 | 5.6) 8.1 21.8 trace May 30 |24 |12/19| | 55| 5.9| 4.4/6.4) [16.7] _|trace May 31 /21*|23/24/21| 68) 5.2*| 8.2/11.0| 5.2/23.4) 2.3/trace June 1 /|22*/30/40|22) 92) 6.4*| 8.2/20.8) 8.9/37.9|16.8) + June 2 (24 |20/32 76} 5.4 | 3.2) 4.4 13.0 trace June 3 50 17.0 trace June 5 |30 |19\20 69| 6.8 | 4.8) 5.0 16.6 trace June 6 /|15*|20/28/27| 75) 6.0*|16.2/34.0/20.6/70.8/49.7| ++ June 7 |29 |24/26 79) 6.5 | 5.9] 5.8 18.2 trace June 8 |30*/31/32/34) 97) 5.2*/14.0/22.3)16.0)52.3/31.2) ++ June 9 |27 (32/25 84) 2.4 | 5.1) 3.3 10.8 trace REMARKS Mixed diet Hemoglobin 94 per cent. Bled 130 ce. Hemoglobin 85 per cent. Bled 262 cc. Hemoglobin 70 per cent. Bled 100 ce. Hemoglobin 68 per cent Hemoglobin 78 per cent. Bled 325 ce. Hemoglobin 59. per cent. Bled 160 ce. Hemoglobin 51 per cent Stools contain no stercobilin Hemoglobin 51 per _ cent Hemoglobin 54 per cent : Hemoglobin 62 per cent * 5 ec. laked red blood cells introduced into the jugular at the end of the second hour. Mean daily output in table = 21.1 mgm. per six hours. Average of twenty control days before anemia is 22.4 mgm 270 G. H. WHIPPLE AND C. W. HOOPER TABLE 48 Simple bile fistula. Anemia after-period. Hemoglobin injections Dog 16-138 BILE : REMARKS i i a8 Srocues 9 ote Bile pigments in milligrams =) : DATE a Ws Z g| §| $a Mixed diet a a §]/ a. oleldlebe le) Bi ett eerie Se Zz q Pam Wee Aa ed Gy ray a fe] i pas S > aD o a [ajcis) 8] 4/4/45) 4) B@/ Es) 2a = STITT ets oy Dae eee | ee ee 1916 Ibs. June 26 |29 |27/27} | 83/2.0 |2.4) 1.8 6.2 0 |33.5| Hemoglobin 83 per cent June 27 |20*|23|26/25| 74/2.2*!9 .9]10.5]10.8/31.2/24 Ojtrace |33.8 June 28 |24 |25|23 72|3.2 |2.8) 2.6 8.6 trace |34.3} Feces contain no stercobilin 9 .4/24.1|16.9]trace |34.1 June 29 |22*\23/25/22| 70/3.0*/5.7| 9.¢ 3.3) 2. 7.0 trace |35.0 June 30 |16 |25)25 66)1.4 wo: *5 cc. laked red blood cells given intravenously at beginning of third hour- Mean daily output in table = 7.2 mgm. per six hours. Average for twenty control days during this anemia after-period, 11. 21 mgm. TABLE 49 Simple bile fistula. Anemia after-period. Hemoglobin injections Dog 16-60 BILE S REMARKS A ti bi 5 > peer gals 1€! Bile pigments in milligrams 5 z DATE a HO g g : Es Mixed diet lgglels| £]£/ 2) 4/3 |e8lebl & TTsele| Fit LT] ] eerie) 1916 lbs. June 26 |37 |24|32 93/3.3 |1.7| 2.8 7:3 0 |34.0/ Hemoglobin 76 per cent June 27 |19*/13|/30/23| 66/5.2*/3.8]/10.0|10.4/34.2/24.3] 0 |34.0 June 28 [39 |20/22) | 81/4.1 |2.2) 2.5 8.8 _ 0 |34.5) Feces contain no stercobilin June 29 |30*|21/28/30| 79/5.2*/2.9 3.2)11.4/17.5) 7.6) 0 |34.3 ; June 30 |39 |31/33]} |103/5.9 13.5] 3.7 13.1 0 |35.0 * 5 ce. laked red blood cells given intravenously at the end of the second hour. Mean daily output in table = 9.9 mgm. per six hours. Average of twenty control days in this anemia after-period is 130 mgm. HEMOGLOBIN INJECTION AND BILE. PIGMENT OUTPUT 271 be retained in the body during active blood regeneration than during normal periods or intervals of acute anemia. It is probable that the body can-not-use hemoglobin as such when injected intravenously, but must break it down into its constituent parts before actual pig- ment construction can take place. How much of this indirect conser- vation of body pigment does actually take place we cannot say at present. There is at least a possibility that some of the hemoglobin TABLE 50 -Simple bile fistula. Hemoglobin injections Dog 16-138 BILE ] REMARKS a = = Amount in cubie| Bile pigments in milligrams ze DATE a rf a a | $| 58 Mixed diet pb te ine .| = /g8| "fe | elles) 2/2) 2) 2/5/28) e213 Titgjzie/2/2\ 5) 2) 2 le4| ee 1916 lbs. July 5 |19 |22/24) | 65/3.4 | 3.0) 2.7 9.1 0 (33.5 July 6 |16*/20)16)22) 58/4.0*| 5.4) 6.1] 7.9|19.4) 4.45 0 (34.0 July 7 |19 |17/20| | 56/6.8*| 7.7| 8.1 22.6 0 |34.0 July 10 |32 |24/26) | 82\2.8 | 2.7) 3.5 9.0 trace |33.8) Hemoglobin 91 percent. R. B. C. 5,856,- f 000 July 11 |23*/15/18/20) 53/3.6*| 2.6) 3.6) 8.6/14.8} 0 | ++ |34.0 July 12 /23 |22/22) | 67\9.4 | 5.4) 5.4 20.2 trace |34.3) Stools contain no stercobilin July 13 |15*|15/31/19| 65)/2.0*) 4.4/11.8) 9.9)26.1)11.1),+++/34.5 July 14 |18 |22)24) | 64) 4.0) 5.4) 4.9 14.3 + (34.5 * 5 ec. laked red blood cells given intravenously at end of second hour. Mean daily output in table —15.0 mgm. per six hours. _ Average for twenty control days during this anemia after-period is 11.2 mgm. _ pigment radicle which does not appear in the bile in some of these in- _ jection experiments may be taken up by the body cells, digested and used for other purposes in the body. There is no evidence that the body cells acquire greater efficiency in this respect after repeated hemoglobin injections. Tables 50 and 51 continue the observations and experiments upon _ these same two simple bile fistula dogs. Regeneration of red cells 272 G. H. WHIPPLE AND C. W. HOOPER following the anemia is still in progress and the general picture is very like that noted in tables 48 and 49. The very low bile pigment output of dog 16-138 after the usual hemoglobin injections is of interest. The urine contains more bile pigment than usual, but at the most only a very few milligrams. The level of blood hemoglobin has returned close to normal, but the mean bile pigment elimination is still much below normal—50 per cent of normal in one dog (16-138). Simple DATE 1916 July July July July July July July July July 5 SCoON Om —_ 11 12 13 14 Amount in cubic centimeters £|2/2/2| 3 a laolelel 23 |19|26 68 22*| 26/23/23) 72 24 |19/21 64 50 19 |23/26| | 68 27*|26/23/25)| 74 25 112/32 69 19*/25|20|32| 77 28 |25/27| | 80 TABLE 51 bile fistula. Hemoglobin injections Dog 16-60 BILE 8 REMARKS Bile pigments in milligrams ae 1 aid E 3 ae Mixed diet of Sa e £4 2 a & = a a a a 3/8 4a g TP T4 21 fA eae ee lbs. = 7.7 | 3.4] 2.4 13:5 0 |34.0) Hemoglobin 94 per cent 6 .4*/12.4/18.8) 9.9/41.1/27.0) + 134.5 6:7 | 453) 4.7 15:7 + |34.3 17.0 trace |34.0 6.0 | 3.1) 4.0 13.1 trace |33.3| Hemoglobin 102 per cent. ie B. Cc 5,- 940,000 4.2*113.4|16.1]10.8/40.3/26.2) + |34.0 6.1 | 3.5] 7.2 16.8 + |34.3) Stools contain no stercobilin 5.2*| 6.7] 9.9|10.8/27.4]13.3} + [33.8 Oo. 2 leer ee 8.9 trace |33.8 * 5 cc. laked red blood cells given intravenously at the end of the second hour. Mean daily output in table 14.1 mgm. per six hours. Average of twenty control days in this anemia after-period is 13.0. CLINICAL AND AUTOPSY SUMMARY Simple bile fistula Dog 16-60. Adult, fat, brindle bull dog, male, 44 pounds. December 21, 1915. Ether anesthesia, bile fistula operation, general condi- tion good during post-operative period. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 273 May 1, 1916. Dog in good condition, 37.5 pounds. See table 43 for details— experimental period. ~ July 14,-1916> Dog in good condition—end of this experiment, 34 pounds. July 21. General condition good, vigorous appetite, normal activity, 33 pounds. Hemoglobin 104 per cent. Ether anesthesia and killed by bleeding. Autopsy at once. Thorax negative. Spleen is normal size, the malphigian corpuscles are perhaps a bit enlarged. Liver shows a definite increase in brown pigment. Bile passages are clear and pale. Common duct is completely ob- structed and no bile can enter duodenum. Its lower end is dilated to about 1 em. in diameter. Stomach and intestines show a normal mucosa. The muscle coat of the small intestine is pigmented a maple sugar color which is character- istic of these dogs. Some of the retroperitoneal lymph glands are large and con- tain calcified foci. The other organs present nothing abnormal. Microscopical sections are negative except from the liver, which shows a slight but definite increase in portal stroma which contains many mononuclear cells. There is a slight increase in pigment in the liver cells. There is no proliferation of the bile ducts. Simple bile fistula Dog 16-138. Adult bull dog, male, 38.8 pounds. March 25, 1916. Ether anesthesia; bile fistula operation; general condition is ~ excellent during post-operative period. April 24, 1916. Dog in good condition, 36.3 pounds. See table 42 for details— experimental period. . July 14, 1916. Dog in good condition, 34.5 pounds. End of experiments cited above. August. Dog remains in good condition. September 6. Dog in coma, bloody urine; blood shows very high urea and non-protein nitrogen. Ether anesthesia and killed. . Autopsy at once. Thorax negative. Spleen is small and fibrous. Liver shows pale ducts and complete exclusion of bile from the duodenum. Stomach and intestines show nothing of interest. A urethral stone is found just below the prostate. Bladder is large and its wall is thick. There is a definite hemor- rhagic cystitis with early ulceration. Both ureters and kidney pelves are dilated. The kidneys show multiple miliary abscesses in the cortex. Microscopical sections add no information of importance to this experiment. SUMMARY It has been generally assumed from the publications of other workers that hemoglobin injection will be followed by a uniform elimination of a corresponding excess of bile pigment—in other words, hemoglobin injection is followed by its quantitative elimination as bile pigment. Our experiments show that the reaction of a bile fistula dog under uniform conditions following a unit injection of hemoglobin intraven- 274 G. H. WHIPPLE AND C. W. HOOPER ously is certainly not uniform. That the elimination of the pigment radicle of the hemoglobin may be quantitative we can not deny, but surely all our experiments speak against any such probability. Some hemoglobin injections may be followed by little or no increase in bile pigment secretion. It is possible that this pigment is taken up by the. body cells and used for some purpose, perhaps to build other body pigments. During periods of acute anemia from bleeding, the bile fistula dogs may show a fall in bile pigment output and at times a lessened elimina- tion of bile pigment following a unit injection of hemoglobin. The latter is quite inconstant, and yet may indicate some conservation of hemoglobin either directly or indirectly. Compare the influence of splenectomy and the associated increase in bile pigments in the next paper. Bile pigment elimination may be greatly depressed during the period of acute regeneration following an anemia, yet the response to hemo- globin injections may be even more irregualr and inconstant than under normal conditions. It is highly probable that we have a cer- tain pigment conservation under these conditions, but the mechanism of this conservation may not be simple and direct but possibly very complex. The whole question of pigment construction is concerned in this problem and much more experimental work is required. BIBLIOGRAPHY (1) Pearce, AusTIN AND EisenBREY: Journ. Exper. Med., 1912, xvi, 375. (2) SELLARDS AND Minor: Journ. Med. Res., 1916, xxxiv, 469. (3) StapELMANN: Der Icterus und seine verschiedenen Formen, Stuttgart, 1891. ¢ (4) Bruescu anp Yosurmoro: Zeitschr. f. exper. Path. u. Therap., 1910-1911, viii, 639. : (5) Bruescu anp Kawasuima: Zeitschr. f. exper. Path. u. Therap., 1910-1911, viii, 645. : 7 (6) Eppincer anp Cuarnas: Arch. f. klin. Med., 1913, Ixxviii, 387. (7) WinBurR anp Appis: Arch. Int. Med., 1914, xiii, 235. (8) WurppLe anp Hooper: This Journal, 1916, xl, 349. (9) WuiprLe anp Hooper: This Journal, 1917, xlii, 544. (10) Hooprr: This Journal, 1917, xlii, 280. (11) WutprLe anp Hoopsr: This Journal, 1917, xlii, 256. (12) Hooper anp WutppieE: This Journal, 1916, xl, 332. BILE PIGMENT METABOLISM VIII. Brite Piement Ovutrut INFLUENCED By HEMOGLOBIN INJECTION ; SPLENECTOMY AND ANEMIA C. W. HOOPER anp G. H. WHIPPLE From the George Williams Hooper Foundation for Medical Research, University of California Medical School, San Francisco Received for publication March 16, 1917 This paper gives the results of experiments upon bile fistula dogs with an added splenectomy. Unit amounts of hemoglobin are injected intra- venously during a control period, during an anemic period and during _ the period of blood regeneration. These experiments supplement those of the preceding paper as well as giving additional controls. The two bile fistula dogs are followed through a normal period, a period of bleed- ing and anemia, followed by an interval of acute blood regeneration. - Here some peculiar differences due to the splenectomy are noted and will be discussed later. A few points bearing on splenectomy and pigment production may be appropriately taken up in this place. It is known that splenectomy in dogs may often be followed by a moderate anemia (Musser (1), Pearce, Austin and Musser (2), Musser and Krumbhaar (3) ), but it is not decided whether this anemia is due to blood destruction or to less active blood formation. It has been claimed by Martinotti and Bar- baci (4) that after splenectomy the output of bile pigments drops to one-half normal. Enough experiments are reported below and in a preceding paper (5) to disprove this claim. We feel safe in asserting that under normal conditions a bile fistula dog will put out the same average amount of bile pigments whether with or without a spleen. It must be kept in mind that bile fistulae with splenectomy plus anemia from bleeding will often show remarkable periods of abnormally high pigment output alternating at times with periods of low bile pigment secretion. This is a very interesting chapter in our splenectomy studies which calls for much more work. It is to be noted that the bile fistula dogs 275 276 C. W. HOOPER AND G. H. WHIPPLE with splenectomy may remain in a normal condition indefinitely and put out the usual amount of bile pigments per six hours. After a moderate — anemia has been produced we may see a striking reaction—not a drop in mean bile pigment output as noted in controls with spleen intact but a period of great increase in pigment output. There is jaundice and heavy pigmentation of the urine. The blood cells may decrease still further and we may assume that all this increase in bile pigment secre- tion is due to red cell destruction. There may be another possibility or several more in fact. We have stated that a bile fistula dog with splenectomy plus anemia may show periods of abnormally high bile pigment secretion during the. long interval of blood regeneration which is greatly prolonged under such conditions. May we say that pigment production is depressed below normal? On the contrary we see’ an abnormal amount elimi- nated in the bile and urine (icterus). Further, it is to be noted that the color index of the blood during such periods is about unity as com- pared with simple bile fistula dogs with anemia when the color index is below one, as is usually observed in a secondary anemia. This may mean that pigment is present in excess, and the corpuscles are satu- rated with hemoglobin. It is not inconceivable that the frame work of the red cell may be formed through some mechanism and as it ap- proaches maturity may be loaded with hemoglobin which it may select from the materials brought to it by the blood. Undersome circumstances the formation of hemoglobin may be in excess of the stroma and the color index may approach unity. These experiments do not show many figures giving the blood color index, and we prefer to leave this point for future consideration. How may we explain the periods of icterus which occur during the regeneration period following anemia in a bile fistula-splenectomy dog? We could show chart after chart in which periods of icterus appear suddenly with no obvious cause and as suddenly vanish with a long slow rise in the hemoglobin curve toward normal. There may be a sharp drop in the number of red corpuscles indicating disintegration of red cells to be in part responsible for the icterus. Yet the spleen is absent and cannot be blamed for this ‘hemolytic icterus.” May we assume that the corpuscles manufactured during such periods are in some way faulty? We may believe that the stroma is less abundantly produced than the hemoglobin; perhaps the stroma is actually defective. The hemoglobin injections give the same remarkable bile pigment fluctuations in relation to the mean output. It will be recalled that 5 HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 277 ec. of laked red cells should be equivalent to 60 mgm. of bile pigment, if there is a quantitative change of the hemoglobin pigment radicle to bile pigment. Do we find any constant output of bile pigments _above the mean following a unit injection of hemoglobin? On the contrary we find all variations from 4 mgm. to 42 mgm. increase above the mean daily bile pigment output—the same inexplicable variations TABLE 61 Bile fistula. Splenectomy. Hemoglobin injection Dog 16-10 BILE ' o REMARES a pene cubic Bile pigments in milligrams a 5 r DATE =| 43 g é | 8| §& Mixed diet slale = a E = 38 | & EISiSiEig| 5) 2) 8) 2) 3/85] 82] 8 Pigiicja) 2 12/2) 2;e las] 821s 1916 Ibs. _ June 26 (24 |29/25) | 7812.9 |12.4/11.3 36.6 trace |29.0| Hemoglobin 106 ~—s per cent June 27 |27*|20|17/23) 60)14.0*/14.9) 8.8|26.8/50.5)17.5) + |29.0 June 28 |23 |22|25) | 70/11.4 | 8.9) 6.7 27.0 trace |29.1| Stools con- tain no stercobilin June 29 /18*/27/21/20| 68) 8.1*/17.0)17.1)16.1/50.2)17.2} + {29.0 June 30 [26 |24/19) | 69/11.0 |11.0) 9.9 31.9 + (|29.5 July 1 46 : 31.8 trace |29.3 July 5 /25 |21/19 65/12.9 | 9.4) 8.6 30.9 trace |29.0 July 6 |21*/26)15/19) 60/11.9*/15.4/19 6/1 .8|53.8|20.8) + 29.5 July 7 /24 {22/20} | 66/11.9 |14.0)14.0 39.9 + |29.5 * 5 ce. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 33 mgm. per six hours. Average of twenty control days 34.4 mgm. per six hours. _ Bile fistula operation and splenectomy done September 30, 1915. in pigment output under as near uniform conditions as can be attained. The reaction following hemoglobin injection is the same in a simple bile fistula as with a combined bile and Eck fistula or a combination of bile fistula and splenectomy. The dogs used in these experiments had been under observation for months in apparent perfect health and under the usual routine described previously (6). The methods used have 278 been described in detail in former papers. Cc. W. HOOPER AND G. H. WHIPPLE These experiments are to be compared with those outlined in the preceding paper. These sple- nectomy animals were kept under the same routine as the bile fistula TABLE 62 Bile fistula. Splenectomy. Hemoglobin injection Dog 16-27 BILE & REMARKS me Rae oa Bile pigments in milligrams 3 : 3H Wes F s| 2|- 38 Mixed diet a a = ee = Go elelggig| 2} 2) 2) 2/3/82] 2818 : we) pe 5.0 48 = SIZiZinQte |:a Pepe). be be) eee 1916 ba. April 24 |11*/20/26} | 57| 8.8 | 7.6) 8.8 25.2 trace |25.8} Hemoglobin 103. Ss per cent April 25 |18*|14/17/21| 52) 5.2*| 6.9/12.7| 9.4/29.0| 4.3] trace |26.5 April 26 |15 |25|25| | 65) 6.8 | 7.9] 7.3 22.0 trace |26.3 April 27 |26}|27|22/18| 67| 5.8t/12.8|14.0| 5.6|32.4| 7.7] trace |26.8 April 28 |23 {18/25 66) 7.9 | 8.9|10.2 27.0 0 26.5 April 29 ‘ 48 22.0 trace |26.0 May 1 |27 |17/16| | 60} 9.7 | 8.4] 6.8 24.9 trace |25.5| Hemoglobin iat | 87 per cent May 2 |14}|22|24|12) 58) 5.67/20.8]18.9] 8.2|47.9|23.2| ++ |26.5 May 3 {13 [14|19) | 46] 5.8 | 5.6) 7.3 18.7 trace |26.8 May 4 /15}/20/17|23] 6011 .57/14.9/13.3]17.0/45.2/20.5|+++|26.8 May 5 /14:/22/12 48| 7.2 |12.8) 7.3 ‘127.3 ++ /26.8 May 6 38 23.8 + /26.3 May 8 /16 |29/16} | 61] 6.4 | 8.6] 5.4 20.4 trace |26.3| Hemoglobin 86 per cent May 9 |12|20/15/21| 56] 5.9T|17.0| 7.4] 4.9/29.3] 4.6] ++ |26.0 May 10 /18 {21/19 58/14.5 12.3] 4.3 31.0 + |26.3 May 11 |19}|20|29/23) 72| 6.0t| 7.2|19.4|16.4|43.0118.3] ++ |26.3 May 12 |12 |29/30|} | 71] 8.9 |11.0| 6.8 26.7 + |25.8} Stools con- tain no stercobilin May 13 54 28.3 trace |26.0 * 4 cc. laked red blood cells given intravenously at end of second hour. T 5 ce. laked red blood ceils given intravenously at end of second hour. Mean daily output in table = 24.7 mgm. per six hours. Average of twenty control days is 23.6 mgm. per six hours. Bile fistula operation and splenectomy done December 8, 1915. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 279 dogs without splenectomy, and the anemia was produced in the same way at the same time as can be seen by comparison of the various | tables. — The three preceding tables (61, 62 and 63) show three bile fistula dogs with splenectomy, with normal blood hemoglobin and normal output TABLE 63 Bile fistula. Splenectomy. Hemoglobin injection Dog 16-41 BILE S REMARKS | Amount in cubic Bile pi tsin milli cj centimeters pigments in milligrams a B i+} DATE : FI 3 * q a g 5a Mixed diet lale m 3 3 a a ‘7 ate Ee B/S/S/Elg] 515) 2/2) 3/85] 21] & Pigiaieie| = | S/S) 2)e Lee s* be 1916 lbs. April 24 |28 |19/24) | 71|10.0 | 7.7|11.5 29.2 trace |36.0} Hemoglobin 111 per cent April 25 |19*|16|34/22) 72|11.0*|12.9/26.0)13.4/52.3/27.3) + |36.3 April 26 |12 {17/16} | 45) 3.8 | 4.6) 5.0 13.4 trace |36.8 April 27 |17*/23/40/32) 95) 3.1*| 8.9)18.8/12.2/39.9]14.9| + (37.3 April 28 (36 |40|33| |109) 9.6 |11.6)11.2 32.4 trace |36.8 May 1 {18 {16/19 53/14.4 |12.0)10.4 36.8 trace |35.5| Hemoglobin 92 per cent May 2 |24*|22/25/32| 79) 9.8*/15.2|23.6|23.8/62.6)17.3} + [36.5 May 3 (36 |29/20 85/22.8 |18.2/16.1 57.1 trace |37.5 May- 4 |32*/34/29/31| 94/13.7*/26.6|30.4/30.£|87 .8|/42.5) + (37.5) Stools con- tain no stercobilin May 5 (23 |19/29 71/12.4 |12.8/17.0 42.2 + |37.0 May 8 {19 |24/21| | 64/11.5 |14.0| 9.4 34.9 trace |35.8) Hemoglobin 103 per cent 27 4*/18. ; + (35.5 May 10 (27 |22/34 93} 9.1 |11.6]10.8 31.5 + (86.3 23/25] 66)11.8*| 9.4/18.8)19.0)47.2)12.4) + (86.5 May 12 [24 [22/3 ; + {85.8 May 13 46}. 35.2 trace |35.3 _ ~I ~J — w Or _ — co _ bo or Ww “J <<) * 5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 35 mgm. per six hours. Average of twenty control days is 32.2 mgm. per six hours. Bile fistula operation and splenectomy done December 2, 1915. 280 C. W. HOOPER AND G. H. WHIPPLE of bile pigments for the unit period of six hours. These experiments are to be compared with those in the preceding paper (tables 41, 42 and 43) in which splenectomy had not been performed. It will be noted that the total output of whole bile as well as of bile pigments is identical in the two groups of dogs. tions does not influence the output of bile pigments. Splenectomy under these condi- TABLE 64 Bile fistula. Splenectomy. Bleeding. Hemoglobin injection Dog 16-27 e BILE g REMARKS a 2m cg ni Bile pigments ir milligrams 8 6 DATE z r a 3 K : ‘ e ra Fy 2 = Mixed diet aiglabele lca lw 1 aa te ee E IE S\5/2la)a|/ 2] 5 |5/4)/9/85| ge | @ eitititie| 2) 24202 leet 6 ee 1916 lbs. May 15 {13 |22|17 5216.2 | 9.9 | 8.8 24.9 trace |25.5| Hemoglobin 90 per cent. Bled 85 ce. May 16 {17 |17|24 58] 5.7| 6.5 | 9.8 22.0 trace |25.8) R.B.C. 4,880,000. Hemoglobin 83 per 1 cent May 17 |12*|/16/23)17| 56|5.8*| 9.0 |12.4| 8.8/30.2) 6.2 126.0 : May 18 /|18 |20/18| | 56/3.2 | 4.6 | 5.2 13.0 trace |26.0| Stools contain no stercobilin May 19 |15*)17|18/17| 52|5.7 |11.4*|16.2)17.6/45.2/21.2} ++ |25.8 May 20 54 36.4 + |25.8 * 5 cc. laked red blood cells injected into jugular at end of the second hour. = 24.0 mgm. per six hours. Mean daily output in table Average for twenty control days is 23.6 mgm. Hemoglobin injections (5 cc. of laked red cells) are followed by the same fluctuations in bile pigment output whether a dog retains his spleen or has been deprived of its services. The same dog under identi- cal conditions may respond to the hemoglobin injection with a rise of 11 mgm. bile pigment above the mean, again with a rise of 42 mgm. Tables 64 and 65 show a continuation of the experiments with dogs 16-27 and 16-41 (tables 62 and 63). from the jugular and later tested again with hemoglobin injections. These experiments are to be compared with the control non-splenec- Each dog is bled a small amount HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 281 tomized dogs (tables 44 and 45). There are no particular differences noted in these tables (64 and 65) as a result of this initial bleeding. The dogs react to the hemoglobin injections as they did before the bleeding. Note particularly the figures in table 65 where it is seen that the first injection of 5 cc. hemoglobin caused a rise above the mean of only 3.4 mgm. bile pigment but two days later the same injection caused an increase of 38.4 mgm. bile pigment. TABLE 65 Bile fistula. Splenectomy. Bleeding. Hemoglobin injection Dog 16-41 BILE 8 REMARKS 7 : : Rn eee cubic Bile pigments in milligrams Fy 5 ° Fs | DATE ai 3 Zz gi $ Bas Mixed diet 2 2h oe Be Bang 5 ees Soe rt elgiglgig| 2/2\2/2)3|28| 28 | Dise| titi Lit |e jae) 4 1 e 1916 lbs. May 15 /14 [16/24] | 54) 9.8 | 8.2|15.8 33.8 trace |34.8| Hemoglobin 106 per cent. Bled 120 ce. May 16 {19 |25|26 70} 6.4 | 8.1/11.0 25.5 trace |34.5| R.B.C. 5,680,000. Hemoglobin 103 per cent May 17 (|33*|30/41/28| 99) 6.7*| 8.2/16.2| 9.4|33.8) 3.4) trace |34.8 May 18 {82 |32|43} |107/10.1 |10.3|/12.8 33.2 trace |35.0| Hemoglobin 100 per cent May 19 |14*/25)/38|31| 94/11..3*/22.6|26.6|19.6|68.8|38.4| trace |34.8| Stools contain no stercobilin May 20 64 29.4 trace |34.8 * 5 cc. laked red bloou ¢ells introduced into jugular vein at end of second hour. Mean daily output in table = 30.4 mgm. per six hours. Average of twenty control days is 32.3 mgm. per six hours. The two preceding tables, 66 and 67, show some very unusual fig- ures as compared with the control tables in the same dogs before the anemia, and compared also with the simple bile fistula dogs in the Table 66 shows a great rise in bile pigment preceding communication. output to double normal with persistent and increasing icterus. The hemoglobin falls after the bleeding has been discontinued, and remains low for weeks with alternating periods of icterus and high bile pig- C. W. HOOPER AND G. H. WHIPPLE 282 TABLE 66 Bile fistula, Splenectomy, anemia. Hemoglobin injection Dog 16-27 BILE : REMARKS Amount in cubic | : . . A ae sintnstete Bile pigments in milligrams a 8 DATE ; FI 2h g g¢ | S| §4 Mixed diet a P| es Bo glfelelel £ | ee | ere ee) we ee 4 |S/4)/4) 3 a a a a 3 £S 4a ee SiTZitie | 2 toe Pee eee 1916 lbs. May 22 |15 |22/20 57| 8.1 |15.2]10.3 33.6 ' trace |26.0} Hb. 90 per cent. Bled 85 cc. May 23 | 7 {16/15 88] 8.4 |21.4/14.8 44.6 + |25.8| Hb. 8&1 per cent. Bled 215 ce. May 24 |19 |17/17| | 53/12.0 |12.7/14.5 39.2 + |25.8| Hemoglobin 59 per cent May 25 |17*/12/13/25) 50/19. 2*/15.2/17.2|37.4/69.8|16.7| +--+ |26.0| Hemoglobin 58 per cent May 26 |17*|19/20/16} 55|20.6*/27.4/32.4/28.8|88.6|35.5| +--+ |26.8) Moderate icterus May 27 |22 |17|17 56|21.8 |18.4/20.6 60.8 ++ |27.3| Hemoglobin 36 per i cent May 29 /|17 {14/19 50|16.8 |16.7|19.8 53.3 ++ /26.0| Hemoglobin 28 per cent May 30 /15 |17|11 43)14.4 |19.2)14.4 48.0}. +++ |26.0| Definite icterus May 31 |11*/24/20/21} 65/10 .0*/29. 2/30. 6/23 8/83 .6|30.5) +++ |25.5| Hemoglobin 27 per cent June 1. |20*/21/19)17| 57|11.0*|17.6|23.0/21.2/61.8] 8.7| +++ |25.3) Distinct icterus June 2 |23 |15/15 §3/20.4 |17.0/17.1 54.5 ++ |24.8 June 3 é 50 54.0 ++ |25.0| Hemoglobin 28 per cent June 5 {11 |24]17} | 52/10.0 |24.8|16.8 51.6 ++ |24.0| Hemoglobin 33 per cent June 6 |17*|19/21/17| 57/15.2*|30.6/34.2/26.6/91.4/38.3|++-++/24.3] Urine contains no urobilin June 7 {13 |14/16 4318.8 |20.8/21.4 61.0 +++ 24.5 June 8 /|14*/16/16)17| 49|16.4*|25 .0/26.0/24.4/75.4/22.3) +++ |24.3 June 9 |20 /19]18 57|24.2 |23.0/21.0 68.2 +++-+/24.5) R.B.C. 1,760,000. Hb. 31 per cent June 10 24 68.8 +++-+/24.5) Marked icterus * 5 cc. laked red blood cells injected into the external jugular at end of second hour. Mean daily output in table = 53.1 mgm. per six hours. Average of twenty control days before anemia 23.6 mgm. TABLE 67 Bile fistula. Splenectomy. Anemia. Hemoglobin injection Dog 16-41 > ' BRE Gy : REMARKS mn Amount in cubic alge : ee Be eohtimeters _ Bile pigments in milligrams =| p iso} age | §| 3% £ £ P Tee Mixed diet . . oe oO 2 ai plo n a a © | 2o e =} s = Elele\Ela| | 8)/2)2)41/38| 221 & alyioiol Ss] ae | wi} oe}|o |] o]s8) 22 | B ry jojo} | & _ co) n B | p z 1916 Tis. May 22 (31 (20/32 & — al bo —_ _ (=) 14.7 38.9 trace |34.3) Hemoglobin 101 per cent. Bled 120 cc. May 23 /20 /|16/15| | 51/14.4 |10.0) 6.8 31.2 trace |34.5| Hemoglobin 94. per cent. Bled 280 cc. May 24 /82 (27/25) | 84| 7.2 | 6.0| 6.7 19.9 trace |35.0| Hemoglobin 71 per cent. Bled 100 cc. May 25 |18*|16)22/23) 61) 7.2*|10.7|19.6/10.4/40.7|11.0} trace |34.8) Hemoglobin 60 per cent May 26 . |27*|28/31|27| 86] 7.9*|1 8.6 .0|28.0/17.6|62.6/32.9] trace |35.3 May 27 (85 |24/20 79 6 6.4 22.6 trace |35.3} Stools contain no stercobilin May 29 |25 13/15) | 53/15.2 |10.5|16.2 41.9 trace |34.8] Hemoglobin 72 per 7 cent. Bled 300 cc. May 30 /|12 |13/12) | 37| 6.2 | 7.3) 8.2 21.7 trace |34.5| Hemoglobin 60 per cent. Bled 160 ce. May 31 |27*|27|34/30| 91] 6.7*| 8.5/21.4/12.8/42.7/13.0| + 34.3} Hemoglobin 48 per cent June 1 |49*/35/45/32/112)12.2*/11 .0|22.6|17.2/50.8/21.1) ++ |35.3 June 2 [32 |29/31 92) 8.7 | 7.2| 6.3 22.2 + |34.5| No icterus June: 3 88 33.6 + |34.8} Hemoglobin 57 per cent. Bled 250 ce. June 5 {23 (22/22) | 67) 10.9) 9.4/10.9 31,2 + |33.5| Hemoglobin 46 per cent June 6 |14*/19/29/23) 71| 9.8*/22.0|/37.6|22.8|82.4|52.7/+++|34.0 Hemoglobin 47 per cent. Urine con- tains no urobilin June 7 [22 |22/21 65]12.8 |14.0/14.4 41.2 ++ 34.5} R.B.C. 3,200,000. Hemoglobin 61 per cent June 8 |31*/30/27|28) 85/14.0*|21 .0|24.2/21.4/66.6/36.9| ++ [35.5 June 9 (35 |31/29 95]12.5 | 9.1] 7.2 28.8 + (|35.8] Stools contain no stercobilin June 10 : 78 24.2 trace |34.8| No icterus * 5 ec. laked red blood cells introduced into the external jugular at the end of the second hour. Mean daily output in table = 29.7 mgm. per six hours. Average of twenty control days before anemia 32.2 mgm. 283 284 Cc. W. HOOPER AND G. H. WHIPPLE ment excretion followed by periods of blood regeneration and low bile pigment output. The other dog (table 67) shows a less striking picture during ‘Gis period, but in later periods showed even more striking icterus and high pigment output which interrupted the regeneration of red cells toward TABLE 68 Bile fistula. Splenectomy. Anemia. Hemoglobin injection Dog 16-27 BILE . REMARKS Amount in cubic Bile ni +68 illi ae png aes pigments in milligrams aS DATE q Pr gi g| 8| 58 Mixed diet aléieletcl «let wos a: ae g Elalslai3| 2/2) 2/2) /88) ge | s Siiidgitiat. © (Spe) fie ae) Be 1916 lbs. ’ June 28 |17 |22/22) | 61| 5.0/6.9} 4.9 16.8 + |24.0) Red blood cells : 3,542,000. Hemoglobin 61 per cent June 29 /10*/16/32/26| 74/3.6*|5.4/15.8/10.5/31.7|25.0) + |24.5) Stools contain no stercobil- in June 30 {19 |27/23 69/4.8 |4.2) 2.6 11.6 trace |24.8 July 1 62 7.0 trace |25.3 July 3 |15 |28/22) | 65)/4.4 |5.6] 4.0 14.0 trace |24.8] Hemoglobin 73 j percent — July 5 {11 {19/17 47|1.7 |3.0] 1.9 6.6 trace |24.5 July 6 |11*/16)/19)20| 55|2.0*/2.9| 5.6] 3.2/11.7| 1.0] trace |24.8 July 7 |23 |21/23) | 67|2.6 |1.9} 3.1 7.6 0 |25.0 July 8 46 12.4 0 |25.3 July 10 |15 |22|20] | 57|4.1 |4.0) 2.8 10.9 0 |24.8) R.B.C. 3,880,000 Hemoglobin 78 per cent July 11 /11*|17|18/17 52|3.5*|/8.0/25.0/11.4/44.4|33.7 ++ |24.5 . July 12 |12 /30/22) | 64)3.7 |3.4| 2.5 9.6 trace |24.8] Stools contain no stercobil- ] in July 13 |15*|23/20/19| 62/1.7*/6.2|11.6|10.8|28.6/17.9 ++4+|24.5 July 14 |12 |25)24 6113.8 |4.5] 2.7 11.0 ++ |24.8 8 * 5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 10.7 mgm. per six hours. Average of twenty control days 23.6 mgm. per six hours. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 285 a normal count. This dog (16-41) required more prolonged bleeding to cause an anemia, but after some delay reacted very much like the The reaction to uniform injections of hemoglobin (5 ce.) is truly _ remarkable, and ranges from 11 to 53 mgm. above the average mean TABLE 69 Bile fistula. Splenectomy. Anemia Dog 16-41 BILE S REMARKS Amount in Bile pigments in a8 ebicienaténe milligrams "8 DATE qu g z# | Se Mixed diet thal | © ms eles elelala/2/2/2) 3 | 28 | 8 eietete [ajahe|s 1s 1916 } = June 19 |35/24/22) 81/31.4/25.8/28.0| 85.2} ++ |34.0) Hemoglobin 63 per cent. : No icterus June 20 29/25/26) 80/35.2/39.4/40.2/114.8) ++ (33.8 June 21 20/32/34) 86/27. 2/28.0\29.0|) 84.2) ++ |34.0|) Urine contains no urobilin June 22 |27/25/30| 82|30.6|27.0/32.4) 90.0) ++ (34.5 June 23 /24/30/31| 85/25.8/50.6/56.7|133.1| ++ |34.3| Feces contain stercobilin June 24 82 125.6} ++ [34.0 , June 26 |37/33|32|102/39 .6/37 .0/43..2|119.8| ++ |33.5| Hemoglobin 67 per cent June 27 |28|30|35| 93/37 .6|34.0)/34.6/106.2) ++ |33.0 June 28 (/30/33/32| 95/25.1/28.6|23.8) 77.5) ++ |33.5| Hemoglobin 69 per cent. R.B.C. 3,440,000 June 29 (29/28/32) 89/21 .0/31.8/33.2|) 86.0) ++ (33.3 June 30 (|26/28/28) 82/37 .6/40.2/35.2/113.0) ++ |33.5 July 1 72 124.0) ++ |34.0| No icterus July 3 |36/33/31|/100/40.8|46 .0/50 0/136 .8|-+-+-+/34.3} Hemoglobin 69 per cent July 5 |28|27\29| 84/56.4/54.6/52.8|163.8|++-+|33.8 | July 6 /24/36|22) 82/38.7/51.2/81.2/171.1|++-+|33.8] Hemoglobin 42 per cent. : R.B.C. 1,920,000 July 7 (|37/27/29) 93/61 .8/72.8)77 .6|/212.2|++-+|34.0 July 8 78 |198 .4|-+-+-++/33.8] Moderate icterus July 10 |24/25|22) 71/64.8/78.8/69.2/212.8|+-+-+|33.5| Hemoglobin 43 per cent. 5 R.B.C. 1,720,000 July 11 |25/23|24| 72/58 .4|77 4/78 .8/214.6|+++/33.0 Urine contains no urobilin July 12 /22/32|28 82/59 .6|86 .4!75 .6|221.6|+-+-+|33.0} Distinct icterus July 13 |29/32|26| 87|68.7|75.9/61 .5/206.1/4++-+ 33.5 July 14 |34/27|25) 86/78 .3|73 2/70 .5|222 .0|+++|33.5 Stools contain stercobilin Average 85 146.3 33.7 286 C. W. HOOPER AND G. H. WHIPPLE level. There is not even any parallelism with the amount of icterus, although the high figures in table 67 are noted in conjunction with icterus. These two tables, 68 and 69, continue the observations upon same two dogs. The younger dog (16-27) has recovered from the spontane- ous icterus (table 68), and shows a low bile pigment output, a rising hemoglobin curve and the same wide fluctuations in bile pigment excre- tion following a unit injection of hemoglobin. This same dog at later periods again developed spontaneous icterus with high bie pigment excretion and a drop in hemoglobin. Tab'e 69 (dog 16-41) shows a truly remarkable picture. Bile pig- ment is constantly present in considerable amounts in the urine and definite icteroid coloration of the skin and mucous membranes devel- ops. The output of bile is normal, but the bile pigments are enor- mously increased—even more than six times normal. Note at this time the low hemoglobin curve and the color index which remains so constantly close to unity. It seems that this dog was putting out about the maximum amount of bile pigment, and perhaps he was manufacturing the maximum amount of pigment substance, which may account for the high color index of the blood. CLINICAL AND AUTOPSY ABSTRACT Bile fistula and splenectomy Dog 16-10. Yellow, mongrel, female, 34.5 pounds. September 30, 1915. Ether anesthesia; bile fistula and splenectomy. Condi- tion remained unchanged after operation. Weight varied between 29 and 31 pounds. y June, 26, 1916. See Table 61 for details. July 15, 1916. Dog in good condition except for mange. Ether anesthesia and killed. Autopsy at once. Thorax negative. Liver is of normal size. There is anin- crease in pigment in the parenchyma. Bile passages are pale and clear. The duct is completely obliterated just above the duodenum. Mesenteric and por- tal lymph glands are slightly enlarged and a bit pigmented a yellowish color. Bone marrow shows slight hyperplasia. Kidney shows a little pigmentation of the cortex as seen in human cases of pernicious anemia. Intestine shows a normal mucosa and considerable pigmentation of muscle coats which are of a maple sugar color. Other organs are negative. Microscopical sections. Liver shows definite increase in the portal stroma which in places contains nests of mononuclears. The bile ducts are normal. The liver cells close to the hepatic veins contain bile pigment in small and large colloid like grains and lumps. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 287 Bile fistula and splenectomy Dog 16-27. Young mongrel terrier, female, 23 pounds. December 8, 1915. Ether anesthesia. Bile fistula operation and splenectomy. Condition excellent after operation. Weight varied between 23 and 26.5 pounds. April 24, 1916. See table 62 for details. July 14, 1916. End of experimental period; 24.8 pounds; condition constantly good until September, when there was some bleeding from the bile fistula. September 30, 1916. Ether anesthesia and killed. Autopsy at once. Thorax negative. Liver is deep greenish and oe pig- mented as may be seen with long standing obstruction; portal tissue not conspicu- ous, except in region of fistula, which is thickened and warty looking. Common duct is cut across and isolated from the duodenum. Intestines and pancreas show a definite pigmentation as described above. Lymph glands about pancreas are slightly enlarged and pigmented. Other organs are negative. ; Microscopical sections. Liver shows a considerable increase in bile pigment deposited in the liver cells close to the hepatic veins. The portal stroma is very slightly increased and contains a few mononuclear cells. Bone marrow of femur shows a distinct hyperplasia which is explained by the period of bleeding during — the two weeks preceding death. About one-half of the marrow is made up of fat cells. The other organs show nothing of importance. Bile fistula and splenectomy Dog. 16-41. Large normal bull dog, male, 38.5 pounds. December 2, 1915. Ether anesthesia, bile fistula and splenectomy. Post- operative condition excellent. April 24, 1916. See table 63 for details; 36 pounds. July 14, 1916. End of this experiment, 33.5 pounds. March 7, 1917. Dog in fair condition, 30 pounds. Dog has shown repeated periods of spontaneous icterus and blood regenera- tion as tabulated above. DISCUSSION At present we can not offer a satisfactory explanation for these pe- riods of icterus and high bile pigment elimination which are observed in bile fistula dogs with splenectomy. It is not certain how much of the bile pigment in dog 16-41 has been built up into hemoglobin and then broken down into bile pigment, but we can say that if this did happen, then the construction of red corpuscles must have been going on at a terrific pace. This dog over a period of two weeks put out on an aver- age of 200 mgm. of bile pigment per six hours, over six times normal. We may safely calculate on a minimum output of 600 mgm. per twenty- four hours, which represents 50 cc. of red corpuscles. This dog weighed 33 pounds, and we may estimate his blood volume as 1200 ce. and dur- 288 Cc. W. HOOPER AND G. H. WHIPPLE ing normal periods about 600 cc. of red cells. His count at this time is two-fifths normal, and we must assume that he possessed about 240 cc. red cells. To account for this amount of bile pigment as coming only from the red cells, we must assume the destruction of his total red cells every four or five days. This is scarcely conceivable when we consider the slow regeneration of red cells which occurs after simple anemia in control dogs. This calculation does not take into considera~ tion the pigment escaping into the urine. Moreover, we wish to point again to the color index of unity wach indicates probably a saturation of the corpuscles with hemoglobin. This fits in perfectly with the tentative suggestion that such animals -are manufacturing pigment substance at a maximum capacity. Some of the pigment substance is formed into hemoglobin until the red cells can hold no more and some of the pigment substance escapes in the urine. Much of it escapes in the bile. We do not believe for an in- stant that all this pigment substance has been built up into hemoglobin and then degraded to other pigment substances. We do not believe the body is capable of doing this to such an extent as is shown in the last table, 69. There is evidently some extremely powerful stimulus present under these conditions of associated anemia, splenectomy and bile fistula which drives the body to maximum speed in its manufacture of pigments for the blood, bile and other tissues. We believe the liver is the mainspring in this mechanism. We hope to report in the near future on some of the interesting poms which are sugeenern by these experiments. . SUMMARY Splenectomy added to a simple bile fistula modifies in no way the se- cretion of bile pigments in the bile. A large number of experiments control this statement. Hemoglobin injected intravenously gives no constant increase in the output of bile pigments in the bile. Splenectomy does not modify this reaction. Variations from 4 to 42 mgm. are noted following a unit injection. . Dogs with bile fistulae, anemia and an added splenectomy may show some remarkable deviations from the control experiments . Experimental anemia in these bile fistula splenectomy dogs may give a very remarkable reaction—periods of spontaneous icterus, blood destruction and high pigment output may alternate with periods of regeneration and low bile pigment output without any demonstrable — oe SE oe Or 7, HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 289 cause. An interaction of the liver and spleen in the construction as _ well as in the destruction of the hemoglobin and red cell stroma may be indicated by these experiments. Regeneration of red cells with consequent recovery from the experi- - mental anemia is very greatly prolonged and may occupy months in the splenectomy experiments as compared with weeks in the — bile fistula experiments without splenectomy. The color index may remain uniformly high during this long period of blood regeneration in the splenectomy experiments. The output of bile pigments may average considerably above normal, and we may suspect an overproduction of blood and bile pigments with perhaps a deficiency of red corpuscle stroma. BIBLIOGRAPHY ~ (i) Musser: Arch. Int. Med., 1912, ix, 592. (2) Pearce, Austin any Musszn: Journ. Exper. Med., 1912, xvi, 758. (3) Musser anp Krumpsaar: Journ. Exper. Med.; 1913, Xviii, 487. (4) Martinorti anp Barsaci: Morgagni, 1890, xxxii, 521, 593, quoted by Austin and Pepper, Journ. Exper. Med., 1915, xxii, 675. (5) WuippLe anp Hooper: This Journal, 1917, xlii, 256. (6) Hooper anp Wutprte: This Journal, 1916, xl, 332. BILE PIGMENT METABOLISM 1X. Bite Piament Ovurrput INFLUENCED BY HEMOGLOBIN INJECTION IN THE COMBINED EcxK-BILE Fistuta Doe C. W. HOOPER anv G. H. WHIPPLE From The George Williams Hooper Foundation for Medical Research, University of California Medical School, San Francisco Received for publication March 16, 1917 The combination of a bile fistula with an Eck fistula has been studied in an earlier communication (1). The Eck fistula shunts the portal blood directly into the vena cava, and limits the blood supply of the liver to the arterial blood coming through the hepatic artery. The Eck fistula liver is smaller than normal, shows central fatty degen- eration and usually a subnormal liver function. The Eck fistula liver secretes less bile pigment than the normal liver—often 50 per cent of normal or less. There is strong evidence that bile pigment formation is dependent in part upon the functional activity of the liver and not solely upon the disintegration of red cells. With these points in mind, we must consider carefully the tabulated experiments given below. Is there any evidence that the Eck fistula dogs have difficulty in supplying the usual amount of blood and body pigment as one might expect from the low bile pigment output? The Keck fistula dog shows periods of anemia which we suspect may be ref- erable to an inadequate supply of pre-pigment material (manufactured in the liver?). This point can not be settled without much further work which we hope to supply in the near future. We have observed that removal of 150+ ce. of blood from an Eck fistula dog may depress markedly the blood hemoglobin level while the same procedure in a control dog may influence very slightly the blood hemoglobin level. This applies to observations made twenty-four hours after bleeding. This observation may be explained in part by: the experiments of Lamson (2) who demonstrated the storage of red cells held in reserve in the liver. This reserve, which is present in the normal liver, may not be found in the Eck fistula liver. 290 HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 291 Combined Eck and bile fistula dogs have less tendency to develop icterus with bile pigments in the urine than do the simple bile fistula dogs. We have no evidence that the Eck fistula liver can not excrete the pigment radicle of hemoglobin as promptly as the normal liver. If the hemoglobin were not removed with some promptness from the blood stream, we might expect more formation of bile pigments in the tissue outside of the liver and more bile pigment in the urine. This is contrary to fact, as there is much less tendency to icterus and pig- mentation of the urine in tle Eck fistula dog than in the control. The injection of hemoglobin intravenously in Eck fistula dogs gives much the same reaction as in simple bile fistula dogs. The unit injec- tion of 5 cc. of laked red cells should give an increase of 60 mgm. of bile pigment above the mean, provided the pigment radicle of hemo- globin is quantitatively eliminated in the form of bile pigment. Com- pare the control table 74 with tables 71, 72 and 73 of the Eck fistula. It is seen that the variations in bile pigment output after hemoglobin injection are perhaps more pronounced in the Eck fistula than in the control. If anything, the Eck fistula dog can take care of more hemo- globin than the normal dog with less bile pigment showing in the urine and less bile pigment increase from the fistula. But one cannot be sure of this point when the same dog will put out an excess of 25 mgm. of bile pigment above the mean following a unit injection of hemoglobin and a few days later put out zero milligrams of bile pigment above the mean following the same unit injection of hemoglobin (tables 72 and 73). EXPERIMENTAL OBSERVATIONS The dogs used in these experiments were also used in experiments previously reported (1). Reference should be made to this paper for more. complete details of control periods. The operative procedures have been reviewed. The general methods and experimental pro- cedures have been described in detail in the first paper of this series (3). The clinical history of each of these three dogs has been, detailed in a recent publication (1), and need not be again outlined. The dogs were all in a uniformly good condition as regards weight, activity and diet. The tables give all experimental data. . The three preceding tables show several points of interest (tables 71, 72 and 73). The same dog with combined Eck and bile fistula shows a slow gain in hemoglobin from 38 per cent to 98 per cent dur- ing a period of three and one-half months. During the anemic period 292 C. W. HOOPER AND G. H. WHIPPLE (table 71) we see a marked reaction to a unit injection of hemoglobin. One injection of 5 cc. laked red blood cells given intravenously causes a rise in bile pigment secretion of 24 mgm. The curve of secretion is very sharp, and the apex falls usually during the second two-hour period after hemoglobin injection, the final period often showing a return to normal. The urine after such hemoglobin injections at the most shows a trace of bile pigment. The Eck fistula animals on the TABLE 71 Eck fistula. Bile fistula. Hemoglobin injection Dog 16-15 BILE & REMARKS i] Amount in cubic ai arcs Peep aer tn aa snnticnetanl Bile pigments in milligrams 2 B r DATE q b| : Bread, bone and z £ p. 55 : milk diet pigisisiel ei ele ete @ B\E/S\8\3|/2)2)2)2)3 | 83 Be 3 Tolle & [on teh ak ise | eer See 1916 | Des. March 27 |15 |12)18) | 45/2.3 | 4.4) 4.0 10.7 trace |20.3| Hemoglobin 45 per cent March 28 |21*/14/13] 9} 36/3.3*|10.4/13.4/10. 2/34 .0/24.3/trace |20.5 March 29 |21 |18|15) |. 5418.3 | 2.4) 3.0 8.7 trace |20.0 April 3 |25 |21/20} | 66/3.9 | 2.9) 2.2 9.0 trace |20.8| Hemoglobin 38 per cent : April 4 |14+/18/16) 6} 40/2.2+| 2.4) 3.2} 2.0] 7.6] 0 |trace |21.3 : April 6 |22 |22|18| | 62/4.9| 4.4] 4.0| |13.3| _|trace |20.3) _ April 7 |25 |16/18} | 59/2.3 | 3.1] 1.9 7.3 0 /20.8 *5 cc. laked red blood cells given intravenously at end of second hour. 1 5 ce. laked red blood cells given intramuscularly at end of second hour. Mean daily output in table, 9.8 mgm. per six hours. Average output for 30 control days is 11.8 per six hours. Eck fistula operation September 13, 1915. Bile fistula operation February 17, 1916. Excellent condition throughout observations. whole seem to store away the injected pigment more effectually than the simple bile fistulas. This statement is made with reservation in the face of wide fluctuations following hemoglobin injections in simple bile fistula dogs and bile fistula-Eck fistula dogs. The after anemia period (table 73) shows even less reaction to hemo- globin injections. Two injections are followed by no bile pigment increase and one injection causes a rise of 19 mgm. in bile pigment HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 293 secretion. We get no evidence in any of these experiments that the Eck fistula liver can not eliminate the pigment radicle as promptly as the normal liver following hemoblogin injections. If anything, the Eck fistula liver seems to react more promptly than the normal liver in pouring out any excess of bile pigments which may be formed as a result of hemoglobin injections. Eck fistula. Bile fistula. Hemoglobin injection TABLE 72 Dog 16-15 BILE Bile pigments in milligrams 7-8 hrs. Total 6 hrs. 1-2 hrs. 7-8 hrs. Total 6 hrs. Increase above mean URINE TOTAL BILE PIG- MENTS, SIX HOURS June June June June ~ 15 16} 13 9f 15 15 14 6 12 7 ll 12 xSS § S & ie 12.0134.9 2.1)19.3 1.7 |0.7) 2.0 4.4 5.4 2.6 {3.0} 1.1 6.7 4.07\7.9| 8.2) 4.4/20.5 8.6 |2.0) 0.6 11.2 6.87|9.8)16.4) 7.0)33.2 5.0 |1.9) 1.0 7.9 27.0 11.4 12.6 25.3 ooo trace 0 trace trace REMARKS WEIGHT ° 21.0 Bread, bone and milk diet Hemoglobin 74 per cent Hemoglobin 78 per cent. R. B. C. 5,352,000 Hemoglobin 72 per cent *7.5 cc. laked red blood cells given intravenously at end of second hour. 7 5 ce. laked red blood cells given intravenously at end of second hour. Mean daily output in table — 7.9 mgm. per six hours. Average output for thirty control days is 11.8 mgm. Another fact deserves a word. It is to be noted in tables 72 and 73 that the color index of this dog is constantly low. Other periods show a very striking deviation from normal, and there is some evidence that the red cells contain a subnormal amount of hemoglobin. A similar condition may be found in simple bile fistulae with anemia, but it is 294 C. W. HOOPER AND G. H. WHIPPLE not as marked nor does it persist with a high red count as we have observed in the combined Eck-bile fistula. We may suspect that this lack of hemoglobin in the red cells may be due to inefficient pigment production in this dog. This point calls for much more work, and we believe more study should be directed toward pigment construction TABLE 73 Eck fistula. Bile fistula. Hemoglobin injection Dog 16-15 BILE % REMARKS -*] . . g ag Pee Bile pigments in milligrams 5 : Fs a é Z FI ax Bread, bone and filet milk di rs es eee Oe Be = Pia it Zo a | E\E\\2)3| 2 |2)2/2| 3 |es] 22 | 8 TIZZFie|] TiLi Lit) e eee ae 1916 lbs. June 19 {17 {15/22} | 54) 4.2 |3.4| 7.4 15.0 0 |20.0| Hemoglobin 78 per cent June 20 |12*|12|11) 8} 31) 5.4*/7.9]14.4/9.2/31.5/19.2itrace |20.5 June 21 | 7 | 7| 6| | 20/12.8 |8.4| 4.5 25.7 trace |19.9 June 26 |17 |12/13} | 42| 2.3 |1.9] 0.9 5.1 0 |20.5| Hemoglobin 80 per cent June 27 |10*/11} 9} 8) 28) 3.6*/5.0} 1.6/1.2! 7.8) 0 0 |20.3 June 28 |13 | 9} 8} | 30] 2.9 |1.2/ 0.8 4.9 0 |20.5| Feces contain stercobilin July 3 [16 {13} 9| | 38] 3.6 |2.6] 1.2 7.4 0 20.5 July 5 /20 |15)14) | 49) 2.8 |1.7| 1.3 5.8 0 |20.3) Hemoglobin 98 per cent. R. B. C. 6,332,- 000 July 6 |14*/11| 9)10} 30) 4.1*/5.7| 5.4/4.9]16.0| 3.7/ 0 [20.5 July 7/91} 711 27| 7.3 |6.6| 2.7 16.6 0 |20.3 July 8 28 18.2 0 |20.5 * 5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table — 12.3 mgm. per six hours. Average output for thirty control days of this period, 11.8 mgm. rather than pigment destruction. We hope to submit experiments which may give information concerning pigment construction in the body. Dog 16-146 (table 74) acts as control to the three preceding Eck fistula tables. This control dog is of about the same weight, activity HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 295 and temperament as the Eck fistula dog (16-15). Both the control dog and the second Eck fistula (table 75) show much variation in bile secretion following the unit injection of hemoglobin. ——— pigment TABLE 74 Simple bile fistula control. Hemoglobin injection Dog 16-146 DATE BILE Amount in cubic centimeters Bile pigments in milligrams 7-8 hrs. | Total 6 hrs. 5-6 hrs. Total 6 hrs. above mean 7-8 hrs. Increase URINE TOTAL BILE PIG- MENTS, SIX HOURS WEIGHT REMARKS Mixed diet 13 14 15 16 17 19 . 20 21 13* 14 16)14/15 12 14 16 SSR BS & RBS 4. 20.8 14.5|53.3/33.5 1) 7.3)37.6)17.8 10.4 21.8 14. 4/46 0/26 .2 6.8 7.2 12.8 14.7 5.4 12.0)34.5 18.2 3.8* 6.4 ee trace ae trace trace trace trace trace 17.8 18.3 18.3 18.2 17.8 18.0 18.3 18.5 18.3 18.0 18.0 17.8 17.5 Hemoglobin 84 per cent Urine contains — no urobilin Hemoglobin 90 per cent Stools contain no stercobilin Hemoglobin 89 per cent * 5 cc. laked red blood cells given intravenously at the end of second hour. Mean daily output in table = 19.8 mgm. per six hours. Average of twenty control days is 18.1 mgm. Table 75 (dog 16-139) shows an unusually high mean bile pigment output, in fact, practically normal for a control dog of equivalent weight. It can be seen in an earlier communication (1), table 36, that this same dog in the previous month put out 14.2 mgm. per six 296 ‘C. W. HOOPER AND G. H. WHIPPLE hours. After the hemoglobin injections were begun, there is a con- stant high level of bile pigment elimination. We do not believe that one depends upon the other, but must admit that there may be some possible connection. Secondly, it is to be noted that this dog devel- TABLE 75 Eck fistula. Bile fistula. Hemoglobin injection Dog 16-139 BILE 8 REMARKS Amount in cubie sane a el ae le er alegraresh Bile pigments in milligrams a B rs vais > a | at aa]. | eo b Pas s g| 22°] i | © milk diet glelglele| ¢ | ¢|e] 212] sel af | & A /a\a/4) 3] 4 S/S) 5/8 188) ga g widlelele| o le pe le ee be ee 1916 lbs. May 31 | 8 |12/11| | 31) 5.8 | 6.4) 5.0 17.2 trace |29.0)Hemoglobin 86 per cent June 1 | 9*| 9/11/11} 31) 4.9*) 3.6/11.4/10.0/25.0) 0.6)trace (29.0 June 2 {10 {12)14 36) 9.2 | 8.2] 5.4 22.8 trace |29.3 June 5 /13 {13/11 37|10.5 | 9.4] 5.9 25.8 trace |28.8|Hemoglobin. 88 per cent June 6 | 7*/10/10/10} 30} 6.6*|10.4/16.2|15.0/41.6)17.2| ++ |29.3/Urine contains no urobilin June 7 | 6} 9j12 27| 7.2 | 8.0} 6.7 21.9 + |29.3 June 8 |14*/12)11) 9} 32/11.2*| 9.7/12.2)11.6/33.5] 9.11 ++ |29.5|Hemoglobin 3 82 per cent. R. B. C. 5,- : 416,000 June 9 | 7 |12)13 32) 9.2 |11.9) 8.8 29.9 trace |29.5 June 10 | 30 25.8} trace |29.8 June 19 /14 {13/12} | 39] 6.4 | 6.2] 6.2 18.8 trace |29.0|/Hemoglobin 91 per cent June 20 /17*/17/14/15) 46) 6.8*|13.0)17.0|16.2/46.2/21.8} ++ |29.0 June 21 {13 |12)14 39)12.8 |10.8) 9.4 33.0 ++ /28.8 June 22 | 7*/11/11/12} 34/16.0*/20.0)15.0)/21.2/56.2/31.8} ++. |29.0 *5 cc. laked red blood cells given intravenously at end of second hour. Mean daily output in table = 24.4 mgm. per six hours. Mean daily output in month previous, 14.2 mgm. per six hours. oped frank clinical distemper within two weeks after this experiment ended, and it is possible that the disease was latent in this animal during this period of experimentation. Distemper is an infection in dogs which can cause a variety of unusual complications. HEMOGLOBIN INJECTION AND BILE PIGMENT OUTPUT 297 SUMMARY We have submitted evidence that the Eck fistula liver secretes less __ bile pigment than the control liver. It is possible that bile pigment formation is dependent in part upon liver function rather than upon the disintegration of red cells. It is probable that the Eck fistula liver can eliminate hemoglobin from the blood stream as promptly as the control liver. The Eck-bile fistula dog has less tendency to icterus and staining of the body tissues with bile pigment. Even with a high red cell count the color index will often be low in the Eck fistula. This is evidence to show that the Eck fistula dog has a subnormal pigment building capacity. The Eck-bile fistula dog shows the same great fluctuation in bile pigment excretion following a unit injection of hemoglobin. It is possible that these dogs can store more pigment substance than the control dogs after unit injections of hemoglobin. BIBLIOGRAPHY (1) WuiprLe anv Hooper: This Journal, 1917, xlii, 544. (2) Lamson: Journ. Pharm. and Exper. Ther., 1915, vii, 169. (3) Hooper anp Wuippte: This Journal, 1916, xl, 332. THE EFFECTS OF ADRENIN ON THE DISTRIBUTION OF THE BLOOD Il. VotumzE CHANGES AND VENOUS DISCHARGE IN THE SPLEEN R. G. HOSKINS anp R. E. LEE GUNNING From the Laboratory of Physiology of the Northwestern University Medical School Received for publication March 17, 1917 ® . In the first paper of this series (1) attention was called to the desira- bility of further investigation of the effects of adrenin on the blood- flow in various organs with adequate attention to dosage and duration of administration of the drug. Apparently in previous studies along this line the spleen has received relatively little attention. Oliver and Schaefer (2) were the first to in- vestigate the effects of suprarenal extract in this organ. In the several cases studied the reaction to the extract was an “‘enormous”’ contrac- tion. Innone was any dilatation observed except for a short time pre- ceding the reaction proper. This was regarded as probably a passive effect. Bardier and Frenkel (3) recorded the results of a study on a single animal, which was, apparently, under the influence of curare. Their extract was made by macerating desiccated or fresh gland for twenty-four hours at 37°C. After three injections they noticed: (a) Dilatation for three minutes followed by constriction while systemic blood pressure rose from 110 to 220 mm.; (b) Contraction followed by dilatation while the systemic pressure varied between 100 and 120 mm.; (c) Dilatation followed by contraction while the arterial pressure var- ied from 80 to 180 mm. Judging from the initial arterial pressures in each case the animal was relapsing into shock. The vasomotor reac- tions in the first and third cases indicate that the dosage transcended physiologic limits. What part was played in the reactions in the spleen by the curare and by protein decomposition products in the ex- tracts was not determined. Falta and Priestley (4) observed that the spleens of animals exposed several hours after subcutaneous injections of large doses of adrenin appeared anemic. Vincent (5) without giv- ing any details states that adrenin administered intravenously causes a 298 EFFECTS OF ADRENIN IN SPLEEN 299 contraction of the spleen. All the foregoing observations indicate that the outstanding effect of adrenin in the spleen is a marked contraction but both primary and secondary dilatations have been recorded. We have extended the investigations as herein reported. Technique. In all cases dogs have served as experimental animals. In most instances ether or morphin-ether anesthesia was employed but a few were decerebrated. The experimental results in both cases were similar. In some instances the vagi were cut but this procedure made no apparent difference in the outcome. The adrenin (Parke, Davis “ Adrenalin’) was introduced into a femoral vein, sometimes in- stantaneously and sometimes slowly. Various dosages were used de- pending upon whether pressor or depressor effects were desired. Si- multaneous records were made of changes of splenic volume or venous outflow or both and of changes of fe- moral arterial pressure. Stl Various types of plethysmographs were tried, including a Roy’s oncome- ter. The most satisfactory type was one improvised in the laboratory; it is somewhat like that of Edmunds. The general plan of the apparatus is eaown 2g egure ne box was selected Fig. 1. Diagram showing con- of suitable size to contain the organ siiction of alate” ploth vane: under investigation. For this purpose graph. a celluloid soap box proved satisfac- tory. Such boxes are light, inexpensive and easily procurable at any shop where toilet accessories are sold. The volume changes were transmitted by means of a plate plethysmograph which served as a lid for the box. It was made as follows: A sheet of copper, gauge 24, was cut the same shape as the top of the box and about 2 cm. greater in diameter. A sheet of rubber tissue was then cut in turn 2 cm. greater in diameter than the copper plate. A hole was drilled through the middle of the plate to communicate with a brass tube 3 mm. in diameter soldered to it to form the outlet of the plethysmograph. The edges of the plate were then beyt up around the four sides to make a shallow box about 0.5 cm. deep. The margin was incised at intervals to permit this bending. Melted collophonium wax or marine glue was introduced as a narrow zone around the outer edge of the box. This served to affix the rubber membrane which was next introduced. The excess circumference of the membrane was taken up in a series of pleats 300 R. G. HOSKINS AND R. BE. LEE GUNNING which were held in place by bending inward and down the sides of the box. A marginal zone of glue or collophonium wax applied just before the copper was bent down served to make the joint hermetically tight. In case this result was not secured at the first attempt a bit more of the cement was introduced to stop the leak or else the rim at that point was simply held for a moment in hot water to melt the cement already applied and allow it to flow into the crevice. In using this type of plethysmograph the organ is isolated and placed on a wet cotton pad in the box. A slit is cut in the side of the box to accommodate the blood vessels of the organ. The plate plethysmo- graph is then applied as a lid to the top of the box and secured in place by two rubber bands. It is then attached by rubber tubing to a re- corder. A T-tube interpolated in the connecting tube permitted the inflation of the plethysmograph with air under a slight pressure suffi- cient to cause the membrane to adapt itself to the inclosed organ but not great enough to interfere in any way with the circulation. In a few instances the Cushny cardiograph was found to give satis- factory results in recording changes in the dimensions of the organ but its use involved a greater degree of exposure of the viscera than did the plethysmograph. Venous outflow was recorded by drops from an oiled caniiula tied in a relatively small vessel. In some cases the spleen was divided and vol- ume change and outflow determined simultaneously. The float recorder previously described (1) was employed in all cases. Results. In determining the effects of adrenin in the spleen 17 dogs were used. In the plethysmograph studies 65 injection and 18 infusion experiments were made while the venous outflow series included 34 injection and 20 infusion experiments. In nearly all cases the effects on organ volume were similar to those described by Oliver and Schaefer for pressor injections, namely, a brief, supposedly passive dilatation followed by marked contraction. The same results were also obtained with depressor injections and infusions as well as with pressor infusions. Figure 2 which was obtained with a slightly depressor infusion will suffice to illustrate all these cases. It was found to be possible to hold a spleen in a state of uniform contrac- tion by adrenin infusion for ten minutes. Longer periods were not tried. The effect in the spleen lasted from a half to five minutes after blood pressure had returned to normal. In no case with either large or small dosage was a secondary dilatation observed during an infusion period. EFFECTS OF ADRENIN IN SPLEEN 301 Occasionally, however, as shown in figure 3, a dilatation occurred after the administration of the adrenin was discontinued. This effect was not passive since it was noted when the arterial pressure was either - normal or depressed. With no dosage was a pure splenic dilatation observed. The threshold for the reaction in the spleen was highly variable but. generally it was lower than that for changes of arterial pressure. In the most sensitive preparation investigated splenic contraction first ap- peared with an injection of 0.5 cc. of a 1:2,000,000 solution. This means of course that the spleen is one of the most sensitive organs in the body and reacts before a sufficiently widespread effect occurs to influence the general blood pressure. Spleen vol Fig. 2. Spleen contraction under the influence of adrenin, depressor infusion. Dose 2.8 cc. 1: 200,000 in 65 seconds. Dog weight 15 kilos. Time, five seconds. In several instances spleens that were previously quiescent began to undergo rhythmic changes in volume after the injection of adrenin. In such cases a new injection would check the rhythmic contractions but they. would begin again as the effect of adrenin wore off. The effect of adrenin on venous outflow was just what would be ex- pected from a consideration of the volume changes. A typical graph is reproduced as figure 4. During the preliminary dilatation period the flow was augmented. The augmentation persisted during the first part of the contraction period as the blood already in the organ was being expelled. Then during the remainder of the contraction period the flow was depressed, reaching the normal rate at about the same time splenic volume was restored to normal. The depressed outflow during the latter part of the period was obviously due partially to retention of the blood in the expanding organ. 302 R. G. HOSKINS AND R. E. LEE GUNNING Spleen vol. Fig. 3. Shows secondary dilatation of the spleen after injection of adrenin, 1 cc. 1: 50,000. Time, five seconds. Ferm. BI. Fi: ar anialinaliate yey em. Pulse. a Hd UNL AR eit POO LUuaaihy art Non NTH AG yt Caray TTT ITT TTT ET 7 TITTTTTTTITT TT TT FT TTT ne TT WT ile TT at Row | eae eat ee cea ae al el ca Sa a el can eae Poa ear ee cea ae oat ey 4) 1 Fig. 4. Shows effect of injection of adrenin, 1: 100,000, on outflow from splenic vein. EFFECTS OF ADRENIN IN SPLEEN ~ 303 SUMMARY Adrenin in all effective dosages whether injected instantaneously or infused causes in the spleen a brief dilatation followed by a contraction. Occasionally the contraction is followed by a secondary dilatation after the administration of the adrenin is discontinued. Aside from the brief preliminary effect in no case was a pure (pri- mary) dilatation observed. The threshold for splenic contraction is lower than for changes in arterial pressure. Occasionally a quiescent spleen is stimulated by adrenin to rhythmic contractions. Adrenin causes a brief increase, then a decrease in the outflow from the splenic veins. BIBLIOGRAPHY (1) Hosxins, GUNNING AND Berry: This Journal, 1916, xli, 513. (2) OtIveR AND ScHAEFER: Journ. Physiol., 1895, xviii, 231. (3) Barbier ET FRENKEL: Journ. d. Physiol. et d. Pathol. Gén., 1899, i, 950. (4) Faura unp Priester: Berl. klin. Wochenschr., 1911, xlviii, 2102. (5) Vincent: Internal secretions and the ductless glands, London, 1912, 164. THE EFFECTS OF ADRENIN ON THE DISTRIBUTION OF THE BLOOD III. VotumE CHANGES AND VENOUS DISCHARGE IN THE KIDNEY R. G. HOSKINS ann R. E. LEE GUNNING From the Laboratory of Physiology of the Northwestern University Medical School Received for publication March 17, 1917 The effects of adrenin on vascular conditions in the kidney, as in several other organs, were first studied by Oliver and Schaefer (1). Without going into details as to any possible differential effects of large and small doses, of the effects of long continued administration of the extracts or of after effects, they reported that suprarenal extracts cause a marked diminution in kidney volume. Largely on the basis of this fact and the observation that the spleen is similarly affected the gener- alization was offered that such extracts cause a marked vasoconstric- tion throughout the splanchnic area. Four years later, in 1899, Bardier and Frenkel (2) investigated the relation of suprarenal extracts to diuresis but included also in their study the vasomotor effects. Their experiments were made on anes- thetized dogs, apparently under the influence of curare. Their extracts were made either from desiccated glands or from fresh glands macer- ated for twenty-four hours at body temperature. Judging from the effects on arterial pressure relatively large doses were employed. The extracts were administered intravenously. These authors described as typical effects a contraction of the kidneys and a depression of urine flow followed by a dilatation of the organ and a diuresis. In certain exceptional cases, however, the injections were followed at once by dilatation and polyuria. Whether the use of curare or the presence of - protein decomposition products in their extracts played any part in the results was not determined. In the same year Gottlieb (3) included the kidneys in a series of ex- periments on the effects of adrenal extracts on the heart and blood ves- sels. He worked on isolated kidneys of hogs and dogs. It was noted that when such extracts were added to the perfusate that was being 304 EFFECTS OF ADRENIN IN KIDNEY ' 305 passed through the organs a marked decrease in outflow resulted. This finding was corroborated by Gioffredi (4). Similar experiments were made by Sollmann in 1905 (5). But in addition to the venous outflow the rate of urine discharge was also studied by this investigator. When adrenin was added to the perfusate to make a dilution of 1: 50,000 _ both the venous and urine flow were markedly decreased and the kid- ney volume as determined by an oncometer was also lessened. The dogs from which the kidneys were obtained had received large doses of morphine. It may be noted that Sollmann used solutions from twenty to one hundred times as concentrated as the adrenin solution in the adrenal veins as determined under ordinary experimental conditions. In three perfusion experiments Pari (6) found that solutions of 1: 20,- 000 to 1: 100,000 caused renal vasoconstriction, but in one case a solu- tion of 1:500,000 caused dilatation for a few minutes followed by constriction. _Jonescu (7) recorded the effects of intravenous injections of adrenin on the blood pressure and kidney volume of rabbits. Doses which caused a moderate rise of blood pressure caused a slight dilatation of the kidneys which was followed by a marked contraction that persisted for some time after the arterial pressure had returned to normal. When smaller doses were used the kidneys showed contraction while the blood pressure remained practically unchanged. From this observation a theory was deduced that the kidney vessels have a special affinity for adrenin. It would be possible, therefore, for the adrenals by a slight continuous overactivity to set up a nephritis without any significant vascular hypertension. The theory is obviously untenable on the grounds cited. The work of Hartman (8) and that reported in the first of this series (9) accounts for the effects observed without invoking any special “affinity” of the renal vessels for adrenin. While the kidneys were. contracting blood-vessels in other parts of the body were expand- ing, hence the systemic pressure was little affected. In a short paper published in 1911 Froehlich (10) reported that both l- and d-suprarenin as well as ‘‘adrenalin” cause a protracted contrac- tion of the kidneys. A more extensive investigation along the same line was reported by Ogawa (11) a year later. He used both l- and d- adrenin and synthesized /- and d-suprarenin. Instead of the oncometer method, however, he utilized perfusions to determine the effect of the drugs on the kidney vessels. Rabbit kidneys in the most sensitive preparations reacted slightly to a dilution of J-adrenin of 1: 20,000,000 showing a diminished outflow. In case of a solution of 1: 1,000,000 306 R. G. HOSKINS AND R. E. LEE GUNNING the primary effect was a sharp decrease in the rate of outflow followed by a rate above normal when the adrenin was discontinued. The aug- mented outflow was noted also as a secondary effect if the adrenin per- fusion was continued for a relatively long period. These secondary di- latations appeared only if relatively strong solutions were used—that is, dilutions of from 1: 1,000,000 to 1: 5,000,000. In two instances pri- mary dilatation was noted with solutions of 1: 40,000,000 and 1: 50,- 000,000, respectively. The same results were obtained with d- as with l-adrenin except that stronger solutions had to be used. The synthetic product also gave qualitatively similar effects. In cats and dogs the same results were obtained but the threshold was higher. In all the foregoing reports renal vasoconstriction following the ad- ministration of adrenin is a prominent feature. In most cases, however, the doses used were probably greater than the quantity that can be discharged from the suprarenal glands in a corresponding length of time. The evidence, so far as it goes, indicates that urine secretion follows part passu the vasomotor effects produced in the kidneys by adrenin. This renders important a definite determination of the question whether the vasodilatation reported by Bardier and Frenkel and by Ogawa is a significant feature of the response to adrenin injections. If it is char- acteristic then adrenin diuresis such as has been described by Kleiner and Meltzer (12) may well be due, partially, at least, to local vasomotor effects in the kidneys. The suprarenal glands might then justly be brought into question as involved in “spontaneous” diuresis. The fact that vasodilatation was observed only as a secondary effect with larger doses when the adrenin would be largely destroyed or the mechanism fatigued, or as a primary effect only when very small concentrations were employed points toward this as a physiologic mechanism, since it is probable that it is only with very high dilutions that the body nor- mally has to deal. This might be correlated with the fact observed by Kleiner and Meltzer that in order to produce diuresis adrenin must be administered as to be slowly absorbed whereas in cases in which it reaches the kidneys promptly it acts as a renal depressant (Cow, 13). In our experiments, the report of which follows, we have accordingly been especially interested in any possible dilator effects that might appear. Technique. In general we have followed the same methods recorded in the two preceding papers of the series. The plate plethysmograph and float recorder therein described were utilized. Etherized dogs have been used exclusively in the work on the kidneys. In most EFFECTS OF ADRENIN IN KIDNEY 307 cases only plethysmograph studies were made but in a few instances venous outflow was also recorded. Results. In this series 16 dogs were used. In the investigation of volume changes 151 injection and 5 infusion experiments were made. In determining the effects on venous outflow 14 injections and 3 infu- sions were given. ii say vil 4 — wa Fig. 1. Kidney contracting under influence of adrenin, 4 ec. 1:.100,000. Blood pressure from femoral artery. Time, five seconds. Dog, weight 18 kilos. The results in the kidney were in general much the same as those in the spleen. The outstanding feature of the reaction was a sharp con- traction such as that shown in figure 1. The graph reproduced is un- usual, however, in one respect: neither the arterial pressure nor the organ volume show the preliminary augmentation that generally appears after adrenin injections. This augmentation Which is short 308 R. G. HOSKINS AND R. E.) LEE GUNNING lasting and inconsequential in degree is regarded as purely a passive effect due to the sudden stimulation of the heart before the outlying tissues are affected. It will be noted that the volume change outlasts for some time that of the arterial pressure. This lag is a characteristic part of the reaction picture. In various cases it was found to persist from a half to two minutes. It was usually longer towards the end of a series of experiments. In one case only was a different type of reaction observed. This is illustrated in figure 2. After the characteristic passive preliminary dilatation as the arterial pressure rose the organ contracted. Then as arterial pressure began to fall the kidney dilated only to return to its Kid. Volume ta “mn, Fem. Blood Py», aad * na ull uaa ne ie LE em. Pulse Time 5S Sec. ee $$ — Fig. 2. Kidney expanding under influence of small dose of adrenin, 0.5 ce. 1: 200,000. Pulse and pressure from femoral artery. Time, five seconds. Dog, weight 8 kilos. initial volume a minute after the normal blood pressure was restored. This result was observed when such doses as 0.5 cc. of 1: 100,000 were administered. When this dose was doubled the ordinary contraction of the kidney appeared and outlasted the change of blood pressure. In no case did a pure dilatation occur such as was reported by Ogawa in two of his rabbit experiments when very small doses were used. Neither in any of our experiments was a secondary dilatation observed either during infusions or after injections or infusions such as was re- garded by Bardier and Frenkel as a characteristic feature of the reac- tion. It should be noted, however, that only five infusion experiments were made. The results shown in figure 2 suggest that in a larger series EFFECTS OF ADRENIN IN KIDNEY ’ 309 animals might occasionally ‘be found which would show a dilatation of the kidneys during infusions with adrenin in high dilution. In view of the fact that infusions gave qualitatively the same results as injection in all cases observed, this possibility was not extensively investigated: It was found that the kidneys could be held for at least ten minutes in a uniform state of contraction. Longer periods were not tried. The threshold for changes in kidney volume and for changes of blood pres- sure were about the same. The reactions in the kidney were quali- tatively similar irrespective of whether pressor or depressor dosages were employed. ! . Our observations consistently support the reports of previous inves- tigators that adrenin decreases the venous outflow from the kidney. Our experiments as a whole do not support the theory that adrenin diuresis is due to a dilator effect of small quantities of adrenin in the kidney. On the other hand they do not definitely exclude the possi- bility that such may be the case in a normal unanesthetized animal. The situation as regards adrenin diuresis may well be not unlike that as regards pituitrin diuresis. From the fact that pituitrin in anesthetized animals often gives a short lasting polyuria a theory has been deduced’ and widely held that this substance is a diuretic agent in the normal organism, whereas, as a matter of fact it is an efficient anti-diuretic (14). In view of the evidence (a) that adrenin in doses which cause renocon- striction depresses urine formation; (b) that adrenin administered sub- cutaneously, and consequently, absorbed slowly, causes polyuria; and (c) that in anesthetized animals renodilatation has occasionally been reported as a result of administering adrenin in high dilutions or as a secondary reaction with larger quantities, the theory is not improbable that in normal animals adrenin in relatively small quantities causes a dilatation in the kidneys. Possibly the matter could be definitely deter- mined by attaching metal guide strips to the poles of a kidney and then after recovery from the operation had occurred studying with a fluoro- scope the renal volume reactions to adrenin when no anesthetic was used. Marshall and Davis (15) have reported that ablation of the suprare- nal glands results in depression of the functions of the kidneys even while systemic blood pressure remains normal. If, as is generally as- sumed, the adrenals keep tlie blood supplied with minute quantities of adrenin, the depression noted might in harmony with the above men- tioned theory be ascribed to a deficiency of circulating adrenin, leaving the renal constrictor factors in the ascendancy. The supposition is, 310 . R. G. HOSKINS AND R. E. LEE GUNNING however, intrinsically improbable. That renal functioning should be dependent upon minute quantities of such a dilator substance, that a hormone reaction should have.been evolved to function in this purely negative way to correct a gratuitous overactivity of some other factor would seem to be a useless complication. SUMMARY AND CONCLUSIONS Adrenin in both depressor and pressor doses ordinarily causes con- traction of the kidneys of dogs and a corresponding decrease in the venous outflow. Instantaneous injections and slower infusions of adrenin give quali- tatively similar results. One animal gave renodilatation with smaller and renoconstrictions with larger doses. The threshold for renal changes and blood pressure changes is about the same. — The observations as a whole do not support but also do not definitely disprove the theory that in normal animals adrenin diuresis is due to ~ renal dilatation. BIBLIOGRAPHY (1) OLIveR AnD ScHAEFER: Journ. Physiol., 1895, xviii, 231. (2) BarpDIER ET FRENKEL: Journ. d. Physiol. et d. Pathol. Gén., 1899, i, 950. (3) GorriieB: Arch. d. exp. Pathol. u. Pharm., 1899, xliii, 286. (4) Giorrrepi: Atti. d. R. accad. med-chir. d. Neapoli, 1904, lviii, 169. (5) Soutmann: This Journal, 1905, xiii, 246. (6) Part: Arch. Ital. d. Biol., 1906, xlv, 209. (7) Jonescu: Wien. klin. Wochenschr., 1908, xxi, 513. (8) Hartman: This Journal, 1915, xxxviii, 438. (9) Hoskins, GuNNING AND Berry: Ibid., 1916, xli, 513. (10) Frorxicu: Zentralbl. f. Physiol., 1911, xxv, 1. (11) Ocawa: Arch. f. exper. Pathol. u. Pharm., 1912, lxvii, 89. (12) Kier1nzr anp Metrzer: Journ. Exper. Med., 1913, xviii, 190. * (13) Cow: Journ. Physiol., 1914, xlviii, 443. (14) Motzretpt: Journ. Exper. Med., 1917, xxv, 153. (15) Marsuauy anp Davis: Journ. Pharm. and Exper. Therap., 1916, viii, 525. FURTHER OBSERVATIONS ON THE DIFFERENTIAL ACTION OF ADRENALIN FRANK A. HARTMAN anp LOIS McPHEDRAN From the Physiological Laboratory, University of Toronto . Received for publication March 17, 1917 In the course of a series of experiments performed by one of us and reported in this Journal (1), it was found that the fall in general blood pressure, which: is caused by the intravenous injection of small doses of adrenalin, is not brought about by dilatation in the vessels of all parts of the body alike. In an animal in which the arteries to the abdomi- nal organs have been clamped, injections of a standard small dose of adrenalin caused much the same fall as had previously occurred in the intact circulation. On the other hand, when the arteries to the limbs were occluded, those to the splanchnic area being intact, the reaction to the standard dose was changed from a pure fall to a rise of blood pressure, as registered from the carotid artery. In that research the only distinction drawn was the broad one as between the reactions of the ‘‘peripheral’’ circulation on the one hand, which included the ves- _sels of bone, muscle, and skin, and that of the “splanchnic” circulation on the other, which, as well as comprising the vessels of the abdominal and thoracic viscera, necessarily included those of the muscles of the thorax and back. The present research was undertaken with a view to following out the subject of the differential action of adrenalin some- what more in detail, and in the hope of arriving at some conclusion as to the mechanism involved in the vascular adjustment caused by it. Oncometric experiments were carried out on intestine, spleen, and kid- ney. While our research was in progress the appearance of the paper by Hoskins, Gunning and Berry (2) made further investigation of the reactions in skin and muscle unnecessary. In every case simultaneous records were taken of the reactions of at least two organs in response to adrenalin, since we considered it of im- portance to determine whether the same range of dose which caused constriction in any one abdominal organ also caused constriction in all the others, and whether the amount giving rise to dilatation in one was 311 312 FRANK A. HARTMAN AND LOIS McPHEDRAN necessarily the same as that which caused another to dilate. Since the reaction of any organ to a given dose may vary not only among different individuals, but also in the same individual during the period of an experiment, it is necessary to record the changes co in the same animal; at the same time. ) The animals used were dogs and cats. In two of the expeniaal on the latter we injected urethane subcutaneously; in all the others the anesthetic was ether. Blood pressure was registered from the right carotid artery. Injections of adrenalin were made with a graduated syringe, the needle of which was thrust through the wall of a piece of rubber tubing fitted to a cannula, which was inserted low in the left jugular vein. The adrenalin solution used was that manufactured by Parke, Davis and Company, 1: 1,000, diluted with distilled water to the required strength immediately before use. In each experiment the first injections were made of .a solution 1: 100,000; when large doses..were ‘required, as was often the case in working with dogs, in, preference to injecting large quantities of distilled water in- to the animal’s circulation, we substituted for the more dilute’ solution one of a strength of Qe ae 1: 10,000. The duration of each. Fig. 1. Gutta percha oncometer for spleen injection was signaled on the : record below the time marker. No special precautions were taken as to absolute uniformity in the rate of injection, but it was kept fairly constant, and was in all cases slow,. as shown by the records. ne In some experiments we left the vagi intact; in the majority they were: cut. We were unable to observe, however, any specific effect of these on the reaction of any organ to adrenalin except the familiar one of cardiac inhibition caused by large doses, with the consequent great: rise in blood pressure. ~The oncometers which we used for kidney and for intestine were gutta percha ones of the ordinary type, fitted with glass lids. The edrly experiments on the spleen were done with the same oncometers; later we had a series of special ones made. These were modelled after the shape of the spleen (see fig. 1) and were provided with two lips for’ stalks, separated by about 0.6 cm. in the smaller and 1.5 cm. in the larger. As recorders in the first few experiments we used Marey’s drums; DIFFERENTIAL ACTION OF ADRENALIN 313 later we substituted for these bellows recorders, which have the advan- tage of recording volume changes without introducing alterations of pressure within the system itself. In several experiments the recorders were calibrated by injections of known volumes of air. The pressure inside the system differed little from atmospheric; in practice we raised the pressure until the bellows were about half filled, and were thus adjusted to give maximum variations in either direction. INTESTINE A loop of the small intestine, about one-third of its total length, was selected, generally that imme- diately above the caecum, since the blood vessels there are long and form a convenient stalk. Two pairs of double ligatures were tied about its lower end, about 2 inches apart, the blood vessels supplying the piece between the ligatures tied off, and the piece of intestine removed. A similar operation was performed at the upper end of the required length, and the loop was ready to be placed in the oncom- eA ce. 1:100,000 eter without the necessity of fur- ther dissection, other than simply slitting the mesentery for an inch 5 peak ; Sa Fig. 2. Constriction of the intestine or two on either side of the stalk. following injection of a small dose of Before putting the loop into the adrenalin (0.4 ce., 1: 100,000). Dog 14 oncometer we washed out its con- (Reduced 3) tents with warm saline. This - prevents the slow formation of gas which otherwise takes place within the lumen, and which interferes with the records of volume changes due to the circulation alone. The whole operation from the time the abdomen was opened until the intestine was put into the oncometer was not longer than fifteen or twenty minutes. During this time the intestinal loop and the other abdominal contents were kept covered with warm saline pads. In the great majority of cases, the effect of doses of adrenalin, both large and small, was to cause constriction of the intestine. In all of 314 FRANK A. HARTMAN AND LOIS MCPHEDRAN these experiments small doses caused only constriction (figs. 2 and 6); as the quantity of adrenalin was increased, however, a prolonged and marked dilatation supervened on the preliminary constriction (see figs. 3 and 7). There were two exceptions, in which the least effective dose caused dilatation. These occurred during the early experiments, before we had fully realized the importance of com- pletely removing gas from the lumen, and we consider it probable that this increase in volume of the loop was caused by the relaxation of the muscles of its walls, under the influence of the adrenalin. The threshold for the constrictor effect was shown, in the six experi- ments in which it was determined, to vary within fairly wide limits, from 0.014 to 0.07 cc. adrenalin 1: 100,000 per kilogram body weight, that is, it was reached by doses such as also caused a slight fall in blood pressure. The general resemblance of these two curves, indeed, make it at first glance appear possible that the one effect may be dependent on the other, and that the constriction in the intestine may be nothing more than a decrease of blood supply to its vessels, caused by the lower- ing of the general blood pressure. Latent periods, duration, and de- gree of decrease of intestinal volume, also bear some relationship to the same changes in the general blood pressure. Closer inspection of the tables (1 and 2), however, shows clearly that this is not the case. Though the diminution in intestinal volume generally occurs several seconds after the beginning of the fall of blood pressure, this is not always the case; for instance in experiments 13, 14, and 20, the records show the intestinal decrease to precede that of the blood pressure by several seconds, and our notes, made during the course of the experiments, corroborate this as actually occurring, and not being due to a possible faulty alignment. The time of the least intestinal volume does not correspond to that of the lowest blood pressure, nor is the duration of the constriction the same as that of blood pressure fall. (See experi- ments 7 and 13, where it is greater, and experiments 3, 18 and 20, where it is materially less.) Above all, the constriction does not take place only when the dose is such as to cause a fall of blood pressure; constriction of the intestine occurs time and time again when the blood pressure is above and not below its normal level (see figs. 3 and 7). The dose of adrenalin necessary to cause a dilatation of the intestine to follow on this constriction is as variable as is the threshold dose for the constriction itself. It varies from 0.04 to 0.31 cc. of a solution 1: 100,000 per kilogram in dogs, and in cats.it is about 0.4 cc. The latent period of the dilatation is longer than that of the constriction, DIFFERENTIAL ACTION OF ADRENALIN 315 AAD y AAR A! reg % mi nei lesegeaente WYMAN AHH likcsbine Fig. 3. Preliminary constriction followed by prolonged dilatation of the in- testine, caused in the same animal as that of figure 2, by a larger dose (0.2 , 1: 10,000) (Reduced 3) Intestine Se Fig. 4. Reaction of same loop of intestine as that of figures 2 and 3, to a dose of adrenalin of much the same magnitude (0.3 cc., 1 : 10,000) as that of figure 3, after the coeliac and superior mesenteric ganglia had been removed. (Reduced 4) 316 FRANK A. HARTMAN AND LOIS MCPHEDRAN and the effect is as if the one were superimposed upon the other. As the doses are increased from the threshold dose on, the resulting con- striction becomes more and more marked and its duration longer. Once the dose is great enough, however, to cause dilatation, this cuts short the first effect, with the result that the latter is reduced by from one-fourth to two-thirds of its former length. The volume change brought about by dilatation is much larger than that caused by con- striction. For instance, in a dog of 25 kilograms the constriction re- duced the volume of the system of the intestinal oncometer by 1 ce., while the dilatation increased it by more than 5 ce. The observations on the resemblance of the curve of constriction to that of the general blood pressure may be applied in much the same way to this case also, for in increasing the doses of adrenalin sufficiently to cause a dilator effect in the intestine, in a few cases we crossed the threshold for pressor effect on the blood pressure. The same argu- ments, however, which prevented our accepting the explanation of a passive effect in constriction, are also valid in this case. The latent periods of intestinal effect are longer than those of blood pressure fall, the time of maximum dilatation never coincides with that of maximum rise of pressure, and its duration is far greater, in many instances two to three times as long (experiments 21, 22, 23, table 2). As before, too, the occurrence of the intestinal effect does not depend on the nature of the blood pressure change; we have several records which, like experiments 3 and 7, show a marked increase in intestinal volume during a fall in general blood pressure. As a possible explanation for the occurrence of increase in the intes- tinal volume, after injections of doses of adrenalin above a certain size, it might be suggested that such doses are just sufficient to affect the intrinsic nervous system of the intestine and to bring about relaxation of its walls, thus permitting expansion of the blood vessels within them. To investigate this, we inserted a rubber balloon into a part of the in- testine immediately adjacent to that which furnished the loop in the oncometer, injected a small quantity of air, and connected it to a small bellows recorder, which made a tracing below the oncometer record. By this it was found that injections so small as to cause only a slight fall in blood pressure were sufficient to bring about a relaxation of the intestinal wall, and that as the dose was increased no well-marked dif- ference could be observed in the reaction of the intestinal wall to that dose which first gave dilatation in the oncometer, nor to any of the succeeding ones. . DIFFERENTIAL ACTION OF ADRENALIN 317 In an attempt to decide upon the origin of this dilator effect of ad- renalin we severed connection between the loop of intestine involved and the central nervous system, by dissecting and removing the two coeliac ganglia and the superior mesenteric ganglion, or by cutting the splanchnic nerves. In all experiments, five in all, after this operation was performed, the dilatation by adrenalin was entirely done away with, and doses which previously had caused a preliminary constriction followed by a dilatation now gave nothing but a simple constriction of the loop (see fig. 4). UC LIE entciriry Gu LabbbOG Fig. 5. Constriction in spleen, followed by a series of waves, after injection of a small dose of adrenalin (0.1 cc., 1: 10,000). Dog 19 (Reduced 3) THE SPLEEN In the dissection of the spleen the gastrosplenic ligament with its numerous fat vessels was ligatured off, bit by bit, and cleared away from the neighborhood of the splenic blood vessels. In the early ex- periments those of the latter which supply the upper half of the spleen were also tied off. After two or three of these dissections, however, we were so dissatisfied with the appearance of the spleen under these conditions that we adopted a splenic oncometer with two lips. This enabled us to leave all vessels supplying the splenic substance intact, except sometimes that to the extreme tip of the upper end, which bound the organ too closely to allow of its being put into the oncom- eter. With the exception of this small piece, the spleen remained in 318 FRANK A. HARTMAN AND LOIS MCPHEDRAN excellent condition, even during the course of an experiment lasting over several hours. During dissection we protected it with saline pads, and we warmed the oncometer to receive it. Of the dogs, ten in all, which we investigated, seven showed only constriction in the spleen in response to the whole range of doses of adre- Intestine RINAIAAARKARG ARK Wi B.P DO CELE REAL TEAL LCG NAN Hany \ fit Hat dy! whan Aa Fig. 6. Dilatation in spleen, and constriction in intestine, after small dose of adrenalin (0.3 ce., 1: 10,000): Dog 21 (Reduced 3) nalin employed. This constriction was more marked and more pro- longed as the doses were increased in magnitude (see figs. 5 and 7). The three others each gave dilatation at some dose of adrenalin. Two of the three showed as a first effect dilatation with small doses; or, to speak more accurately, small doses of adrenalin set up in these DIFFERENTIAL ACTION OF ADRENALIN 319 spleens (which had previously been relatively inactive) a series of waves, of which the first was in the direction of dilatation (see fig. 6). In both these organs the effect of increasing the dose was to increase the constriction in the waves at the expense of the dilatation. until large doses caused only decrease in volume. In the third of these spleens doses of adrenalin also set up series of waves, but its reaction differed Intestine BP 1 f rh Lh ppaapes A it iF 1N ' LANNE ey Spleen Fig. 7. Effect on same animal as that of figure 6, of a larger dose (1 cc., 1: 10,- 000); slight dilatation followed by marked constriction in spleen, preliminary " . . . . . . x . i 1\ constriction and marked dilatation in intestine (Reduced }) from that of the other two in that, on administration of relatively large doses (0.2 cc., 1 : 100,000 per kilogram) the constriction was followed by dilatation. KIDNEY The upper part of the ureter and the kidney vessels of one side throughout their entire length were dissected out to formastalk. The mesentery was removed as gently as possible from the surface of the 320 FRANK A.. HARTMAN AND LOIS MCPHEDRAN kidney. A few of the larger veins running from it into the capsule were ligatured. During the dissection the kidney was protected as completely as possible with warm pads. Four experiments in all were performed, two on cats and two on dogs. In every case injections of adrenalin caused constriction in the kidney (see figs. 8 and 9); with small doses of low concentration this Kidney Intestine RS ae tis Se ane ere 5 Sec. Fig. 8. Constriction of kidney after a dose of adrenalin 0.2 ec., 1: 100,000 (Re- duced 3) was the only effect (see fig. 8); in two cases the preliminary constric- tion caused by large doses (e.g., 0.32 ec. 1: 100,000 per kilogram), such as occasioned a preliminary rise followed by a fall of blood pressure, was followed by a dilatation of the organ. The curve of this dilatation was similar in form to that familiar to us in the reac- tion of the intestine to large doses of adrenalin. It showed a rise of "(# peonpey) (000'0T: T "99 @°9) eSOp 19FAV] 104J8 ‘UOTJOTIYsUOD poZuojoid pus uoTyeze[Ip Arvurutoid ‘g ‘Sy ul se Aoupry oMBs Jo MOTJOBVOY “6 ‘B1q a ADRENALIN OF Z o = a oO < DIFFERENTIAL ~ ee haupry 322° FRANK \A. HARTMAN AND’ LOIS McPHEDRAN the lever, which only gradually returned to the base line, not before 170 to 180 seconds had elapsed, and thus continued long after the blood pressure had regained its normal level. The cause of the occur- rence of this dilatation we were not able to determine. : In conclusion we wish to point out that the reactions of the various organs, though they may be of a similar nature, do not necessarily take place at the same time, nor for the same dose. Thus, for in- stance, in dog 28, table 2, though both kidney and intestine give con- striction in response to small doses and dilatation in response to large ones, nevertheless that dose of adrenalin (0.4 cc., 1:100,000 per kilogram) which is enough to cause transition from a dilatation to a constriction in the intestine, still gives rise to nothing but a constriction in the kidnev. Numerous other examples may be found by reference to that table, That the output of adrenalin, which has been shown by the work of the last few years (3), (4), (5), to be so small during normal quiet life as:to have no appreciable effect on blood pressure, is augmented during conditions of mental excitement, as well as by the asphyxia attendant on violent exercise, and by sensory stimulation, has been shown in a series of experiments by Cannon and the workers in his laboratory (6). Theexactextent of thisincrease in secretion has not been determined, nor is it known whether it is sufficient to effect a rise rather than a fall in blood pressure. Elliott (7), in working on thesecretion from the adrenal glands whichis brought about by stimulation of the splanchnic nerves supplying them, has shown that in this case the quantity se- creted is within the range of doses which have a depressor effect on the general blood pressure; whether this is also the case during the reflex stimulation of normal life, we are still ignorant. In any case, as shown by our experiments, the first effect of the outpouring of adre- nalin during excitement must be to cause a constriction of the intes- tine and kidney, and generally, though not always, a similar constric- tion in the spleen. By this means there is brought about a shifting of ‘the blood from these organs to the muscles, which, as Hoskins, Gunning and Berry (2) have shown, are at the same time actively dilated. If, as may prove to be the case, the output of adrenalin increases till the concentration in the arterial blood is of the order of about one-half or more that necessary to bring about a rise in blood pressure, a dilata- tion of the intestine, and perhaps also of the kidney, must take place, through the agency of some central mechanism, the location of which, and the source of stimulation to which, are as yet unknown. 323 DIFFERENTIAL ACTION OF ADRENALIN 24} UI 0OO‘OOT :T UIpeUeApE “99 ¢ *‘sorqey Y}0q 04 sarjdde siz, ‘quoull1edxe 94} UI QOO‘OT :[ Ul[vUeIpe ‘90 ¢'Q AT][vOL SBA 9B[qQBY ‘UOT}NTIP OOO‘OOT :1 JO SUITE} UI MAAS o1v AdYy So[qv} 94} UT o[qvaeduIOD VsOp 94} ByvUI 04 qnq ‘WOTPeIZUIDTOD YOO‘OT :T B UT [VUTU OYy OFUT poqoofur AT[eNZOw o10M YOO‘OOT :T Ul[eUoIpe ‘do Z UBYy 10}¥013 SesOp ITV TOT IST 0% TOT vA YA Nema: (peg ede oy | a al Poyley 6 00°% 0g 98 99 ZZT 94 9OT 8 6% poteyl Il 19°0 “17 Bod Ig OFT OF OT ‘¥I Iv poyteyy oe Z0°0 CT. OZ Z8T 0} BE: l ZI poyreut A19A | 6 80'% 18 ZZT 94 OST Il 1Z poyrvyy | OT 12°0 "0% oq 09 OST 94 O9T TI €€ POYTBAT | | ZT ¢c0"0 ¢’6 COT ' OST 99 96 CT poyteypy «9G T 8% 18 94 66 8 poyreur AI9A | = OT r1'0 “ET Bod €8 06 93 ZOT os PoayTBAy II €0°0 y) +r | TST 09 Gez 07 OLT 8z peyvyy | Zr £9 SIZ 09 POT SI Teas bles ye @) 16 CPI 9F PSI 94 SLT 16 [[Bag CIZ OLT 99 SST [PA OIL prt’ 9T'0 \ ine. € 7 4%9 09T OTT 99 OST TIP O9T pote] 90°0. CT's f - * spuooas : spuoras spuovas spuoras spuogas *suby | IHDIGM AGOG cea auos xAUQOUAW JO i mou, eery a wih aes “waa Nowa't od Olu -Id rf co'nowtena ['"aanseaus coow | S401 |ovormasnos | Rowoeuaxoo | 40 saaatwovm | axaava | sxatmmnieay | Smo" | Twmxy 10 seis iekiesde -NI NITVNGY atalip: | a ~ay 40 ENAOWKY . Uovnzn) Up fo 99U9L4N990 =? wouorssuod pounsoqus fo ac ay} fo buzuariousy Lél eo ; { ; } uf | | ; SOT} ; } \ OUS -nlib ay “09 boalsard TTAVL f O13 bantts aoitas Oik Beir TCT eae eee em | uorzey | | 1) Set GST 94 ZOT O9T “SIIp Peyreyl | Olz Pexe ZI PexB | 00°% w01j901138 -u00 peyieur : Aq pemoy > 12 Bog -]9} u0y ; ¢8 82T 07 PST “ByETIP [TRU 0gT 1s auoN | 20°0 uo 48 ; 400} 02 SPT 97 O9T o¢ “GIP pexey 7 ON.) ‘T0°0 ST |) uOT}OTI4SU09 qua } £9 €1Z 97 9OT| G8z | pexTwur AOA “FISI9T | POV | ZL UOTPOLIGSUOD gaa ; » ST 8D SIT (89 9926 9908 | O1Z =| poyreur Aro, oLt Teug | ¢e°0 WOTPOII4S 49 ~—-|60T 97 OFT 94. ZET| 08 “W100 | Pore £8 Teug ; 200 6S |) WO1}911}SU09 poyreur ) GILT OST 0F SLT OF TOT Seg poyreur AIaA) — OZT AI9 AQ ST Teas | 29°0 UWOTPOIIYS Ty foes 48 [GPT 97: L8T 94 PLT 0&% “G09 (Pomel | «= OT peyrte suON | 22°0 que UWO0T}OII38 SP LZyT °F CST] ~ISI9g | -uo0o TTeug 08 TTeug suo N €0°0 ST J spuosas spuosas spuovas spuovas spuosas “suby . IHDIGM a aon AMoJeUr Jo 819}0Ur ytd esueyo esueyo esueyo u01787e[Ip |u01;8z"[IP Jo} uOTo1IysU0O pepe Xa0a dO “nC “Yyyar uresueyD | 2a a joopnyluzey_Y = jo uoMeing joopnyusey = jo uoNeing| epnzuse_ [jo uoMeing epnyruseyy Petite fact TE 0000131 corsa] vnc mor7eze[I WOTzOIIZSUOD -ILN@O o1a.09 @uNSeaud GOOTs NO Load AGNGIX NO LOgada N@G1ds NO Loadaa a GNIISALNI NO Loads 40 INDORY. qonpiuipur aups ay} ur supbio puasaffip uo uypuaspo fo szoaff[a ay fo uosiupdwoo y o ATaVL 821 F2Z 07 SZT 0} FET TST 94 SPT IZ 09 SFT au 00T WOTyEIE Aq. pomoy]o} uorqordys -u0d poxleyy worqOTI4s “009 poyIeyy u01}01148 -u0d poyxeyy eo que4sisied £q pomoy]oy uoljo1ys (“uo0d |[BUIg : wOorTyey 8p yeuis qua. ‘-4s1s10g SIT ole pA pemeHo}- poyxreyy [sag poyavur ai &T 81 auoN WITS or -13d90 | -10d Ajorvg [eug g'1Z > 8% 30 326 FRANK A. HARTMAN AND LOIS McPHEDRAN In investigating the dilatation occurring in various organs in the body under the influence of different concentrations of adrenalin in the blood - circulating through them, we hoped to gain some light on the vexed question of the existence of dilator fibers in the sympathetic nerves to the blood vessels. The existence of these has been denied by many authorities, notably by Brodie and Dixon (8), and by Cannon and Lyman (9), On the other hand, evidence deduced from experiments of widely differing character has been brought forward in support of the theory that dilator fibers are present in the blood vessels, and are sensitive to adrenalin in solutions too. dilute to stimulate the endings of the constrictor fibres. Dale’s experiments on the reversal of the effect of adrenalin by ergotoxine (10), are interpreted to this effect. by him. The pioneer work of Brodie and Dixon (8) on the lung already — cited, and the later results of Desbouis and Langlois (11) and especially those of Enid Tribe (12) seem to point in this direction, the dilatation observed by Park (13), Elliott (17), and Cow (15), in coronary vessels, and that in vessels of other organs by Pari (16) and by Ogawa (17), } in response to adrenalin offer further proof of the possibility of the ex- istence of vaso-dilator fibres in the sympathetic system. On this subject our éxperiments have the value only of a negative finding, but as far as they carry us we have found no evidence of a direct stim- ulation of vaso-dilator endings by adrenalin, in any concentration | approaching that which occurs under physiological conditions. SUMMARY 1. Small doses of adrenalin cause constriction of the vessels of in- testine, of kidney, and generally also of spleen. . 2. The minimal dose necessary to produce this constriction is in much the same order of magnitude as that required to cause a fall in blood pressure, but it is not necessarily identical with it, nor is it the same for every organ in the same animal. 3. Increase of the dose of adrenalin causes in all cases marked dila- tation of the intestine. This dilatation: is brought about by doses materially less than those which are necessary to cause a rise in general blood pressure. 4. This dilatation in the intestine is under control of the central nervous system, and is done away with by severing connection with the central nervous system. 5. Adrenalin in the majority of cases has in all doses a constrictor © es eS ae 7 DIFFERENTIAL ACTION OF ADRENALIN 327 effect on the spleen; in some, minute doses cause first dilatation, which is the initial change of a series of rhythmical splenic waves. i ———. BIBLIOGRAPHY — (1) Hartman: This Journal, 1915, xxxviii, 438. (2) Hosxrns, GuNNING AND Berry: Ibid, 1916, xli, 513. (3) O’Connor: Arch. f. Exper. Path. and Pharm., 1912, Ixvii, 195. . (4) Srewart: Arch. Exper. Med., 1912, xv, 547. .(6), Hosxins anp McCuvre: Arch. Int. Med., 1912, x, 343. (6) CANNON AND DE LA Paz: This Journal; 1911, xxviii, 64. CANNON AND Hoskins: Ibid., 1911, xxix, 274. Cannon, SHOHL AND Wricut: Ibid., 1911, xxix, 280. (7) Exuiotrr: Journ. Physiol., 1912, xliv, 376. (8) Bropiz anv Drxon: Ibid., 1904, xxx, 476. ~(9) CANNON AND LyMAN: This Journal, 1913, xxxi, 376. (10) Dae: Journ. Physiol., 1906, xxxiv, 169. (11) Dessouts anp LanGtots: C. R. Soe. Biol., 1912, Ixxii, 674. 12) ‘TRIBE: Journ. Physiol., 1914, xlviii, 154. ) ‘Parx: Journ. Exper. Med., 1912, xli, 532 and 538. (14) “Exxrorr: Journ. Physiol., 1905, xxxii, 443. (15) Cow: Ibid., 1911, xlii, 125. (16) Part: Arch. Ital. de Biol., 1906, xlvi, 209. (17) Ogawa: Arch. f. Exper. Path. and Pharm., 1912, Ixvii, 89. AN HYDROLYTIC STUDY OF CHITIN! SERGIUS MORGULIS Witn THE COLLABORATION OF EB. W. FULLER From the United States Fisheries Biological Station, Woods Hole, and the Depart- ment of Physiology of the College of Medicine, Creighton University Received for publication March 17, 1917 Since its discovery in 1823, chitin has aroused a great deal of interest. This substance has a very wide distribution throughout the inver- tebrate kingdom, particularly in the skeletons of arthropods, but the mode of its formation and its nature are still shrouded in mystery. For the solution of this problem the first and essential thing is a clear understanding of the composition of chitin. Although much atten- tion has been given to. this matter by previous investigators neither the constitution of chitin nor indeed its empirical formula may be said to be beyond dispute, and this uncertainty is due to the fact that the subject has not been approached from the proper point of view. In general it may be said of chitin that it is characterized by extreme resistance to strong reagents. Thus it is soluble in fairly concentrated mineral acids only, and is not affected by boiling with strong alkali, in which respect it resembles glycogen. In most analytical studies of chitin, the substance was dissolved in sulphuric acid. The nature of the chitin has been argued from the products thus obtained. No at- tempt has ever been made—so far as I am aware—to appraise the de- composition products by a strictly quantitative method, except per- haps for the sugar which is yielded as a glucosamine. The material for this investigation was prepared from thirty-three lobsters with the total weight of about 11 kilograms. The animals, carefully weighed, were arranged in three groups; the average weight of the lobsters in the first group was 344.1 grams, in the second group 270.2 grams and in the third 337.8 grams. The lobsters were decal- cified in 10 per cent nitric acid, which for this purpose was found more effective than the hydrochloric. The decalcification was continued, the acid being frequentiy changed, until there was no further reaction 1 Published with the permission of the Commissioner of Fisheries. 328 HYDROLYTIC STUDY OF CHITIN 329 for calcium. The material was then washed in running water. After this preliminary treament the shell can be very easily separated from _ the soft parts..The large bulk of the soft tissues was therefore re- moved mechanically. The decalcified shells were then boiled with 20 per cent potassium hydroxide which completely digested all fatty stuff as well as every trace of meat. The hydroxide solution was renewed - several times and the boiling maintained until the KOH remained colorless. The material, which at this stage is practically white, Was again washed in running water. To insure complete renewal of traces of pigment it was left over night in dilute potassium permanga- nate, washed again in running water and bleached with sodium bi- - sulphite. The chitin thus obtained is of a clear, snow white color. Tt is now washed with distilled water for several days, until tests for sulphate or chlorides are negative. The chitin is then dehydrated by treating it with several changes of alcohol, followed by several changes of dry ether. When removed from the ether, the substance dries very quickly. The material thus prepared is dazzlingly white, retains every structural detail of the original shell, and is tough as leather. The material is chemically pure chitin, leaving no ash residue on incineration. The purity of the several samples was further tested by determining its nitrogen content, as this has been found to be very constant. The final desiccation of the chitin was accomplished at a temperature below 100°C. until its weight remained unchanged. The total yield of chitin was 358.63 grams. In the first group I found 11.54 grams chitin per lobster; in the second 9.588 grams; and in the third group 11.68 grams. AVERAGE WEIGHT OF LOBSTER AVERAGE WEIGHT OF CHITIN PER CENT OF CHITIN 270.2 9.588 3.08 344.2 11.54 3.35 337.8 11.68 3.45 It is evident, therefore, that the larger animals contain not merely an absolutely greater quantity of chitin, but also a relatively larger amount. Dry chitin contains 6.39 per cent nitrogen. The elementary analy- sis? yields 6.7 per cent hydrogen and 44.94 per cent carbon. The en- tire composition of chitin (lobster) may therefore be presented thus: 21 avail myself of this opportunity to express my sincerest gratitude to Dr. P. A. Levene for the courtesy of making this analysis for me. 330 SERGIUS MORGULIS C H H O 44.94 6.70 6.39 41.97 The empirical formula which corresponds to this elementary compo- sition is CsHi;NO,s. This formula may represent a monacetylglucosa- _ mine of the following constitution: . - CH,OH—(CHOH);—CHNHCH; CO—CHO From the fact that upon decomposition chitin yields acetic acid, which may even be detected by smell, and glucosamine, the theory has been developed that chitin represents nothing else than a polymerized acetylated glucosamine. The isolation of an acetylglucosamine from a sulphuric acid solution of chitin by Frankel and Kelly*® would thus seem to furnish experimental proof of the theory. We shall see later that the substance isolated by these authors probably has nothing to do with chitin, and the correspondence between the empirical formula and that for acetylated glucosamine is entirely fortuitous. The iron chain of logical arguments upon which the theory is based has always seemed to me to possess a weak link; namely, that the solution of chit- in in concentrated sulphuric acid may so modify the original mole- cule as to completely obscure its relation to the chitin. | Following Krawkow’s* method, I attempted to prepare chemically pure chitin by dissolving the material obtained by the above described procedure in concentrated sulphuric acid, with every precaution to prevent heating of the mixture, and precipitating it with a great excess of water. Although I made numerous trials to get chitin by precipi- tation—primarily because I wished to have the material in the form of a powder—I succeeded only once in getting a fine white precipitate from an hydrochloric acid solution of chitin. I will subsequently refer to this precipitated material and its probable relation to chitin. I wish merely ‘to mention here that even in concentrated sulphuric chitin goes into solution somewhat slowly, and by the time the substance has dissolved the solution invariably gave a reduction with Fehling’s, showing that the chitin was already decomposed. In searching for a method by which the chitin could be dissolved without immediately causing it to break down, it was found that much more dilute acid (60 to 80 per cent) would dissolve it as quickly as concentrated acid with the * Frankel and Kelly: Sitzungsb. der klin. Akad. d. Wissensch. in Wien, 1901, 110, Abt. 2. * Krawkow: Zeitschr. f. Biol., 1892, xxix, 177. HYDROLYTIC STUDY OF CHITIN 331 added advantage that decomposition (as evinced by development of a brownish color) is delayed much longer. It was found still further that by raising the temperature to about 60°C with even 35 to 40 per cent acid large quantities of chitin may be easily dissolved, the solu: tion remaining colorless for a considerable length of time. The study of the solubility of chitin in various strengths of HsSO, showed that 20 per cent of the acid is still effective and I availed my- self of this weak reagent to determine the following points: (1) Will chitin dissolve in it completely if sufficient time be allowed: (2) Will this treatment cause the splitting off of the acetyl groups; (3) Will it hydrolyze the theoretical amount of glucose? With these points in mind a series of experiments was performed and the technique em- ployed will be described here fully inasmuch as it has been followed throughout this work. _ A weighed quantity of the dry chitin was put in a large Erlenmeyer flask with 200 cc. of the acid mixture. The flask was closed with a rub- ber stopper having four holes. Through one hole a tube passed almost to the bottom of the flask, while a short tube passed through another hole and the tubes were bent on the outside. The former was connected with the compressed air, the air bubbling through sulphuric acid and potassium hydroxide before passing through the chitin solution. The second short tube was connected to another tube reaching down to the bottom of a receiving flask containing standard sodium hydroxide, so that any volatile acid produced in the acid-chitin’ mixture was carried off by the air current and absorbed. Through the other two holes in ° the stopper were inserted a thermometer and a separatory funnel from which water could be admitted from time to time to maintain a con- stant volume in the flask. The flask was placed on an electric plate and the heat so regulated that the temperature never varied more than 50° to 60°C. In the experiment which I shall describe here 1.8828 gram of chitin was used. On the basis of the structural formula quoted above this quantity of chitin should yield 0.5112 gram. CH;COOH, 1.5336 grams. CsH 0, and 0.1193 gram nitrogen. This amount of CH;COOH _ is equivalent to 85.2 cc. 4 acetic acid. The hydrolysis was continued uninterrupted for nearly five days. At the end of that period most of the material was still undissolved, and the mixture was very turbid owing to the suspension of fine particles which have been detached mechan- ically through the agitation. Of the total quantity of standard alkali ‘used 88.5 cc. 3% were neutralized by volatile acid, driven off from 332 SERGIUS MORGULIS the hydrolysis mixture, as was determined by titrating with 7 HCL, using phenolphthalein as indicator. This represents 0.531 gram. of CH;COOH. In view of the fact that not all of the chitin had yet been dissolved, the recovery of 104 per cent of the total theoretical amount of acid made the computation on the basis of the structural formula appear erroneous. The mixture was then filtered, the clear filtrate made up to volume and analyzed for sugar and nitrogen. This contained 0.2210 gram glucose, or 17.6 per cent. The residue was further hydrolyzed with 40 per cent H2SO,. It dis- solved completely within a very short time. The hydrolysis was car- ried on for thirty-six hours, during which time an amount of acid was produced equal to 50.1 ce 4, or 0.3006 gram-CH;COOH. The total amount of acid thus formed was (reckoned as acetic) 0.5310 + 0.3006 = 0.8316 gram, or practically 63 per cent more than could be expected according to the formula. In the second hydrolysis there was produced also 1.1650 grams glucose and 0.989 gram nitrogen. The total amount of glucose yielded was, therefore, 1.3860 grams (97 per cent), and that of nitrogen 0.1240 gram (103 per cent). This and similar experiments gave weight to my skepticism towards the hypothesis that chitin is an acetylated glucosamine. It also strength- ened my belief that the acetic acid detected when chitin is treated with violent reagents is not a primary hydrolytic product but rather a by-product resulting from the decomposition of the glucose mole- cule itself. In view of these results it was desirable to follow the dif- ferent phases in the hydrolysis to get a more exact understanding, ex- pressed in quantitative terms, of the entire process. After many trials and errors it was found that a concentration of about 35 to 40 per cent sulphuric acid (in 30 per cent the chitin is not completely dissolved) at a temperature of 50° to 60°C. are the most ideal conditions under which it is possible to carry on hydrolysis for days without causing the mix- ture to become much colored, and the mixture usually lacks the cara- mel smell. In 50 per cent sulphuric acid it is no longer safe to con- tinue the hydrolysis very long as the mixture not merely turns a dark reddish brown, but an insoluble charred residue is formed and the whole mass smells strongly of caramel. ° In all such hydrolysis the amount of sugar recovered was invariably smaller than the theoretical, the glucose evidently having undergone decomposition. The production of vola- tile acid under the circumstances was usually very large. It is there- HYDROLYTIC STUDY OF CHITIN 333 fore clear that in hydrolyzing the chitin a great deal of care had to be exercised to guard against secondary decomposition. In this respect the results obtained with 40 per cent sulphuric were as nearly free from objection as is possible under the circumstances, and the account will be confined to these results exclusively. The hydrolysis was performed by the method already described, the volatile acid being driven off by a slow but continuous current of air and absorbed in standard alkali. The hydrolytic mixture was made up to a definite volume of which aliquot portions were analysed for glucose by the Berthrand permanganate titration method. It was noted already in the early experiments that if the acid mixture was made alkaline, ammonia would distil off. When this was collected in standard acid it was discovered that the amount of ammonia nitrogen thus recovered was less than the theoretical amount of nitrogen in the chitin hydrolyzed. If the mixture was first digested with an excess of acid, made alkaline and distilled as is usual in Kjeldahl determinations, the ammonia nitrogen distilled off was equal to the theoretical amount. In other words, some of the nitrogenous groups were easily split off by the weak acid forming ammonium sulphate, while other groups were present in the chitin in a more stable combination and could only be set free by the complete oxidation of the molecule. It was a natural conclusion that the easily detachable nitrogenous group was, the NH: group of the glucosamine and this was still further corroborated by the observation to be recorded later that the formation of glucose and ammonium sulphate keep pace with each other during hydrolysis. This fact suggested immediately that the empirical formula fails to convey an adequate idea of the complexity of the chitin molecule, which is certainly more intricately constituted than the adherents of the acetylated glucosamine hypothesis suspect. In view of the significance of the results, the records of the experi- ments are fully reproduced in what follows: I. August 18. 2.0339 grams dry chitin. 200 cc. of 40 per cent H,SO,. Temper- - ature, 50° to 60°C. 9.20a.m. Hydrolysis started. 72.24 cc. 7y NaOH in receiving flask. 9.50a.m. Chitin entirely dissolved. Solution colorless. 2.00 p.m. Straw yellow color. August 19. -9.20a.m. Hydrolysis discontinued. Solution clear, dark yellow. Faint acid smell. Alkali in receiving flask titrated with 16.13 ce. N zo HCL. 334 72.24 — 16.13 = 56.11 ce. 7 used up. SERGIUS MORGULIS Duration of hydrolysis, twenty-four hours. Solution made up to 500 cc. Two 25 ec. samples carefully neutralized. Eliiedae! determined: (1) 22.60 (2) 22.70 22.65 ec. 7y KMn 0, used. 22.65 X 0.00663 = 0.1502 gram Cu 0.0826 X 20 = 1.652 grams glucose. 0.0826 era glucose. Two 50 cc. samples made aklaline; distilled into standard acid (11. 28ee. to). (1) 6.42 (2) 6.33 6.38 cc. gy NaOH required. 11.25 3.19 8.09 cc. 7 HCL used up. 8.09 X 10 X 0.0014 = 0.1133 gram N. II. August 19. 2.0063 grams dry chitin. perature, 50° to 60°C. 9.40 a.m. Hydrolysis started. After twenty-four hours. 4,23 cc. 55.97 ec. ty NaOH neutralized. August 20. 9.40 a.m. Color light yellow. 60.20 cc, August 21. 9.40 a.m. Color dark yellow. After twenty-four hours. August 22. 9.40 a.m. Color same. After twenty-four hours. August 28. 9.40 a.m. Hydrolysis discontinued. crim 4OOTI ry “£205 200 cc. of 40 per cent HSO.. em- 60.20 ce. iy NaOH ‘used. 10 yond )10.05 ce. To HCL required. tb’ ai 50.15 ee. NaOH heubea laa it 1 NaOH pis 75 HCL required. 60.20 ce. 17.15 ce. 43.05 ce. ®- NaOH used.) | To required. TOT *. NaOH neutralized. i+ NaOH used. | iy HCL required. *. NaOH neutralized, iti 60.20 ec. 20.87 cc. 39.33 ce. 188.50 cc. 4; NaOH ad up. Duration of hydrolysis, ninety-six hours. Solution made up to 500 ce. Two 25 ce. samples neutralized; glucose determined. (1) 21.40 ee. (2) 21.30 ee. 21.35.cc. 7; KMn0, used. 0.00663 < 21.35 = 0.1416 gram Cu = 0.0773 X 20 = 1.5460 gram glucose. 0.0773 gram glucose. Two 50 cc. samples made alkaline; distilled into standard acid (11.28 cc. 75). HYDROLYTIC STUDY OF CHITIN 335 (1) 6.60 ae ‘ 6.65 cc. 5 NaOH required. 11.28 B.30 7.95 ec.75 HCL used up. 7.95 X 10 X 0.0014 = 0.1113 gram N. _ Two 50 cc. samples, digested and distilled into standard acid (1. 28 cc. i5)- (1) 4.50 (2) 4.50 4.50 cc. 35 NaOH required. 11.28 2.25 9.03 ec. 74 used up. 9.03 X 10 X 0.0014 = 0.1264 gram N. III. August 21. 2.0492 grams dry chitin. 200 ec. of 40 per cent H,SO,. Tem- perature, 50° to 60°C. _ 9.30a.m. Hydrolysis started. 60.20 cc. 7> NaOH used. 11.30 a.m. Slight turbidity. 4.00 p.m. Solution complete; very slight color. 9.30 p.m. Hydrolysis discontinued. 26.40 cc. 7 HCL required in titra- tion. 60.20 — 26.40 = 33.80 ce. 44 NaOH used up. Duration of hydrolysis, twelve hours. Solution made up to 500 cc. Two 25 cc. samples neutralized ;.glucose determined. (1) 22.40 (2) 22.30 22.35 ec. 7 KMn0, used. 0.00663 < 22.35 = 0.1482 gram Cu = 0.0813 gram glucose. 0.0813 X 20 = 1.6260 grams glucose. Two 50 cc. samples made aklaline and distilled into standard acid (11.28 cc 7y)- (1) 6.35 (2) 6.45 6.40 ec. 35 NaOH required. 11.28 3.20 8.08 ec. 7 HCL used up. 8.08 X 10 X 0.0014 = 0.1131 gram N. Two 50 cc. samples digested and distilled into standard acid (11.28 ec. 7). (1) 3.85 (2) 3.90 3.88 ec. gy NaOH required. 336 SERGIUS MORGULIS IV. 1.28 1.94 9.34 cc. 7g HCL used up. 9.34 X 10 X 0.0014 = 0.1308 gram N. August 23. 1.9990 grams dry chitin. 200 cc. of 40 per cent H.SO,. Tem- perature, 50° to 60°C. 4.00 p.m. Hydrolysis begun. 30.10 ec. 7> NaOH used. 4.20 p.m. Material completely dissolved. 11.00 p.m. Hydrolysis discontinued. Solution slightly colored. Titr&ted with 10.20 ec. 7 HCL. 30.10 — 10.20 = 19.90 cc. 7; NaOH used up. Duration of hydrolysis, seven hours. Solution made up to 500 ce. Two 25 cc. samples neutralized; sugar determined. (1) 20.50 (2) 20.50 20.50 ec. #5 KMnO, used. 0.00663 X 20.5 = 0.1359 gram Cu = 0.0738 gram glucose. 0.0738 X 20 = 1.4760 grams glucose. Two 100 ce. samples made alkaline and distilled into standard acid (22. 55 cc.7y). (1) 22.70 (2) 22.50 22.60 cc. 35 NaOH required. 22.55 11.25 11.25 ec. 75 HCL used up. 11.25 X 5 X 0.0014 = 0.0788 gram N. V. August 23. 2.2024 grams dry chitin. 200 cc. of 40 per cent H2SO,. Tem- perature, 50° to 60°C. 4.30 p.m. Hydrolysis started. 60.20 cc. 75 NaOH used. August 24. 11.30 a.m., Titrated with 1.10 cey HCL. 60.2075 NaOH used. x 59.10 cc. fy NaOH used up. August 26. 8.30 a.m. Hydrolysis discontinued. Titrated with 19.75 cc.7g¢ HCL. 60.20 — 19.75 = 40.45 ec. 7) NaOH used up. . 99.55 cc. 7p used up. Duration of hydrolysis, forty hours. Solution clear, dark yellowish. Made up to 500 cc. Two 25 cc. samples neutralized; sugar determined. * (1): 23:25 (2) 23.25 a 23.25 ec. 7; KMnO, used. a: HYDROLYTIC STUDY OF CHITIN 337 0.00663 X 23.25 = 0.1541 gram Cu = 0.0850 gram glucose. 0.0850 X 20 = 1.7000 grams glucose. Two 50 cc. samples neutralized and distilled into standard acid (11.28 ee.75). (1) 5.30 (2) 5.34 5.32 ec. % NaOH required. 11.28 2.66 8.62 ce. 7p HCL used up. 8.62 X 10 X 0.0014 = 0.1207 gram N. August 24. 1.8550 dry chitin. 100 cc. of 40 per cent H»SOx. Temperature, 60°C. No aeration, flask stoppered. Material dissolved in about 45 minutes. At the end of one hour solution made up to 500 ce. On standing a slight tur- bidity appeared. Two 25 cc. samples neutralized; glucose determined. (1) 11.05 (2) 10.95 EE 11.00 ec. 4} KMn0, used. 0.00663 X 11 = 0.0729 gram Cu = 0.0375 gram glucose. 0.0375 X 20 = 0.7500 gram glucose. _Two 100 cc. samples neutralized and distilled into standard acid (11.28 e¢.75). (1) 8.10 (2) 8.20 8.15 ec. 75 NaOH required. 11.28 4.08 7.20 ec. 74 HCL used up. 7.20 X 10 X 0.0014 = 0.0504 gram N. The results just presented are summarized in the following table: s.1 21s [882 bss eee] e |= [88 [See ie a ° B Hy e#SZ |Beo a iS e 8 $2 he a a ZAO, |2h@ 5 ae ws a aA ag igs Curtin | ©°9 5 ASEaiaz< 68° | as | a Boa lez l-azle ee | 3 | 72 |CEegicee lage | 82 | 88 | SBF lceagia= BR | € | Ba lazezlasaznede] 22 | BR | A838 |aSaSE < a Z fs is Ip 3 cc) fe Be 5 gm. hours gm. gm. | gm. gm. | 1.8550 1 |0.1185)0.0504) 42.5 | 2.72 | 3.67 |1.5044/0.7500) 49.9 40.4 ? 1.9990; 7 |0.1277\0.0788| 61.7 | 3.94 2.45 |1.6202/1.4760) 91.1 | 73.8 | 19.90 2.0492! 12 |0.1309|0.1131) 86.4 | 5.52 0.87 |1.6619|1.6260| 97.84 79.4 33.80 2.0339} 24 (|0.1300|\0.1133| 87.1 | 5.57 0.82 |1.6495)1.6520|100.15) 81.4 50.10 2.2024) 40 |0.1407|0.1207) 85.8 | 5.48 0.91 |1.7861|1.7000) 95.2 | 77.2 | 99.50 2.0063; 96 (0.1282/0.1113) 86.8 | 5.55 | 0.84 |1.6271|1.5460| 95.0 | 77.1 188.50 338 SERGIUS MORGULIS This table brings out the following points: As the duration of the hydrolysis increases, more of the NHe groups are split off being convert- ed into ammonium sulphate, so that after twelve hours practically the maximum NH.-nitrogen is obtained. This maximum represents 5.5 per cent of the chitin, while the stable nitrogen which cannot be cleaved off by the hydrolysis is about 0.8 to 0.9 per cent of the chitin. It was found that merely dissolving the chitin in 40 per cent sulphuric and allowing it to stand for one hour is sufficient to cause 49.9 per cent of the total glucose to be split off. An examination of the table will show that after 24 hours of hydrolysis a maximum yield of glucose was ob- tained equal to 81.4 per cent of the chitin. This coincides with the largest recovery of NHb»-nitrogen. The amount of glucose found at this stage is practically 100 per cent of the theoretical expectation. Beyond this stage further hydrolysis is accompanied by a loss of glu- cose and a very rapid increase of the volatile acid fraction. The data contained in the above table are presented graphically in a series of curves (fig. 1) which bring out the various important points of this study at a glance. We may say, therefore, that the maximum yield of NHo-nitrogen coincides with the maximum yield of glucose and is independent of the acid formation but that if the hydrolysis is continued beyond 24 hours the great increase in the acid production is coincident with the oxida- tion and destruction of glucose. If the volatile acid is the product of oxidation of glucose—in other words, if it is not a primary constituent of chitin—is it possible that it is represented by acetic acid only? It seemed to me that if the volatile acid produced in the hydrolysis was a mixture of the lower fatty acids such a finding would take away the ground from under the hypothesis that an acetyl group is present in the chitin molecule, for there would be then just as much justification—or just as little—to consider every one of those acids as a constituent part of the chitin. Owing to the dif- ficulty of identifying various volatile acids in small quantities this side of the question has not yet been fully worked out. In one experiment the air current was passed through a solution of AgNOs and the presence of formic acid was indicated by the reduction of the silver. To verify this observation and to gain some approximate idea as to its relative amount, about 2 grams of chitin were hydrolysed in the usual way for forty-two hours and the distillate was collected in standard alkali. The amount of acid absorbed, as determined by titration, was 60.65 ce zy. To the neutralized solution a mixture of mercuric-chloride and HYDROLYTIC STUDY OF CHITIN 339 etate was added; the solution was then boiled for 6 hours, taken to prevent evaporation. A fine crystalline precip- ious chloride was formed. This was filtered off, dried RY yg Sf -4 Me P 7 7 va , 4 on v Zz vd pa if , i 3 4 4 - fs rs a | / / f / f / f / / : 40 : 5 chitin recovered as glucose iene > sAehrecorered os Ml,-nitrogen ighed. There was thus obtained 0.0480 gram HgCl which cor- to 0.0047 gram HCOOH, or approximately to 1 cc. 7. The acid thus formed 1.65 per cent of the total acidity produced. these experiments must be further extended it is evident, none- 340 SERGIUS MORGULIS theless, that in the hydrolysis of chitin acetic acid is not the only low fatty acid liberated. The question next to be considered is the relation to chitin of the acetylglucosamine which has been separated from acid solutions of chitin and which has the same empirical formula as the latter. The sub- stance investigated by Frankel and Kelly was obtained in the following manner: Chitin was dissolved in 70 per cent sulphuric acid and this was left standing for several days. When this solution was poured into an excess of water a fine powder precipitated out which on analy- sis proved to be monacetylglucosamine. It must be pointed out, how- ever, that acetyldiglucosamine was likewise found (Offer5). In one of my experiments, already referred to, I succeeded in getting a fine pre- cipitate by pouring a solution of chitin in hydrochloric acid into ten times its volume of distilled water. The precipitate formed very slowly, a mere turbidity appearing at first, but over night it accumulat- ed in a fair quantity on the bottom of the beaker. It was filtered, washed free of chlorides with distilled water, then with alcohol and ether and dried. Unfortunately the elementary composition of this precipitate was not determined. But its nitrogen content throws much light on its nature especially when this is considered in conjunc- tion with the data gained from the hydrolysis. This substance was found to contain only 5.8 per cent of nitrogen instead of 6.4 per cent which was found in the lobster material. From the hydrolysis experi- ments we learned already that about 0.8 per cent of chitin is in the form of a stable nitrogenous combination quite different from the re- maining 5.5 per cent which represents the amino groups. It would thus seem likely that the substance I got by precipitation from the acid solution of chitin consisted of the latter moiety of the nitrogen in com- bination with glucose while the other nitrogenous portion remained in solution. In other words, the precipitate which was formed from chitin was really no longer chitin, and I believe that the substances which have been studied by other investigators were derived from chitin but were not chitin. How then is this acetylated aminosugar formed? This, too, it seems to me, is very easy to understand. The fact must be clear- ly borne in mind—and reference to experiment VI will substantiate this—that leaving chitin in sulphuric acid will lead quickly to its break- ing up with the liberation of glucose and ammonia. Acids, too, may readily be formed through the oxidation of the glucose by the sulphurie > Offer: Biochem, Zeitschr., 1907, vii, 117. HYDROLYTIC STUDY OF CHITIN 341 acid. The stronger the acid the more extensive and rapid may these changes be. But glucosamine and acetic combine in the presence of a dehydrating agent, and in our mixture all conditions are present (the sulphuric acid acting also as the dehydrating agent) for the recombi- nation of these substances. This synthesis of the hydrolytic and de- composition products of the chitin molecule may account for the cir- cumstances that the new substances thus produced may either be acetylglucosamine, or acetyldiglucosamine or diacetylglucosamine. The acid medium will hold all these in solution, but being insoluble in _ water they are thrown down when the acid is greatly diluted. _ Having conclusively shown the fallacy of the assumption that chitin . isa polymere of acetylglucosamine, a different interpretation of its com- position may be attempted on the basis of the above hydrolytic studies. Referring again to the data recorded in the table, it will be noted that the two kinds of nitrogen present in chitin bear a definite ratio to each other asi1to7. Starting from this fundamental fact, it is clear, there- fore, that chitin, whose empirical formula is CsHi;NO¢, must be a poly- mere of no less than eight such groups, and its composition must there- fore be Cg:HizoNsOus. If we designate the nitrogenous portion of the chitin, the true character of which is still unknown to us, by the symbol Xn we may represent the stile brought about by ies by the following equations; (1) CesHizoN 304s + 8H20 = (2) 7CsHisNO; + CoeHi2Os + Xn + 7H20 = (3) 7NH3+7CeH2O0¢ + Cs5H20¢ + Xn. 8CeHw2O¢ According to this chitin should yield 81.1 per cent by weight as glu- cose; 5.54 per cent of nitrogen in the form of ammonia (which can be distilled off directly without preliminary digestion); 0.8 per cent of nitrogen which is not split off by hydrolysis and can only be converted into ammonium sulphate by digestion. The amino nitrogen, therefore, would constitute 87.5 per cent of the total nitrogen. The agreement of the experimental findings with these theoretical expectations is truly remarkable. It reveals the great complexity of the chitin molecule which consists not merely of glucosamine but also of,glucose and a nitrogenous substance of a still unknown nature That the glucose is present in the molecule in two forms is borne out by the study of the relation between the total amount of the sugar and the total NH:- nitrogen recovered. If all the sugar were in the form of glucosamine 342 SERGIUS MORGULIS the ratio of N:G would be 1:12.8, but in the experiment with the max- imum yield this ratio N:G is equal (0.1133:1.6520) to 1:14.6, thus proving beyond any doubt that some of the sugar does not possess the amino group. What the nature of the unrecovered nitrogenous residue is, is still a mute point. It may very well be that it forms a nucleus around which the glucose molecules are grouped. Attempts to isolate this — substance for purposes of identification and detailed study have thus far been unsuccessful, but further efforts in this direction are continued. I think that we may say with certainty that this residue cannot be of a protein nature because its nitrogen content is toolow. If the reactions involved in the hydrolysis have been correctly represented in the preced- ing scheme—and this would seem to be thoroughly corroborated by the experimental data—this residue should have the formula CjgHssNOis. This formula is suggested without any intention on my part to forestall what should be based on careful study. It may serve as a guide in the research which is to follow. SUMMARY 1. The nitrogen of the chitin molecule is partly in the form of the NH: group of the glucosamine, which is readily split off in hydrolysis — with dilute acid, and partly in the form of a stable combination from: which it can be released by digesting with concentrated sulphuric acid only. This latter portion constitutes one-eighth (12.04 to 12.45 per cent) of the total nitrogen. 2. The volatile acid produced in hydrolysis of chitin is probably a mixture of lower fatty acids. Its production is associated with a de- composition of the glucose molecule. 3. The maximum yield of glucose is about 81 per cent by weight of the chitin. 4. The evidence presented in this paper is against the accepted view that chitin is a polymerized acetylglucosamine. 5. An hypothesis is suggested, based upon results of a quantative study of the hydrolytic products, according to which chitin is composed of glucose amine, glucose and a nitrogenous moiety of still unknown nature. THE FATE OF SUCROSE PARENTERALLY ADMINISTERED SHIGENOBU KURIYAMA _ From the Sheffield Laboratory of Physiological Chemistry, Yale University, ' New Haven Received for publication March 19, 1917 - It has been stated by many investigators that some substances which are introduced into the organism in excessive amounts, whether they are foreign to the organism or not, can be eliminated through other channels than the kidneys, namely through the gastrointestinal tract, mammary glands, etc. Various _ elements such as iron, copper, strontium, barium or radium are eliminated in part, at least, through the bowel (1), (2), (3). Among organic compounds, the excretion of antipyrine, curarine, diphtheria toxin, egg protein, etc., through the intestine has been reported. (4) Many dyes, e.g., Sudan III,. Biebrich scarlet, methylene blue, phenolsulphonephthalein, phenoltetrachlorphthalein, ~ alkali blue, indigo carmine, have been reported to be eliminated into the alimen- tary canal, either by way of the liver or through the intestinal wall (5), (6), (7), (8), (9). _ Késsa (10) reported that in phlorhizinized hens sugar is eliminated in the in- testine, the proportion of the sugar in the urine to that in the feces being as 1:0.3. Résseler (11) also reported that feces of diabetic patients contained a demonstrable quantity of sugar and in an increased quantity after sugar feeding. In experimental diabetes, however, Allen (12) could find no sugar in the feces, even after feeding a large amount of dextrose. No sugar was found in the saliva of his diabetic animals. _On the other hand, Naunyn (13), Brauer (7) and Wood- yatt (14) were able to find reducing sugar in the bile of the animals, which were made diabetic by pancreas extirpation, piqire or phlorhizin injection. After injecting a large quantity of sodium chloride solution intravenously into rabbits, MacCallum (15) reported that sugar is eliminated into the alimentary tract. Investigating this problem further, Fischer and Moore (16) demonstrated that rabbits which are made to secrete sugar in the urine by a sugar puncture or by intravenous injection of dextrose or sucrose solution, do not excrete sugar into their gastrointestinal tract. But when a sodium chloride solution is injected intravenously at the same time, sugar is eliminated by the small intestine. They ascribed this phenomenon to an increased permeability of the intestinal mucous membrane resulting from the salt injection. Injecting a dextrose solution alone into rabbits, intravenously, in a sufficient quantity, Kleiner proved that hyper- _ glycemia alone can cause the sugar elimination into the intestine and stomach. The amount of sugar excreted, however, is incomparably smaller than the amount eliminated through the kidneys. A preceding double nephrectomy increases the gastrointestinal elimination of the sugar. After injecting 112 to 162 grams of dextrose in a 25 per cent solution intravenously into dogs, Grigaut and Richet 343 344 - SHIGENOBU KURIYAMA (17) also confirmed that the hyperglycemia suffices to cause sugar elimination | into the gastrointestinal tract. They recovered 1.0, 3.5 and 10.2 grams of glu- cose, respectively in the intestinal content, while 14.0, 11.0 and 14.6 grams of glucose, respectively were found in the urine. They also demonstrated that sodium chloride and urea can be eliminated in the intestine in the same manner. ‘According to the reports of Mendel and Kleiner (18) and others, su- crose, introduced into the animal body parenterally, can not be recovered quantitatively in the urine. In my own previous experiments, an average of 75.5 per cent of sucrose injected was recovered in the urine. In trying to explain the fate of the rest, the question of the possible production of invertin, which has been reported by Abderhalden and others to appear in the serum, was taken up by some investigators. My previous experiments failed to demonstrate the presence of invertin in the serum. Recently, Folkmar (19) also has been unsuccessful in finding invertin in the serum after parenteral administration of sucrose. He considered that the damage of renal function, resulting from sugar injection, might have some relation to the retention of the sugar in- ‘jected. Jappelli’s (20) experiments suggested that part of the sucrose injected might be kept in the liver, gradually appearing in the circula- tion or in the bile afterwards, or might be excreted by the stomach or . Salivary glands into the alinienbai tract. It has already been briefly indicated that under certain oinburhetanaee other organs than the kidneys can eliminate substances which appear abnormally in the organism. At the suggestion of Prof. Lafayette B. Mendel, I have tried to find out whether sucrose parenterally admin- istered, is in part eliminated through the liver or the intestinal mucous membrane. METHODS Full-grown dogs were used. In all experiments, a 10 per cent su- crose solution, sterilized by boiling, was injected into a jugular vein, sometimes very slowly (in the course of an hour), sometimes a little more rapidly (in twenty to thirty minutes). The amount of the solution injected was 150 cc. to 480 cc. The bile was collected by a biliary fistula, through a cannula in the common bile duct. After completing the experiment the bile obtained through the fistula was united with that in the gall-bladder, and this mixture was used for sugar determination. It is a very difficult problem to ascertain whether in- jected sucrose is eliminated through the intestinal membrane or not. The sugar may be eliminated there, but almost immediately be inverted and absorbed. In the present experiments, a loop of the small intestine was isolated between two ligatures, and a large glass cannula inserted SUCROSE PARENTERALLY ADMINISTERED 345 at each end. The upper end of the loop was a few centimeters from _ the beginning of-the duodenum, the length of the loop varying between 22cm.and79cm. The intestine was replaced in the abdominal cavity and the abdominal wall closed, only the rubber tubes connected with the cannulas being kept outside. The animal was kept warm during the experiment. Physiological saline solution, warmed to the body temperature, was introduced into the upper end of the loop, and the solution coming out from the cannula of the lower end of the loop col- lected. Every precaution was taken not to damage the intestinal mucous membrane and the blood vessels supplying the intestinal tract. Before the sucrose injection, the intestinal loop was washed out very gently with the saline solution (about 300 cc.), so that there was no reducing substance in the last portion. After sucrose injection, the sa- line solution was introduced drop by drop and the fluid coming out from the lower end of the loop analyzed for sugar. The saline solu- tion was not permitted to be contaminated with blood. In some ex- periments, the blood vessels of both kidneys were ligated before su- crose injection. For analysis sucrose in the bile or in the saline solution which was used to irrigate the intestinal loop, was first hydrolyzed by invertin. The reducing sugar, obtained, was then determined by Allihn’s gravi- metric method, colloidal iron being used to remove the protein and other disturbing substances. Invertin was prepared from compressed yeast, following the suggestions of Hudson. The reaction of the prep- aration was very slightly acid. Concerning the occurrence of reducing sugar in the normal bile, there are many contradictory reports. In discussing those which claim the absence of sugar in normal bile, Naunyn (13) stated that a minimal amount of sugar is a quite common occurrence in the normal bile, ob- tained from a fistula in the common bile duct. Brauer, (7) however, demonstrated, that normal human or dog bile does not contain any re- ducing sugar. Woodyatt (14) reported that in his dog experiments no reducing substance was found in the bile before phlorhizin injection. Okada (21) reported that the hydrogen ion concentration of normal dog’s-bile is variable within the ranges of P,, 5.34 to 6.97 in the case of bladder bile and 7.49 to 8.15 and 7.54 to 8.01 in liver bile during fasting and digestion respectively. On the other hand, the activity of the invertin is very sensitive to the hydrogen ion concentration of the medium. Following Michaelis and Davidsohn’s (22) report, the optimal zone for invertin is P,, 3.67 to 5.25, the medium between P,, 3.16 and 8.30 still allowing invertin to be active. Sdérensen (23) SHIGENOBU KURIYAMA 346 “UISUI $°6Z “WSU 8° 1h ‘wa l'¥Z a“ ‘ “uIsUI 0) “USUI Q°FT | “su yO, | “mau 9. ‘USUI 9°g UOISI9AUT B1OJag Rares (tae See : bd bs | 09 €°¢ "90: 3° 9OT “09 ¢°9E *00 9°6 09 0°01 JOppeyq [Tes wo1y ‘oon | °00 gt] ‘o0¢%%| ‘200'9T| “20 FIT BINIsy WOT me “UIUL OE “uIUI Og inoy T “UIUL OF “UTUL 0Z uoryoefur jo uoreing uoroef SUIBIS (Q'0z | SUIBIS Q' CZ | SUIvIZ Q' QF | SUIBIZ G' zz | SUIvIS CCT poyelur yunoury 2 eS "s]98 ‘sos ; -S0A [BUdI |-SdA [BUOL jo uonjes |jo woes "S98 -11 + doo, |-1] + doo] doo, |-SeA |Buel [@UI9so}y |[Veul9yso4 jeurysey | jo uones -ul + B[ny |-Ur + Bin, j-ur + eng |-1] + Bn Biny “sg Arerig|-sy Arerfig| -sy Arerig| -sy Areyig| -sy Areyig wory}e10dQ) “UTUL “UIUL "s1q F CI “Aq T ‘sy ¢ ‘S1y fF 0z “S1y Z qyuoulliodxs jo uor1yving royqye + zoqyo + reqge + reqye + reqyeo + (arv13 Q'9) |(weIS Q'9) jureaZQ°eT) |(meIs QZ) |(uIeI3 T'Q) sUBY}OI| SUOJOIOTYO} ouoqer0[yH| osuvyjoig|] osurydioyy soljoyysouy ‘UsyzO'S | “WBYO'ET | “ByOzE | “wWSYS'zI | “wW3xO OT yystom Apog : od SoA © AI © III © «II roa | qoquin Ny pe.ajsiuipn fizjvsajuasnd asouons fo a70f ay J, 347 SUCROSE PARENTERALLY ADMINISTERED ‘UOISIOAUT 10978 YUOD Iod GF'( 0} BUISvOIOUT ‘UOISIOAUT BI0JoG YU90 10d QO’ SBA ‘B8O1}XOP SB PoyB[NOTwo “yu0e4 -u00 IB3Ns OY], “poyjyoul yorpousg-stmoT oy} AQ PpouTUIvxe SBM 4JUEZUOD ABZNS poo;q oy} ‘yuoUTTIOdxe oY} JO pue ou} IV « UOISIOAUT 104 V Ss Shen £°v — oA U014BYOY 6°01 + 8°9T + oA MOPIOAUT G10jOg = _ +++ - +--+ 4- UOISIOAUT IO4FV oun) ‘uoTpONpeyy _ —_ + -- + UWOISIOAUT 910JOgT "90 OT ‘09 0°% *09 LZ 00 G°Z ‘09 FET SCE A. “wise O'Zzg) “TABor "eg, | “wasor 0: sh TOISTOAUL 1OSV | reGns yAOAUT SV Po}%| “ut Z SIP] “Wu F'gg | ‘ws CO uorsoaut orojog, [°T¥? raed mapepey doo} : [vurysoqur ‘99 089 "09 OT "09 0&9 P0}09][09 WOTYNIOS oUTTwS JO OUNJOA |jo UOTyeB III] “Ud 6 "UID £9 "Wd ZZ doo] [eurjysezur Jo yysue'T 348 SHIGENOBU KURIYAMA ascribed P,, 4.4 to 4.6 as the optimal point for invertin, the medium between P,, 2.55 and 7.30 still allowed some activity of the ferment. As preliminary experiments, therefore, the reducing-power of the normal bile and the activity of the invertin preparation added to the bile were examined. ‘The bile, obtained from a normal dog, a cat and two pigs never showed any reducing power. Sucrose added to those samples of the bile was easily hydrolyzed by the invertin preparation under the experimental conditions employed. When the invertin prep- aration was previously boiled, or the active invertin preparation alone was added, no reducing substance was produced in the bile. The procedure for sugar determination in the bile was as follows: After adding 5 ce. of the invertin preparation and 0.1 cc. of toluene to 10 cc. of bile, the mixture was kept 23 hours at 38° to 40°C. Sixty cubic centimeters of distilled water and 25 cc. of colloidal iron solution were then added and the mixture was shaken vigorously for a few min- utes. Half an hour later, about 0.5 gm. of sodium sulphate was added. Seventy-five cubic centimeters of the clear and slightly yellow filtrate were used for sugar determination by Allihn’s method. For control, the same procedure was always carried through with boiled invertin. When the amount of bile was not enough, less than 10 cc. of it was used for one procedure, the added substances being lessened in the same pro- portion. Determination of sucrose in the saline solution, which was used to irrigate the intestinal loop, was performed in the same manner. Twenty cubic centimeters of the invertin preparation were added to 100 cc. of the saline solution. Five cubic centimeters of colloidal iron solution were enough to remove protein, The control with boiled in- vertin was also performed. The amount of reducing sugar, both before and after inversion, was always calculated as invert sugar. After completing the experiment, the small intestine and the ligature of the renal yessels were examined. The mucous membrane of the intestinal loop showed a slight edema in some cases. Otherwise no noteworthy changes were observed, either in the loop or in the rest of the intestinal tract. The urine obtained from the bladder at the end of the experiment was tested for sugar. When both kidneys were shut out from the circulation, the urine obtained was only a few cubic centimeters and contained no sugar either before or after inversion. When the kidney vessels were left intact the urine contained a large amount of sucrose and only a small amount of reducing sugar. When the sucrose was hydrolyzed with invertin, the urine acquired a marked levorotatory and strong reducing power. The results of the experi- ments are shown in the preceding table. : SUCROSE PARENTERALLY ADMINISTERED 349 CONCLUSION From the results of the present experiments, it will be seen that su- crose, injected intravenously into dogs, is eliminated very quickly through the kidneys, but at the same time, a minimal amount of it can be eliminated through the liver and the intestinal mucous membrane. The ligation of the renal vessels does not necessarily facilitate its ex- cretion through the liver and the intestinal mucous membrane. The permeability of the kidneys and intestinal mucous membrane for dextrose is said to be increased by parenteral administration of sodium chloride (24), (16). Phlorhizin glycosuria and _ glycocholia are also considered to be due to the change of the permeability of the kidneys and liver respectively. It is not improbable that sucrose may behave similarly. Parenter- al administration of sucrose, in sufficient quantities, may increase the permeability of various organs, not only for sucrose itself, but also for dextrose in the blood. This may explain why a small amount of re- ducing sugar preéxisted in the bile and the urine together with sucrose. In my previous work (25) already, it was noticed that reducing sugar sometimes appears in the urine after parenteral administration of su- crose, (no anesthetics were used). At that time, it was suggested that part of sucrose might be inverted somewhere in the body, so that the reducing sugar, preformed, appeared in the urine. The change of renal permeability for normal blood sugar offers another explanation. The preéxisting sugar in the saline solution collected from the intes- tinal loop may also be explained by change of the permeability for dextrose. At the same time, however, in this case, the activity of invertin in the intestinal juice must be considered. It can not be as- serted that the irrigation of the intestinal loop, however carefully it may be performed, is absolutely harmless to the mucous membrane. It is very unlikely, however, that this is the only cause of the appear- ‘ance of sucrose in the intestinal canal. On the other hand, part of the sucrose, which appeared in the intestinal loop, might be absorbed: immediately, notwithstanding the constant irrigation. If such absorp- tion did not exist, the amount of the sucrose, recovered in the loop, might be much larger. When sucrose is injected parenterally, 20 to 30 per cent or more of the amount injected, fails to be recovered in the urine. As the amount of sucrose excreted in the alimentary tract is extremely small, this channel for sucrose excretion seems to be insufficient to explain the fate of the missing part of the sucrose injected. The question as to 350 SHIGENOBU KURIYAMA whether sucrose is rendered utilizable by being inverted in its passage _ through the intestinal wall remains unanswered. I desire to express my thanks to Prof. Lafayette B. Mendel, to whom I am greatly indebted for his suggestions, help and criticism; also to Prof. F. P. Underhill for his advice. BIBLIOGRAPHY (1) Menpret AND TuacuER: This Journal, 1904, xi, 5 (2) MenpDEL AND Sicuer: This Journal, 1906, xvi, 147. (3) Bera AND WELKER: Journ. Biol. Chem., 1905-06, i, 371. (4) Lanamr: Zeitschr. f. exper. Path. u. Therap., 1906, iii, 691. (5) MenprE.t AND Dantrets: Journ. Biol. Chem., 1912, xiii, 71. (6) Satant AND Benets: Journ. Biol. Chem., 1916, xxvii, 401. (7) Braver: Zeitschr. f. Physiol. Chem., 1903-04, xl, 182. (8) ABEL AND RowntrReEE: Journ. Pharm. and Exper. Therap., 1909, i, 231. (9) Kurryama: Journ. Biol. Chem., 1916, xxvii, 377. (10) Késsa: Arch. internat. de Pharm. et de Thérap., 1906, xvi, 33. (11) Résseimr: Zeitschr. f. Heilk., 1901, xxii, Abt. f. int. Med., 302. (12) Auten: Glycosuria and diabetes, Cambridge, 1913, 19. (18) Naunyn: Arch. f. exper. Path. u. Pharm., 1875, iii, 157. (14) Woopyratt: Journ. Biol. Chem., 1909-10, vii, 133. ~ (15) MacCautium: Univ. of California Publ., Physiol., 1903-04, i, 125. (16) Fiscoer anp Moors: This Journal, 1907, xix, 314. : (17) Grigaut AND RicuEer: Compt. rend. Soc. Biol., 1912, lxxii, 143. (18) Menpex anp Kuerner: This Journal, 1910, “eel, 396. (19) Fotxmar: Biochem. Zeitschr., 1916, lxxvi, 1. (20) JAPPELLI: Jahresberiche d. Tetohein., 1905, xxxv, 79. (21) Oxapa: Journ. Physiol., 1915, 1, 114. (22) Micnar.is anD Davipsoun: Biochem. Zeitschr., 1911, xxxv, 386. (23) SdreENsEN: Biochem. Zeitschr., 1909, xxi, 131. (24) UNDERHILL AND Ciosson: This Journal, 1905-06, xv, 321. (25) Kurryama: Journ. Biol. Chem., 1916, xxv, 521. THE PERFUSION OF THE MAMMALIAN MEDULLA: THE EFFECT OF CARBON DIOXIDE AND OTHER SUBSTANCES ON THE RESPIRATORY AND CARDIO-VASCULAR CEN- TERS ; D. R. HOOKER, D. W. WILSON ann HELENE CONNETT : From the Departments of Physiology and Physiological Chemistry of the Johns Hopkins University ; Received for publication March 23, 1917 INTRODUCTION It is generally accepted at the present time that the hydrogen ion . concentration of the blood is alone concerned in the chemical regula- tion of respiration. Haldane (1) and his co-workers laid the founda- tions for this conception of the respiratory function but Winterstein (2) apparently was the first to present the facts in the form of a concrete hypothesis. Winterstein reported experiments on the perfusion of new-born rabbits which led him to the conclusion that neither oxygen- want nor carbonic acid-excess as such stimulated respiratory activity; rather the hydrogen ion concentration of the perfusate was the deter- mining factor in the stimulation. This hypothesis has received funda- mental support in the work of Hasselbalch (3), Barcroft (4) and others, and has been successfully employed as a basis for the treatment of dis- ease and for the explanation of numerous physiological processes. While this hypothesis has received wide application and has stimu- lated. a renewed interest in respiratory problems, it is nevertheless open to criticism chiefly as regards its exact experimental basis, a basis which has not been strengthened by Winterstein’s subsequent contri- bution (5). Laqueur and Verzér (6) using Winterstein’s method of perfusing new-born rabbits, believed they were able to show that car- bonie acid acts asa specific respiratory stimulant. The evidence which they offer, however, is not convincing. Rona and Neukirch (7), studying the behavior of isolated loops of rabbit’s intestine in dif- ferent solutions, found that the addition of sodium bicarbonate mate- rially improved the rhythmic contractions of the preparation. Fur- thermore, they were able to show that the effect thus produced was 351 352 D. R. HOOKER, D. W. WILSON AND HELENE CONNETT apparently in no wise associated with the concomitant hydrogen ion concentration of the solution. We have here an instance, therefore, in which the HCO; ions apparently act specifically. Consequently it has seemed advisable to put the original hypothesis as advanced by Winterstein to a new experimental test. This test was to perfuse the mammalian medulla by a method previously described (8) using blood perfusates of which the hydrogen ion concentration and carbon dioxide tension were Known, and to study the effects produced by these two factors upon the activity of the respiratory center. Our results would indicate that an adequate statement of the regulation of respi- ratory activity cannot be fully summarized in the brief statement of Winterstein’s hypothesis. We have obtained evidence that carbonic acid exerts a specific influence upon the respiratory center independent of its effect upon the hydrogen ion concentration of the blood perfusate. From these results the conclusion seems justifiable that either carbon dioxide acts as a specific stimulus or alters the irritability of the ‘ center to the normal stimulus. EXPERIMENTAL A number of preliminary experiments were performed which demon- strated that the method was adequate to the problem in hand. Ref- erence to table 1 will make it clear that the respiratory center responds to changes in the blood perfusate in accordance with current ideas on this subject. These experiments show the general nature of the response. Sodium bicarbonate and sodium hydroxide depress respiratory activity; carbonic acid, hydrochloric acid, lactic acid and oxygen-want increase respiratory activity.. The effect upon the cardiac and vasomotor centers appears to be of the same order as that upon the respiratory center. That is to say, those substances which increase respiratory activity tend to increase the arterial blood pressure and heart-rate, and those substances which depress respiratory activity tend to decrease the arterial pressure and heart-rate. The response of the latter centers is what we should expect from our knowledge of the functional interaction of the respiratory and cardio- ‘In the single experiment that we have carried out increased respiratory activity was very apparent when blood poorer in oxygen was perfused. ‘The rapidity of response precludes the possibility of the formation of acid meta- bolic products by decreased oxidation and suggests that oxygen-want may cause increased irritability of the respiratory center. Further work along this line is contemplated. 353 EFFECT OF CARBON DIOXIDE ON RESPIRATORY CENTER % % eT PPL PPT 04 6ZT 0} 9ZT 04 SOT 2 0 OZT °F 18 ) @ & A A ZL 04 U8 poddoig Giva TaVaH > 1g GsZ‘T \orT‘T | Tg 1g og 02 | 9°2 | 9°2 > GZI_ |osr‘% jO8o‘T | #¢ a4 oF 02 | é9°2] 92 | Bu01y4s “uemM—to > SI— |oZP‘T |esz‘T | a 1¢ ee Sain ip UP A gk "O16L ‘Tg Arenuve > GL G66‘T |OFT‘T | 2g Lg eg 02 | éo°L | G2 > 09 o9r'% joce‘T | gF #S Ch G | 2o'L | GL | yweoom Quem—*o | | 102 |¢h6~—s(STE €9 | $9 | SI of eres ) og cIz‘T |ors 18 18 cI OL L172 © plow o1oeT 18 GZL‘T |0z8 GL 28 ST Or OL) FL SLI |Szt‘T |Sor Gl 18 cI G OL) FL IOH 19 Ose‘T |0F8 69 8 02 01 2/2 62 09¢‘T 1028 SL 18 0 Or TL) 02 °° O16T ‘gE Arvnuve TSt ozs‘ jeze¢ 99 GL 0z L SLES 69 0z8‘T |082Z 99 SL 02 Or GL) FL OFT |Gz8‘T |082 | gz SL GS OL SLL plow orpow'y 6 gz9‘% joce‘T | ¢z 06 Gg GI | co. | #2 Z8tjose‘T joze Sg 1g GZ OL | ca'2| FL 10H |J a €8 099 joo¢ ee | 81 02 | 02 |-6'9'| 9°2 IOH |) = Zt =|006~—s« SOF 08 1Z 0g cl ZLI9'L yued ted OT 709 = 0 Ozr \0zF 1Z 12 02 02 2.4) 9'2 a SZT |gce |68T 18 lz cl L SLIGO ‘OH |} * ‘SI6L ‘Iz sequie0eq = S4I 092 022 0g lz c% OF: 1: Ss. FO > 618 ¢ess‘e joes 1g 68 9 Or 2L4)9'L | Juoa sod OT *O0 < Sg O28 0 01 9 18 OF |-F'219'2 | = lz FOS = OF 8 9 88 OP |-FL/9'2 quod sed ¢ FOX |" GTET ‘ZT taquieseq < P OFZ =—|1&z 2 ee Or nA < b— |88% |008 ¥% 0g ZI Or ‘OOH®PN |" ** GI6I ‘9 toquMaAON = OOF‘T |Osk jog cl ¢ 0g 01 \ > oss jo9g [eZ ¥Z cI | StI g West ws cc lee soit < |oor-|o loee | 0 | ee | 0 | ot HO®N | f pics ana’ “‘HNSOY | O1OJog_ | 4[NSoy | atojogy | *y[Nsexy | eroyogyT | quo | : those [1O1U0D “joo | “2 | -wexdmye | aoderey | upopeaidary VERRY act ava NoIVulasan = Hq qoord — Suamisadxa hunurmyasd fo synsau pasuapuoy : 354 D. R. HOOKER, D. W. WILSON AND HELENE CONNETT vascular centers. Nevertheless, the action of these several substances on the cardio-vascular centers has not been discussed and the results are, therefore, of interest. It will be noted that the response of the cardiac and vasomotor centers is not so definite as that of the respiratory center. This may be due in part to a lesser sensitiveness of the centers themselves, but it seems more probable that the difference is primarily due to the con- dition of the peripheral mechanisms. In these experiments circula- tion is maintained in the trunk to preserve the diaphragm as a record- ing mechanism for respiratory activity and no effort is made to retain normal conditions as to temperature, etc., for the peripheral vaso- — motor system. The trunk, therefore, may be regarded as being in “shock.”’ The cardio-motor fibers are, on the other hand, almost certain to suffer some injury in the technical procedure of inserting the perfusion cannula in the common carotid artery close to its origin on the aortic arch. It is not surprising, therefore, that cardio-motor responses should be weak and often entirely absent. In both table 1 and table 2 the figures in the column under per- centage change in respiration represent the change in the factor rate X amplitude of respiration expressed in percentage of the same factor before the experimental blood was tested. The minus sign (—) be- fore the figure indicates a decrease in the factor; its absence indicates an increase in the factor. This factor is used to give an approximate indication of pulmonary ventilation. It is of significance because, as will be noted, variations of rate and amplitude do not run parallel; sometimes the one and sometimes the other determines the value rate xX amplitude and sometimes one is increased while the other is de- creased. But their product gives quite consistent results throughout the table. To this there are but two exceptions, namely, in the experi- ment of November 6, 1915 and of January 31, 1916, and in these cases the deviation from the control values is relatively small. The necessity of using this factor (rate X amplitude) as a basis of comparing the results obtained, raises the interesting question as to whether or not sensory impulses from the lungs alone control the rate of respiration. Scott (9) has presented evidence to show that when carbon dioxide is inspired the increase in rate of respiration is wholly dependent upon vagus function; that when the vagi are sectioned and carbon dioxide is administered, the only result is an increase in depth of respiration. In the experiments here reported, the rate as well as the amplitude of respiration is variously affected. The vagi are in- FEFECT OF CARBON DIOXIDE ON RESPIRATORY CENTER 355 tact but the thorax is open and the lungs are being regularly expanded with _artificial~ respiration. Vagal action cannot, therefore, be of influence in determining the respiratory rate, and the .conclusion appears inevitable that variations in the blood supplying the respira- tory center may determine not only the amplitude but also the rate of respiratory discharge. In one of these preliminary experiments we used lactic acid to render the blood more acid to see if its effect differed from that of hydrochloric acid yielding the same hydrogen ion concentration. The experiment is pertinent in that lactic acid is a normal product of intermediary metab- TABLE 2” Condensed results in comparison of CO2 with HCl BLOOD Py= RESPIRATION g = ss “pe Rate —— : 2 DATE TEST MATERIAL |- «> |millimeters — BS = Per Be 8 2 cent 3 RieclLeleiet® 2 bi | camer aa e/Eleleis/e] 2] 2 35 OlfmimiAlalaA -) QA Fy Read te COz5percent |7.6 |7.31| 3] 35) 69} 87| 207] 3,045] 1,366) 40 1916 : HCl 7.6 7.31) 45 40| 48 | 75) 2,160) 3,000 39) «5 CO, 5 per cent |7.6 |7.31) 35 | 55) 39 | 84! 1,365) 4,620) 165) 25 June 21 CO, 5 per cent |7.55/7.25| 18 | 50) 75 | 93) 1,350) 4,650/ 244) —8 1916 F HCl 7.55|7.25| 19 | 25) 78 | 96) 1,482) 2,400) . 62) = CO, 5 per cent |7.55/7.25) 25 | 25) 45 | 72) 1,125) 1,800 60} 15 HCl 70°14 -2or oe 5) 75: |}. (Si 220n ee a00 73} +6 June 22, HCl 7.6 |7.23) 5 9| 78 | 99) 390) 891 128} —9 1916 CO: 5 per cent |7.6 |7.23) 5 | 100) 72 114) 360)11,400| 3,069} 3 ||COz 5 per cent |7.6 |7.23| 55 | 70) 18 | 48} 990) 3,360} 239) 5 olism and is formed in excessive quantities under certain conditions which lead to lowering of the carbonic acid content of the blood. Under these conditions it might presumably be as efficient as carbonic acid in calling forth a respiratory response and more efficient than other acids, for example, hydrochloric. The results of this experiment indi- cate that there is no marked difference in the action of the two acids and since hydrochloric acid was somewhat more convenient, we used it in the subsequent experiments designed to investigate the specificity of carbonic acid as a respiratory stimulant. Our experiments were directed primarily to a comparison of the - respiratory response elicited by blood, the hydrogen ion concentration 356 D. R. HOOKER, D. W. WILSON AND HELENE CONNETT of which was altered to the same degree by the addition of carbon dioxide and hydrochloric acid. To this end we used as a control, blood which had been shaken to remove the excess of carbon dioxide and ob- served the physiological reaction produced by the substitution of bloods with higher hydrogen ion concentration. In this procedure a number of precautions had to be observed. (1) General character of the several perfusates. To insure the utmost simi- larity in regard to possible unknown factors, the defibrinated blood was collected, mixed and aerated in a porcelain-lined vessel after which it was subdivided in clean stoppered flasks. These samples were then submitted to the appropriate change of hydrogen ion concentration and placed in a constant temperature water-bath until used. (2) Tem- perature. Variations in temperature doubtless influence the behavior of the medullary centers. To guard against possible errors due to this factor, the sample, the response to which was next to be investigated, was placed in the substitution reservoir of the apparatus some ten min- utes before being used so that its temperature, if not actually that of the control in circulation, was at least constant as compared with any other similar sample investigated in the same experiment. (3) Change an wrritability of the respiratory center. The medullary centers un- doubtedly lose in functional capacity in the course of perfusion and this rate of loss is not comparable in different experiments. It is evi- dent, therefore, that the comparison of the response produced by two stimuli so closely alike as, for example, carbon dioxide and hydrochloric acid bloods, is unsafe unless they are both tested in the same experi- ment. And even then the progressive loss of irritability may establish conditions such that the test of the second specimen of blood is, for purposes of comparison, invalid. If, however, the perfusion of carbon dioxide blood produces the greater increase of respiratory activity when tested after as well as before hydrochloric acid, we may believe that the carbon dioxide blood is the more effective stimulus. This we were able to show in the present experiments. PREPARATION OF PERFUSATES The defibrinated blood used as control was agitated in air until its hydrogen ion concentration had reached a minimum. One portion of this blood was placed in a flask and agitated while being exposed to an atmosphere of 5 per cent carbon dioxide in oxygen. The hydrogen ion concentration of this sample was determined independently by two EFFECT OF CARBON DIOXIDE ON RESPIRATORY CENTER 357 of us using the indicator and electrical methods. A second portion was brought to the same reaction by treating with hydrochloric acid and shaking to remove the excess of carbon dioxide liberated. In this way two specimens of blood were prepared; one in equilibrium with a high tension of CO. (5 per cent of an atmosphere) and therefore con- taining a considerable quantity of free carbonic acid and a relatively high concentration of total carbonate; the other in equilibrium with a low tension of CO, (0.04 per cent of an atmosphere) and containing relatively less carbonic acid and total carbonate but with the same hydrogen ion concentration as the former. The reactions of both specimens were determined by the colori- metric and the electrical methods. The former could be used advan- -tageously to obtain approximate results in a short time, while the lat- ter method furnished the more trustworthy results especially with _ the bloods containing carbonic acid. These determinations were car- ried out at room temperature, using the improved Hasselbalch hydro- gen electrode, McClendon potentiometer and a recently standardized Weston cell. A portion of the blood specimen was placed in the elec- trode to establish CO.-equilibrium with the hydrogen gas and then replaced with a second portion on which the determination was made - in the usual manner. Three experiments comparing the efficacy of carbonic acid and hy- drochloric acid bloods were performed. The results are presented in table 2. The headings used in this table are the same as those used in table 1 and require no further consideration. We wish, however, to draw particular attention to the figures in the column under per- centage change of respiration. These figures bring out forcefully the difference in response to the two experimental bloods. Carbonic acid * blood when used before as well as after hydrochloric acid blood pro- duces much the greater effect. The apparent exception to this state- ment in the second test of carbonic acid blood in the experiment of . June 21 may, we believe, be properly explained by a progressive de- _ terioration of functional activity in the center. In this case the re- sponse is at least as great as it was in the preceding test with hydro- chlorie acid blood.. The influence of such a progressive loss of func- tion was considered earlier in the paper and we do not believe that our conclusions are weakened by one instance in which carbonic acid blood failed to elicit a greater response than that produced by hydrochloric acid blood tested previously. Portions of the graphic record of one of these experiments are pro- NE CONNETT ND HELE ILSON A . W. W D ? HOOKER R. D. 358 *poolq 9prxolp UOGIvO Y4IM UOTJOVOI OY} UI ADU94R] 10J1OYS OY} OSV 94ON *O[Qe4yIIAI o10U ATQqevuINseid 910M S.190}U90 ALC][NPoUl 9Y} USYA OUIT? JOBUO] B IOJ 40" 0} POMOT[V SBA 41 YSNOYYI ‘pooyq prov oIMOTYDopAy 9Yy} Aq poonpoid yoyo qyST[s AJOATZVIOI OY} BION “EZ'L = Fg poolq oprxorp uoqivo oy} Jo 4VYy} pu Czy = Bg poo;q plo¥v o1I0TYyDoIpAY oy} jo yey} eg’, = Bq SBM pooTq [BVUIOU 94} JO UOTJOveI BY, “poo;q [eyUeUIIOdxe oY} YIM UOISNj10d 9Yy} Jo UOTZe]duUI0D pu ZuruuIZeq 9Y} p109eI oY} JO UOT}OOS YOR UI O}VOIPUT OUT] STY} UT SYvorq OMY OYJ, “spoOlsed puodDes OM} UT OUTT} YIM ainssoid poo]q 010% ‘p109e1 JAEMOT ‘JOJEWIOURUT AINDIOUL YIM Poepso0de1 vinsseid poorq ‘p1ooe1 oIPpI “wWMNtIyseBIdae jo syuoumoAou A10;vIIdsei ‘plover 1oddy, “gT6T ‘Zz eUuNL JO QUOUTIIEdxe UI poUTeyqo piodd1 oTydeAS JO UOT}IOg “TT “B17 | Uadhelt pALE, nhl a seat ACT POPE EFFECT OF CARBON DIOXIDE ON RESPIRATORY CENTER 359 duced in figure 1, the protocol of which may be given as an example of all. Experiment of June 22, 1916. Comparison of effects of blood brought to the same hydrogen ion concentration with CO, and HCl on dog’s medullary centers. Large dog bled several times after intravenous injections of warm Ringer’s solution. Total volume of defibrinated blood thus diluted about 3000 ce. While still warm the blood was aerated by rapid stirring in a large vessel for half an hour to lib- erate excess of CO.. Four hundred cubic centimeters were submitted to an atmosphere of CO. 5 per cent + O2 95 per cent by bubbling the gas through a trapped flask containing the blood for about fifteen minutes. Two hundred cubic centimeters were treated with M/7 HCl in 0.4 per cent NaCl and the blood shaker until the hydrogen ion concentration ceased to diminish, i.e., until the excess of free carbon dioxide was removed. The reactions of the three speci- mens were: INDICATOR ELECTRICAL METHOD METHOD Normal blood...... ae Ra ae aes, p, = 7.6+ | p, = 7.55 oy ey race ces cece elect cues 7.25 7.22 MEM og oe es Begs. et ae 7.25 7.22 Normal blood after repeated perfusion.......... 7.51 Blood in stoppered flasks placed in water bath at 36°C. The experimental animal was operated under chloretone anesthesia. The perfusion was at 80 mm. Hg, and 37°C. The venous outflow was about 180 ce. per minute (30 cc. in 8.5 seconds). Probably the normal blood will tend toward an acid reaction as the rapidity of flow made it technically difficult to segregate the experimental blood. During perfusion of the experimental bloods they were of necessity partly exposed to room atmosphere. This will not affect the HCl blood but may result in lowering the hydrogen ‘on concentration of the COs blood by diffusion of the gas. Inspection of the figure shows, furthermore, as indicated in the elapsed time between the breaks on the time record that the hydro- chloric acid blood was perfused for periods of eighteen and ten seconds without eliciting a sharp reaction in either case, while the perfusion of CO, blood for two periods of four seconds each produced not only a much more decided reaction but one which exhibited itself very much more promptly. The fact is significant as further indicating the greater efficacy of carbon dioxide. The other experiments in this group showed this sharp response to carbonic acid blood to a lesser degree probably because the rate of perfusion was slower. 360 D. R. HOOKER, D. W. WILSON AND HELENE CONNETT DISCUSSION The experiments thus outlined appear to prove conclusively that blood with a comparatively high tension of carbon dioxide causes a greater stimulation of the respiratory center than does blood with a lower tension of carbon dioxide but with the same hydrogen ion con- centration. The method of stimulation is unknown though. several theoretical considerations are of interest. The reactions of the bloods were within physiological limits. As the bloods with higher hydrogen ion concentration were still alkaline, the diffusion of free acids other than carbonic may be considered neg- ligible. The similarity in the results obtained by the use of bloods to which either hydrochloric or lactic acids were added suggests that there is no appreciable difference in the diffusion of the two anions. As it would seem that lactic acid ions could easily penetrate the respiratory center on account of their great diffusibility the above observation is opposed to the conception that hydrochloric acid ions elicit a relatively. small response because of their failure to readily diffuse into the same cells. The variations in the concentration, of the salts in the perfu- sates, incident to the alteration in the reaction of the blood do not have an appreciable inhibiting action on the respiratory center, as unpub- lished experiments have shown. The cause for the smaller stimulation by the bloods to which hydrochloric acid was added (and the excess of carbon dioxide shaken out) must therefore be due to the decreased concentration of carbon dioxide or carbonic acid. As the tendency for diffusion of carbon dioxide from the cells into the blood depends on the difference in the tension of carbon dioxide of the two systems, the higher the carbon dioxide content of the cir- culating blood, the slower the diffusion from the cells and the higher the concentration in the cells when a new equilibrium is established. The tension of carbon dioxide in the respiratory center would presum- ably, therefore, be higher in our experiments in’ which blood with a high tension of carbon dioxide was employed and, vice versa, low when blood with low carbon dioxide concentration was used. It would seem that the hydrogen ion concentration of the two systems would like- wise maintain an equilibrium. If this is true, the hydrogen ion concen- tration in the center would be the same in both series of experiments because the reaction of the two bloods was the same. The greater activity produced by the bloods containing the high tension of carbon dioxide must therefore be due to some specific action of the carbon dioxide. | EFFECT OF CARBON DIOXIDE ON RESPIRATORY CENTER 361 Whether carbon dioxide as such stimulates the respiratory center or whether variations in the carbon dioxide concentration alter the irri- tability of the respiratory center can hardly be demonstrated. If we are to adopt an explanation for the results obtained in these experi- ments, the most satisfactory point of view- would seem to be that the hydrogen ion concentration of the environment of the respiratory center is its effective stimulus but that the irritability of the center to this stimulus may vary and be influenced by many factors. The normal irritability of the respiratory center is doubtless the summa- tion of a number of effects including those produced by carbonic acid, oxygen and various ions as well as changes in metabolic activity, the nature of which is not understood. CONCLUSIONS 1. An increase in alkalinity of the perfusing blood (produced by adding sodium bicarbonate or sodium hydroxide) tends to depress the medullary centers. 2. A decrease in alkalinity of the perfusing blood (produced by add- ing hydrochloric acid, lactic acid or carbon dioxide) tends to stimu- late these centers. 3. A specimen of blood containing a high tension of carbon dioxide causes greater activity of the respiratory center than another speci- men of the same hydrogen ion concentration but with a low tension of earbon dioxide. Carbonic acid thus acts specifically upon the respiratory center. . BIBLLOGRAPHY (1) Hatpane: Organism and environment, New Haven, 1917. (2) Winterstern: Pfliiger’s Arch., 1911, exxxviii, 167. (3) HasseiBatcu: Biochem. Zeitschr., 1912, xlvi, 403. (4) Barcrort: The respiratory function of the blood, Cambridge, 1914. (5) WintTerRsTEIN: Biochem. Zeitschr., 1915, lxx, 45. (6) Lacqueur AND Verzir: Pfliiger’s Arch., 1912, exliii, 395. (7) Rona anp Nevxircn: Pfliiger’s Arch., 1912, exlviii, 273. (8) Hooxer: This Journal, 1915, xxxvili, 200. (9) Scorr: Journ. Physiol., 1908, xxxvii, 301. “THE AMERICAN J OURNAL OF PHYSIOLOGY VOL. 43 JUNE 1, 1917 No. 3 THE ey N OF UREA FROM THE KIDNEY TO THE BLOOD : T. ADDIS anp A. E. SHEVKY From the Laboratory of the Medical Division of Stanford University Medical School, San Francisco Received for publication March 20, 1917 - During unsuccessful attempts to measure the rate of flow of blood through the kidney, we found that the concentration of urea may be higher in the renal vein than in the renal artery. The rate of flow of blood through the kidney can be calculated if the urea concentration in the blood of the renal artery and vein is determined over a period during which the rate of flow. of urine and the rate of urea excretion is imated. But we found that we could not obtain these data without adopting measures designed to prevent or lessen vasoconstriction of the renal arteries. For after tying off one kidney and exposing the other, the manipulations required for the removal of blood from the renal vein were always attended by a cessation of the flow of urine which appeared to be due to vasoconstriction, since we were able in some instances to observe that the renal artery grew smaller. Renal vein blood obtained under these conditions contained more urea than the blood entering the kidney. The experiments cited in this paper ex- tend and confirm this initial observation. This additional urea found in the renal vein must have had its origin in some accumulation of urea within the kidney. The kidney con- tains more urea than other organs and is an exception to the rule of the approximately even distribution of urea throughout the tissues and fluids of the body (1). The microchemical work of Leschke (2) and of Oliver (3) demonstrates that the reason for this high urea content is the special concentration of urea in two separate situations in the kidney, in the cortex within the cells of the proximal convoluted tubules 363 4b 364 T. ADDIS AND A. E. SHEVKY and in the medulla in the urine lying within the collecting tubules. It is not possible to determine which of these stores of urea is the source of the urea returned to the renal blood, nor which contributes the greater part if the returned urea comes from both. We found that the medullary portion of the kidney contained somewhat more urea than the cortex, but on the other hand the cellular store in the cortex is in direct contact with the blood, while the urinary accumulation in the medulla is separated from the blood by a layer of renal cells. But whatever the exact source of the urea added to the renal blood may be, the fact of its return from the kidney whenever the secretion of urine stops, throws some light on the mode of action of that force which enables the kidney to prepare a concentrated solution of urea such as the urine from a dilute solution of urea such as the blood. The accumulation of urea at certain locations within the kidney in much higher concentration than exists elsewhere in that organ must be accom- plished by a force which is able to annul or overrule the physical laws governing the diffusion of urea. Such a force might act through the medium of some physical or chemical configuration which would re- main passively operative even when the active functions of the kidney were in abeyance, just as the valve of a machine will continue to pre- vent the reflux of fluid when the power is shut off. But the immediate return of urea from the kidney to the blood indicates that this force has to be in continual operation to maintain the high urea concentration. When the kidney stops working this force relaxes its hold upon the heaped up urea, so that it again becomes subject to the laws of diffu- sion, and falls from the site of high concentration in the kidney to the lower levels of the blood, just as a weight held in the hand will fall to the ground when the grasp is relaxed. THE RELIABILITY OF THE METHOD USED FOR DETERMINING THE UREA CONTENT OF THE BLOOD Triplicate determinations were made on each blood except in a few instances in which only duplicates were obtained because of accident to one of the samples or failure to obtain enough blood. From such material an expression of the probable error in the meas- urements might have been obtained by finding the standard deviation for each set of triplicate or duplicate results, and multiplying the average of these standard deviations by 0.67. This procedure however is not only cumbersome but has been shown by Otis (4) to give a value. RETURN OF UREA FROM THE KIDNEY TO THE BLOOD 365 less than the true probable error. He has demonstrated that the me- dian of all the differences between triplicate or duplicate measurements divided by the square root of 2 gives the probable error of a single determination. The probable error of the average of three determi- nations is this probable error divided by the square root of 3. The differences between repeated determinations on a blood with a high urea content were no greater than those found when the urea content of the blood was low. All such differences are therefore di- rectly comparable. There were in all 168 differences. The median of this series was 0.9 mgm. Applying the formula we obtain a prob- able error of 0.64 mgm. for single determinations and 0.37 mgm. for the averages of three determinations such as are given in our tables. In other words, in half of our figures the quantity recorded, e.g., 100 mgm., might have been any value between 100.37 mgm. and 99.63 mgm. In the other half of our figures, the error is greater than 0.37 mgm. Mr. Otis pointed out to us that if the frequency distribution of errors may be assumed in this case to be ‘‘normal,” that is, in ac- cordance with the law of the distribution of errors, then theoretically the errors will be less than twice the probable error in 82 per cent of the cases, less than three times the probable error in 95 per cent, and less than four times the probable error in slightly over 99 per cent of cases. In only one case in a hundred, therefore, will the error reach 1.5 mgm. The urease method of Marshall was used, carried out in much the same way as is recommended by Van Slyke and Cullen. The quantity of blood taken for each determination was 1 ec. measured with an Ostwald pipette. Care was taken that the bloods whose urea content was to be compared were treated in an exactly similar manner as re- gards the amount of soy bean extract added, and the length of time of incubation and aeration. The acid was measured with an automatic pipette. In titration those refinements were used which were intro- duced by Barnett (5) in connection with his method for determining small quantities of ammonia. An increase in the urea content of blood from the renal vein at a time when the secretion of urine had stopped, was observed in a num- ber of the unsuccessful attempts to measure the rate of flow of blood through the kidney referred to at the commencement of this paper. Those figures are not given since the determinations were carried out on such small quantities of blood that the reliability of the method 366 T, ADDIS AND A. E. SHEVKY was considerably less than when the larger quantities available in the experiments cited here were used. + The animals used were rabbits. In some cases urea was given by stomach tube before operation in order to increase the urea content of the blood. THE DECREASE IN THE UREA CONTENT OF BLOOD FROM THE RENAL VEIN WHEN PRECAUTIONS ARE TAKEN TO DISTURB THE FUNCTION OF THE KIDNEY AS LITTLE AS POSSIBLE As soon as the animal was fully under the influence of ether, the left kidney was fully exposed through an incision in the flank, the renal vein cut with scissors and the blood collected in a vessel containing a little powdered oxalate. In two rabbits a direct comparison was made between the urea con- tent of the renal vein blood and the blood of the renal artery. Two ligatures were placed in position round the renal vein, the one next the vena cava tied, the swollen vein snipped with scissors, and after enough blood had collected the one next the hilus of the kidney was tied. Im- mediately afterwards the renal artery was cut. A decrease was found in the venous blood, 12 and 39 mgm., as against 15 and 41 mgm. in the arterial. The venous blood was slightly more concentrated as judged by the relative volumes occupied by red blood cells and plasma, but not to such a degree as appreciably to affect the urea content. Such direct comparisons between blood from the renal vein and artery have the disdavantage of involving a- cessation of the flow of blood through the kidney and therefore carry with them a. tendency to inter- ference with kidney function. Further, even such brief manipulations as are required in clamping or ligaturing the renal vein in two places are apt to induce a constriction of the renal artery. In our other experiments we have therefore taken the blood of the jugular vein as representing blood from the renal artery so far as its urea content is concerned. In a few experiments blood from the femoral artery was used. That this is justifiable is shown by the result of comparisons of the urea content of blood from the renal artery with the urea con- tent of blood from the jugular and carotid. Such differences as are recorded are not significant. (table 1.) In twelve rabbits the renal vein blood was compared with the jugu- lar. The jugular was first bared, the kidney quickly exposed and the RETURN OF UREA FROM THE KIDNEY TO THE BLOOD 367 renal vein snipped with scissors. Immediately thereafter, without waiting to tie-the renal vein, the jugular was cut. The whole pro- cedure did not take more than one or two minutes, and there was no mechanical interference with the circulation through the kidney. The results are given in table 2. TABLE 1 Comparison of the urea content of blood from the renal artery, jugular vein and carotid artery RENAL ARTERY JUGULAR VEIN CAROTID ARTERY mgm. per 106 cc. mgm. per 100 ce. mgm. per 106 ec. 40.47 40.00 21.00 22.37 24.50 24.47 83.35 83.90 19.57 19.74 24.75 25.47 TABLE 2 Comparison of the urea content of blood from the renal vein and the jugular vein when the blood was taken quickly RENAL VEIN JUGULAR VEIN LESS IN RENAL VEIN MORE IN RENAL VEIN mgm. per 100 cc. mie: per 100 cc. mgm. per 100 ce. mgm. per 100 cc. 47 57 10 27 27 180 189 9 97 102 5 22 23 1 42 53 12 184 196 12 212 223 11 197 196 1 38 38 139 : 149 10 211 216 5 There can be no question from these. figures that in some of these eases the kidney must still have continued to remove urea from the blood passing through it. In seven out of the twelve experiments the decrease in the urea content of the venous blood is considerably greater than could be accounted for on the basis of technical error. 368 T, ADDIS AND A. E. SHEVKY A DECREASE IN THE UREA CONTENT OF THE BLOOD OF THE LEFT RENAL VEIN WHEN THE BLOOD: IS TAKEN QUICKLY FOLLOWED BY AN INCREASE IN THE UREA CONTENT OF THE BLOOD OF THE RIGHT RENAL VEIN OBSERVED IN THE SAME ANIMAL AFTER A PERIOD OF OPERATIVE MANIPULATION DURING WHICH THE SECRETION OF URINE CEASED In seven rabbits, blood was obtained quickly in the manner already described from the left renal and jugular veins. The vessels of the left kidney were then clamped and ligatured. The bladder was opened and a catheter inserted into the right ureter. This was done in order to make sure that the secretion of urine had stopped. We did not TABLE 3 Comparison of the urea content of blood from the left renal vein and the jugular vein when the blood was taken quickly and comparison of the urea content of blood from the right renal vein and the jugular vein when the secretion of urine had stopped BLOOD TAKEN WHEN THE SECRETION B00? Thee eee OF URINE HAD STOPPED RABBIT NO. Left Jugular Difference Right Jugular Difference renal vein vein renal vein vein More Less More Less mgm. per | mgm. per | mgm. per | mgm. per | mgm. per | mgm. per | mgm. per | mgm. per 100 cc. 100 ce. 1€0 cc. 100 ce. 100 cc. .| 100cc. 100 ce. 100 cc. 11 169 178 9 194 188 15 88 93 5 100 100 38 92 96 4 105 105 39 180 187 7 201 189 12 40 130 129 1 136 131 5 41 130 140 10 145 139 6 42 7 78 1 87 87 obtain urine from the catheter in any of these animals. The right kidney was then exposed and blood collected from the right renal vein and immediately afterwards from the jugular. In-a few cases the comparison was made with blood from the femoral artery, as enough blood was not obtainable from the jugular. The urea content of the renal vein could thus be compared with that of blood corresponding in urea content to the blood sent to the kidney in the renal artery under two conditions, first at a time and under cir- cumstances in which the kidney might still be functioning, and sec- ondly at a time when we had evidence that kidney function had ceased. The results are given in table 3. RETURN OF UREA FROM THE KIDNEY TO THE BLOOD 369 It will be noted that there is in general a decrease in the urea con- tent of the renal vein when the blood is taken quickly, and an increase when the blood is taken at a time when no urine is being secreted. ‘That there should be considerable variation in the amount of the de- crease or increase is to be expected, since neither the degree of kidney activity nor the time at which that activity ceased was known or exactly controlled. But that there should. be in any of these cases a clear decrease in the urea content of the renal vein blood can only be ac- counted for by the passage of urea from the blood into the kidney. — ‘Similarly a definite increase under other conditions is only to be ex- plained by the return of urea from the kidney to the blood. SUMMARY When blood is taken from the renal vein so as to disturb the function of the kidney as little as possible, it usually contains less urea than the blood sent to the kidney in the renal artery. This is explained by the passage of urea from the blood to the kidney. When blood is taken from the renal -vein at a time when the secre- tion of urine has stopped, it usually contains more urea than the blood sent to the kidney in the renal artery. This is explained by the passage of urea from the kidney to the blood. The return of urea from the kidney to the blood whenever the kidney ceases to secrete urine is taken as indicating that the force or forces which store urea in high concentration within the kidney must remain in continued operation to hold that accumulated urea, and do not act through the medium of any physical or chemical mechanism which would passively maintain its hold upon the urea even when the active ~ concentration of new urea from the blood had ceased. BIBLIOGRAPHY (1) Marswatt Anp Davis: Journ. Biol. Chem., 1914, xviii, 53. (2) Luscuxe: Zeitschr. f. klin. Med., 1914, Ixxxi, 14. (3) Orrver: Journ. Exper. Med., 1916, xxiii, 301. (4) Oris: (Stanford University). To be published in the near future. (5) Barner: Journ. Biol. Chem., 1917, xxix, 459. ADDENDUM Some time after we had sent our manuscript for publication a paper by Cushny (Journ. Physiol., 1917, li, 36) reached us, which contains data demonstrating in another way the return of urea from the kidney to the blood after the activity 370 T. ADDIS AND A. E. SHEVKY of the organ ceases. One kidney was removed while urine was being secreted, and its urea content per gram of tissue determined. At the same time the cord was cut in the cervical region so that a pronounced drop in blood pressure was produced and urine secretion stopped. After an interval of one to one and one- half hours the remaining kidney was removed. It was found to contain less urea than the kidney removed while still active. A COMPARISON OF THE EFFECTS OF BREAKFAST, OF NO BREAKFAST AND OF CAFFEINE ON WORK IN AN ATHLETE AND A NON-ATHLETE I. H. HYDE, C. B. ROOT ann H. CURL From the Physiological Laboratory of the University of Kansas Received for publication March 22, 1917 INTRODUCTION - The results pertaining to this investigation were obtained in 1912 and 1913, and are reported under two sections. The first section deals with experiments secured with a modified Lombard type of ergo- graph; the second section deals with tests obtained with an ergometer. _ The experiments were conducted on two men in perfect health but of very different physical training. ~ Subject “‘B,” the athlete, was 5 feet, 8 inches in height, weighed 196 pounds, was 29 years of age and an instructor in physical education. During the two years preceding these tests, in addition to his duties, he had been accustomed to take daily exercise, especially of the arms, chest and back muscles. Subject “A,” the non-athlete, was 5 feet in height, weighed 140 pounds and was 26 years of age. Before beginning these tests he had taken no special physical exercise. - The object of the experiment was to compare the pulse rate, blood pressure, ergographic and ergometer work in both men under the fol- lowing conditions: of certain doses of caffeine without breakfast, break- fast without caffeine, neither breakfast nor caffeine, and of different intervals of time following the partaking of caffeine or breakfast. The literature bearing directly on this problem is limited. The references will be confined to reports that may aid to a better under- standing of the subject. The ergographic work, calculated from the eyclometer that recorded the sum of the heights of the contractions multiplied by the weight lifted, was performed by the flexor muscles of the second finger of the right hand, lifting a weight of 5 kgm. every two seconds. To prevent the neighboring flexors from participating, the first and third fingers were placed in an adapted clamp, and the 371 372 I. H. HYDE, C. B. ROOT AND H. CURL hand held securely in pronation to the ergographic support by means of a narrow non-elastic bandage, placed back of the metacarpal-phalan- geal joints. With this arrangement the subjects were able to work more than an hour without discomfort or interference with the circulation. The experiments were performed in the morning between 8 and 10 o’clock, and under similar conditions. Breakfast at 7.15 a.m., for each man consisted of one soft boiled egg, 2 ounces wheat bread and 2 ounce of butter. For the experiments with caffeine, no breakfast was taken, instead 7 ounces of coca-cola,! containing a total of 1.42 grains of caffeine, being the average amount in a strong cup of coffee or of tea. The experiments with caffeine alternated with those obtained either with or without breakfast. The athlete did not drink coffee and the non-athlete had taken it for breakfast, but several weeks be- fore and during the time that the experiments were in progress, of course, both of the subjects gave up the use of articles of diet that had caffeine in them. Records were kept of the hours and conditions of sleep, of the pulse and systolic blood pressure on rising, before and after exercise, and the intervals elapsing between exercise and partaking of food or of caffeine. The pulse and blood pressure were secured with a Tycos sphygmo- manometer while the subject was seated. SECTION I. ERGOGRAPHIC WORK WITH FLEXOR MUSCLES Preliminary to the main experiments, the flexor muscles in both sub- jects were daily exercised on the ergograph under like conditions. After six weeks of training the results became more constant, and for practical purposes the muscles were considered in training. Before that time, therefore, the muscles were considered untrained. Part 1 deals with the average results of the untrained; part 2, with those of the more trained flexors. Part 1. For purposes of comparison, only the average of all the results observed were tabulated in table 1. A study of this table re- veals that at the beginning and after eating breakfast, ‘“‘A” could contract the untrained flexors two hundred and seventy-nine times until utterly fatigued in nine minutes and eighteen seconds, doing 12.65 1 Analysis of coca-cola syrup: Sugar 53 per cent, caffeine 1.42 per cent, water 44 per cent, citric and phosphoric glycerine and alcohol qualitative test. One ounce of syrup equals about 7 fluid ounces of coca-cola. EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 373 kgm. of work; but by the end of the month he actually trebled these figures.On the other hand, it is seen that from the first “B” did one and one-half times as much work: 18.75 kgm. of work in eight minutes, fifty-five seconds, and in less time than “A” did his, and that he more than trebled his power for work during the month’s training. Now instead of doing one and one-half times as much, he did one and two-thirds times as much as did ‘‘A,” in one and one-fifth the time. TABLE 1 Average results of experiments with untrained flexors on the ergograph. (a) The first and twenty-fourth test. (b) Average of eight experiments after eating break- fast. (c) Without breakfast. (d) After taking caffeine. DATE 1912 CONDITION SUBJECT DURATIONOF WORK IN MINUTES NUMBER OF CONTRAC- LIFTED IN CENTIMET- 8 ¢ | ZERe| ¥ i] 2 = fe) = e (a) | November 19 | Breakfast A 9’ 18” 279 253 | 12.65 December 20 | Breakfast A 26’ 54” 798 726 | 36.30 (a) | November 19 | Breakfast B 8’ 55” 273 375 | 18.75 December 20 | Breakfast B 30’ 18” 918 | 1,205 | 60.25 (b) Breakfast A 14’ 17” 402 377 | 18.85 (b) Breakfast B |15'24"| 462 733 | 36.65 (c) No breakfast A 12’ 00 360 357 | 17.75 (c) No breakfast B 15’ 03” 398 665 | 35.25 (d) Caffeine A 25’ 18 759 939 | 46.95 (d) Caffeine B 38’ 43”| 1,154 1,421 | 71.05 The average hours sleep for both subjects = 73. The lapse of time between breakfast and caffeine and exercise = 1 hr. The average room temperature = 59° F. The number of contractions until fatigued = until unable to lift the 5 kgm. weight. These experiments were begun November 19 and ended December 20, 1912. - Comparing the amount of work done, without breakfast, one hour after breakfast, and also after a dose of caffeine, it was noticed that both subjects did almost as much without, and in practically the same time, as after having eaten breakfast. ‘‘B’’ moreover, felt better than when working after having eaten the meal. _ When both subjects took a dose of 1.42 grains of caffeine, “A” was able to do 46.97 kgm. of work which was two and one-half times as much work, and “B” 71.05 kgm., which was twice as much work 374 I. H. HYDE, C. B. ROOT AND H. CURL as when working one hour after eating breakfast. The after effect for both, but more observable in “‘B,’”’ was a heightened sense of irri- tability and weariness not noticed except when working after taking caffeine. The reason that the same dose of caffeine stimulated the working power of the athlete less than it did the non-athlete was prob- ably because the athlete is the larger and heavier man. The dose per kilo weight was less, therefore, for him than it was for ‘A.” A consideration of this set of experiments demonstrates that the flexors and probably other untrained muscles in an athlete, are more efficient and more readily trained, than are the untrained muscles in a non-athlete. Part 2. Ergographic results with flecor muscles in training. After six weeks of training the muscles, the results became more constant, and the second set of experiments was begun, and continued for two months. The object of this set of experiments was to compare the effects of no breakfast, of breakfast one hour and one and one-half hours before work. The average results are tabulated in table 2. It is there shown that when working without having eaten breakfast, ““B,” the athlete, continued as before (table 1) to do twice as much work in one and one-half the time, as did ‘‘A.’”’ But when working one hour or one and one-half hours, after having eaten breakfast, al- though both of the men increased their power for work enormously, they did more after eating than when not eating breakfast, and more in one hour, and in less time, than in one and one-half hours after eat- ing breakfast. Nevertheless “‘B’”’ no longer did twice as much work as “A”? but only one and one-half as much. It may be that the greater increase in power in ‘‘ A” is due to the fact that he attained his maxi- mum power more gradually than “B’’ did his. The ergographic work had practically no effect either on the pulse or blood pressure. The normal blood pressure was much higher in ‘B” than it was in “A” but there was little difference in their pulse rate. It appears, and this agrees with Lombard’s (1) results, that exercise of this character can be performed better one hour than one and one-half hours after eating breakfast. SECTION II The influence of neither breakfast nor caffeine, breakfast without caf- feine, and caffeine on ergometer muscle work in a trained athlete and a — non-athlete. The ergometer consists of an adjustable grip bar, con- nected through two pulleys to a weight of 25 kgm. The recorder, EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 375 similar to the one attached to the ergograph consists of an endless tape, 1.5 em. wide and 500-cm. in length that passes around two pulleys, each 50 cm. in circumference. A cyclometer attached to one of the pulley records, therefore, 50 cm. for each revolution. These are noted, and thus the whole height of contraction can be directly ascertained, and the work in kilograms determined. In conducting the experiment, the subject stands on a line, a definite distance from the bar. The bar has been properly adjusted to the height of the subject, so that with arms fully extended he is able to grip it firm- TABLE 2 Average results of fleror muscles in training on ergograph, no breakfast (Ai Bi) one hour (Az Bz), and one and one-half hours (A; Bs) after eating breakfast a 2 vi = z } i i aoe pn EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 381 with greater force, and work longer before fatigued than they could at that interyal-after taking either of the other doses. “B” did 5429 kgm. work in nine minutes, fifty-four seconds, lifting the 25 kgm. one hundred and ninety-eight times. “A” did 2680 kgm. work in five minutes, fourteen seconds, lifting the 25 kgm. one hundred and four times. With the strongest dose of 3.58 grains, both subjects did less work than they did with the medium dose. In fact “A” did even less after taking the strongest and ‘“‘B” only one-seventh more than after taking the weakest dose of 1.42 grain. It is evident that the strongest dose proved more depressing to both men than did the medium, and far more so to “A,” the lighter weight man, than to “B” the heavier weight man. TABLE 7 Average effect of different doses of caffeine on work Mere isl, ia]: (8 |elele 2 = - 3 & = z Bs 5 a ° < = | a e a g Zz Pa E.| a meee; - Fe} 8 ee |e | 8 | 8 | BH | oy & & ° & 4 z 4 > gO m wel & pecpeme |e. |S “2 18,18] & | ge] 8 E = & a z 2 2 g 5 z P a q = a = a aa Deemer se | & |&&| # | 42 | g8 | 861] 8 | § | ee] se a ee a | & E : hk peeopa °F | grains A 12| 66 | 1.42} 30’ |3’ 33") 71] 8,511 | 2,138] 72] 110 | 118 | 138 B 10 | 66 | 1.42 | 30’ |4’ 53”| 97 | 13,906 | 3,476) 72} 112 | 142} 160 A 8 | 70 | 2.24} 30’ |5’ 14”| 104] 10,700 | 2,680) 69 | 119 | 125 | 137 B 8 | 70 | 2.24 | 30’ |9’ 54”| 198 | 21,716 | 5,429) 71 | 116} 145 | 165 A 6} 70} 3.58 | 30’ |3’ 27"| 69 7,109 | 1,777; 71 | 118 | 124 | 131 Bin: 3} 70 | 3.58 | 30’ |7’ 20”, 147 | 15,728 | 3,932) 78 | 116 | 136 | 156 The strongest dose of 3.58 grains greatly accelerated the pulse rate, but depressed the blood pressure below the normal level after work in “A,” while it had only a little more influence on the pulse and practi- cally no more on the blood pressure in “‘B’’ than the medium dose had on these activities. | The effect of caffeine when the dose is taken per kilo body weight. Two sets of experiments were undertaken for the purpose of ascertaining the effects of caffeine thirty minutes after each subject received equal amounts of caffeine per kilo body weight. In the first set of experi- ments “A,” who weighed 66.6 kilos, took 1.42 grains, and “B,’’ whose weight was 93.3 kilos took 2.24 grains of the drug. Each subject was then receiving practically 0.21 grain per 9.3 kilo of his weight. In the 382 I. H. HYDE, C. B. ROOT AND H. CURL second tests, subject ‘A’ took 2.24 grains and “B” 3.58 grains. The results obtained from these experiments, and which are summarized in table 8 add another viewpoint to the facts obtained from the experi- ments that dealt with equal doses of caffeine for each subject, without respect to thei weight. At the thirty minute interval after each subject received the weaker dose of 0.21 grain of caffeine per kilo of his body weight, “‘B’”’ was able to do two and one-half times as much work, and work almost three times as long before becoming fatigued, as was “‘A.” His heart rate at the same time was only seven more beats per minute than was that in TABLE 8 Effect of caffeine when given per kilo body weight ee ee ee 4 ae % b i ee a re 5 = ze a 3 5 -| a Q p SI ro) =) 9 = fa a 5 E i - (2) a Q . z < As ° ag Zi g |. = 5 5 fe = ot ge ee rd & & « . 8 z | Bu ° By & fo) ° & nw 2 a @ Qe o|.s ) a z ° ng z a ee | BO e |86| & ai cee) E ap oe Ag | ® a mB a |mel « : af & mB a E 4 PB a | ge | Se = gs| 2 e £9 7 5° oa fe a 4 oo? E D pe } = < S) pr Q° oo 5 Bp S= | os D Z mR < 4 =) m E Pe Aa a Q 1= 0.2 grains of caffeine per 9.3 kilo body weight = A 1.42 grains, B 2. grains. 2= 0.2 grains of caffeine per 5.9 kilo body weight = A 2.24 grains, B 3. grains. S « grains 1 A} 12] 66 | 1.42 | 30’ |3’ 33") 711} 8,511 | 2,138) 72 | 110 | 118 | 138 B) 8| 70 | 2.24 | 30’ |9’ 54”| 198 | 21,716 | 5,429} 71 | 116 | 145 | 165 B} 3] 70 | 3.58 | 30’ |7’ 20” 147 | 15,728 | 3,932} 78 | 116 | 136 | 156 Bt: 8 | 70 | 2.24 | 30’ |5’ 14”| 104 | 10,700 | 2,680} 69 | 119 | 125 | 137 These experiments were begun March 28 and ended June 6, 1913. “‘A,” while the increase in blood pressure above the normal level was the same in both. Thirty minutes after each subject received the stronger dose of 0.21 grain per 5.9 kilo of their body weight, ‘‘A”’ did one-fourth more work, and worked two-thirds longer before becoming fatigued than he did with the weaker dose. . ‘‘B,’’ however, was able to do only about two-thirds as much work, and work about two-thirds as long as he could with the weaker dose. Therefore ‘“‘B’’ now did only one and one-half as much work and worked only one and two- thirds as long as ‘‘A’”’ before becoming fatigued. His pulse rate was EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 383 | accelerated thirty-eight and “A’s” fifty beats per minute. On the other hand his blood pressure was 20, and “A’s” only 12 mm. Hg above the normal level. . Comparing the results obtained from these two doses of caffeine, the fact is brought ou that in “B,” the heavier subject, the weaker | dose of 0.21 grain per 9.3 kilo body weight was more stimulating to the muscular and cardiac activity than was the stronger dose of 0.21 grain per 5.9 kilo per body weight. On the other hand in “A” the stronger dose of 0.21 grain per 5.9 kilo body weight proved the more stimulating. Therefore we conclude that doses given per kilo weight exert for each individual a specific effect, and that at the same interval after taking the drug the effect may be different for different individuals. The after-effect of caffeine. It became of interest to ascertain whether caffeine exerted a prolonged influence on the system. For this purpose twelve experiments were performed consisting of three sets of four each; with 1.42 grain, 2.84 grains caffeine, and control tests respectively. For the control tests no breakfast was eaten. Each experiment in- volved five observations on the effect of ergometer work, namely, at 8.20 and 9.20 a. m., and at 4.20 and 8.20 p.m., and after twenty-four hours at 8.20 a.m. Between these intervals the subjects rested or did light routine work. The average results for the control tests are re-- corded in table 9. They show first, that when working without break- fast, both of the subjects worked longer, with greater force, pulse rate and blood pressure the first period, than they did one hour later, indi- cating that they had not fully recovered from the previous hour’s fatigue. Second, that both subjects worked longer, did more work, and their pulse and blood pressure increased more after the 4.20 and 8.20 p.m. periods; that is, four and two hours following the meal, than they did the first hour in the morning. This indicates that they had recovered from the fatigue of the previous work and were benefitted by the preceding meal. Third, repeating the tests the following morning at 8.20 it was observed that practically all the conditions and data were the same as they were the previous morning. Fourth, “B” did one and one-third more work, had a higher pulse and blood pressure than “A” at the beginning of work, but at 8.20 p.m. his power had lessened, although his pulse and blood pressure had not. Since at this period “A” did his best work, it happened that at this hour his power for work practically equaled that of “B.” A study of the data secured on the duration of the effect of 1.42 grain of caffeine, brings out the interesting results that now both I. H. HYDE, C. B. ROOT AND H. CURL 384 POT | ZI | OTT zs |¢-sor'z |re96 | ss [cz +) udozst| 8 § SBS ae oeieg oh ‘urd 0g ZI 3% qounyT ee ae es CASP 0 30M OUTNOT || losqu0g GST | OT | 66 ¥9 |O'€89T |Z8L9 89 aSG ,§|B 0S 6] iT Z YB FIOM > q] uvseq ‘“inoy T peyseyy SFI | 8ZI | SOT 92 |9°$22'% |F06‘'S os 480 ,F) U8 0Z'8 t cect Bepesy ata ea ey ye FIOM i uvZeg ‘ysByyvorq ON |) OZT | SOL | OOT 92 |2'FH9'T |92¢°9 89 a¥% ,€| B® 0S 8] FZ G rt HIOM UBB : “qsuyyeoiq ON ‘doojs ; simoy 2) Suru10ul yxXoN SI 9II | 80T 92 |2°88%'% |ee6'8 | 06 |.08 4) 02'8| ZI | ae Wetter yIom uvsog ‘BuIpvol yysrT “0g'9 | qe rouurqg “urd 9 [[19 é Aep jO YIOM oUTNOYy joryu0y Off | 9IT | ZIT | 92 |0'996'T |FOs'z | 82 [eo ,e/ td 0Z'F| 8 TO IS yom ues By ; | P -og ‘ABp Jo YOM our} ; “noy ‘OE cl 7% YounyT OZT | FIL | 26 ZL |¢ LOFT jogg’¢ 6¢ aLG ,G) UB’ OZ’ 6) iT “esta 4 7 a i ae rea 9 aa yom uvsogq ‘Inoy [ peysexy OZI | SOT | OOT 92 |L'PPO'T l6lo’9 | 89 |,8% ,g|/rd 028 1 igi) a ch 6 ire Stir ta YIOM uvdogq ‘ysvyyvorq ON J sunoy | surpvi6 y oy ~ ss F:| hy Z iS 3 o y ° 2 > oe oe s q q Zz HO af ad q Z 8 4 8 = 18/8/88 | 36 | seh | ge] GB] ge | & | ee] F | gs : Dw al A * Aw oS 3 ie] =z 4 Sn AS re] y S ° - 3 22 Be 8 od oa Se ) z Z 9 4 NOILIGNO sh lag) C8 ee) Gee eee g |e Ese ; Aa | oe " Z BS Q , z & 5 ° & Ag ‘a S z mR Z 1 al sunoy 4nof-fiqzuam 07 auo sazfo auraffvo fo woafq 6 WIdVL 385 EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 2ST Zot b9T cel 98T OFT vel S&T S&T 8éI aa OIT OI OI OI 82T 9IT Ott sor 90T 601 92 92 GL GL v8 GL vl Z'LS1'% L 801% L°t28'% 1° 882% ¢°88o'Z 1° 90L'T 0’ P82 '% 0°GL2'% SLI ‘ZI 620‘Z1 629'8 cIF's 201 raat 8L 18 al 98 raul «0G SE a9$ AU aL0 a8 «80 0G #80 40% W ‘urd 0%." F “Ur"® 06 “UW'B 0Z°8 ‘urd 0%°8 ‘urd 0Z°8 ‘und 0% 'F “Ul'’ 0Z'8 ‘ur’ 0Z'°8 “u'8 02" 6|: £% II VG at GPT 4 a | a i a ee . “48 YOM ‘ard og Zr 98 youny - Aep JO YIOM OUTPNOYY sce Naw elwieieg a" > 1048] son -UIU .0Z YOM UBZOg ‘OUIOYVO SUTBIZ ZH T HOOT = 6 TIN poqsoy PP EAP et ete tare 4@ YOM uBsog “4ysBpyvorq ON ““YIOM “4S8BpyvOIG OU Burusi0W 4xoN “days Sinoy J/BYy-9u0 puBw U9aAVG ““yIOM “Burpvor yy 3] ‘gouurq ‘we'd g j1yun Avp jo YOM ourynoyy “und Og ZT #8 youny ‘Aep jo yom ourynoyy Sethald eke. 0 whe 108] soqn -uIU 0Z YIoM uvsog ‘OUIOIJVO SUIBID ZH] FOOL, «6 [Hun popsey seeetesesssagpryBaIq ON “""MIOM “4sBpywerq ou Surui0ur 4xeN ‘doojs sinoq = 4yarq sarees + = Fett Seeed tf Re rouulq “wed g [yun Aep jo yom ourynoyy ~ , eee aT; eee > 4 atl eT JoryU0Kg QI] I. H. HYDE, C. B. ROOT AND H. CURL 386 | : 8ST | tet 09 |¢'2e8‘E [OSes | OST |,0€ ,2/"m'e CG's] Fz ¢ | yeyIOM “ysBjyeorq |) 8% ON “SuIUIOUL 4XON S91 OFT OZI | 88 |e’ FEe"e [sets | S6I [ee ,6|"He CF'g #8'Z | 1 | YOM uvseg -ouToy | -J89 SUIvIZ FQ°Z YOO} { ail / CTS dV “ysByyBoIq ON ShT eT 9OL | 82 [0198'S jhrP'TT | 2OT [Se S7me oo) Ff fp ySBJyYBVoIq ON | | cel = PIT «| «OTT =| «(08 «9 €22'% 1706'S | 8 [ET .F/ re O's] FZ ¢ YB YOM “YsVpyvoq = |) ou ‘Surmi0ul 4xeN OFL OZt OGL | 89 jo TSS [ORIOL | 26 |.88 HH eOe Ss! F | PWS) TL po W'S g§ 78 | OUIOVO SUIBIS FQ°Z vul YOoT, “yweyyvoiq ON SEE | SIT | SOF ZL |¢°998'T |99F'2 | €8 {OT ,F/ He 0O'Z sr essss ss ageryBarqg ON |) trT | 8ZI = SOT S| «92 «JO FFs'S 19268 | 29 |,02 .F/ E028] EZ ere, ea qsvyyvorq ou Sut |) -UIOUI JXONV “YOM ‘desis sunoy 4ysIy Sot OFL «Oct | O8 j0°¢6z'% j081'6 | 2 [ee Fr urd OZ'8| ITI fp [9B YOM “Burpvor = | > qysarT ‘urd 0¢'9 qouutq «= ‘urd 9 [1 Aep jo yIoM ourqnoy | | dll sunoy | suns6 WUOM uaLiIv asiond MuUOM auOdaad aASTAd ugdaWoan MUOM asoa Loaraas fo NOILVuUod NOILIGNOD asod wal MUOM YALAV auassaud dood MUOM DAUNOAGTA aunossaud dood SNVUSO TIA NI @NOG WHOM SUALAWIINGD NI GaLdIT LHOIGH SNOILOVUL -NOOD dO -~AV AWIL dO @SdvVT MuOM JO uaaUuoO MUOM NVDAE AWLL panuyuoj—bh ATAVL EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 387 subjects did more work the second period than they did the first; that is, one hour after working without breakfast and twenty minutes after taking caffeine, and there was no indication of fatigue due to the pre- vious hour’s work but rather signs of stimulation, or inhibition of fatigue. Their pulse rate was much greater, but their pressure was little changed. Moreover, even after twenty-four hours, they still did more work than they did at the same hour of the previous day, and their pressure and pulse rate were no higher than at that time. As without, so now with caffeine for the periods chosen for observation “A” displayed his greatest power for work at 8.20 p.m. He did now one and one half times as much work as was possible without the drug, and what is remarkable, did at least as much work at that particular time as ‘‘B” could perform. Concomitant with the great amount of work on “A’s” part, there was an enormous acceleration of fifty-eight heart beats per minute and an increase of 26 mm. Hg in blood pressure which was a fall of 8 mm. Hg from the increase in pressure due to work without breakfast. “B” now did his best work, as before, without eating breakfast with the 1.42 grain of caffeine at 4.20 p.m., or seven hours after taking the dose and four hours after luncheon. He did more work than at any of the other periods tested. His heart rate increased forty beats per minute, but his blood pressure increased only 14 mm. Hg which was much less and his heart beats considerably more than when work- ing without eating breakfast. With the stronger dose of 2.84 grains of caffeine, ‘‘A” did one and one-third as much work as was possible without the drug, but did no more than he was able to do twenty minutes after taking the weaker dose of 1.42 grain. ‘‘B” did twice as much as he did without the drug, almost twice as much as he did at his best with the weaker dose, and twice as much as “A” was able to do with the stronger dose. ‘B’s” pulse and blood pressure were increased, the former more so than with the weaker dose, while ‘“‘ A’s” pulse rate fell slightly after taking the drug but was almost doubled after the work, and his blood pressure showed less rise than it did with the weaker dose. After twenty-four hours, however, both subjects did much more work than they did at their best without eating breakfast, showing that the after effect of the strong dose not only persisted twenty-four hours, but that it still had a very powerful effect that would probably continue considerably longer. The strong dose, after twenty-four hours, caused a great depression of ‘‘B’s’”’ pulse rate far below normal 388 I. H. HYDE, C. B. ROOT AND H. CURL and an increased rate above normal in “‘A’s’”. In both subjects the pulse was greatly accelerated by the work, indicating a heightened irritability to stimuli produced by the work at that time. Therefore the stimulating influence of the caffeine persisted at least twenty-four hours after the doses were taken. There seems also to be an optimum period when the effects were more pronounced, and a maximum dose limit beyond which the power for muscular work is lowered. This period and limit varies in different individuals. When these investigations were undertaken, ‘‘B’’ complained of weakness of his eyes, and on November 3, 1912, was fitted with glasses. On June 3, 1913, during the time he had been experimenting with the larger doses of caffeine and after he had taken one of the strongest doses, the external rectus muscle of his left eye became paralysed. This may have been a weak muscle and readily affected by the drug. The paralysis of the muscle incapacitated him for further work. At about this time “A’’ also complained of being very irritable and feared that the caffeine was detrimental to his health. It seemed to him that he could work indefinitely without fatigue ‘yet his muscles failed to contract and his heart beat at a tremendous rate. ’ "The experiments were therefore discontinued and the study of the after effects of caffeine, which had been planned and just begun, had to be abruptly abandoned. ‘The indications were that there were after ef- fects that interfered with efficiency of physical and mental activities. ‘‘B’s”’ friends, among them two physicians, charge it to the influence of the caffeine that ‘‘B”’ failed in an athletic exhibition in which he took part in the spring dur- ing the time that he was conducting the caffeine experiments. Before he began the experiments he had trained himself so that he was able to hit the punching bag with his head, feet and hands alternately on its rebound. It required speed, accuracy and control of muscles, and concentration of thought. He had become an expert in this feat. But his power of concentration, accuracy and precision in his muscles had been greatly impaired so that he was unable to repeat the athletic demonstration with any credit during the time he was taking the strong doses of caffeine. DISCUSSION In the two subjects on which the tests were conducted, the pulse rate, blood pressure and power and duration of muscular contraction, in the trained as well as the untrained muscles, were more or less dif- ferently affected by the following conditions: the same doses of caf- feine as well as the doses of caffeine per kilo body weight; no break- fast; breakfast without caffeine; and the interval following either breakfast or a dose of caffeine. Lombard (1), by repeating his ergographic tests several times daily, quadrupled the power of his untrained flexor in twenty-two days. Food increased his power for muscular contraction, and he accomplished EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 389 more work one hour after than one and one-half hour after a meal. He found thatthe usual amount of coffee did not affect his power for muscular contraction. The results obtained by the athlete and the non-athlete corroborate those found by Lombard, but do not agree with his conclusions in reference to the effect of the usual amount of a cup of coffee. The difference may possibly be due to the fact that the effects of caffeine as a rule do not disappear in twenty-four hours, and therefore drinking this daily would keep the subject constantly under its influence and he could not compare its effect with those obtained with no caffeine. The athlete had seldom in his life, and the non-athlete only moderately, partaken of coffee, which facts may explain why the athlete was affected more strongly and more rapidly by the drug, than was the non-athlete or Lombard. The athlete felt generally better when work- ing without breakfast, and both subjects did more ergometer work two and one-half hours after than one hour after eating. The pulse rate was much more accelerated in the athlete, in fact often more than double that of the non-athlete, and this may account for the greater fatigue of the athlete from work after eating breakfast. The conclusion of previous workers that an optimum dose of caffeine increases the capacity for muscular work and inhibits the sense of fatigue, and that a larger dose decreases the power for muscular contraction was confirmed.. It was also found that an optimum dose of caffeine may double the capacity for work over that produced at any time after eating breakfast; that it inhibits the sense of fatigue for many hours after the most exhaustive work with the ergometer; and that when the optimum dose, which was different per kilo weight in the two subjects was increased the working power and blood pressure were decreased, the heart rate accelerated, and with a very large dose was inhibited. Rivers and Webber’s (2) results are in harmony with these. They found that 1 to 3 grains of caffeine stimulate and 4 to 6 ‘retard the speed of typewriting, and that the relation between blood pressure and pulse rate is not constant, and that strong doses accelerate the pulse but depress the blood presgure. These results, according to Sollmann and Pilcher (3) and other investigators, are held to be due to the stimulating effect of the optimum dose on the muscle and cardiac tissue, and the depressing effect on the nervous system that controls the sense of fatigue. The large doses stimulate the muscle and cardiac tissue more, and if increased above a certain amount produce a pro- gressive depression of these tissues. The large dose also inhibits the peripheral vasomotor mechanism, causing dilation of the blood vessels, 390 I. H. HYDE, C. B. ROOT AND H. CURL and at the same time stimulates the vasomotor center. These two effects neutralize each other more or less, and thus produce changes in the blood pressure that are proportional to the extent of neutraliza- tion. Sollmann and Pilcher (3) also observed that large doses may be fol- lowed by paralytic phenomena, causing fatigue and depression of mus- cular contractions. It was interesting to find that the effect of the same dose varies con- siderably with the interval of time after the drug is taken. In the athlete, the effect of the weak dose of 1.42 grain of caffeine gradually increased from twenty to forty-five minutes, and in the non-athlete from twenty minutes to three hours after the drug was taken—that is, this dose was only five-eighths as strong per kilo body weight for the athlete as it was for the non-athlete, but had its maximum effect in the athlete in two-eighths of the time that it did in the non-athlete. It was of shorter duration, less stimulating to muscular contraction and less ef- _ fective in inhibiting fatigue; but it had a greater stimulating effect on the pulse rate of the athlete, and but little more on his blood pressure forty-five minutes after it was taken, than it did at the same interval of time on the non-athlete. Therefore, for each indivdual, different doses of caffeine may exert their optimum influence at definite intervals after the dose is administered. When taken per kilo body weight, the effect of doses of caffeine taken at definite intervals before beginning work was different in the two subjects. Of the two doses, 0.2 grain per 9.3 kilo body weight, and 0.2 grain per 5.9 kilo body weight, and thirty minutes after the drug was taken, the weaker dose proved more stimulating to the work- ing power of the athlete, and the stronger dose to that of the non- athlete. At the same time, the pulse rate was enormously increased in both subjects, while the increase in blood pressure was the same in the athlete as with the stronger dose, and less in the non-athlete than with the weaker dose. It was shown throughout the experiments that the athlete was more sensitive to the effects of the caffeine than was the non-athlete, and therefore, a,dose that would prove stimulating to the athlete’s power of muscular contraction and heart action might prove less so for these functions in the non-athlete at that particular time. For comparative work it therefore seems advisable to study the ef- fects of the dose irrespective of body weight as well as of definite doses taken per kilo body weight of the subject. EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 391] . Another factor that deserves consideration is that the observations ought to be made at certain periods of the day. Lombard and other investigators found that there were diurnal variations that had an in- fluence on the power for work and recovery from fatigue. Lombard observed that his power for muscular contractions was greater from 5.30 to 6.30 p.m. than from 3.30 to 4.30 p.m., and greater at 4.30 p.m. than 11.30 a.m. Maggiora (4) held that recovery from fatigue was more rapid before 10 a.m. than at 11 a.m. These observations are strengthened by those obtained in studying the influence of caffeine and no breakfast at definite periods of the day. It was found that without eating breakfast, and also with the weak dose of caffeine, the power for muscular contractions in the athlete was greater at 4.20 p-m. than at 8.30 a.m. or 8.20 p.m. and in the non-athlete under the ‘same conditions at 8.20 p.m. than at any of the other periods of the day tested. In both subjects, moreover, the effects of the caffeine con- tinued at least twenty-four hours after the drug was taken. In Hol- ° . lingsworth’s (5) subject the stimulating effect of caffeine was notice- able even after three days. Kraeplin also observed that strong doses of caffeine retarded the transformation of intellectual conceptions into actual movements. These after effects of caffeine are explained by Cushny (6) as the results of stimulation of the central nervous system that is associated with psychical activity. After a strong dose of the drug connected thought is rendered more difficult and impressions fol- low each other so rapidly that attention is destroyed, and it requires more effort to limit it to a single object. These views would explain why the athlete, during the time he was taking strong doses of caffeine, was unable to perform the athletic feats in which he had excelled be- fore conducting these experiments. SUMMARY The general results of the experiments recorded in this paper are as follows: The working power of the untrained flexor muscles in a trained ath- lete may be increased at least three and one-half times, and the same muscle in a non-athlete three times in one month of daily training. From the first the athlete did one and one-half, in less time, and later at the same stage of training, twice as much work as was done by the same muscles in the non-athlete. The lack of breakfast had at first a slightly less favorable effect upon the amount of work done, although the athlete always felt less fatigued 392 I. H. HYDE, C. B. ROOT AND H. CURL working without breakfast than when working one hour after the meal. Both subjects were able to do more work on the ergograph when the muscles were in training after eating breakfast, and more one hour than one and one-half hour following the meal. After taking 1.42 grain of caffeine, both subjects did more than twice as much work than they were able to do after eating breakfast. The after effect, however, was a heightened degree of irritability especially noticeable in the athlete. The ergographic work had practically no effect on blood pressure and only a slight effect, if any, on the pulse rate, when working either without or after eating breakfast. The normal pulse rate was practically the same, but the normal blood pressure was higher at all times in the athlete than in the non-athlete. After both subjects had had equal preliminary training for one month of the arm and trunk muscles on the ergometer, the athlete did more than twice the work done by the non-athlete in one and one-half ‘the time without, and more than one and one-half as much work, after eating breakfast. The efficiency of both subjects grew in proportion as the interval between the meal and beginning of the work increased from one to two and one-half hours. The non-athlete did one-half and the athlete one-third more work two and one-half hours after than they were able to do one hour after the meal. The increase above their normal blood pressure after working either with or without breakfast was the same for both subjects, notwithstand- ing that the athlete did more work. But under the same conditions the pulse rate in the athlete was practically double that in the non- athlete. The increase in heart rate was least in both subjects when working two and one-half hours after eating breakfast, that is, the time when the greatest amount of work was accomplished. A weak dose of 1.42 grain of caffeine, without work or breakfast, gradually increased the pulse rate during the first hour, but in the non-athlete as a rule only after a slight initial fall. In both subjects the pulse returned to the normal rate within three hours. With the — larger dose, 2.24 grains, under the same conditions, the increase in pulse appeared more promptly, but in thirty minutes was depressed below normal in the non-athlete,.and accelerated above the normal rate in the athlete. The blood pressure rose above the normal level in one hour and frequently had not returned to the level in three hours after taking either of the doses of caffeine. The effects of caffeine taken at different intervals before work, varied with the dose and the individual. In the athlete the maximum influ- EFFECT OF FOOD AND OF CAFFEINE ON MUSCULAR WORK 393 ence of a dose of 1.42 grain was manifested in three-quarters of an hour, and in the non-athlete three hours after the dose was taken. The athlete did but little more work forty-five minutes after than he did twenty minutes after taking the drug. But the non-athlete did two and one-half times as much work three hours after as he did twenty minutes after taking the dose. ) _ Power and endurance for work, and cardiac activity and increase in blood pressure do not keep pace with increase of dosage. The maxi- mum power for work in both subjects was attained with the dose of 2.24 grains of caffeine. With this dose both subjects did two and a half times as much work as they were able to do one hour after eating _ breakfast. In the athlete with this optimum or with the weaker dose — of caffeine, the blood pressure was no greater than after the maximum work done either with or without breakfast, and the heart rate was only slightly more accelerated. In the non-athlete the pulse rate was increased almost three times as much, but the blood pressure was no higher than it was after the maximum work following the meal. A stronger dose of 3.58 grains depressed the muscular power for work in both men, but very markedly so, as well as the blood pressure and pulse rate in the non-athlete. In the athlete the blood pressure was no different, but the heart rate was less after the work following the weaker dose. When the dose was taken in proportion to the body weight, e.g., 0.2 grain of caffeine per 9.3 kilo body weight, or a stronger dose of 0.2 grain per 5.9 kilo weight, the results presented another viewpoint to those obtained when the dose was taken irrespec- tive of body weight. The facts showed that of these two doses thirty minutes before beginning work, the weaker dose and not the stronger stimulated the working power in the athlete most. But in the non- athlete the reverse was the case. With the stronger dose the athlete did one-fourth less and the non-athlete one-fourth more work than with the weaker dose. At the same time the pulse rate was enormously increased in the non-athlete and less so in the athlete who did double the work done by the non-athiete. On the other hand, the blood pres- sure fell slightly in the non-athlete, and fell also or remained unaltered in the athlete after work and after taking the stronger dose. There- fore, for each subject there was a definite optimum dose which, when increased, proved depressing for muscular work, blood pressure and pulse rate. One hour’s rest did not remove the sense of fatigue produced by the ergometer work, but when caffeine was taken the fatigue of the previous 394 I. H. HYDE, C. B. ROOT AND H. CURL hour’s work was inhibited and both subjects did more work then, and even twenty-four hours after taking caffeine, than they did before tak- ing the drug. With the same dose of caffeine and also without eating breakfast, the power for muscular work in the athlete was greater at 4,20 p.m. and in the non-athlete at 8.20 p.m. than at 8.20 a.m. That is, the athlete did his best work eight hours after taking the caffeine and four hours after luncheon, and the non-athlete did his best work twelve hours after taking the dose, and two hours after dinner. At these respective periods, the pulse rate and blood pressure increased greatly in the non-athlete, and the pulse but not the pressure in the athlete. The after effect of the larger dose was a heightened condition of irritability that persisted many hours after the drug was taken. The power and endurance for work were increased, and the cardiac activity greatly affected, but the blood pressure less so than with the stronger dose. It was not possible to state how long the after effect would endure, because the experiments were suddenly interrupted by the paralysis of the rectus muscle of the left eye in the shea i and the nervous condition of the non-athlete. BIBLIOGRAPHY (1) LomBarp: Journ. Physiol., 1892, xiii, 1. (2) Rivers anp WesBeErR: Journ. Physiol., 1907, xxxvi, 33. (3) SottMaNN AND PitcHerR: Journ. Pharm. and Exper. Therap., 1911, iii, 19. (4) Maaarora: Arch. f. Anat. u. Physiol., 1890, 191. (5) Hotyineswortu: Psychol. Rev., 1912, xix, 73. (6) Cusuny: Pharm. and Therap., 1915, 283. i elo THE EFFECTS OF ADRENIN ON THE DISTRIBUTION OF THE BLOOD 2% DY: EFFrect OF Massive Doses on THE OUTFLOW FROM MUSCLE R. E. LEE GUNNING From the Laboratory of Physiology of the Northwestern University Medical School Received for publication March 26, 1917 That intravenous injections of adrenin in all dosages cause dilatation of the vessels of the muscles of the limb was recently reported from this laboratory (1). The largest dosage used in that investigation was 5 ec. of a 1—25,000 solution. Inasmuch as such dosages as these are probably many times greater than any single discharge from the nor- mal adrenal glands, dosages of higher concentration were not studied. Also we saw no reason to believe that dosages of higher concentrations than those used would alter the reaction. Recently Cannon and Gruber (2) published some muscle contraction graphs of animals under the influence of large doses of adrenin. Fol- lowing the injections the contractions were diminished in height—a fact which suggests that vasoconstriction in the muscle might be oc- curring. Indeed the authors postulate this condition. Later we were informed by Dr. W. J. Meek that he had observed vasoconstriction in the muscles due to very large doses of adrenin.. Gruber (3) reported that in perfused muscles, with the nerve supply cut, adrenin in small doses produces vasoconstriction. Dilatation to be sure could not be expected with the vessels in a state of extreme relaxation due to the loss of the tonic effect of the constrictor nerves when the vessels are sup- posedly dilated to their limit. Cannon and Lyman (4) have observed that after the blood pressure has reached a certain low level, intra- venous injections of adrenin will no longer exert a depressor effect. In view of the foregoing observations it was decided that further experi- mentation on the effect of adrenin on the circulation in the muscles was desirable by way of supplementing our former communication. 395 396 R. E. LEE GUNNING METHOD The same methods of investigation were used as were described in the first paper of the series (1). As before, dogs were used for experi- mentation. Volume curves were’ not recorded. The observations were all made on venous outflow. RESULTS In investigating this phase of the problem, small doses were used at first and the amount gradually increased. When the dosage reached j\’ it Da a) Outflow, . ME TT TTT) Fig. 1. Successive segments of graph showing effect of massive dose of adrenin on blood pressure, femoral pulse and venous outflow from muscle. Interval omitted in each case two minutes. Dose 1 ec. 1: 1000 ‘‘Adrenalin.’’ Dog; weight, 11 kilos. No reduction. 1 to 2 ce. of a 1—-1000 solution the reaction of pure dilatation, such as was previously reported, changes as is shown in the accompanying figure. There is a short preliminary increase in the outflow following the injections which lasts about ten seconds and occurs during the rise in blood pressure. This is probably due to the vasoconstriction fore- ing the blood out of the vascular spaces into:‘the veins. At the time, ADRENIN AND BLOOD FLOW FROM MUSCLE 397 or shortly after the blood pressure reaches its crest, an active vaso- constriction takes place in the muscle circulation. This is evidenced by a diminished outflow which lasts, when the injections do not pro- duce a too long maintained blood pressure reaction, until some time after the blood pressure has again returned to normal. In the long maintained blood pressure reactions the diminished outflow begins its return to normal several seconds after the blood pressure commences to drop. The outflow, after returning to normal, invariably showed a tendency to “over recovery” by a second increase in the rate of out- flow. This secondary increase might be due to the release of blood dammed back by constriction in the veins of the muscles as was ob- served in the mesenteric circulation by Henderson (5) from mechanical stimulation of the intestines. It was observed, however, that there was no recovery from this secondary dilatation and also that the ten- dency to it was more marked toward the end of an experiment. This “over recovery” then is due most probably to a fatigue of the vascular musculature. That the vasoconstriction was not taking place in fatigued or over- dilated vessels was proven by the facts that the blood pressure was maintained, that the vasoconstriction would take place early in an experiment and that small dosages of adrenin would produce pure dila- tations at any time during the experiment. The threshold for the vasoconstrictor effect was found to vary both in the same dog and in different dogs. It was found to be slightly lower with the prolongation of an experiment and with an increase in the concentration of the injection. It was observed that 5 cc. of 1-5000 solution might produce in a given animal pure local dilatation without a marked general blood pressure reaction, whereas 1 cé. of a 1-1000 solution would produce the vasoconstriction reaction as de- scribed. The quantity of adrenin in both cases was the same. These results are probably due to the fact that the concentrated solutions reach the tissues still in more concentrated form. The solutions of lower concentration, since they are greater in volume, amount in fact to a short lasting infusion. DISCUSSION These results hardly have a bearing on our main thesis which is concerned with suprarenal physiology. The high dosages necessary to produce vasoconstriction in the muscles are not physiologic. They are rather to be considered as pathologic. A study of the blood pres- 398 R. E. LEE GUNNING sure curves, induced by these massive doses, gives the impression of a tetanus. The fact that a maintained ‘‘over recovery” occurs sug- gests a toxic effect which has injured some part of the vasomotor ap- paratus. It is a well known fact that adrenin in very large doses does produce toxic effects. It is extremely unlikely that the normal adrenal glands could at any one time pour such quantities as these into the circulation. The normal discharge, judging by all data now available, is similar to an experimental infusion of low concentration. SUMMARY 1. Adrenin in massive dosages produces a diminished circulation in the skeletal muscles. 2. There is a preliminary increase in the outflow due supposedly to the blood present in the vascular spaces being forced into the veins by the vasoconstriction. 3. There is a maintained secondary dilatation due onohably 4 to a fatigue of the vascular musculature. 4. The vasoconstrictor threshold was found to be lowered by an_ increase in the concentration of the adrenin and by the repetition of injections. BIBLIOGRAPHY (1) Hoskins, GuNNING AND Berry: This Journal, 1916, xli, 513. (2) CANNON AND GRuBER: Ibid., 1916, xlii, 36. (3) Gruper: Proc. Amer. Physiol. Soc., Ibid., 1917, xlii, 610. (4) CANNON AND Lyman: Ibid., 1913, xxxi, 376. (5). Henprerson: Ibid., 1917, xlii, 589. A correction. In the first paper of this series (1) the statement. was made that hydremic plethora serves to convert a vasoconstrictor effect of adrenin in a limb to a vasodilator effect. Subsequent investigation has failed to corrobo- rate our earlier observations on this point. EFFECTS OF ADRENIN ON THE DISTRIBUTION OF THE BLOOD Y. VoLUME CHANGES AND VENOUS DISCHARGE IN THE INTESTINE eee : R. G. HOSKINS ann R. E. LEE GUNNING From the Laboratory of Physiology of the Northwestern University Medical School Received for publication April 19, 1917 Although a number of investigators have studied the effects of adrenin on the activity of the intestines relatively little attention, appar- ently, has been given to the effects on vascular conditions. Oliver and Schaefer (1) made no direct determinations but concluded from inspection that vasoconstriction in the gut wall follows the adminis- tration of adrenin. The generalization was offered that this substance, or, more specifically, suprarenal extracts, cause vasoconstriction throughout the splanchnic area. Elliott (2) has reported a single instance in which he painted adrenin on the wall of the intestines of a fowl and also gave an intravenous injection. The result as determined by inspection was an intense vasoconstriction. Froelich (3) in a brief Centralblatt article has reported that both d- and I/-suprarenin as well as ‘“‘adrenalin’”’ caused long lasting contrac- tion of a segment of gut enclosed in a plethysmograph. He used both eats and dogs. Few details as to the experiments were given. Vin- cent (4) states that sometimes the intestine expands under the influ- ence of adrenin and publishes a plethysmograph curve that shows a slight degree of expansion that might be interpreted as passive. Brodie and Dixon (5) noted that the addition of adrenin to a per- fusate materially decreased the flow through the vessels of the isolated intestines. Ogawa (6) included the gut in aseries of perfusion studies on the vaso- motor effects of adrenin. Rabbits, cats and dogs were utilized as experimental animals. It was reported that J-adrenin in greater con- centration than 1:5,000,000 caused a clean-cut constriction in the intestinal vessels, as was shown by a decreased venous outflow. With the higher concentrations occasionally a secondary dilatation was ob- 399 400 R. G. HOSKINS AND R. E. LEE GUNNING served. In higher dilutions, e.g., 1: 50,000,000, the effect was primary dilatation. It was found that d-adrenin had a similar effect but larger doses were required. Technique. In our investigations both plethysmograph studies and determinations of the outflow from the opened intestinal veins have been made. Ether anesthesia was used in all cases. The adrenin solutions were made with Parke, Davis ‘‘ Adrenalin’”’ in distilled water. The technique employed has been described in sufficient detail in the first two papers of this series (7). It need only be added that in the plethysmograph work segments of small intestine about 20 cm. in length were used. These were coiled or folded into the box with due care, of course, to prevent occlusion of the blood vessels. The prompt volume changes and frequently the appearance in the tracings of vas- cular pulsations showed that success in this respect had been achieved. In all some three hundred and twenty-five determinations were made on forty-five dogs. In the infusion experiments the duration varied from one-half to nine minutes. Results. The results of the experiments are summarized in the ac- companying table. In many instances a brief inconsequential prelim- inary dilatation was observed but as this is a common feature in all the organs studied and probably is merely a passive effect, it is ignored in the table and subsequent discussion. The most prominent features in the results as a whole were augmentation of the gut volume and of the discharge from the opened veins. Often, however, the dilatation phase was preceded by a more or less pronounced contraction and in some instances contractions only were obtained. The only possible combination of these two conditions that was not encountered was dilatation followed by contraction. In case of the venous outflow the results were somewhat more consistent an augmentation always hav- ing occurred but this was not infrequently preceded by a decrease during the first part of the reaction. e There was no very definite correlation between the dosage and the reaction in the gut. In some animals smaller quantities caused con- traction which was succeeded by dilatation when the amount of the _adrenin was increased. In other animals, however, the reaction re- mained constant except as to degree whatever quantity was injected. The doses varied in the injection experiments from 0.25 to 8 ce. of 1: 100,000 solution. ‘The effects of massive doses were not investigated but the upper range actually employed very materially transcended physiological limits so far as these can be determined from data now EFFECTS OF ADRENIN ON INTESTINE 401 available. The data in the accompanying table are reduced roughly to a quantitative basis. Since the dogs used were not greatly dissim- ilar in size (averaging about 12 to 13 kilos), while the variations in reactions among various animals were much greater, nothing would Effects of adrenin in the small intestine NORMAL AFTER ECK FISTULA Number of| Average |Number of| Average cases dose cases dose mgm. mgm. 1. On volume a. Injections Pure contraction............... 9 0.013 0 Pure duatation................ 52 0.021 7 0.034 Contraction followed by dila- ae a A Sears 90 0.023 52 0.026 Dilatation followed by con- 0 SS ee eae 0 0 minute minute b. Infusions Pure contractions.............. 2 0.011 0 Pure dilatations............... 15 0.030* 8 0.043 Contraction followed by dila- MEIN a ans a eh Fons oe ee .....| 0.069 112.0 8.0 2 21 1.59 22 18.0 0.138 114082 1000 2:9 17.0 0.129 5 D5 17.6 0.116 25 18.0 0.120 ‘Average....... AL PLAT iPS ty Sa :...|° 0.124 114.0 10.0 - 3 18 2.60 1.6 18.2. | 0.190 | 110.0 10.0 1.8 18.0 0.167 1.8 18.0 0.167 1.4 18.5 0.220 16 18.2 0.201 1.8 18.5 0.166 AVOERRG 0.05... 5s oc cog eine eh eee es alee 0.185 110.0 10.0 4 19 221 1S 18.2 0.201 122.5 11.0 1 17.0 0.236 12 18.0 0.250 3 18.0 0.250 13 18.0 0.200 1.6 18.5 0.193 1.6 18.5 0.193 AVETAQO.. oss gc ss cece + vs pcr Rene ee 0.217 122.5 11.0 5 15 0.85 Aer 1805 0.114 90.5 8.5 25 18.0 0.1205 :1} 2.6 18.0 0.115 Average 2. iy Neate es Si yh 0.116 90.5 8.5 VASCULARITY OF ADRENAL BODIES 411 TABLE 2 Average values of flow and pressure WEIGHT OF - BLOOD PRESSURE IN MM. HG NUMBER OF : BLOOD FLOW EXPERIMENT PER SECOND ney It Dog Gland Arterial Venous kg. grams ce. . 1 20 1.38 0.069 112.0 8.0 “f va 21 1.59 0.124 114.0 10.0 i 3 18 2.60 © 0.185 110.0 10.0: “2 4 19 2.21 0.217 922.6 ot) on AO E-2 - 15 0.85 0.116 |- 90.5 8.5. “Average ....... | 18.6 chive 0.142. | >: 109.88] —° g7Ber9 8.5 cc. ina minute. At a pressure of about 100 mm. Hg, 100 grams of adrenal tissue are supplied with about 490 cc. of blood in a minute. _ If this value is now compared with the succeeding values of the blood- flow through other organs (4) it will be evident that the adrenal gland ‘is a very vascular organ; its blood-supply being exceeded only by that of the thyroid vot. cc. in a minute EE ea ee Pec Pat Wee - 5.0. _. Skeletal muscle.............. a ere prrere ae ee a Baey 2. aie hese ean 5 Aes tipi ao BRB she Pre 20.0 Se ae PPAR See SE SE sg. MEO oo eek tee! 0 GCE RPT ore eae bode: 2 26.0°"> TS) | tat ae AD cake SEEING: .......... 2. a SRN ee AE aL ee MC, Sores eee ted a" 5 31.0; to Dpleen......... ik il: n'a waie'o's «34 eee 58.0 5. pc vielen te series r= ds So ee een oe 2. =, EI Coes rg 2. eee 84.0° I Se. eo te ee “id Ls deeper ss 136.0 I gots. ce cicada oe eben oe dnd ee Va bibiedieai pied 150.0, os oo. Ss a tie ne eaisle- «oto meee 490.0... Thyroid oe sik pres sh. wigs ws Nee ee 560.0. In this connection it is interesting to note that Neuman (5): ihe found: an oxygen-consumption of 0.045 cc. per gram of substance’ and per minute. These tests were made upon the suprarenal of-the eat, the Barecroft-Roberts differential. blood-gas apparatus being used. This value indicates a flow of about 6 cc. per gram per minute.” ' 2 The current number of the Proc. of the Soc. for Exper. Biol. and Med. which has been issued since the preparation of this manuscript, contains a preliminary notice by Stewart and Rogoff in’ which it is stated that the caval segment: fills at the rate of about 8 cc. per minute, in a dog weighing 10 kilos. . 412 R. BURTON-OPITZ AND D. J. EDWARDS By employing the method previously described, Biedl has found that the number of drops escaping from the reservoir may be consid- erably increased by stimulation of the splanchnic nerve. This accel- eration, however, does not set in immediately, but about ten seconds after the beginning of the stimulation and lasts for some time after its close. Moreover, while it pursues a course which is practically parallel to the rise in blood-pressure, the fact remains that it outlasts the stimulus as well as the increase in pressure and this seems to show that it is dependent upon an active enlargement of the suprarenal _ blood-vessels. These alterations in the blood-supply of this gland we have suc- ceeded in registering a number of times, but as these curves present only quantitative and not qualitative differences, it may suffice to illustrate them by a single example taken from experiment 3. The second and third phases of this experiment are reproduced in the ac- companying figure. The blood-flow (St) showed in this case the nor- mal value of 0.155 cc. ina second. The arterial pressure (A) amounted to 100.2 mm. Hg and the venous pressure (V) to 9.8 mm. Hg. Shortly after the beginning of the excitation of the greater splanchnic nerve at S, the blood-flow increased gradually until it assumed its maximal value of 0.244 cc. in a second at about point 7. The latter coincides with the cessation of the stimulation and also with the maximal rise in the arterial pressure. Subsequent to 7 the arterial pressure de- creases gradually and assumes its normal level early in the course of the third phase of the stromuhr, about fifty seconds after the cessation of the stimulation. The venous pressure exhibits a slight rise during the period of the increased flow. The blood-flow pursues a very similar course. It is true, however, that the flow does not always return to normal with the pressure, but remains augmented for some time after the stimulation. In the present case, for example, the normal minute-volume for 100 grams of suprarenal substance amounts to 260 cc., while the minute-volume . during the stimulation measures 560 cc. It appears, therefore, that the excitation of the splanchnic nerve has resulted in an increase of flow of 200 cc. in a minute for 100 grams of substance. This fact in itself, however, does not seem to us to justify the deduction that the suprarenal blood-vessels are equipped with a dilator mechanism, be- cause the dissociation between the flow and the pressure here noted, could also be caused by the rather slow return to normal of passively distended blood-vessels. It also seems singular that the excitation of *(spuooes Aquemy ‘UID GT) SALOU OTUYOUB]dS 1048013 O44 JO UOIZB[NUITYS ZuliMp puvyls [vuorsidns oy} YAN0IY, MOY poolq ey, *[ "By AL BODIES N = on a) < bit titad | Mani HIT M Aili lilt WW i Hi Wy \| At a te ta wD i Hine el Wt OF It. My nT nly | VASCULARITY 414 R. BURTON-OPITZ AND D. J. EDWARDS the splanchnic nerve with currents of medium strength and high fre- quency should dilate the blood-vessels of the suprarenal gland and con- strict those of the other organs innervated by this nerve. The possibility that the splanchnic rise in blood-pressure is the es- sential factor in the increased blood-supply of the suprarenal gland we have attempted to test by the simultaneous stimulation of the distal and central ends of the splanchnic nerve, the former with a current of ordinary strength and frequency and the latter with a current of low frequency. This procedure necessitated the division of the left thoracic sympathetic nerve a short distance above the diaphragm and the ap- plication of two shielded electrodes. The inductoriums were then adjusted in such a way that the fall in blood-pressure obtained in con- sequence of the excitation of the central end, was about balanced by the rise resulting from the stimulation of the distal end. In this way, we succeeded ‘in retaining the blood-pressure practically at its normal level, because the rise in pressure resulting from the constriction of the blood-vessels in the different splanchnic organs, was counterbalanced by the vasodilatation produced elsewhere. We have, however, not been able to satisfy ourselves that under this con- dition the blood-flow through the suprarenal gland is markedly in- creased. But naturally, this result cannot justly be regarded as prov- ing the absence of vasodilator fibers, because the stimulation of the central end of the splanchnic nerve may have destroyed this effect reflexly through possible nervous connections between the suprarenal gland and the solar ganglia. If the central end of the thoracic nerve alone was stimulated, the resulting fall in blood-pressure. was invari- ably associated with a diminution in. the blood supply of this organ. BIBLIOGRAPHY (1) Brepu: Pfliiger’s Arch., 1897, Ixvii, 443. (2) Dreyer: This Journal, 1899, ii, 203. (3) Stewart: Proc. Soc. Exper. Biol. and Med., 1916, xii, 187. (4) Burtron-Oprirz: Quart. Journ. Exper. Physiol., 1911, iv, 113. (5) Neuman: Journ. Physiol., 1912, xlv, 189. ‘THE INFLUENCE OF SECRETIN ON THE NUMBER OF ERYTHROCYTES IN THE CIRCULATING BLOOD ARDREY W. DOWNS anv NATHAN B. EDDY Tot the Physiological Laboratory of McGill University, Montreal, Canada Received for publication April 2, 1917 In 1895 Dolinski (1) showed that acids brought into contact with the mucous membrane of the duodenum set up a secretion of pancreatic juice. Pawlow and his co-workers (2) further decided that the acid acts reflexly through a nerve center. Later Popielski (3), working under the direction of Pawlow, showed that if acids are introduced into the duodenum the pancreatic secretion appears after resection of both vagus and splanchnic nerves, after extirpation of the solar plexus and even after destruction of the spinal cord. He concluded that the secretion arose from a peripheral reflex through scattered ganglia of the pancreas, situated mostly near the duodenum. The same results and conclusions were reached by Wertheimer and Lepage (4). At this point Bayliss and Starling (5) demonstrated the true explanation of the phenomenon. They showed that the acid acts on a substance in the duodenal mucous membrane, prosecretin, and changes it into another substance, secretin. This is carried by the blood and activates the pancreatic cells. Bayliss and Starling (5) also showed that secretin increases .the secre- tion of bile. We have confirmed this by noting the rate of flow of bile incidentally in the course of other experiments. Sir Edward A. Schafer (6) states that secretin increases the flow of bile and of succus entericus but to a less extent than it affects the flow of pancreatic juice. He also states that intravenous injection of duodenal extract (evidently secretin from the context) has been shown by Cow (7) to cause the appearance of the pituitary autacoids in the cerebro-spinal fluid. Beveridge and Williams (8) in their very ingenious exposition of what they call the proteomorphic theory of immunity claim to have the records of over two hundred cases of diabetes and exophthalmic goiter in which the number of red corpuscles per cubic millimeter of 415 416 ARDREY W. DOWNS AND NATHAN B. EDDY blood was increased by the administration of secretin. Their theory of the production of immunity depends greatly on the power to hydro- lyze proteins which they attribute to the red blood corpuscles. If we grant that these premises are correct, then any agent capable of bring- ing about a sufficient increase in the number of the red corpuscles becomes of therapeutic value. We have been unable to obtain details of the records to which reference has been made. If secretin is to exert any influence as an immunizing agent by increasing the number of red corpuscles in the circulating blood, it is obvious that a single dose must be capable of causing a great and fairly prolonged rise in the red corpuscle count. As a means of ordinary treatment, hypodermic medication is preferable to intravenous, and if it can be shown that secretin administered hypodermically is able to increase the number of red corpuscles, then again, in order to be of service, a single dose, or at most three or four successive doses, should produce and maintain a largely increased erythrocyte count. Acting in accordance with the ideas thus suggested we determined to try first, the effect of a certain arbitrarily fixed dose of secretin given - intravenously and second, the effect of the same dose when introduced hypodermatically. In our selection of the animal to be used we were guided by the recent work of Lamson (9) on acute polycythemia in which he has shown that adrenalin, fright, pain, etc., cause sudden and very marked elevation of the red corpuscle count in the dog and cat but that these agents are without effect on the erythrocyte count of the rabbit. Therefore, that we might avoid the use of an anaesthetic, especially in those experiments . which were to be continued over several days, in order that the attend- ing conditions might be as uniform as possible, rabbits were employed in all of the experiments recorded in this paper. The secretin was in all cases prepared from the intestine of the dog. The animal was anaesthetized by ether alone and the upper half of the small intestine removed. This intestine was carefully washed in run- ning water and the mucous membrane scraped off with a dull knife. The scrapings were rubbed up in a mortar with sand, covered with 50 ec. of 0.4 per cent hydrochloric acid and allowed to stand for an hour or more. The mixture was then boiled actively for several minutes, neutralized with strong potassium hydroxide while boiling and again ~ rendered faintly acid with glacial acetic acid. Finally the propananos was strained through muslin and filtered. We found that this preparation when kept in the dark retained its INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 417 potency for about five days; but if glacial acetic acid were added to the filtrate in sufficient quantity to make this 2 per cent acid by volume and the solution evaporated to dryness, the residue was found to re- tain its potency for months at least. When required, a weighed _ quantity could be dissolved in distilled water and neutralized, thus giving a preparation of the same effectiveness as the original solution. Over two hundred determinations of the red corpuscles per cubic millimeter of blood were made in the course of these experiments, the blood being obtained from the ear of the rabbit and the count made in the usual manner with the Thoma-Zeiss apparatus. The dose of secretin solution selected for the first experiments was 1 ce. per kilogram of body weight. Five rabbits were taken, the eryth- rocytes per cubic millimeter of blood counted, and the proper dose of secretin injected into the femoral vein. The results of these experi- — ments are recorded in table 1. TABLE 1 Dose: 1 cc. secretin solution per kilogram of body weight - PERCENT- eee faire couwe | MATDE™ | Ameer | age aee | Mme | omarion eae Gad 5,320,000 | 7,510,000 | 2,190,000 | 41.16 30 60 2 4,560,000 | 6,990,000 | 2,430,000 | 53.29 15 70 3 4,450,000 | 6,350,000 | 1,900,000 | 42.69 40 65 4 5,610,000 | 8,210,000 | 2,600,000 | 46.35 45 90 5 4,830,000 | 5,630,000 800,000 | 16.56 25 45 Averages....| 4,954,000 | 6,938,000 | 1,984,000 | 40.04 31 66 As a type of this series of experiments the first one is given in detail: Experiment 1, November 6, 1916 10.05 a.m. Red blood corpuscles, 5,320,000 per cubic millimeter 10.10 a.m. 1 ce. secretin per kilogram of body weight given intravenously 10.25 a.m. Red blood corpuscles, 6,940,000 per cubic millimeter 10.40 a.m. Red blood corpuscles, 7,510,000 per cubic millimeter 10.55 a.m. Red blood corpuscles, 7,120,000 per cubic millimeter 11.10 a.m. Red blood corpuscles, 5,350,000 per cubic millimeter 11.30 a.m. Red blood corpuscles, 5,290,000 per cubic millimeter The next thing to be determined was the effect of the same dose upon the number of red corpuscles when it was introduced beneath 418 ARDREY W. DOWNS AND NATHAN B. EDDY the skin. A tabulated report of the results obtained in this way will be found in table 2. * TABLE 2 Dose: 1 cc. secretin solution per kilogram of body weight PIPERIMENT. | peretay, couwe | MAZIMOM | 4MounT C? |( aan re) Ace ae 6 5,120,000 | 6,305,000 1,185,000 23.1 60 90 7 4,548,000 | 5,844,000 1,296,000 28.4 30 30 8 4,536,000 | 5,545,000 1,009,000 22.2 30 90 9 4,906,000 | 5,619,000 713,000 14.5 55 95 10 5,840,000 | 7,197,000 1,357,000 23.2 60 90 Averages....| 4,990,000 | 6,102,000 1,112,000 22.2 47 79 As typical of this series of experiments the first one is here presented in detail: Experiment 6, November 20, 1916 1.30 p.m. Red blood corpuscles, 5,120,000 per cubic millimeter 1.35 p.m. 1 cc. secretin per kilogram of body weight given hypodermatically 2.05 p.m. Red blood corpuscles, 5,336,000 per cubic millimeter 2.35 p.m. Red blood corpuscles, 6,305,000 per cubic millimeter 3.35 p.m. Red blood corpuscles, 5,601,000 per cubic millimeter 4.35 p.m. Red blood corpuscles, 5,006,000 per cubic millimeter A comparison of the results obtained in these two groups of experi- ments shows certain features‘in favor of the intravenous method of | administration. The most striking difference is in the percentage in- crease—an average of 40 per cent when the secretin is given intrave- nously and 22.2 per cent when given hypodermatically. However, if we note the average actual increase in number of red corpuscles the differ- ence is only slightly over one-half million in favor of the intravenous method—1,984,000 in the former case and 1,112,000 in the latter. As might be expected the intravenous method gives the effect in a shorter time than the subcutaneous,—the maximum effect obtained in thirty- one minutes in one instance and. in forty-seven minutes in the other; but when we compare the duration of the effect we find it almost iden- tical in the average of the two groups—sixty-six minutes was the average time that the increase lasted when the secretin was given intravenously and seventy-nine minutes when it was given hypodermatically. INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 419 The second group of experiments had convinced us that secretin injected subcutaneously was capable of exerting an influence, at least so far as affecting the number of red corpuscles in circulation was concerned. Moreover, the greater effect obtained by giving the secre- . tin intravenously was not sufficiently pronounced to make it the method of choice so far as any therapeutic application was concerned. There- fore, we decided to adhere to the method of hypodermic administra- tion in the remainder of our experiments, particularly as these two series of observations appeared to furnish sufficient data from which to deduce the probable action’ of any particular dose of secretin when given intravenously if we had determined its effect when given hypodermatically. To determine the most effective dose we made several series of experiments using the following doses per kilogram of body weight: 0.75 ec., 0.5 cc., 0.25 ec., 1.5 ec., 2 ee. In this way we tested the ef- fect of amounts of secretin less than and greater than the original and arbitrary dose of 1 cc. per kilogram of body weight. Each group table gives the summarized results of each experiment in the group and the averages for that particular series. Following each group summary is a report giving the details of one experiment in the group, serving as an example of all. TABLE 3 Dose: 0.75 cc. secretin solution per kilogram of body weight eee fester cove | Met | amount oF | “sae mm | MAEM | Ore 13 5,327,000 6,224,000 897,000 16.8 60 90 14 6,334,000 7,168,000 834,000 13.1 30 30 20 6,384,000 7,098,000 714,000 11.0 60 60 ~ 21 5,324,000 6,961,000 1,637,000 30.7 90 90 Averages. 5,842,250 6,862,750 1,020,500 17.9 60 67.5 Experiment 13, January 12, 1917 9.30 a.m. Red blood corpuscles, 5,327,000 per cubic millimeter 10.20 a.m. 0.75 cc. secretin per kilogram of body weight given hypodermatically 10.50 a.m. Red blood corpuscles, 5,206,000 per cubic millimeter 11.20 a.m. Red blood corpuscles, 6,224,000 per cubic millimeter 12.20 p.m. Red blood corpuscles, 6,042,000 per cubic millimeter Red blood corpuscles, 5,100,000 per cubic millimeter 420 ARDREY W. DOWNS AND NATHAN B. EDDY TABLE 4 Dose: 0.5 cc. secretin solution per kilogram of body weight TABLE 5 Dose: 0.25 cc. secretin solution per kilogram of body weight : PERCENT- mrmanemer | omaconwr| Magne | aMouwe oy | “cua | Maun) Somer 16 6,251,000 6,837,000 586,000 - 9.3 30 30 19 5,741,000 7,875,000 2,134,000 37.1 90 90 24 6,048,000 | No effect 30 5,804,000 | No effect 31 6,760,000 7,046,000 286,000 4.2 60 60 Averages....| 6,120,800 | 6,722,000 601,200 10.1 36 36 Experiment 16, January 22, 1917 12.55 p.m. Red blood corpuscles, 6,251,000 per cubic millimeter 1.00 p.m. 0.5 cc. secretin per kilogram of body weight given hypodermatically 1.30 p.m. Red blood corpuscles, 6,837,000 per cubic millimeter 2.00 p.m. Red blood corpuscles, 6,079,000 per cubic millimeter 3.00 p.m. Red blood corpuscles, 6,426,000 per cubic millimeter PERCENT- mmnemt | oaeay couws| Magee | AMODEE OF | “ieee | ee 11 5,888,000 6,810,000 922,000 15.6 30 30 17 5,754,000 | No effect 18 5,280,000 6,144,000 864,000 16.3 30 30 25 4,075,000 5,015,000 940,000 23.1 30 30 27 5,970,000 | No effect Averages....| 5,395,000 5,940,200 | 545,200 11.0 18 18 Experiment 11, November 28, 1916 1.10 p.m. Red blood corpuscles, 5,888,000 per cubic millimeter 1.15 p.m. 0.25 ce. secretin per kilogram of body weight given hypodermatically 1.45 p.m. Red blood corpuscles, 6,810,000 per cubic millimeter 2.15 p.m. Red blood corpuscles, 5,988,000 per cubic millimeter 11.30 a.m. INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 421 TABLE 6 eisesy,, Dose: 1.6 cc, secretin. solution per kilogram of body weight ee ertcours | Moree | “moeniad” | sauce | MAtnevee| sunacom P minutes minutes 12 6,376,000 7,440,000 1,064,000 15.1 60 90 22 7,188,000 | No effect — «28 4,959,000 7,162,000 2,203,000 44.8 120 120 32 5,936,000 7,416,000 1,480,000 24.9 60 90 Averages... 6,114,750 7,301,500 1,186,750 21.2 58 60 . Experiment 12, December 20, 1916 10.00 a.m. Red blood corpuscles, 6,376,000 per cubic millimeter 10.30 a.m. 1.5 cc. secretin per kilogram of body weight given hypodermatically 11.00 a.m. Red blood corpuscles, 6,937,000 per cubic millimeter 11.30 a.m. Red blood corpuscles, 7,440,000 per cubic millimeter 12.30 p.m. Red blood corpuscles, 6,014,000 per cubic millimeter TABLE 7 Dose: 2 cc. secretin solution per kilogram of body weight Se peta comme) Me™ | Atouw or | “Lae me, | MAINO | oomsion 15 4,112,000 | 5,188,000 | 1,076,000 | 26.1 30 60 23 5,928,000 6,935,000 1,007,000 16.9 30 90 26 5,858,000 6,471,000 613,000 10.4 30 90 29° 5,400,000 | 7,502,000 | 2,102,000 | 38.9 60 120 Averages. 5,324,500 | 6,524,000 | 1,199,500 | 22.4 37.5 90 : Experiment 15, January 12, 1917 9.20 a.m. Red blood corpuscles, 4,112,000 per cubic millimeter 9.30a.m. 2 cc. secretin per kilogram of body weight given hypodermatically 10.00 a.m. Red blood corpuscles, 5,188,000 per cubic millimeter 10.30 a.m. Red blood corpuscles, 4,507,000.per cubic millimeter Red blood corpuscles, 4,207,000 per cubic millimeter As we proceeded with our observations the results pointed to a dose of 1 ce. per kilogram of body weight as being the most efficient. In order to assure ourselves on this point we made seven more experiments in which the dose was 1 cc. per kilogram of body weight. Taken to- 422 ARDREY W. DOWNS AND NATHAN B. EDDY gether with the five original experiments recorded in table 2 we have a total of twelve such determinations. For the purpose of computing an average with as large a number of experiments as possible these have been brought together in table 8. TABLE 8 Dose: 1 cc. secretin solution per kilogram of body weight SPEEDMENE | nermzatcovxt| “oomyu" ||| Suowaiaot | ee 0) Mtge minutes minutes — 6 5,120,000 6,305,000 1,185,000 23.1 60 90 7 4,548,000 5,844,000 1,296,000 28.4 30.52 30 8 4,536,000 5,545,000 1,009,000 22.2 30 90 9 4,906,000 5,619,000 713,000 14.5 55 | (95 10 5,840,000 7,197,000 1,357,000 23.2 60 90 33 5,778,000 6,894,000 1,116,000 19.3 30 90 34 5,872,000 6,579,000 707,000 12.0 50 70 35 6,200,000 6,850,000 650,000 10.4 50 90 36 5,456,000 5,955,000 499,000 9.1 70 80 37 5,200,000 7,240,000 2,040,000 39.2 25 105 38 4,720,000 7,264,000 2,544,000 53.8 75 30 39 5,684,000 6,752,000 1,068,000 18.7 35 30 Averages....| 5,321,666 6,503,666 1,182,000 22.2 47.5 73.3 Reviewing the average percentage increase with each dose we find the results to have been as follows: 0.25 cc. secretin per kilogram of body weight, 11.0 per cent; 0.5 ec. secretin per kilogram of body weight, 10.1 per cent; 0.75 cc. secretin per kilogram of body weight, 17.9 per cent; 1 cc. secretin per kilogram of body weight, 22.2 per cent; 1.5 cc. secretin per kilogram of body weight, 21.2 per cent; 2 cc. secretin per kilogram of body weight, 22.4 per cent. These records indicate a dose of 1 cc. per kilogram of body weight as the most efficient dose of secre- tin. We also find that the longest average time the effect lasted was ninety minutes—where the dose was 2 cc. secretin per kilogram of body weight. With a dose of 1 ce. per kilogram of body weight the average duration was 73.3 minutes. It seemed worth while to test the effect of repeated doses of secretin on the increase in the number of erythro- cytes per cubic millimeter of blood, both as to the amount of increase and the duration of the increase. Is it possible to produce a sum- mation effect? To answer this question the following experiments were performed: One in which a dose of 1 ec. secretin per kilogram of - INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 423 _ body weight was followed in two hours by a second dose of 1 cc. per kilogram of body weight; two experiments in each of which four suc- cessive doses of 1 cc. secretin per kilogram of body weight were ad- ministered at intervals of one hour; one experiment in which five _ successive doses of 1 ce. secretin per kilogram of body weight were bial at intervals of one hour; one experiment in which five successive of 1 ce. secretin per run of body weight were given at inter- vals of twenty-four hours. The results of these observations are eapeced. Experiment 40, February 14, 1917 9.45 a.m. Red blood corpuscles, 4,548,000 per cubic millimeter 10.00 a.m. 1 ce. secretin per kilogram of body weight given = Alege 2 10.30 a.m. Red blood corpuscles, 5,844,000 per cubic millimeter 11.00 a.m. Red blood corpuscles, 5,338,000 per cubic millimeter 12.00 noon Red blood corpuscles, 5,025,000 per cubic millimeter 12.30 p.m. 1 cc. secretin per kilogram of body weight given hypodermatically 1. 00 p.m. Red blood corpuscles, 5,042,000 per cubic millimeter 2. 00 p:m. Red blood corpuscles, 5,600,000 per cubi¢ millimeter 3.00 p.m. Red blood corpuscles, 5,787,000 per cubie millimeter February 15, 1917 10. 00 a.m. Red blood corpuscles, 4,286,000 per cubic millimeter February 16, 1917 10.00 a.m. Red blood corpuscles, 4,457,000 per cubic millimeter kei ; Experiment 41, February 19, 1917 9.55 a.m. Red blood corpuscles, 5,765,000 per cubic millimeter 10.00 a.m. 1 ce. secretin per kilogram of body weight given: hypodermatically 10.55 a.m. Red blood corpuscles, 7,094,000 per cubic millimeter 11.00a.m. 1 cc. secretin per kilogram of body weight given hypodermatically 11.55 a.m. Red blood corpuscles, 6,756,000 per cubic millimeter 12.00 noon 1 ce. secretin per kilogram of body weight given hypodermatically 12.55 p.m. Red blood corpuscles, 7,459,000 per cubic millimeter 1.00 p.m. 1 cc. secretin per kilogram of body weight given hypodermatically 2.00 p.m. Red blood corpuscles, 8,327,000 per cubic millimeter 3.00 p.m. Red blood corpuscles, 6,749,000 per cubic millimeter 4.00 p.m. Red blood corpuscles, 6,195,000 per cubic millimeter February 20, 1917 9.00 a.m. Red blood canadien: 6,156,000 per eubie millimeter February 21, 1917 9.00 a.m. Red blood corpuscles, 5,832,000 per cubic millimeter 424 8.55 a.m. 9.00 a.m. 9.55 a.m. 10.00 a.m. 10.55 a.m. 11.00 a.m. 11.55 a.m. 12.00 noon 1.00 p.m. 2.00 p.m. 3.00 p.m. 10.00 a.m. 10.05 a.m. 10.30 a.m. 10.55 a.m. 11.00 a.m. 11.30 a.m. 11.55 a.m. 12.00 noon 12.30 p.m. 12.55 p.m. 1.00 p.m. 1.30 p.m. 1.55 p.m. 2.00 p.m. 2.30 p.m. 3.00 p.m. 4.00 p.m. ARDREY W. DOWNS AND NATHAN B. EDDY Experiment 42, February 20, 1917 Red blood corpuscles, 5,245,000 per cubic millimeter 1 cc. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 5,645,000 per cubic millimeter ‘1 ce. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 6,723,000 per cubic millimeter 1 cc. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 6,659,000 per cubic millimeter 1 cc.-secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 6,384,000 per cubic millimeter Red blood corpuscles, 5,553,000 per cubic millimeter Red blood corpuscles, 5,284,000 per cubic millimeter Experiment 43, February 21, 1917 Red blood corpuscles, 5,825,000 per cubic millimeter _ 1 ce. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 6,522,000 per cubic millimeter ; Red blood corpuscles, 6,764,000 per cubic millimeter 1 ce. secretin per kilogram of body, weight given hypodermatically Red blood corpuscles, 6,004,000 per cubic millimeter Red blood corpuscles, 7,402,000 per cubic millimeter 1 ce. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 8,883,000 per cubic millimeter Red blood corpuscles, 6,111,000 per cubic millimeter 1 cc. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 5,472,000 per cubic millimeter Red blood corpuscles, 7,016,000 per cubic millimeter 1 cc. secretin per kilogram of body weight given hypodermatically Red blood corpuscles, 7,094,000 per cubic millimeter Red blood corpuscles, 8,089,000 per cubic millimeter Red blood corpuscles, 6,912,000 per cubic millimeter February 22, 1917 10.00 a.m. Red blood corpuscles, 5,446,000 per cubic millimeter February 28, 1917 10.00 a.m. 10.00 a.m. 10.05 a.m. Red blood corpuscles, 5,296,000 per cubic millimeter Experiment 44, February 22, 1917 Red blood corpuscles, 5,829,000 per cubic millimeter 1 cc. secretin per kilogram of body weight given hypodermatically February 23, 1917 10.00 a.m. 10.05 a.m. Red blood corpuscles, 5,245,000 per cubic millimeter 1 ce. secretin per kilogram of body weight given hypodermatically INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 425 February 24, 1917 10.00a.m. Red blood corpuscles, 7,710,000 per cubic millimeter 10.05 a.m. 1 cc. secretin per kilogram of body weight given hypodermatically February 25, 1917 10.00 a.m. Red blood corpuscles, 6,961,000 per cubic millimeter 10.05 a.m. 1 cc. secretin per kilogram of body weight given hypodermatically February 26, 1917 10.00 a.m. Red blood corpuscles, 5,523,000 per cubic millimeter 10.05a.m. 1 cc. secretin per kilogram of bédy weight given hypodermatically March 1, 1917 10.00 a.m. Red blood corpuscles, 5,975,000 per cubic millimeter March 4, 1917 10.00 a.m. Red blood corpuscles, 6,218,000 per cubic millimeter These experiments, numbers 40 to 44 inclusive, show that successive doses of secretin at short intervals are capable of causing a progressive inerease in the number of red corpuscles per cubic millimeter of blood, but the increase is not maintained from one dose to the next, so that between the doses there is a diminution from the maximum count re- sulting from that dose before the next dose exerts an effect. In other words, two doses do not give twice the effect of one dose or three doses three times the effect of one dose. Moreover, when the administra- tion of the secretin is stopped the number of red corpuscles in the cir- culating blood reverts to normal almost as quickly as after a single dose. In the case of the rabbit which received a daily dose of secretin for five days the increase in the number of erythrocytes per unit volume of blood on the eighth day as compared with the initial count was 146,000, showing that secretin has no ability to produce a permanent increase in the red corpuscle content of the circulating blood of the normal animal. Three main conclusions are inevitable from the observations that have been recorded: first, secretin, even when injected subcutaneously, is capable of producing an increase in the number of red corpuscles in the circulating blood; second, the increase thus effected is not great as compared with the increase that may be obtained by the action of other agents; third, the length of time that this larger number of red cor- puscles persists is comparatively short. 426 _ARDREY W..DOWNS AND NATHAN B. EDDY Let us consider briefly the possible therapeutic benefits that may be derived from the exhibition of secretin. If we grant the correct- ness of the theory of Beveridge and Williams (10) that the red cor- puscles constitute one of the chief defensive agencies of the animal organism against the invasion of pathogenic bacteria or the products of such bacteria, then, in order that aid may be given to the establish- ment’ of immunity by this means, we must be able in some way to bring about a marked augmentation in the number of red corpuscles. and an augmentation that will continue for a time sufficiently long to be of service. We. haye shown that. the most efficient dose of secretin is in the proportion of 1 ce. per kilogram of body weight, which means for the average man 70 ce. of secretin subcutaneously or 38.5 ce. intrave- nously. Furthermore the effect of this dose disappeared on an average 73.3 minutes after it was given. Moreover, it was not possible to pro- duce a lasting increase in the number of red corpuscles by giving suc- cessive doses, either at intervals of one hour, two hours or twenty-four hours. The facts that have been adduced militate against any thera- ‘peutic value for secretin but do not detract in the slightest from its physiological significance. On the contrary, they appear to give se- cretin added importance in the normal organism. _ It is possible that one ‘of the means by which the normal number of red corpuscles 1 is main- tained in the blood stream is the action of secretin. Naturally the next question that presents itself is: How does secretin produce this increase in the number of the red corpuscles in the cir- culating blood? This is a question which we are not as yet prepared to answer, but several suggestions can be offered. The first. and sim- plest explanation that presents itself is that secretin exerts a direct stimulating influence upon the red inarrow of the bones thus leading to the formation of new cells. Such a conclusion is entirely i in accord with the known activities of secretin. - If this substance i is capable of ‘promoting the activity of the pancreatic cells, of the hepatic cells, and of the cells of the pituitary body, it is entirely reasonable to assume that it may also have the ability to increase the formation of red blood corpuscles by the red marrow of the bones. i pene ‘A second way in which secretin might bring about an increase in the number of circulating erythrocytes is by causing variations in ‘their unequal distribution. This might be effected by a direct constricting action'on the capillaries of some large area, such as the liver, or by an indirect action through stimulation of the adrenals. Lamson (9) has shown that in the cat and dog an increase in the number of red cor- INFLUENCE OF SECRETIN ON NUMBER OF ERYTHROCYTES 427 puscles per unit volume of blood may be obtained by the administra- tion of adrenalin. Inthe same animals fright raises the number of red corpuscles per unit volume of blood an average of 80 per cent. In both cases there is no increase in the number of erythrocytes if the hepatic artery be ligated. It has been shown by Cannon (11) that fright stimulates the adrenals, and Lamson attributes the presence of a greater number of red corpuscles in the blood stream to a constric- tion of the capillaries of the liver caused by adrenalin. Mautner and Pick (12) inform us of the presence of an extremely sensitive nervous mechanism in the liver of the dog, reacting to epinephrin by constric- tion of the capillaries, and the absence of such in the liver of the rabbit, or the presence in this animal of a much less sensitive mechanism. Lamson (13) has shown also that excitement or the intravenous injec- tion of adrenalin causes no polycythemia in rabbits. Therefore, it seems that we can rule out the suggestion that secretin increases the number of red corpuscles in the circulating blood of the rabbit by stim- ulating the adrenals. As to whether the secretin acts directly to pro- mote capillary constriction or not we have no evidence and do not.know of any work that has been reported on the subject. One other explanation that should be mentioned is the possibility that secretin causes a decrease in plasma volume and thus gives rise ~ to a higher erythrocyte count per cubic millimeter of blood. The means by which secretin acts to produce an increase in the - number of red corpuscles in a unit volume of blood is a question outside of the scope of the present investigation. In undertaking these ex- periments we were actuated by the desire to know whether secretin increased the number of erythrocytes in the blood stream or not; and, if so, how much of an increase could be hoped for, and how long it would be possible to maintain this increase. These questions have been answered, we believe, and the solution of the mode of action will be found later. , CONCLUSIONS 1. It is possible to produce a considerable increase in the red cor- puscle count per cubic millimeter of blood by the administration of secretin even in small doses and by subcutaneous injection. 2. The most efficient dose is 1 cc. of secretin per kilogram of body weight. 3. The increase in the count appears quickly and is very transient. 428 ARDREY W. DOWNS AND NATHAN B. EDDY 4. By repeating the dose of secretin at short intervals the increase in the erythrocyte count can be kept up for several hours but drops promptly after the administration of the last dose. 5. The administration of secretin over a period of five days, in daily doses of 1 cc. per kilogram of body weight, has very slight, if any, lasting effect on the red corpuscle count in the normal animal. BIBLIOGRAPHY (1) Douinsxr: Arch. d. Sc. Biol., 1895, iii, 399. (2) Pawtow: The work of the digestive glands, 1910, 135. (3) Porrensxi: Pfliger’s Arch. f. Physiol., 1901, lxxxvi, 215. (4) WeRTHEIMER AND LepaGce: Journ. d. Physiol., 1901, iii, 335. (5) Bayxiss AnD Staruine: Journ. Physiol., 1902, xxviii, 325. (6) Scuirer: The endocrine organs, 1916, 125. — (7) Cow: Reported by Schifer. (8) WiLtL1AMs AND BeveripGe: Amer. Med., 1914, xx, 702. (9) Lamson: Journ. Pharm. and Exper. Therap., 1915, vii, 169. (10) WinL1ams AND BevertpGe: Amer. Med., 1914, xx, 621. (11) CANNON AND DE LA Paz: This Journal, 1911, xxviii, 64. (12) MautnEr anp Pick: Minch. med. Wochenschr., 1915, xxxiv, 1141. (18) Lamson: Journ. Pharm. and Exper. Therap., 1916, viii, 167. THE ACTION OF ULTRA-VIOLET RADIATION IN KILL- ING LIVING CELLS SUCH AS BACTERIA vs Sac W. E. BURGE From the Physiological Laboratory of the University of Illinois (From experiments carried out at Nela Research Laboratory) Received for publication April 4, 1917 Several theories have been advanced in an attempt to explain the mode of action of ultra-violet radiation in killing living cells. One theory is that the short wave lengths of the spectrum act by destroy- ing the intracellular enzymes. So far as I have been able to find, the only basis for this theory is the fact that intracellular enzymes in common with other enzymes are destroyed by exposure to ultra-violet radiation. The object of this investigation is to show that the destruc- tion of living cells, such as bacteria, by ultra-violet radiation .is not due to the destruction of the intracellular enzymes but-to the coagula- . tion of the protoplasm of.the cells by the radiation. The. bacteria used were B. liquefaciens, B. ‘prodigiosus, B. fluores- cens, B. proteus vulgaris, B. pyocyaneus and B. subtilis. These bacteria were chosen because they possess the property of liquefying gelatine, this property in turn being dependent upon the intracellular proteolytic enzymes. Twenty-five cubic centimeters of liquid contain- ing great numbers of B. liquefaciens were exposed in an open vessel to the radiation from a quartz-mercury vapor burner, operating at 140 volts, 3.3 amperes, at a distance of 10 cm., until they were dead as ‘was indicated by negative results on plating. By means of a centri- fugalizing machine the dead bacteria were thrown down and subse- quently ground up in a mortar with sand and 30 per cent alcohol. © In this way the intracellular enzymes were extracted from the dead bac- teria. All of the bacteria named above were treated after this manner. Ten cubic centimeters of the alcoholic extract of the different kinds of dead bacteria were introduced into- separate test-tubes containing gelatine. Ten cubic centimeters of liquid containing the different kinds of living bacteria were also introduced into tubes containing gelatine. These tubes were permitted to stand at room temperature 429 430 | W. E. BURGE for ninety-six hours. At the end of this time the extent to which the gelatine had been liquefied in the different tubes was measured. The measurements are given in table 1. TABLE 1 Under the different kinds of bacteria are shown the extent of liquefaction ef gelatine by the living bacteria and by extracts of the dead bacteria LIQUEFA+ | PRODIGIO-| FLUORES-| PROTEUS | PYOCYA- BACTERIA CIENS sus cens |vunearts| nevus | SUBTILIS mm, mm. ‘mm, mm. mm. mm. Extent of liquefaction by living bacteria........... 12 8 6 6 5 4 Extent of liquefaction by Ges extract of dead bacteria... 10 7 6 5 4 4 It may be seen that the gelatine in the tube containing living B. liquefaciens was liquefied 12 mm.; that containing living B. prodigio- sus, 8 mm.; B. fluorescens, 6 mm.; B. proteus vulgaris, 6 mm.; B. pyo- cyaneus, 5 mm.; and B. subtilis, 4mm. It may also be seen that the extract of dead B. liquefaciens had liquefied 10 mm. of gelatine; B. prodigiosus, 7 mm.; B. fluorescens, 6 mm.; B. proteus vulgaris, 5 mm.; B. pyocyaneus, 4 mm.; and B. subtilis,4 mm. If the amount of gela- tine liquefied by the living bacteria be compared with that liquefied by the extract of the corresponding dead bacteria, it will be found that there is very little difference in the extent of liquefaction. This is taken to mean that, while the ultra-violet rays had killed the bacteria, it had affected very little their intracellular enzymes. These experi- ments would seem to render untenable the theory that ultra-violet rays kill living cells by destroying their intracellular enzymes. / The following experiments were carried out to show that ultra- violet radiation kills living cells by coagulating their protoplasm. Several drops of water containing great numbers of paramecia were introduced into a shallow glass vessel and covered with a quartz plate. The glass vessel was then placed on a block of ice under a quartz mer- cury vapor burner operating at 140 volts, 3.3 amperes, at a distance of 5 cm., and in this position the organisms were exposed for twenty minutes. A drop of the liquid containing the dead paramecia was mixed with a drop containing living ones on a glass plate and covered with a cover glass. Having located under a microscope a dead or- ganism and a living one lying close together a micro-photograph was ACTION OF ULTRA-VIOLET RADIATION ON LIVING CELLS 431 made of them. Similarly micro-photographs were made of paramecia killed by heating to 45°C. and 90°C. respectively. These photographs are shown in figure 1.—‘The upper organisms under A, B and C are living transparent animals; the lower one under A was killed by heat- ing to 90°C., the lower one under B by heating to 45°C. and the lower one under C by exposure to ultra-violet radiation. By comparing the lower organisms under B and C it may be seen that there is no dif- ference in the appearance of these two organisms, both being slightly more opaque than the living organisms. The lower organism under B was killed by the coagulation of its protoplasm by heat and since _ there is no difference in the appearance between this one and the lower Fig. 1. Microphotographs of paramecia. The upper ones under A, B and C are the normal transparent living animals; the lower one under A was killed by heating to 90°C.; the lower one under B by heating to 45°C; the lower one under C by exposure to ultra-violet radiation. organism under C, which was killed by exposure to ultra-violet radia- tion, it would seem to be fair to assume that the latter was killed by the coagulation of its protoplasm by the radiation. By comparing the lower organism under A with that under B it may be seen that the lower one under A is very much more opaque than the lower one under B. This greater opacity is explained by the fact that proteins are more firmly coagulated at a temperature of 90°C. than at a temper- ture of 45°C. It will be noticed also that the lower organisms under B and C which were killed by heating to 45°C. and by exposure to ultra- violet radiation respectively had begun to disintegrate while the lower one under A had not begun to do so because of the firmer coagulation of the protein of this organism heated to the higher temperature. 432 W. E. BURGE Henri (1), Hertel (2) and others observed that when protozoa were exposed to ultra-violet radiation the body became swollen, water drops appeared on the surface and the organisms finally disintegrated, but they did not observe any coagulation produced by the radiation. The failure of these observers to obtain conspicuous coagulation in the organisms was due to the fact that the radiation to which they ex- posed the organisms was not of sufficient. intensity to coagulate firmly the protoplasm. If paramecia are heated to 40°C. they are killed after a time, but very little indication of coagulation is produced as is indi- cated by the fact that there is very little decrease in the transparency of the organisms thus killed. By increasing the temperature, however, at which the organisms are killed, the protoplasm becomes firmer and the animals more opaque. Similarly by killing the organisms by ex- posure to ultra-violet radiation of low intensity, a very inconspicuous amount of coagulation is produced, and hence there is very little change in the transparency of the organisms. If the intensity of the radiation is increased, however, the coagulation of the protoplasm, as well as the opacity of the animals, becomes more marked. CONCLUSION Exposure of living cells to ultra-violet radiation of sufficient inten- sity to kill the cells does not decrease to any appreciable extent the activity of the intracellular enzymes. Evidence is presented in this paper to, show that ultra-violet radia- tion kills living cells by coagulating their protoplasm. BIBLIOGRAPHY (1) Henri, Henri, pes Bancets pt WurMser: Etudes de photochimie bio- logiques, Paris. (2) Herrex: Ueber physiologische Wirkung von Strahlen vershiedener Wellen- lange. Zeitschr. allg. Physiol., v, 1 and 95. ‘THE EFFECT OF THYROID FEEDING ON THE CATA- LASE CONTENT OF THE TISSUES W. E. BURGE, J. KENNEDY anp A. J. NEILL From the Physiological Laboratory of the University of Illinois Received for publication April 4, 1917 Tt has been observed that when animals are fed thyroid ’or an extract of the gland they lose weight and strength. The increased oxygen intake and carbon dioxide output of such animals show that oxidation is augmented while the increased nitrogen elimination indicates an _ inereased tissue destruction. Magnus-Levy (1) observed an increased carbon dioxide output in a man fed upon thyroid extract and an in- ~ creased oxygen intake in cases of exophthalmic goiter. Fritz Voit (2) found that thyroid feeding increased protein metabolism in dogs. Anderson (3) observed a decrease in metabolism in cases of myxedema as was indicated by a decreased oxygen intake and carbon dioxide output, and that the metabolism was increased to normal by thyroid feeding. As a result of these and similar observations it is generally accepted that the effect of thyroid feeding or of hypersecre- tion of the glands as in exophthalmic goiter is to increase oxidation and tissue destruction, while in myxedema there is a decrease in metabo- lism. It has been observed that when oxidation is increased or de- creased in a tissue the catalase content. is correspondingly increased or decreased (4). Since thyroid feeding causes an increase in oxida- tion a corresponding increase in catalase should be found if the rela- tionship between oxidation and. catalase is to hold. It has also been observed that when the catalase content of a tissue is decreased the tendency of that tissue to undergo autolysis is correspondingly increased (5). Since thyroid feeding increases autolysis in muscular tissue, for example, as is indicated by a loss in weight and strength of the muscles, there should be a corresponding decrease in catalase in the muscles and in all other tissues in which autolysis is increased. It is known that the autolyzing enzymes in common with all the ordinary enzymes are easily oxidized and destroyed. 433 434 W. E. BURGE, J. KENNEDY AND A. J. NEILL The object of this investigation was to determine if thyroid feeding increases the catalase content of certain tissues, which would account for the increased oxidation in animals fed thyroid, while in other tis- sues, such as the muscles and fat, it causes a decrease in oxidation which would account for the increased autolysis in these tissues. The ani-. mals used in these experiments were cats. They were placed in sepa- rate cages and fed twice a day 5 grams of desiccated thyroid mixed with 60 grams of ground meat. It was found necessary to vary the diet frequently by mixing the thyroid with different kinds of meat in order to induce the cats to eat. Fresh white fish, canned salmon, beef sausage and liver were among the meats used. The control or normal cats were fed the same kinds and amounts of food as the thyroid cats except that there was no thyroid added to their food. Even with every inducement it was found that certain cats refused to eat after the first day or two. The data given in this paper were obtained from cats that had eaten thyroid for at least five days. After this period of feeding the cats were used as soon as they refused ‘to eat or when they showed great emaciation. None of the animals were fed eee longer than two weeks. After etherizing the cats approximately 25 cc. of bide were col- lected from each cat and allowed to clot. The blood vessels of the animals were then washed free of blood by the use of large quantities of 0.9 per cent sodium chloride at 38°C., as was indicated by the fact that the wash water gave no test for catalase. The heart, the back (latissimus dorsi, trapezius) and leg (biceps, semi-membraneous) mus- cles were removed and ground up in a hashing machine. The clotted blood was pressed through several thicknesses of cheese cloth and ground up in a mortar. The catalase content of the muscles was de- termined by adding 1 gram of the hashed muscle to 45 cc. of hydrogen peroxide in a bottle and as the oxygen gas was liberated it was con- ducted through a rubber tube to an inverted burette previously filled with water. The volume of gas was read off directly from the burette where it had displaced the water. After reducing this volume to stand- ard atmospheric pressure the resulting volume was taken as a measure of the catalase content of the gram of material. In determining the catalase content of the blood, 10 drops of blood were added to 500 cc. of hydrogen peroxide in a large bottle and as the oxygen gas was liberated it was conducted through a rubber tube to a large inverted graduated cylinder previously filled with water, After reducing the volume of gas which was read off directly from the cylinder to stand- 4a EFFECT OF THYROID FEEDING ON TISSUE CATALASE, 435 ard atmospheric pressure, the resulting volume was taken as a measure of the amount—of-catalase in the 10 drops of blood. The hydrogen peroxide used in all these determinations was prepared by diluting commercial hydrogen peroxide with an equal volume of distilled water. It was found very necessary to use the same make of hydrogen perox- ide in all the determinations since the different makes gave different results. For this work a stock supply of about 200 liters of hydrogen peroxide was purchased and kept in a container in a dark and cool place. The results of the determinations for the blood and heart of the normal and the thyroid-fed cats are given in table 1. TABLE 1 After blood and heart are given the number of cubic centimeters of oxygen liberated in ten minutes from hydrogen peroxide by 10 drops of blood and 1 gram of hashed heart respectively CAT AVERAGE AMOUNT OP $12 ge) ee Be Oe eg 11d eee ee ce. Blood OS 560} 430} 970|1040) 640) 720) 630) 630) 750} 890 726 Cat fed thyroid...,..... 1280) 2285) 1850) 1520) 1940/2220/2560|2200/1570|1460} 1888 Heart Deorar Gat...) Ss... 210} 219} 228} 248} 228) 220) 228) 212} 204} 213 221 Cat fed thyroid......... 154} 126} 162} 198} 198) 184) 190) 160) 115) 120 161 It may be seen in the table that the average amount of oxygen lib- erated from. 500 cc. of hydrogen peroxide in ten minutes by 10 drops , of blood of the normal cats was 726 cc.; that of the cats fed thyroid, 1888 cc. of oxygen. The average amount of oxygen liberated by 1 - gram of the hashed heart of the normal cats was 221 cc. of oxygen, while the average for the hearts of the cats fed thyroid was 161 ce. From these results it is evident that thyroid feeding increased the cata- lase content of the blood by approximately 160 per cent as is indi- cated by the increase from 726 cc. to 1888 cc. of oxygen, while it de- creased the catalase content of the heart by approximately 30 per cent as is indicated by the decrease from 221 ce. to 161 cc. of oxygen. The catalase content of the leg and back muscles and in some cases of the fat was determined, but the results were not very uniform and for that reason they were not included in the table. On the whole, how- ever, it might be said that the catalase content of these tissues and hence oxidation was probably decreased by thyroid feeding. 436 W. E. BURGE, J. KENNEDY AND A. J. NEILL It has been shown that the catalase content is an index to the amount of oxidation in a tissue, being greatest where the amount of oxidation is greatest and least where oxidation is least. From this it follows that thyroid feeding increases oxidation in the blood and decreases it in the heart and probably in the fat and skeletal muscles. Furthermore, it has been observed that when oxidation is decreased in a tissue the ten- dency of that tissue to undergo autolysis is increased. This was shown in starvation, for example, where oxidation is decreased in the skeletal muscles and fat, with a corresponding increase in the rate of autolysis resulting in the carrying into solution of these less vital tis- sues, while oxidation in a more vital organ, such as the heart, remains normal, with a corresponding high resistance to autolysis. The mech- anism by which thyroid feeding produces a loss in skeletal muscles and fat would seem to be the same as that which causes the loss in starva- tion, namely increased autolysis made possible by decreased oxidation in these tissues. The effect of starvation on the heart, however, is different from that of thyroid feeding in that starvation does not de- crease oxidation in this organ while thyroid feeding does, with the resulting increase in autolysis. This decreased oxidation with result- ing increase in the rate of autolysis of the heart may explain the harm- ful effect on the heart encountered in thyroid feeding for cbesity. It may also account for the characteristic heart disturbances in exoph- thalmic goiter where there is’a hypersecretion of the thyroid glands. The increased oxidation in animals to which thyroid is fed may be accounted for by the great increase in catalase and hence in oxidation in the blood. Many more cats were used in this investigation than the numbers indicate in the table. Most of these were the animals that refused to eat the thyroid after a day or two of feeding. Some of these animals - were used as soon as they refused to eat, while others were kept and starved until they began eating again. The results from these cats indicated that thyroid feeding began to decrease the catalase in the heart after about two days, while it required four or five days’ feeding to increase the catalase content of the blood. CONCLUSIONS 1. Thyroid feeding increases the catalase of the blood and decreases it in the heart and probably in the fat and skeletal muscles. 2. The increased catalase of the blood may account for the increased oxidaticn in animals to which thyroid is fed, while the decreased cata- _ EFFECT OF THYROID FEEDING ON TISSUE (CATALASE d set e heart, skeletal ss ogy and:fat may account for the srioreniiae : : , the idea being that when oxidation is de- n these ie: a femaiher amount of the autelyzing enzymes is ind destroyed, resulting in an increase in the rate of autolysis. BIBLIOGRAPHY y: Berl. klin. Wochenschr., 1895, xxx, 650; von Noorden’s dbuch der Pathologie des Stoffwechsels, 1907, 325. itschr. f. Biol., 1897, xxxv, 116. ie oo Stockholm, 1898 ten | in Tigerstedt’s Leni eee der on Journal, 1917, sti 58. CONTRIBUTIONS FROM THE BERMUDA BIOLOGICAL STATION FOR RESEARCH NO. 67 RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS BLOCH HOVEY JORDAN Ss Received for publication April 4, 1917 CONTENTS I. Introduction: «.. .......... 20.0. peeinarele gamete « BU ne ~ 438 A. The. problem... «...0:5:..2.5:45 Sc Guy siapee pie ie = eae pla 488 B. Review of literature... ... ¢is¢.isueues cake + ee 440 II. Description of experiments... >... 2:52, c4: -u56 5 + te ss esas 1 442 A. Posterior and lateral orientation. 255 os so. «ss (sips os ie 442 1, In:groups of fishes) 57s a Veins we en 442 . 2:,In mdividual fishes: ..c2.c8ss casa ea 443, B. Experiments on regional sensitivity..................0.ceceeeeeee 445 C. Summary of normal rheotropiem. ji... 5. 6250... 6s +i doe 446 D. The end organs concerned in rheotropism.....................+.- 446 1.-Method of determination... 200.....5 0 4... 0... see 446 2. Observations and experiments...........-5.5 20000 senceeeeee 446 a. Observations. >. .cuiasaes oa. eee 446 b. Experiments, Gisgest Got px os)e sald’s 3 | | lar & < Pe umber of the ‘ observation. . .|1/2/3/4|5|6/7|8)9|10/11/12)13) 14/15/16) 17|18)19|20/21)|22|23/24/25|26/27/28|29|30/210| 7 {106 “entation q : I. Posterior... . |6/6/4/3|5)6|5|5/4 4| 5} 5) 5} 5) 6) 5] 6} 5} 4] 6} 7| 5} 2) 3) 5) 2) 3) 5) 4) 5/141) 4.7167. 2. Lateral... . .|1/1/2/4/2|1/2/2/3) 3} 2| 2} 2} 2) 1} 2) 1) 2) 3} 1} O} 2} 4! 4) 2) 5) 4) 2) 3) 2) 67) 2.2)31. 3. Anterior.... 1 | 2/0.06} 0 the exact nature of this reaction—whether the two positions are due to differences between individual fishes or are phases of one reac- tion. Single groupers, however, when studied continuously showed that both positions are assumed in the course of one reaction. 2. In individual fishes. Individual hamlets were tested in a spacious oblong aquarium (fig. 2) provided with plane glass front and back to allow uninterrupted observation from the sides as well as from above. A glass tube 1 cm. in diameter and so directed as to make about equal angles with two adjacent sides of the aquarium, delivered a strong current (C) diagonally across the tank. The volume of this current was approximately 0.1 liter per second. It is the only one which is significant for the purposes of this investigation. Those peripheral to this were not of sufficient strength or regularity to influence the fish in any con- sistent manner. These currents ' : E x N — Ay ag a U Fig. 2. Diagram of aquarium (20 by 30 inches) and currents as seen from above. C, Main diagonal current. were studied by the use of floating and suspended objects, and the results plotted show their main features. The fish under investigation (fig. 3) seemed to prefer the region of the strongest current; that is, it remained near the source of the cur- rent, shifting from one position of quiet to another, settling to the bot- tom, or remaining suspended at a fixed place.in the current for a few moments, and then again changing position slightly. All the while, however, the fish assumed either posterior or lateral orientation to the 444 HOVEY JORDAN main diagonal current (C). It held this position in the tank indefi- nitely. When left in the current for two or three hours no change of general position was noted. When the current was shut off, the fish under investigation would swim to the bottom or to one of the corners of the tank, where it would remain relatively quiet. In figure 3, one out of several records is reproduced to show a typical . reaction. The diagonal line (C) shows the direction of the main current. The elapsed time in min- utes, from the beginning of ob- servations, is given in table 2 for each of twelve successive positions. These positions are indicated in figure 3 by straight lines, the an- terior end of the fish being denoted by a short cross line, and the suc- cessive positions, by consecutive numerals placed near the end rep- resenting the tail. It should be noted, however, that in this particular experiment, contrary to the most of them, the percentage of posterior and lateral orientations is nearly equal. That posterior orientation is the purpose Fig. 3. Twelve successive positions occupied by a normal fish with refer- ence to the main diagonal current, C. To avoid confusion only six positions are indicated on eath of the two dia- grams. Position 11 was retained longer than any other; it is therefore re- garded as the most significant orienta- tion. One complete reaction is re- garded as occupying the interval of time between two successive assump- tions of such a position. of shifting the position, is sug- gested by the fact that the fish © remained stationary for a con- siderable time only when it was almost directly tail into the cur- rent, position 11. After this period (about two minutes) it again be- gins a series of changes in position, like those shown in figure 3, which lasts for about four minutes, whereupon another period of rest ensues. What I have called one complete reaction, then, requires about seven * minutes. time head into the current. It is most important to note that the fish did not at any The two different experiments, one with groups and the other with individuals, are consistent in showing that posterior and lateral orien- tation to a current is the normal reaction of Epinephelus striatus, _ Elapsed time in minutes.............| 0 | 3 RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS 445 TABLE 2 ed in assuming the positions, 1 to 12, shown in figure 3 , Time elaps NUMBER OF THE POSITION 1p SPS Fe, SEO Ee ae 110 Ib Tis woo 1 | 13) 13} 2 | 3 | 33} 32) 43) 7 Time intervals between positions in cei. $43 4 ek 1 ee B. Experiments on regional sensitivity A study of the relative sensitivity of various parts of the body was undertaken with the hope of finding some evidence as to the nature of —-—------ + = 2 ’ "a AUTHOR’S CORRECTION 444 (Insert opposite page 444, Volume xliii, 1917) The sentence (page 444, paragraph 2) beginning: “It should be noted, how- ever,”’ should be interchanged with the first part of a sentence on page 450 (paragraph 1), viz., with “It should be noted that . . . . the positions were posterior.” . reaction, a swimming-or-packimg-away trom we curren. 2v a pus- sible, however, to divide the body into five regions based on their rela- tive susceptibility to stimulation by such a current. In the order of promptness of reaction these are as follows: lip region (seven seconds) ; caudal fin (sixteen seconds); dorsal fin, posterior part (twenty-two sec- onds) 3 cheek and operculum (twenty-five seconds) ; sides of body(about thirty seconds). The belly was not tested because of its inaccessi- bility. Thus it appears that the lip region is by far the most sensi- tive part of the integument tested. If stimulation of the lips is pro- longed, the hamlet becomes very vigorous in its attempts to escape. 3 That the fins are not essential in rheotropism is indicated by the fact that when either dorsal or caudal fins are removed, the normal reaction is unaltered. It was also noticed that fishes whose fins had become badly frayed by long cap- tivity were normal in their responses to currents shown in figure 3. Laps 446 HOVEY JORDAN When ‘‘cornered” by the current it literally stands on its head, a ter- mination of the negative reaction which is extremely unusual among fishes. This high sensitivity suggests at once an explanation of pos- terior and lateral rheotropism, for it may be that stimulation of the lips by the current in the aquarium or cage was so strong as to produce decided irritation, and thus to cause the fish to place. that portion of its skin in a less exposed position. C. Summary of normal rheotropism The foregoing experiments show: 1. That to a moderate artificial water-current a normal orientation of Epinephelus striatus, in groups or individually, is posterior or lat- eral, as phases of one complete reaction, but almost never anterior. 2. That the lips are the most sensitive integumentary region, other regions being less sensitive in the following order; tail>, dorsal fin>, side of head>, middle of body. 3. That the peculiar posterior and lateral reaction to a current is perhaps an attempt to protect from the current the highly sensitive — lips. D. The end organs concerned in rheotropism 1. Method of determination.. In searching for the end organs con- cerned in rheotropism, it was, of course, necessary to consider all pos- sible sensory cells. My conclusions relative to the significance of equi- libration (semi-circular canals), muscle sense, and pressure sense in these reactions are, for the most part, based upon observations only. The lateral-line organs, eyes and cutaneous receptors, on the other hand, were experimentally eliminated, and each operated fish was care- fully studied to detect any resulting variation of the response from that of the normal individual. It was established by these experiments that the end organs concerned in rheotropism in the hamlet are located in the integument and are probably the tactile corpuscles. 2. Observations and experiments: a. Observations. When confined in large volumes of still water, groupers are seen usually to lie inactive on the bottom of the tank. In captivity they swim about very little, seem- ‘Tt is important to note that the same reaction can be induced by the use of a fine glass rod (tactile stimulation), and also that the variation in regional sensitivity to such stimulation corresponds exactly to that described for stimu- lation by the current. RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS 447 ing to prefer muscular repose to exertion, the fin movements being few and slow; pectoral fins are vibrated about twenty-three times per minute. ‘The application of a localized current of little force was suf- ficient to start the fishes from a position of complete rest, but when they were beyond the range of the current, they again settled to the bot- tom. This behavior indicates that the agreeableness of muscular ef- fort is not sufficient to cause any prolonged swimming. On the other hand, the exertion and possible fatigue involved in maintaining a rela- tively constant position in a current which is broad enough to cover the whole fish might be expected eventually to produce a negative re- action. Instead of this, however, the fishes remained for an hour or two in the strongest part of the current (fig. 3), seeming to prefer it to the quiet water. This leads one to the conclusion that the muscular effort necessitated in these reactions is not in itself a deterrent factor. Many times, too, when a localized current was directed against the side of the fish, lying at the bottom near the wall of the aquarium, it was observed that one of the pectoral fins—extended horizontally to the wall—served to keep the body of the fish from contact with the side of the tank. In these cases the effort involved in maintaining this position did not cause the reaction time to vary. With a view to ascertaining what importance, if any, attaches to the pressure sense, I made use of a current of water directed through the glass tube (1 cm. in diameter) already referred to in other experi- ments. If the pressure sense is a factor in rheotropic response, it is to be expected that the response to a very strong current would be more prompt than to a weaker one. Accordingly I repeatedly subjected the same fish successively to a weak current (about 1/28 liter per second) and to a stronger one (about 1/8 liter per second). Though the latter was of sufficient force to produce an appreciable indentation of the skin and musculature of the body, the reaction time was not shorter than in the case of the weaker current. I may also add that fishes can be pressed by the hand against the side of the aquarium with considerable force without causing any definite response. In his study of the pressure sense as a possible cause of rheotropism, Lyon (2) (p. 154) enclosed fishes (silver sided minnows) in long stoppered bottles which were floating down the stream. He found that under these conditions, with all pressure stimulation thus eliminated, the fishes responded normally, by swimming in a direction opposite to the drift of the bot- tle, in an attempt to keep the visual environment constant. My ex- periments, though not of such fundamental importance as Lyon’s, 448 HOVEY JORDAN tend to substantiate his conclusion that [considerable] pressures do not cause or influence the rheotropic reactions. Observations were also made upon the equilibration of hamlets. They were studied in still water and when subjected to a current suffi- ciently strong to cause a change in the direction of the dorso-ventral axis. Any differences between the rheotropic response of fishes whose dorso-ventral axis is normal and those in which this axis has been dis- placed would suggest that the organs of equilibrium may be involved. An individual fish when in quiet water usually lies in such a position that its dorso-ventral axis makes an angle of 10° to 15° with the normal (fig. 4). When this angle was doubled by a current from a glass tube directed against the side of the fish there was no variation in the time required to produce a negative reaction. Moreover.the motion which restores the fish to an approximately vertical position usually follows, and never precedes, this rheotropic response; whereas, if the fish is carefully put in the same oblique position by a slow displacement with - the hand, instead of by the current, the righting movement takes place much more promptly. It seems, then, that any slight disturbance in the hamlet’s equilibrium which the current might cause would neither produce nor in any way affect the observed behavior. This conclu- sion is in accord with Parker’s conclusion (6) (p. 203) to which reference has already been made. b. Experiments. Lateral-line organs can be excluded from the list. of possible rheotropic receptors for two reasons: First, when the cur- rent is directed immediately against the lateral-line canal upon a lim- ited mid-body area, the slow response (thirty seconds) characteristic of regions both dorsal and ventral to the lateral-line, but excluding it, is neither quickened nor retarded. Secondly, a hamlet in which all of the lateral-line nerves had been severed responded normally to the current. ‘This experiment confirms Parker’s results (4) (6) from a similar test made upon Fundulus. In order to determine whether the visual organs are essential to these reactions, experiments were also performed successively on several individuals after enucleation of both eyes. Fishes thus blinded were subjected to conditions of stimulation identical with those in the tests which were made upon unoperated hamlets (fig. 2) and the succes- sive positions which they assumed were recorded. One of these records is reproduced (fig. 5, table 3) for comparison with figure 3, which shows the consecutive orientations of a normal fish in a like énvironment. RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS 449 Fig. 4. Photographs of resting fishes tipped at a characteristic angle from the vertical. A, Side view; B and C, front views; D, side view of fish, showing the whole aquarium. 450 HOVEY JORDAN The two records, while showing slight variations, are remarkably alike in the time of response and the number of different positions as- sumed in changing from an almost lateral orientation to an approxi- mately posterior one. It should be noted that in this series of orien- Fig. 5. Twelve successive positions assumed by a blinded fish in response to the current C. As in figure 3, only six positions are shown in each diagram. tations 75 per cent of the positions were posterior, and that, as in figure 3, no anterior positions were as- sumed. It seems, then, that the eyes of the hamlet are not the es- sential rheotropic end organs. In an effort to locate the cells which are stimulated by the cur- rent, the skin was stripped from one of the more sensitive body areas. It was impossible to obtain any response from a current which was directed against the subcuta- neous structures (muscles, etc.) thus exposed. In a few cases in- definite reactions, which were much slower than normal, were observed, but they were not characteristi- cally rheotropic in nature. These results lead one to the conclu- sion that the end organs concerned in rheotropism are located in the integument. TABLE 3 Time elapsed in assuming positions 1-12 (fig. 6) NUMBER OF THE POSITION |3|4|s 6|7|8| 9 |10| | 12] 18 Time elapsed in minutes.........| 0 Time intervals between positions in Minttesewens sak oo ee $ 3) 1] 13] 13] 2 | 23) 3 | 3a} 4 | 4a] 7 4040 4).8 49 £9, 2 2 eee The problem of determining the particular type of sense organ which is sensitive to these currents is resolved, then, into a physiological RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS 451 study of the hamlet’s cutaneous end organs. Of these only thetactile corpuscles are significant, because the receptors for chemical, photic and thermal stimuli plainly can not be involved in these rheotropic responses. It has been stated previously that the areas of greatest cutaneous sensitivity in the case of both touch and current stimulation have the same distribution. Cocaine is known to inhibit the functioning of the end organs of touch. It was, therefore, used to ‘eliminate, func- tionally, the tactile corpuscles, in order that their relation to rheo- + tropism might be determined. Of all available areas the lips were chosen for the application of this reagent because the experiments had shown that their stimulation gave the most marked and peculiar re- sponses. Whether the lips are stimulated by a fine glass rod or by a moderate water current, the fish performs the very curious reaction of either backing away violently or of standing on its head. The method of treatment with the cocaine was as follows: the fish was removed from the éxperiment tank and the lips were immediately immersed in a 0.1 per cent solution of sulphate of cocaine for about ten seconds; this was supplemented by bathing the lips with the same solution ap- plied by means of a soft cloth at about ten-second intervals for fifty seconds. After this treatment the fish was returned to the tank and its responses to stimulation by a glass rod and by a weak current (about 1/28 liter per second) were noted. By a repetition of the treat- ment it became evident that there was a slowing down in time of re- sponse with increased exposure to the solution and that this was pre- cisely the same for both types of stimulation. Two repetitions of the . first treatment—in all about three minutes—were usually sufficient to inhibit completely all responses to either type of stimulation; neither the glass rod nor the localized current then produced any reaction. In subsequent trials the lips were treated continuously—without. periodic subjection to stimulation—for a period of about three minutes. The effect was the same as in the preliminary treatment just described. After such administration of cocaine the fish swam about in a slightly abnormal manner, manifesting an irritation due, doubtless, to the drug. ‘This insensitivity of the lips lasted about a minute, sometimes a few seconds longer. Then a gradual functional recovery occurred until, at the end of three to four minutes, normal responses could be obtained by the use of either stimulus. It is most significant that the time of disappearance of normal sensitivity, as well as that of its re- appearance, was absolutely the same for both kinds of stimulation. 452 HOVEY JORDAN This fact indicates that those sensory cells which are stimulated by touch and defunctionized by cocaine are also the cells which are the primary end-organs of the rheotropic response. | 3. Summary of end-organ determination. 1. The end organs of the hamlet essentially concerned in rheotropism are located within the integument. 2. The regional distribution of sensitivity to a water-current and to touch is the same. 3. Cocaine applied to the lips for about three minutes renders those organs insensitive both to touch and to currents. 4. These facts indicate that the end organs of touch serve also as the essential end organs of current stimulation in the hamlet. 3 5. Other sensory cells may be more or less affected by currents in some fishes, but they appear to be only accessory end-organs of rheo- tropism, and in the responses described in this paper they evidently play no part at all. ‘ III. DISCUSSION There is much evidence to show that the rheotropic responses of the hamlet, as suggested for Fundulus by Parker (4) (6), are effected chiefly by the tactile corpuscles. His attempt to prove this by immersing the entire fish in a solution of cocaine did not succeed because the general action of the drug entirely destroyed all sensitivity and movements of the fish; but the great sensitivity of the lips of the hamlet has afforded an excellent opportunity to study changes in the fish’s behavior result- ing from the local application of this narcotic. In this case regional anaesthesia, producing insensitivity to touch, is as satisfactory as gen- eral narcotization would be, because it causes a most unique rheotropic response totally to disappear. Some of Lyon’s experiments (2) on rheotropism, from which he con- cluded that the reaction of Fundulus to currents is chiefly to compen- sate the transporting effect of the current, seem to furnish evidence that the integumentary cells (tactile corpuscles) were directly concerned in the rheotropism which he observed. Among these is experiment 9 (p. 157), in which a blinded fish, without touching any solid object— often required for orientation by fishes without eyes—headed into the rushing current. This certainly may be interpreted as a response to direct tactile stimulation of the integument by the water, and Lyon himself admits that this may be called a true rheotropism. In his opinion, however, it is due to the “sliding contact’’ (stereotropism) RHEOTROPIC RESPONSES OF EPINEPHELUS STRIATUS 453 between fish and water, although he admits the possibility of another interpretation involving the idea of unequal pressures on different parts ‘of the body of the fish. Whether the rheotropism induced by this “sliding contact” is the equivalent of stereotropism, with which Lyon believes it is closely related, is a question needing further investi- gation.® The comparative importance, in the behavior of Fundulus, of these two types of impressions—optic and cutaneous—is, I think, suggested by some of Lyon’s interesting experiments. An example of this is his experiment 3 (p. 153). Here a normal fish, surrounded by a rapidly moving artificial environment, is immersed in a current of water flow- ing in the same direction as the environment, but less rapidly. The fish swims in the direction of, but faster than, the current flow, follow- ing the moving environment in rate as well as direction. This is un- questionably a case of optical response: When the environment. stops moving, the current still flows on in the same direction, but with a gradual decrease of speed due to friction; but the fish, having been carried passively by the current, turns, without a reference point, and faces [swims against?] the current. This may be due, as Lyon says, to an apparent reversal of the visual field; but, since blinded fishes. (experiment 9), without touching a reference point, also orient against the current, it seems equally logical to interpret the turning of fishes with eyes (experiment 3) as the result of the normal rheotropic re- sponse to direct tactile stimulation of the integument, which had, during the movement of the optical field, been subordinated to the sight-reflex. If this be a proper interpretation of the results with Fundulus, we have in the hamlet a reversal of the relative importance of the two kinds of stimuli. Here, under the experimental conditions described, all optic stimuli were, apparently, subordinated to tactile impressions; the direct effect of the current being predominant and able to cause entirely normal reactions in the absence of eyes. It is, however, not quite satisfactory to compare the two sets of experiments (those by Lyon and by myself) from this point of view, because the relative amounts of cutaneous and optic stimuli in each are indefinite and vari- able. It is certain that, in my experiments, there was relatively little stimulation of the eyes, because the fish remained in an approximately 5 It seems probable that the currents of a narrow trough would not be suffi- ciently strong nor distinct from one another to simulate solid objects, but that there would be an almost insensible gradation between them. 454 HOVEY JORDAN constant position, and that, in Lyon’s movable-environment experi- ments, their total stimulation was much greater. It can not be said whether the tactile corpuscles were subjected to a proportionate stim- ° ulus or not. It may be that the hamlet, too, would orient to a movable — environment regardless of contemporaneous tactile stimulations’ by the current; but the facts remain that the tactile-corpuscles do, of them- selves, effect orientation, and that this orientation under the above cir- cumstances is unaltered by the presence of eyes. This orientation, it seems to me, is caused by a direct stimulation of the integument by the water currents as such, and to it we should apply the term rheotropism. The response so well described by Lyon as due to optic reflex might then be called a rheoscopic response in view of the fact that it is due to the optical effect of a flowing or moving environment. How the current stimulates these tactile end organs is still a matter of speculation. It may be that differences of velocity in different por- tions of a current provide slight local variations of foree sufficient to cause a definite response on the part of the fish. How the stimulation results in orientation is a further question, for the mechanism and sensation may or may not be the same for rheotropism that they are - for stereotropism. The author wishes to express his appreciation to Dr. E. L. Mark for the privilege of working at the Bermuda Biological Station, and to Dr. Mark and Dr. W. J. Crozier for valuable assistance IOs ad- vice in the work, BIBLIOGRAPHY (1) Verworn: Allg. Physiol., 2te Aufl., 1897, xi, 606. . (2) Lyon: This Journal, 1904, xii, 149. (3) Scuuutze: Arch. f. mikr. Anat., 1870, Bd. vi, 62. (4) Parker: Bull. U. 8. Fish Com. for 1902, 45. (5) TutiserG: Bihang K. Svenska Vet. Akad. Handlingar, Stockholm, 1903, Bd. xxviii, No. 15. (6) Parker: Amer. Nat., 1903, xxxvii, 185. 6 It would be interesting to determine the relative importance of the eyes and cutaneous elements (tactile corpuscles) as sense organs in the orientation and rheotropic motions of many other fishes. FURTHER EVIDENCE REGARDING THE ROLE OF THE VAGUS NERVES IN PNEUMONIA W. T. PORTER anp L. H. NEWBURGH From the Laboratory of Comparative Physiology in the Harvard Medical School Received for publication April 7, 1917 A former communication! recorded the discovery ‘that secti>n of the vagus nerves protects the respiratory cells and prevents their exhaustion in pneumonia. In pneumonic dogs in which both vagus nerves have been cut, the rate of respiration remains normal through- out the disease. In such animals there is no dyspnoea.” It was necessary in that research to cut the right vagus nerve within the chest and to obtain complete recovery from that operation before cutting the left vagus nerve in the neck. Some days after this second operation the dogs were inoculated with pneumonia. The disease ran its usual course, except for the extraordinary fact that the typical dyspnoea was entirely absent. Obviously, the section of the vagus nerve within the chest and the recovery of the animal without infec- tion are troublesome and time-consuming procedures. This method ~ is essential to the demonstration that the exhaustion? of the respiratory mechanism characteristic of pneumonia does not take place when the path of afferent impulses from the lungs to the bulbar respiratory cells is severed by the section of the vagi. It is not, however, essen- tial to the demonstration that the dyspnoea in pneumonia depends on impulses passing from the lungs through the vagus nerves. The dependence of the dyspnoea upon vagal impulses may be shown by cocainizing the vagi while the pneumonia is at its height. The follow- ing protocol is evidence of this. Experiment June 15,1916. At4p.m., 24cc. ofa broth culture of Friedlinder’s bacillus were injected into the right bronchus of an anaesthetized dog weighing 8 kilos. 1 Porter and Newburgh: This Journal, 1916, xlii, 175. 2 Discovered by Newburgh, Means, Porter: Journ. Exper. Med., 1916, xxiv, 583. 455 456 W. T. PORTER AND L. H. NEWBURGH June 16, 9 a.m. Rectal temperature, 40°C. Respiration labored; 80 per minute. The dog was placed on the operating table and lightly but sufficiently etherized. The vagus nerves were exposed in the neck and were surrounded by a layer of absorbent cotton. The absorbent cotton was wet with a 1 per cent solution of cocain. 9.30 a.m. 10.00 a.m. 12.00 m. 1.00 p.m. 2.00 p.m. 3.00 p.m. 3.15 p.m. 3.30 p.m. 5.00 p.m. 6.00 p.m. 6.30 p.m. 7.00 p.m. 7.10 p.m. Respiration, quiet, ‘‘easy,’’ 20 per minute. Respiration, 16. Respiration, 16; temperature, 40°. Respiration, 15. Respiration, 30. Respiration, 60; temperature, 40°. A few drops of cocain solution were dropped on the cotton surround- ing the vagus nerves. Respiration, 16. Respiration, 14; temperature, 39° (the beginning of the fatal fall). Respiration, 14; temperature, 37.5°. The dog lies upon his side; he cannot stand. Respiration, 14; temperature, 37°. Dog semi-conscious. Respiration, 14; temperature, 37°. Dog in coma. Death. Autopsy. Red hepatization of the right middle and both lower lobes of the lungs. CONCLUSION Cocainizing the vagus nerves changes the violent dyspnoea of pneu- monia into quiet, normal breathing. OROKINASE AND SALIVARY DIGESTION STUDIES IN THE HORSE! C. C. PALMER, A. L. ANDERSON, W. E. PETERSON anp A. W. MALCOMSON From the Veterinary Research Laboratories, University Farm, Saint Paul. Minnesota Received for publication April 13, 1917 INTRODUCTION The work reported in this paper may be divided into four main groups or parts, as follows: I. Orokinase—an enzyme produced by the glands in the mouth, which activates the saliva. Il. Bacteria of the mouth as activating agents. Ill. Amylolytic action of mixed saliva obtained from the mouth and esophagus. ; , IV. The amount of complete starch digestion in the mouth. The work was planned and outlined, the tissue extracts prepared and the operations were performed upon the experimental horses by Palmer. Anderson carried out the bacteriological studies and assisted in the operations and preparation of the tissue extracts. Malcomson studied the amounts of starch conversion in the mouth. Peterson studied the amylolytic action of mixed saliva. The entire group par- ticipated in the studies with the activating substances. I. OROKINASE Orokinase is the name proposed by Palmer for the enzyme produced in the mouth and found in the saliva, which makes active the inert saliva emptied into the mouth from the salivary glands. We believe, and our experiments prove, that this enzyme, orokinase, is produced by the buccal glands, and possibly by the lingual glands. 4 Published with the approval of the Director as Paper No. 65 of the Journal Series of the Minnesota Agricultural Experiment Station. 457 458 PALMER, ANDERSON, PETERSON AND MALCOMSON That we should find such an activating substance was suspected before we began our studies. The idea was presented through a study of the literature. Ellenberger and Hofmeister (1) state that mixed saliva of the horse has a very powerful diastatic action, whereas saliva collected from the parotid ducts is inactive. Mathews (2) in reviewing this fact suggests the possibility of a co-ferment or kinase produced by the mucous membrane of the mouth. He also adds that ‘‘the late Dr. Cook told the writer that if the human mouth is carefully washed out with a sterile solution of water or dilute antiseptic, the saliva col- lected from the ducts may be inactive, whereas the saliva which has been in contact with the mucous membrane of the mouth is very active.” Our first work, then, consisted in confirming Ellenberger and Hof- meister’s work, and especially since some workers do not accept this view but are of the opinion that horse saliva is inactive under all con- ditions. Why these investigators have failed to demonstrate the amy- lolytic activity of mixed horse saliva, will be discussed under part III. We have been able to confirm Ellenberger and Hofmeister’s state- ment that saliva is inactive before it reaches the mouth, by studying the amylolytic activity of parotid fistula saliva of three horses, and the glycerine or water extracts of the parotid, submaxillary and sublin- gual glands of eight horses. The same methods of study were em- ployed as those used by Palmer (3) in his studies on ox saliva. Basing our conclusions on approximately one hundred examinations of the substances named, we conclude parotid fistula saliva when stim- ulated under natural conditions (feeding oats) is without trace of amylolytic activity upon cooked or raw starch at least within a period of several hours incubation. Glycerine and water extracts of the three salivary glands are also without action as indicated by clearing of the starch solution, loss of blue color with iodine, or the presence of reducing sugars. We also agree with Ellenberger and Hidémeinior that mixed saliva is very powerful. This conclusion is based upon the study of seventeen positive cases. In eleven horses active mixed saliva was collected from the mouth, and in six it was obtained from the SENET The de- tails of this work are discussed in part III. Another method of demonstrating that mixed saliva is very powerful, is by feeding raw corn and oats; these substances contain no reducing sugars before feeding, but show heavy reduction when caught from an esophageal fistula. This work is discussed in part IV. OROKINASE AND SALIVARY DIGESTION IN HORSE 459 From this work it is evident that in the mouth the previously inactive saliva becomes active, and our work was now directed towards locating this activating substance. Ellenberger and Hofmeister suggested bacteria of the mouth as the activating agents, but we were unable to confirm this, as shown in our bacteriological studies. Our efforts to locate the activating substance were first rewarded on December 7, 1916, when on this date a 50 per cent glycerine and water extract of the mucous membrane of the buccal region gave excellent results. Only a few drops of this extract were required to activate parotid fis- tula saliva and gland extracts from the three salivary glands. If we used a large enough quantity of this extract, it would not only activate the fistula saliva and gland extracts, but it would digest starch itself. We could, however, so reduce the amount until we had a quantity suf- ficient to activate the fistula saliva or gland extracts, but which would not alone digest starch within a period of two hours incubation. Further studies with this activator “‘I’’ as we called it, revealed the following facts: Its ability to activate the saliva was destroyed by boiling. When a few drops of this activating substance were added to 50 ce. of fistula saliva, and incubated for several hours, we could not demonstrate that a small amount could activate an indefinite amount of fistula saliva if given time enough. Our next step was to demonstrate this activating substance in a number of animals. The next three buccal mucous membrane ex- tracts gave negative results, so we began to search anew for the acti- vating substance. In carefully dissecting the mouth, the small buccal glands are found under the mucous membrane of the cheeks and lips, ‘and many of the lobes are quite adherent to the mucous membrane. In fact it is difficult to dissect the mucous membrane away from its underlying parts without removing some of these glands with it. These glands form quite a large mass, just anterior to the commissures in the lower lip, and another large group is found in the buccal region just posterior to the commissures. In addition to these locations, the glands are numerous under the mucous membrane of the buccal region, and here two rows of small ducts present themselves. In some subjects there is a small deposit of dark pigment around the opening of each duct, and the ducts number about one hundred or more. The glands are also quite numerous under the mucous membrane of the lower lip and the openings of their ducts can be easily made out. In a fresh specimen a small quantity of the secretion of these glands can be squeezed out through the ducts and onto the mucous membrane where 460 PALMER, ANDERSON, PETERSON AND MALCOMSON it can be collected. In two specimens this buccal juice would activate fistula saliva, would not digest starch itself, and was destroyed by heat. We have now demonstrated in ten horses that extracts of the buccal glands contain the activating substance. In several of these cases the mucous membrane from the buccal region and the lips was carefully removed, but they were negative in all cases. Our results in the first case (activator I) can probably be accounted for by the fact that some of these glands were removed with the mucous membrane. The acti- vators obtained from these ten cases have not all behaved alike. Some of them would digest starch as well as activate fistula saliva; some would activate the saliva, so that the digestion was carried to the maltose stage, while others would only carry the digestion to the soluble starch or dextrin stage. They were all destroyed by heat. Our best extracts have been obtained from horses freshly killed, but even some of these were unstable and lost their activating power after thirty-six to forty-eight hours: For example, the most powerful activator (No. III) was obtained from a horse a few hours after death; this activator was very powerful and table 1 shows in detail one experi- ment using this activator, but after standing twenty-four hours this extract was inactive. This activator would not digest starch when used alone and was destroyed by heat. We did not succeed in obtaining a potent extract in every case exam- ined. We invariably failed in cases where the horse had been dead for a few days or when the head had been frozen, and in our positive cases the activators seemed to vary in strength, even though we tried to use the same relative amount of extracting material in each case. This may be accounted for by the fact that the activator is unstable, or our methods of extracting are not ideal, or that after death sub- stances are present in the buccal glands which destroy this activating substance. These buccal glands and the openings of their ducts are located and pour out their secretion in places which are well adapted to activate saliva coming into the mouth. The parotid duct in the horse opens opposite the third upper cheek tooth, and as this parotid saliva flows down over the buccal mucous membrane, it directly comes in contact and mixes with the secretion of the buccal glands. The ducts of the submaxillary gland opens opposite the canine tooth, and here in this region just anterior to the commissure of the lips we find another group of glands very similar in gross structure to the buccal and very likely belonging to the same class. Similar glands are also found under the 00 ¢ ‘Youe4s [°09 ¢°T ‘OF pel[l4siq :surezuoo ¢ 00 g ‘YOIRe4S [°09 O'T ‘O*H pet[tysip [09 gg “yoRryxe [BoONg :suTBeyUOD T 09 G ‘yore4s £09 GO ‘OFF peT[I}stp {'00 T “4ovs}xo. puvyls [ensul[qng :surezuo0yD “fy "09 g ‘yoreys £199 6° ‘O*H poTystp f*99 [ “yovajxo puvls Areyixeurqng :sureyuoo Sy 09 ¢ ‘youeys £°09 G'O ‘OF PeT[stp f*09 [ “qouryxo puvls pryoieg :sureyuoo 87 "09 ¢ ‘Yore4s £99 G°O ‘O*H Pe[gsip f'00 [ “BATTeS BlNySTY :sureyuoo Sy ‘00 G ‘qore4s {00 G'Q ‘JoRryXO [BOON {'99 [ ‘JoRI}XO puBIS [VNSuUI[GnY, :surefuod §q ‘00 ¢ fyoue4ys [°00 G'CQ ‘yoRryxO [BOON {'90 [ “JouI}xe puBls Arv][IxvuIqng :suTeyuOD 8 ‘20 @ ‘Yore4s {199 ¢°Q “JoBI}XO [BOON {*90 [ “YoRI}XO pUvIs pryoIvg :surezUOD *q_ "00 G ‘YOre4S [00 ¢'Q ‘JoBI}XO [BOON {°00 [ “BAITVS BlINISIY :surBzUOD *W worjonp Sul -91 ON|- en{d|: -1B9]0 ON c moronp : sul -01 ON enid -1B9]) ON I UOTJONp} UOTJONp} WoTyoNp] woTZoNp) uoToONp - Sul sul sul Sul Sul -O1 ON| -91 ON] -O1 ON} -O1 ON| -01 ON] ONTG] ONTG] ONT] eNTg| oeN[_|-1vo]9 ON|-1BaTD ON|-1v]D ON|-Iv9TD ON|-IBETD ON| SH UWOTJONp} UOTJONp} UoTZONp| uUoTZONp} uoTyoNp sul sul sul Sul Sut -01 ON| -O1 ON} -01 ON] -O1 ON] -01 ON] ONT] ONT! oN[_] eng] eN[_|-1vO]0 ON|-1vOT) ON|-1B9[O ON|-IB]D ON|-IBOTD ON] SH woryonp| WoTJONp} UoTyONp| uoTZONp} uoroNnp Sul sul Sul sul Bul -01 ON| -O1ON| -0O1 ON] -91 ON} -01 ON] ONTG| ONT_] eng] eN[g| oN] q]-1Be]0 ON|-1v9]0 ON|-1B9]D ON |-IBE[D ON|-1BOTD ON] 8 uoTjONp] WOTJONp} UoTJONp} uoToNp| uoTyonp sul sul sul sul Bul -01 ON| -O1ON| -O1 ON] -O1 ON} -01 ON] ON[G| ONT_] ONT] eN[g| oN[_]-1Be]D ON|-1v9[O ON|-IBO]D ON |-1B9]D ON|-IBOTD ON] 8 aq eae Aavoy| AABvoP] a0] Ivy} 9YSIIS| ong! eng] eng) ong) oni IBID IBID IBID IBID} -pisuog) *q \ aqe1e Aavoy] Avo] o107 eq) yySIIg] ong] ontq| oentg| oentg| ontq IvI]O IBID IBID IBID} -pisuog| % Ieayo Aavoy|] Aavoy]] 10] eq} 9ysyg| ong! onjg| onjg] onjq] ent IBID IBID IBID IBIJO} ATIVON| Sq | . Ieo[o AAvoy 10 ITB eq) F4Sg} eng] enfg! enjg_) onfg| eng IBID IvIIQ IBID Ivajp| Apwen| ®y “O81 06 09 08 a =| oO | 06 09 0g st a 06 09 og st a SQLONIN NI NOILONGaH j . SULANIN NI ANIGOI HLIM HOTOO SGLONIN NI ONIUVATI sjon.yza, puns hunayos pun nayos vynjsyf uo asput 1yo.lo fo uoys ay) buynuysuowap yuamrsadxa auo fo .wjaq T ATAVL 461 462 PALMER, ANDERSON, PETERSON AND MALCOMSON mucous membrane of the lower lip, and the secretion of these glands mixes with the sublingual saliva. The lingual glands on the base of the tongue may also produce oroki- nase, since mixed saliva coming from the esophagus is more powerful than any mixed saliva which we have obtained from the anterior part of the mouth. Extracts of the lingual glands (located in the base of the tongue) have been prepared from five horses, and in all cases the extract alone would digest starch either to the soluble starch, dextrin or maltose stage, but they did not activate fistula saliva. We have succeeded in procuring from the mouth of one animal what we believe to be almost pure buccal juice. A grey mare which nor- mally was a willing animal to salivate, had a parotid fistula on the right side. We would stimulate secretion by teasing with oats and corn and from the right cheek obtain a secretion much more viscid than the fistula saliva and which resembled buccal juice.2 We know this was not parotid saliva because it did not appear like parotid secretion and because no parotid saliva was being poured out onto this cheek, and unless saliva crossed the mouth from the left side, we had pure buccal juice. In a few trials, this secretion gave good results, a small amount would activate fistula saliva. A few drops would activate 50 cc. of fistula saliva after several hours incubation and if a large enough quantity was used, it would digest starch. Orokinase can also be demonstrated in the mixed saliva of man and horse. To demonstrate'this action with mixed horse saliva, the saliva obtained from an esophageal fistula should be diluted 1 to 10 or 15, with distilled water and two drops of such a mixture used. Two drops of this mixture will not digest starch, at least within a period of two hours incubation, but when added to 1 ce. of fistula saliva, good digestion results. Table 2 shows in detail one experiment demon- strating the activating properties of two drops of a mixture of mixed horse saliva, 1 to 10 in distilled water. To demonstrate orokinase in human saliva, the human saliva must be diluted 1 to 50 in distilled water. Two drops of this mixture will be inactive or very slightly active (varying with individuals) but when added to 1 ce. of horse fistula saliva the resulting mixture gives good digestion. Table 3 shows in detail one experiment demonstrating the activating properties of two drops of a mixture of human saliva, 1 to 50 in distilled water. * Fistula saliva is clear but not viscid, and flows like water; buccal juice is clear and very viscid; mixed saliva is clear and viscid but not as viscid as buccal secretion. . ‘09 ¢ ‘Youeys {°09 T “10¥BM PoT[I}SIC, :Sureyuo0o 00 G ‘Yoreys [°09 T “BAITS BINISTY :SsureyUoD &o ‘00 ¢ ‘yoreys ‘sdoip Z ‘(OI[-T) BAT[VS eSI0Y poxIUI f*90 T ‘107BM PoT][T}SIG :sureyzuOd Sq ‘09 g¢ ‘youeqs {sdoap Z ‘(QI-[) BAT[BS OSIOY poxtul {'d0 [ “BAITBS B[NYSTY :sureyUoD SW uoronp 3ul -01 ON oniq -IB9]) ON ral worpoONp} UOTJONp} WoTjoNp} uUoTZanp} uUoTyONp sul sul sul sul ’ Bul -O1 ON} -01 ON] -01 ON} -01 ON] -91 ON] ONT] ONT_] ON[_} eN][_] enN[_|-«v9]0 ON|-IvaTD ON|-IB9[D ON |-1vaTD ON |-1B9TD ON] 8H UOTJONpP} UOTyoNp} UoTyoNp} uoTJoONp| UoTyonNp sul Sul sul sul sul -01 ON} -01 ON} -01 ON| -OL ON] -O1 ON] ONTG! ONTg}] ON[g}] ong] oN[_|-1vo]9 ON|-IvOTD ON |-TvO]D ON|-IvOTD ON|-1voTD ON| Sq on qYSI[s} Uoryonp : — pooy} = aregq] «= 4y81g} AoA] -or ON! onTg] ontg}! ong) enTg} ong IBID IBID] AoIVID}] JOIVIID| BSS) ®y oer 06. 09 0g s1 ras 06 09 08 st a 06 09 08 gt oe qaoL SALONIW NI NOIWOAaGH SALONIW NI ANIGOI HLIM HOTOO SHLONIW NI SNIBVATO ti { DAYDOS aSLoy pax ur aspuryo1o Buryynusuowap yuamrsadxa auo fo spwyaq "09 G ‘YouBys [90 [ ‘19yBM PdT]SIG, :SuUIeIUOD CG ‘00 g ‘Youe4s 1°00 [ “BATTeS BINISTY :suTezUOD Sy ‘09 ¢ “youeys ‘sdorp Z “(Qg-1) BAI[eS UvUINY {90 T ‘“19yVM paT[ysiq :sureqzUuoD Sq ‘09 ¢ ‘youeys {sdoip Z ‘(Q¢-[) BAITBS UBUINY {99 T ‘BAITS BINYSIA :surejzUOD Sy uorjonp - Bul -0l ON ent -IB9]9 ON uoljonp| WoOTjoNp| uoTyONp} uoTyonp| woyonp sul sul sul Sut sul -01 ON] -01 ON| -01ON| -01 ON] -04 ON) ONT} ong! onjg| onfg| on[_)|-1ve[> ON|-ave]9 ON|-rve[D ON -IB9][D ON |-IBE]D ON wOoTJONp} UOTyONp} UoTJoNp} WoTyoNp} uoToNp sul Sul Sul sul sul ~91 ON} -01 ON| -OLON| -O1 ON) -O1 ON] ON[G] ONG] oN[g} eng] oen[_)-1va]o ON|-1vo]9 ON|-1BapD ON|-Tva[D ON|-IvI]D ON Ia Aavoy| AAvoy] Aavoy] pooy|) yySyg] ong] ong] eng] eng) ong IBID VIO] Aeva[D| JeIvE]D ayn? 0z1 06 09 08 I a 06- | 09 08 gI a 06 09 08 SI S@LONIW NI NOIWWOONaGa SGLONIW NI ANIGOI HLIM HOTOO S@LONIN NI ONIBVEIN gaoL DayDs UDUUNY paxiuUL UL asDUryolo BuYynsjsuowap quauisadxa auo fo s]wjaq € HIAVL 464 OROKINASE AND SALIVARY DIGESTION IN HORSE 465 _ When a few drops of mixed horse saliva are added to 50 ec. of fistula saliva, and the-mixture incubated for several hours, there seems to be - greater digestion than in a fresh mixture of the same amounts. In _ the few tests we have made with human saliva, we have failed to demonstrate this property. , All attempts to artificially activate horse fistula saliva or the gland extracts have failed. Magnesium salts failed in our experience. Cal- cium chloride probably-does not, or if it does, its action is very slight. Various degrees of acidity and alkalinity, using various acids and bases have also failed. Letting saliva evaporate at room temperature, ‘and using a water solution of the precipitate, failed. Injecting fistula saliva into the mouth failed to activate the fistula saliva. We could not demonstrate orokinase in human saliva, which had been slowly heated in a water bath until the ptyalin had been destroyed. The heating probably killed the orokinase as well as the ptyalin. Inactive horse saliva very likely becomes self-active with age. Glycer- ine extracts (50 per cent) of the salivary glands were kept in the ice box and tests made every few days until putrefactive changes set in. _ We were unable to demonstrate activity except in the case of sublingual extract. This became active at forty.days. There was no evidence of putrefaction at this time. Fistula saliva covered with tolucl was kept at incubator temperature and no activation had occurred at sixty days. . _ There are a number of points concerning orokinase which remain to be worked! out, and our studies are being continued. A few experiments with the pig strongly point to a similar phenomena. II. BACTERIA OF THE MOUTH AS ACTIVATING AGENTS Ellenberger and Hofmeister (4) apparently did not associate the glands of the mouth with the process of saliva activation, but suggested the bacterial flora of the mouth as being the activating agents. In five experiments using cultures from five different horses, we failed to demonstrate activation. The bacteria were obtained from the oral cavity by means of a sterile swab and grown on plain agar for thirty-six hours at 37.5°C. The growth was washed off with sterile water into a sterile bottle. No attempt was made to differentiate between the various bacteria, and the entire growth was used. Various amounts of the bacterial emul- sion were added to 1 cc. of parotid fistula saliva and salivary gland ex- 466 PALMER, ANDERSON, PETERSON AND MALCOMSON tracts. To this was added 5 ce. of a 1 per cent starch solution, and the mixture incubated at 40°C. Digestion was studied by noting changes in the viscosity, the color with iodine and the presence of a - reducing sugar. The. results obtained were negative in the five experiments and we can conclude from this work that bacteria of the mouth do not possess the property of activating the inactive horse saliva which is emptied into the mouth. If positive results had been obtained in these studies, it was our intention to isolate the various organisms found in the mouth, and attempt to determine the relative activity of the different bacteria. Ill. AMYLOLYTIC ACTION OF MIXED SALIVA Investigators differ in their opinion regarding the presence of amylo- lytic enzymes in the saliva of the horse. Ellenberger (5) is of the opin- ion that mixed saliva has strong amylolytic properties. Mills (6) states that horse saliva has a very feeble action upon starches. R. Meade Smith (7) found that horse saliva would convert crushed raw starch to sugar in fifteen minutes, and that mixed saliva had a more ~ marked amylolytic power than individual secretions. .Fred Smith (8) states that according to his observations on the horse, saliva has no chemical action on the raw starch of its food. We have demonstrated in seventeen horses that mixed saliva possesses amylolytic: properties, and that when a good sample can be obtained | this amylolytic action is about as powerful as human saliva. Methods of collecting saliva. As a rule it is difficult to stimulate salivary secretion in the horse. A few horses are found, which are habitual slobberers and from which a considerable quantity of saliva can be obtained without much difficulty. We have been unable to find a satisfactory chemical stimulant. Pilocarpine will stimulate a profuse flow of saliva, but the drug itself will digest starch and enough of it will be excreted in the saliva to demonstrate this action. This fact was shown to be the case by Palmer (9) in his studies on ox saliva, and in repeating this work with pilocarpine, we have been abie to con- firm his statements. In one horse, possessing a parotid duct fistula, 0.5 grain of pilocarpine hydrochloride was administered subcutaneously. A proncunced flow of saliva occurred in about ten minutes and this saliva showed marked amylolytic properties. Fistula saliva from this horse had been previously tested many times when secretion had been stimulated by feeding oats, and was always found to be negative. OROKINASE AND SALIVARY DIGESTION IN HORSE ~ 467 In some cases we were able to stimulate a fair amount of secretion by irrigating the mouth with dilute acetic acid, and giving inhalations of strong acetic acid. Teasing with grains or forage, or placing a bit in the mouth also assisted in some cases. In two horses saliva was being secreted, at the time of collecting the sample, and these cases gave good results. When secretion was stimulated, it was necessary to collect the saliva by means of a spoon, as the animal would swallow as soon as a small quantity collected in the mouth. In the majority of cases only 5 to 10 ce. of saliva could be obtained, and in some animals it was very difficult to collect. In the cases in which it was very difficult to collect the saliva, the small amount of material collected seemed to be a mix- ture of buccal secretion and mucus, and these samples invariably gave poor results. There was a direct relation between the ease with which the saliva was collected and the amount of digestion. When it was difficult to stimulate and only a small quantity of mouth secretion was obtained, the results were either negative or very poor. This fact in our opinion very largely explains or accounts for the negative or poor results reported by some workers. We obtained negative results in a number of horses, but if we could procure these same horses at a time when the flow of saliva was easily stimulated, we invariably ob- tained good digestion. . The collection of the mixed secretions from the esophageal fistula was, of course, attained without any difficulties. Every five to ten minutes the horse would swallow from 5 to 20 cc. of mixed mouth secretions. : Methods of study. The usual methods of noting changes in viscosity, color with iodine and the presence of a reducing sugar were employed. The cooked starch used in our studies consisted of a 1 per cent solution of Dakomin corn starch, and the uncooked starch solution was prepared by grinding whole corn or oats in a mill, adding water, straining through cheese cloth, and using the liquid portion. The tubes containing the mixtures under observation were invariably incubated at 40°C. Results. In eleven horses a potent secretion has been obtained from the mouth, but in several horses we have failed to demonstrate amylo- lytic activity in the mouth secretions. We not only failed to obtain positive results in some horses, but even the eleven positive cases gave varying results. The degree of activity was directly proportional to the ease with which we obtained our sample. The following table gives a summary of the relative activity in the eleven positive cases. 468 PALMER, ANDERSON, PETERSON AND MALCOMSON TABLE 4 ; Summary of eleven positive cases showing when changes became pronounced CLEARING IN MINUTES REDUCTION IN MINUTES 60 | 90 | 120] 15 30 | 60 | 90 | 120 2} 1] 1| 2}—| 4] 2] 38 Number of positive cases........ | 3 | 4 Table 5 shows in detail one experiment demonstrating the amylo- lytic activity of mixed horse saliva, obtained from the mouth. The saliva in this case was obtained from an aged grey mare. The secre- tion was obtained without much difficulty, and was stimulated by teas- ing with oats. Digestion was not as strong as that obtained in some horses, neither was it as weak, and the results in this case are suggestive © of what can normally be expected. It will be noticed that clearing was not recorded until fifteen minutes, and that at this time slight reduc- tion had occurred. Digestion was not complete in this case at two hours, as indicated by the blue color with iodine. The amount of digestion in this case was similar to that reported by Meade Smith. With the exception of one horse, the mixed secretions obtained from esophagus were very powerful. The particular animal which gave the digestion below par was a large grey mare, with a parotid fistula on the right side. The secretion swallowed by this mare differed from that in any of the other. horses, and was of a consistency similar to that of egg white. This secretion possessed weak amylolytic action, but had strong activating properties when mixed with the inactive fistula saliva. This secretion was evidently largely composed of buccal and lingual secretions. Masticated corn or oats obtained from the esophageal fistula in this mare, however, contained about as much sugar as in the other cases studied. . In five horses the secretions collected from the esophageal fistulae were of a uniform potency, and were somewhat more viscid than the parotid fistula saliva. Even in the horses with parotid duct fistulae, (on one side) the secretions collected from the esophagus were not like those found in the above grey mare, but were similar to those obtained from the horses possessing only esophageal fistulae. In these five horses the mixed secretions possessed very powerful amylolytic action. The starch solution became clear in one to five minutes, the blue color with iodine disappeared within this time, and at these intervals the re- duction was very heavy. Table 6 shows in detail one experiment demonstrating the powerful amylolytic activity of mixed saliva, col- lected from an esophageal fistula. 00 ¢ “Yore4s [00 T ‘1048M PoT[T}SICG :sureyuOD ‘00 ¢ ‘YOI4S [00 [ ‘BAITVS BINS pryored osi0zy : Sure}UOD 85 ‘09 g ‘YOrB4s [00 T “BAITeS UBUINY poxTP :SuTeyUOD *q "09 g ‘Yoreys [09 T “BAITRS O810Y Pox :SuTeyUOD *W uorjonp Bul -01 ON oni -189]9 ON a UoTJONp| WOTONp}, woTjONp} uoTyoONp| uoryonp 3uy Bul But Bur Zul “OI ON} -O1ON| -O1 ON] -O1 ON] -OL ON] ONTG! oNnTg] ontg] eng] on[g_|-rvo]> ON|-1v9]o ON|-IvO]D ON|-TBOTO ON|-IvOTD ON| 8H IO]OO}] IO][OO] IOTOO] OOO} AOTOO Aavoy| Aavoy] Aavoyy] Aavoy] Aavoy] on; ON} ON} ON] ON IBID IBID IBID IBID Iwao) " elalé|/élelalalé gm.| gm. | gm.| gm gm. gm. |gm.| gm. Firstiday 3. GiGi ic. ake. Daa 3.0/4.7 |3.0/9.0 |3.0/9.0 |3.0|8.1 Second: day idseu jes winds eae Ged 3.0/4.7 |3.0/2.6 |3.0/9.0 |3.0|4.0 Third day’ .iisc5042 0 ie. a , . s n ‘ _ f yy Pitz ’ . : . ' 4 ¥ t . ‘ ~ * CONTRIBUTIONS FROM THE BERMUDA BIOLOGICAL STATION FOR RESEARCH, NO. 68. THE BEHAVIOR OF HOLOTHURIANS IN BALANCED ILLUMINATION W. J. CROZIER Received for publication May 3, 1917 I. The behavior of phototropic animals when so situated as to be illuminated from opposite directions gives important information re- garding the nature of the mechanism of stimulation. It was held by Loeb (1) that a phototropic organism, when placed at the center of the line joining two equal sources of illumination, should move in a direc- tion perpendicular to this line. The accuracy with which this kind of ~ response is obtainable in a suitable organism is demonstrated by Pat- ten’s (2) experiments with the negatively phototropic larva of the blowfly. Patten further found that the angular deflection of the larva’s path from the perpendicular was so related to the percentage difference in intensity between the two lights, when these intensities were caused to alter in a graded series of ratios, as to produce.a smooth curve when these values were plotted on the axis of ordinates and abscissae, respec- tively. On the basis of this graph, and taking into account the blowfly larva’s method of locomotion, Patten was able to show (3) that the responses appear to be determined by the presence of two photosensi- tive surfaces inclined to each other at a definite angle. The definite path of progression adopted by the larva under the influence of two op- posed unequal beams may be predicted on the assumption that the ani- mal ceases to deflect from its original locomotor path—which in these experiments is a straight line normal to that connecting the sources of light—as soon as the luminous intensity on these mutually inclined sur- faces is made equal by the larva’s position. II. There may be deduced from this principle the corollary that, if the photoreceptive surfaces of an animal are so arranged as to be sen- sibly parallel, no definite position should be assumed by this organism when bilaterally illuminated. For, in such a case, if the balanced lights were equal in intensity, it would be impossible for the animal to place itself in a position which would result in unequal illumination of the two sides; while, if the lights were wnequal, the amount of light received by 510 BEHAVIOR OF HOLOTHURIANS IN BALANCED ILLUMINATION 511 either side could not in any position be made equal to that incident upon the other. It is assumed, of course, that the longitudinal axis of the animal remains straight when orientation is completed. Thus an animal of this kind might be clearly phototropic according to other criteria, and yet fail to satisfy the requirement of orientation under bilateral illumination which Loeb’s contention demands. The existence of such organisms, however, would add further support to thé assumption made by Patten, namely, that the exhibition of a definite > path of locomotion under these conditions depends upon the final equal- ity of illumination on each of the two photosensitive surfaces. III. Certain pedate holothurians have been found to be different from most other echinoderms, in that they exhibit a pronounced bilaterality in their structure, coupled with a fixed direction of progression (4), (5), (6). Typical tropistic behavior is thus made possible. It has been shown that some of these holothurians (Holothuria surinamensis, H. captiva) are photo-negative, that they are sensitive to light over their whole surface, that they give well defined reactions to shading, and are non-reactive to a sudden increase. in light intensity, yet orient away from the light with conspicuous precision. It is impossible to conceive, as I have already pointed out (4), that these animals are ori- ented by light in any way other than through the direct action of illumi- nation upon their integument. One of the holothurians referred to, H. captiva, is the most outstand- ingly bilateral of all the species which I have studied. Its lateral sur- faces are so nearly straight and parallel that, to all intents and purposes, they may be regarded as strictly so; that is, longitudinal elements of its surface are parallel. I have previously described the accuracy with which it orients away from the light. Experiments were subse- quently made to determine the behavior of this species when illumi- nated from opposite sides. Inasmuch as the whole surface of H. captiva is sensitive to light, the higher sensitivity of the oral end should not greatly obscure the inter- pretation of such experiments. When horizontal light is employed, the sides of the animal may be regarded as equivalent to two photosensitive areas. This particular species of holothurian, fortunately, tends to preserve a straight position of the body. In my present study upon this animal, the tests were made in a dark room, the animals being placed in a flat-sided aquarium with their long axes perpendicular to the line connecting the two sources of light. The lights used were two tungsten incandescent filaments, placed at 512 W. J. CROZIER various distances from the holothurian in different experiments. In some cases the animal was induced to begin crawling in a direction per- pendicular to that of the light beams before the latter were turned on. The intensity of the lights was varied by using filaments of different wattage (25 to 100 watts). Since the characteristic features of the re- sults soon became evident in these experiments, no great refinement was necessary in the physical precision of the light adjustments. IV. In no case did Holothuria captiva assume a path of locomotion | definitely related to the intensities of the acting lights. In no case was it constrained to move, after the fashion of the blowfly larva, along a path inclined to the direction of the beams. A number of animals were found to move first toward one light, then toward the other; this was the case when the lights were of nearly equal intensity. Many individuals circled in an aimless, indeterminate way (as when illumi- nated from above), while-others became curled into the shape of the letter S, the oral end, and frequently the cloacal end as well, being bent away from the stronger light. When the intensity of the stronger illumination was made eight or nine times that of the weaker one, the holothurians almost invariably oriented precisely and away from the stronger light. But when they had moved somewhat toward the weaker light, their movements be- came quite irregular. To meet the criticism that possibly the incandescent filaments as sources of light were too small as compared with the size of the ani- mals, the latter being 6 to 8 em. in length, experiments were tried with sunlight reflected from two mirrors, each measuring 15 by 20cm. The holothurians were-in a dark chamber, an opening on either side (10 by 8 cm.) allowing the sunlight reflected from the mirrors to enter the box from opposite directions. One or several lightly smoked glasses were then interposed between one mirror and the corresponding opening in the dark chamber. The behavior of the holothurians under these con- ditions was qualitatively identical with that in the trials with electric lights. Hence I conclude that H. captiva does not maintain a pre- dictable path of locomotion under the influence of bilateral illumination. V. The validity with which this result may be applied to the criti- cism of photic stimulation theories depends largely on the share taken by the general surface of the animal in photoreception. That this share is a considerable one, is shown by the following: - a. Excised pieces of the “dorsal”? integument (including skin and muscle layers) which have been cut from the midbody surface of H. BEHAVIOR OF HOLOTHURIANS IN BALANCED ILLUMINATION 513 surinamensis or H. captiva exhibit local wrinkling contractions when a spot of intense light is thrown upon them. The wrinkling movements cease when the strong light is removed. When first prepared, the pieces of the body wall are strongly contracted; but if they be allowed to lie in seawater for half an hour or so, or if fhe be subjected to gentle _ traction, they become-somewhat relaxed, although the cut edges remain curled. It was on such relaxed preparations that the light test was made. Tube feet on isolated portions of the “ventral” surface also ' bend away from the light. b. A phenomenon closely akin to “circus movements,’ and probably in principle identical with them, appears when these holothurians are exposed to general illumination from above, provided one side of the -animal’s surface has been rendered insensitive to light. In the narco- tizing, cocaine seems to give the best results, but chloretone was also ’ used. In the case of either narcotic, a strong solution (in seawater) was employed and the holothurian, while being held out of water, had one side painted over with the solution several times. The parts so treated soon become quite insensitive to light, while the untreated side is still sensitive, and its tube feet all remain normally functional. On being returned to seawater the holothurians soon assume a curved atti- tude, the narcotised side being the relaxed one.(convex). If the ani- mals are then strongly illuminated from above, they rapidly become curved in the opposite direction, the narcotised side being then the inner, contracted-(concave) portion of the bend. Thus the holothur- ian bends toward the wnstimulated side, whether the non-stimulation is determined by the relative absence of light, or by the enforced inca- pacity of its photoreceptors. VI. These experiments afford evidence: (1) that the amount of light received by the photosensitive surface of Holothuria surinamensis and H. captiva determines their behavior in an illuminated field; (2) that the assumption of a definite path of progression, with respect to bal- anced illumination, depends on the non-parallel relation of photosensi- tive surfaces. BIBLIOGRAPHY — (1) Loxs: Studies in general physiology, Chicago, 1905, Part I, 13. (2) Parren: Journ. Exper. Zool., 1914, xvii, 213. (8) Parren: This Journal, 1915, xxxviii, 313. (4) Crozier: This Journal, 1914, xxxvi, 8. (5) Crozer: Zool. Jahrb., Abt. Physiol., 1915, xxxv, 233. (6) Crozier: Sci., N. S., 1916, xliti, 148. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 43, NO. 4 THE EFFECT OF DEXTROSE GIVEN INTRAVENOUSLY ON BLOOD COMPOSITION AND URINARY SECRETION! | DAVID M. DAVIS From the Laboratories of the James Buchanan Brady Urological Institute, Johns Hopkins Hospital, Baltimore Received for publication May 5, 1917 In a previous paper (1) the writer has summarized the data at hand concerning the relations between the composition of the blood and diuresis. In addition to studies directed toward a broad and general explanation of these relations, the attempt has frequently been made to concentrate attention upon some single substance and its behavior in such a connection. It is desired to present here an addition to these studies in the form of a series of injections of different dextrose solu- tions, using the same animal throughout. Only a few previous studies will be mentioned. Lamy and Meyer (2) could find no relation between sugar diuresis and systemic blood pressure, the viscosity, molecular concentration (as measured by the depression of the freezing point), or dry weight of the blood, or the volume of the kidney in chloralized dogs with oncometers in place. The diuresis according to them may begin at a time when the viscosity of the blood is unchanged or even increased. It is noted that in gen- eral the polyuria runs parallel to the concentration of sugar in the blood, but there are numerous exceptions to this rule. These excep- tions usually take the form of a high blood sugar without diuresis. Fisher and Wishart (3) working in Lusk’s laboratory, find that after the administration of dextrose per os, the hemoglobin of the blood is much decreased, but has returned half way to normal before diuresis begins. The hemoglobin concentration is taken as a measure of blood volume, although its accuracy in this réle is open to some question (4). Ewing (5) determined the specific gravity of the blood, confirming Fisher and Wishart’s observations for glucose administered intra- . 1 Read in abstract before the Society for Experimental Pathology, New York, December 29, 1916. 514 BLOOD SUGAR AND URINARY SECRETION 515 -venously. He adds that diuresis may occur without change in the con- centration (specific gravity) of the blood, or may be absent when the blood shows a well defined dilution. Diuresis is usually dependent on blood sugar level, but one experiment is a notable exception to this tule. When the blood sugar exceeds 0.4 per cent, the urinary. sugar concentration appears to remain fairly constant, while as the blood - sugar falls and diuresis decreases, the urinary concentration goes up. Ewing’s injections were made rapidly. . Epstein (6) studied the relation of glycosuria to blood sugar and concluded that it depends on an increase in the total amount of sugar in the circulation, which must be determined by noting simultaneously the blood sugar concentration and the blood volume (as indicated by its hemoglobin content). When a glycosuria is precipitated by a hyperglycemia in this sense, the urinary sugar concentration must show a tendency to remain constant, since “diuresis in diabetes mel- litus plays an important réle in determining the total amount of sugar _ eliminated in the urine, but has no influence on its concentration or percentage.” This relation is found to be so constant that Epstein thinks that there is a glycosuric constant, like that of Ambard for urea, which is higher in individuals with nephritic kidneys. In such cases a greater hyperglycemia must be present to produce the same glycosuria. _ Arrous (7) working with intravenous injections, contrived a coeffi- cient expressing the relations of the volume of fluid injected and the volume of urine excreted, which remained constant for each concen- tration of dextrose, regardless of the size of the dose. Blumenthal (8) brought forward very clearly the necessity for in- cluding time in the conception of glucose tolerance and utilization, - and Woodyatt (9) with his method of precisely timed injections made possible a more accurate study of the questions above outlined. It was planned, therefore, to have the present work consist of two series of injections, using the same animal for all, that each injection might be better compared with the others. In the first series, the same quantity of glucose, per kilogram-hour, was to be administered each time, but dissolved in varying amounts of water, while in the second, the same amount of water per kilogram-hour was to be in- jected, but containing varying quantities of glucose. In each experi- ment the volume of the urine and the sugar of the blood and urine were to be followed. The general outline of experimental technique was the same as that described in a previous paper (1). No anesthétic of any kind 516 DAVID -M. DAVIS was used. Blood was drawn from the jugular with pipet and needle. This procedure caused in the animal, which was of a very quiet and phlegmatic disposition, no perceptible disturbance, and the blood sugar curves give no evidence of fluctuations from this cause. Urine was removed by the suction catheter which has been described. The injections were made with a small motor-driven pump embodying the same principles as that of Woodyatt (10). The glucose used was purified by several recrystallizations from crude dextrose. For this product the writer is deeply indebted to Dr. George Peirce (formerly of this laboratory). The blood’ sugar determinations were according to the method of Lewis and Benedict. The heating was done in’an autoclave. The urinary sugar was estimated by means of the polar- imeter. The use of small tubes made it necessary to dilute only the very small samples. The urine was cleared with alumina cream, which gave a beautifully clear supernatant fluid. It was found how- ever that there was a slight adsorption of sugar by the aluminium hydroxide; from 1 to 5 per cent, according to the concentration. Cor- rections were made for this adsorption. ~ TABLE 1 DEXTROSE GRAMS WATER CUBIC CENTIMETERS PER KILOGRAM—HOUR PER KILOGRAM— ; cote 40.0 20.0 10.0 2.5 per cent per cent per cent "per cent 0.7 3.5 1.0 5 2.0 5 10 20 80 3.0 15 Table 1 gives an outline of the injections made, with the quantities of water and sugar administered, per kilogram-hour. The percent- age concentration of the injection fluid in each case is shown by the figures in the boxes, each figure representing an experiment actually performed. The duration of the injections varied slightly, but all ran four hours or more except the 0.7 gram to 20 cc. experiment, and in the comparative figures all have been calculated merely for the first four hours, except this experiment. The injection referred to lasted three hours, and is included merely for reference, since a different dog was used for it. It is felt that presentation in the form of curves will bi the most advantageous to‘the reader for grouping quickly the results. These BLOOD SUGAR AND URINARY SECRETION 517 curves have all been drawn to the same scale. The abscissae repre- sent time, the ordinates the amounts of the various substances. The solid line represents the volume of urine secreted, calculated as a rate, Le., cubic centimeters per fifteen minutes. The dotted line shows the blood sugar in per cent. The interrupted (dash) line gives the concentration of sugar in the urine, again as per cent, while the dash and dot line combines the first and third as the rate of excretion, or grams per fifteen minutes for the sugar. The vertical solid line repre- sents the end of the injection. Some of the features of these injections may be best displayed by bringing together the curves from a number of experiments into com- _ posite charts. . _ Figure 8, the first of these charts, has four curves representing the water output rates in the experiments where the sugar injected re- - mained constant. Each is labeled according to the concentration of the injection fluid used. As is to be expected, the more dilute solu- tions produce a greater diuresis, but that from the 80 per cent solution is practically as great as that from the 20 per cent solution. The diuresis begins in each case at the same time, namely about twenty to twenty-five minutes after the beginning of the injections. Figure 9, the second composite chart, gives the rates of sugar output in the same four experiments. Here it is seen that, although the sugar intake is exactly the same per kilogram-hour in each case, the injec- tion of 5 per cent solution, accompanied by a great diuresis, leads to a much smaller excretion of sugar than that of an 80 per cent solution, accompanied by very little diuresis. The other concentrations fit in well between. It must however be observed that this smaller excre- tion, in the case of the dilute injection, is principally a delayed excre- tion, since while with the 80 per cent injection, the sugar output shoots up most rapidly, it has, at about the fifth hour, atttained much the same level with all four concentrations of the injection. This delay in excretion means, naturally, that there has been greater retention in the body of dextrose with the dilute than with the con- ‘centrated injections. The extent of this retention is shown numeric- ally in table 2 in grams per kilogram-hour, for a four-hour period of injection. This makes it seem that a copious diuresis interferes with the excre- tion of glucose by the kidneys, but it is probable that the true expla- nation is somewhat different. Zuntz (11) in metabolism experiments designed to show the protein sparing action of carbohydrate noticed H,0 CC. PER EHR. 250}. 200}. 150}. 1oo|2. Boop Sucar % 6 5% Sor. iced Pen Kein liv ) ee —-— TOTAL SUGAR --- SUGAR% BLOOD SUGAR Urine Sucar | Sucar cm.rengHR| Yo IZ “TNLEGNG as ig _--e-- = ~— 30 I 30 IL 30 Il 30.W 30 ¥ 50- Fig. 1 Urine H,0 | Broop cen Sena Re 300.6 6) 12 10% Sor. ; 2504.5 b Gn ten Ke. Hr. In. i; 510 — H,0 : 7 “ TOTAL SUGAR i ZOGA ee fog ee AE os dea | / =} | ISO}.3 Pa , 376 a Mf rs. ore OG 2. fh ae | 4 | er s ye 4 oh Be 2B cues rs if 56.1 | 1}2 i F| 0 Fig. 2 518 30 I 30 £f 0 DH 3 W WO V 30 20% So. Boop af mM) Per Ke.H ER G. R. Inv. Sucar% | *. Fo 2 a , Beals : ws ae re hae s . Pod 7 s-- 7 * . Eases oe: 7 eich ee : ) ae . =; ; fe pie C= 8a m > er / —— TOTAL SUGAR = / —-- SUGAR % o Ad ‘ “= BLOOD SUGAR S e. ee = Urine Sucar cm.perzHr| % & 150}. 100 50k 30 I 30 I 30 M 30 W 30 Fig. 3 1 B 0% Sor i Loop C M. > Per Ke.Hr_! Sucar % } ates ot ay. i "i ° di Ce. ! 1 oOrnssesne i se aed . tenga err Mitt eeece Py ! ! r H SS . aia H,0 3 eee i J TOTAL SUGAR # > —-— SUCAR % As i i “7 BLOOD a eae eS 3 #- : er es arm me . ; Pe ma! ‘ ae Frag z\\ in es ° | ; . \ >] bye : i prt Z i A at? ae i y ea ¥ ° 4 ’ / F : 3 : | / ) Rey : 1 / | 2 | / Wes bey { : Y 30 Uaine Sucar | Sucar. GM.PERZHR. % 6} 12 5].40 30 I 30 I 30 D 30 W 30 V 30 Fig. 4 519 Urine HO: Boop Sucar | Sucar THR] SUGAR % a GM.PERLHR| % 300.6 6) 12 i 3% Sor 3 2501.5 | fe Goh Per Ke.Hr. Ins. 5 | 10 140 —-— TOTAL SUGAR 5 200.4 a Pore rhst te = 413 ja isk oo | 316 100.2 A|4 50}."i ne OS ; WD i Te we I 30 W 30 VY 30 Fig. 5 Urine fhO Broop Sucar, Sucar LHR] OUcAR% CH.PERZHR| % 300.6 611k 5% Sor. 250).5 iow | Per Ke.Hr. Inu. 5|10 —— 1,0 —-— TOTAL SUGAR ae 3 cpm pele m 4\8 : i50|.3 A ge ola f 24 So) I}2 b “ 30 I 30 IT 30 BD 30 W 30 V 30 Fig. 6 520 BLOOD SUGAR AND URINARY SECRETION §21 IB Seu., ae SOC. Per Kotla Ins. 7s oA s; Z , ~ 10] 20 - l r ' / | i \ 7| 18 / , ie. i “Fl : een =e -_ ¢ 3 \6 ios ° |i t Ps . a eed ' if c 14 at LY = ls i a i —H,O Bs | . 6 Im —— TOTAL SUGAR . —-- SUGAR % 1 ——- BLOOD SUGAR cay ee ; $ 5]10 L. ‘ a ~ ae WW 3 oo” 0 130 130 M3 WV 30 Fig. 7 % Urinary Waren Water Oureut ih ai Constant sucan INTAKE (26m. PER KG-HR.) 300 > —-— 10% Roi = INJECTION 50 200}: 150 ae. 100 eta See ee ore 50 ze eS nOfe” igo Be: ok 30° I 00 WT MIS Eo ee Fig. 8 Urine $ oacutee ca. Totat Sucar Ourrut se HR. Constant Oucar INTAKE (RGM. PER KG-HA.) — 5% =: 10% eT non} INJECTION 5 4 3 L | JOT 30 Tl So mh oe: a 0 ae Fig. 9 522 BLOOD SUGAR AND URINARY SECRETION 523 TABLE 2 th f: WATER CUBIC CENTIMETERS PER KILOGRAM—HOUR DEXTROSE GRAMS 40 20 10 2.5 Sugar grams Sugar grams Sugar grams Sugar grams per kilogram— | per kilogram— | per kilogram— | per kilogram— hour hour hour hour 20° Excreted . 0.43 0.49 0352's) 0.61 4 Retained 1.57 1.51 1.42 1.39 that when, in his animals, carbohydrate was added to the diet, there was a sudden great increase in weight, due to a retention of water. This finding has been confirmed by Wimmer (12), Chauveau?, Bischoff and Voit? and others. These experiments offer further striking con- firmation by an entirely different method of investigation. The following table (table 3) shows the fate of water injected with the sugar. TABLE 3 WATER CUBIC CENTIMETERS PER KILOGRAM—HOUR SUGAR GRAM FRR — 40 20 10 2.5 Water: Water Water Water ce. cc. cc. cc. 2.0 { Excreted 972 718.4 328 364 : Retained 788 178 94 — 258 Although the diuresis was much greater with the dilute solutions, the amount of water retained was, like the sugar, also greater. The retention decreased steadily until, with the 80 per cent injection, there was an actual loss of water to the organism. Although one need not go as far as Zuntz in assigning a numerical value to the relation be- tween the amount of water necessary for glycogen formation and the glycogen formed, yet it is clear that the retention curves of sugar and water run parallel to each other in these experiments. What has occurred, then, is an increased storage of glucose in the body, and one must seek further to make clear the effect of dilution of the injection on kidney function considered separately. The data from the blood sugar determinations throw additional light on the mechanism by which this result is accomplished. They 2 Quoted by Allen: Glycosuria and diabetes. 524. DAVID M. DAVIS appear in figure 10. No blood sugar figures are available from the 10-per cent injection. When the strong solution (80 per cent) is in- jected, the blood sugar rises very sharply, nearly to 0.6 per cent, in striking parallelism with the rate of total sugar output in the urine and then remains approximately constant. With the weak solution, however, the initial rise is only in the neighborhood of one-half as great to be followed by a leisurely but steady rise through the five and one- - Broop . ee Broop Sucar Concentration t Constant Sucar Intake (2 Gm. PER KG-HR.) 50%, Bee Bore s *: R; Py ae 7 a ry / ; “e 5 Rt 4 i) age —-- 20% > INJECTION een wenee Yor rhe DT D Te oo Fig. 10 _ half hours of the experiment. If figures for the water content of the blood were at hand, the interpretation of these findings would be as- sisted, but unfortunately they are not. The influence of dilution of the blood, with its attendant plethora, must be considered by indirect approach. This can be done after a glance at tables 2 and 3 and figure 10. Inthe 40 ce. (5 per cent) experiment, 788 cc. of water were retained, while a total of 18.7 grams sugar was excreted (see protocols). In the 2.5 cc. (80 per cent) experiment, 258 cc. of water were lost to the BLOOD SUGAR AND URINARY SECRETION 525 body, the output being greater than the intake, and 26.6 grams sugar were excreted, the same quantity of sugar (88 grams) having been in- jected in-each case. The animal then contained some 1046 cc. more of water and 7.9 grams more of sugar at the end of the first than at the end of the second experiment mentioned. The animal weighed about ‘11 kilograms, which would indicate a blood volume of roughly 770 ce. Now at the end of the 2.5 cc. (80 per cent) experiment, the blood sugar stood at 0.56 per cent, and even if one assumes—which is manifest- ly an exaggeration—that the entire 258 cc. which were lost during the experiment came from the blood and were not replaced, the circula- tion then contained (770-258) X 0.0056 = 2.87 grams sugar, in 512 ce. of blood. Proceeding on the assumption that all of the 1046 ce. of fluid and 7.9 grams of sugar retained in excess in the 40 ec. (5 per cent) experiment remain in the plethoric blood, there would be in circulation at the end of this experiment (1046 + 512 = ) 1558 ce. of fluid con- taining (7.9 + 2.87 = ) 10.8 grams of sugar. If this were the case the concentration would be (10.8 + 1558 = ) 0.69 per cent, but the actual observed concentration at the end of the experiment was 0.476 per cent, so that it is evident that at least part of the 7.9 grams re- tained in excess has gone to the tissues along with the bulk of the re- tained sugar. In addition, to assume that 1046 cc. of fluid remain in - the circulation is going far beyond the necessities of the situation, since it would be an increase in blood volume of 136 per cent, and it is doubtful if this can occur. ; In the experiments of Fisher and Wishart, it was calculated from the hemoglobin content that the blood volume might be increased by as much as 50 per cent after the ingestion of glucose. It is unfortu- nate that no accurate method of determining blood volume is avail- * able. Similar calculations for the 10 cc. experiment demonstrate the same facts, even more conclusively. . One can then conclude definitely that hydremia does not account for all of the sugar which is retained in the body in the dilute injec- tion experiments in excess over that retained in the concentrated in- - jection experiments, and that a considerable part of this sugar leaves the circulation for the tissues. This is equivalent to saying that the presence of a greater quantity of water in the injection fluid acceler- ates the storage or at least the taking over of dextrose by the tissues. These considerations now enable one to return to the question of the relative diuretic efficiency of concentrated and dilute solutions. If it is true that the presence of a plentiful excess of water assists the 526 DAVID M. DAVIS flow of both sugar and water from the vessels to the tissues, what effect has it on the flow of the same substances from the blood into the urine? Sugar in the body can scarcely affect the kidney except as it comes in contact with the renal cells while circulating in the blood. To make a fair comparison, therefore, times should be chosen during’ the various experiments at which equal concentrations of sugar in the blood occurred. Since more sugar has been removed from the circulation by other means in the case of the dilute injections, the kidney, having to deal with a blood of lower glucose concentration, does not receive as great a stimulus as in the case of the concentrated injections until much later in the experiment. If, for instance, the even figures for blood sugar level of 0.1 per cent, 0.2 per cent, 0.3 per cent, etc., are arbi- trarily chosen, and the rate of sugar excretion noted in each experi- ment at the time each of these levels was reached, a table like the following rough example (table 4) can be made. TABLE 4 WATER CUBIC CENTIMETERS PER KILOGRAM—HOUR BLOOD SUGAR 40 20 10 2.5 PER CENT poles Sugar gram per Sugar gram per Sugar gram per Sugar gram per 15 minutes 15 minutes 15 minutes 15 minutes va 0 0 0 0.2 0 0 0 0.3 0.4 2:40 1.2 1.3 0.5 3.0 2.25 0.6 4.5 3340 Glycosuria begins in each experiment when the blood sugar is be- tween 0.2 and 0.3 per cent. At 0.4 per cent the 40 cc. injection ex- periment is excreting 2.75 grams of sugar per fifteen minutes, while much less is being put out in the case of the more concentrated injec- tions. From this standpoint, then, at a given level of blood sugar, the presence of an excess of water accelerates the passage of sugar from the blood into the urine, just as it does from the blood to the tissues. ca Comparing figure 10 and figure 9, it will be seen that the nearest approach to parallelism exists between blood sugar and total sugar BLOOD SUGAR AND URINARY SECRETION 527 output. This relation is maintained in all the experiments, the total sugar curve showing a tendency to hold its course more or less regard- less of changes-in the quantity of urine, especially the minor ones, _ which are compensated for in a truly remarkable way by changes in the concentration. This is represented graphically in some of the curves—the concentration changing one way, the quantity the other, and the total sugar line continuing its original course. sore). Usine en Urine Sucar Concentration 4 Constant Sucar Intake (26m. Per KG-HA) — 5% ee Oe INJECTION —~ 80% 10 Be ea 0%... ‘ K eee Was ee-- te ee 2 Seedy en ail ; i ae ¥ “ : es ee eae a a 2 00° E030. TT 30 TE: 30. TE30. YO Fig. 11 Figure 11 represents the urinary sugar concentration in the first four experiments. In those cases with copious diuresis the concen- tration falls off as the diuresis increases. The concentration in the 80 per cent (2.5 cc.) injection is below that in the 20 per cent (10 cc.) injection at first, because the diuresis commenced earlier with the 80 per cent injection. The concentration does not seem to run parallel with the blood sugar, but varies inversely as the quantity of water available. Thus in figure 7 the steady increase in the sugar output 528 DAVID M. DAVIS is obtained by a steady increase in the concentration, the urine quan- tity remaining constant, while in figure 3 exactly the reverse is true. Epstein’s (6) statement, as noted above for diabetics, that the concen- tration is constant, the sugar output varying with the quantity of urine excreted, is certainly not true for a normal dog excreting sugar under these conditions. The concentration may continue to rise when the blood sugar goes above 0.4 per cent (fig. 7) contrary to the state- ment of Ewing (5). The series of injections in which the amount of water given remained constant (20 cc. per kilogram-hour), figures 5, 6,2 and 7, shows that the water output in such a case is entirely dependent on the sugar content. This is in accordance with the well known fact that dis- tilled water given intravenously causes no diuresis. SUMMARY It is felt that the experimental conditions described are unusually free from disturbing influences. Under these conditions the sugar output per unit of time shows a direct relationship with the blood sugar level. Upon intravenous injection of a given quantity of dextrose, the amount of diuresis produced depends on the amount of water available, within certain limits. If a very concentrated solution is given, the body is robbed of water. | If large quantities of water are given with the dextrose, its percent- age retention in the body is increased and water is retained along with it, so that less sugar is excreted, despite diuresis. . This retention is not confined to the blood, but occurs also in the tissues. If more water is available, more dextrose is excreted into the urine at a given level of blood sugar than when less water is available. I am much indebted to Mr. J. H. Janney of the fourth year class in the medical school for his valuable assistance in carrying out these experiments. ae BLOOD SUGAR AND URINARY SECRETION 529 BIBLIOGRAPHY (1) Davis: Journ.-of Urol., 1917, i, 113. (2) Lamy anp Meyer: Journ. de Physiol. et Pathol. Gen., 1904, vi, 1067. (3) Fisher anp WisHart: Journ. Biol. Chem., 1912, xiii, 27. (4) Lamson: Proc. Natl. Acad. Sci., 1916, ii, 365. (5) Ewrne: Proc. Soc. Exper. Biol. and Med., 1916, xiii, 69. (6) Erstern: Proc. Soc. Exper. Biol. and Med., 1916, xiii, 67. (7) Arrous: Compt. Rend. Soc. Biol., 1904, ii, 258; 1907, i, 649, 807, 845. (8) BiumenrHa: Hofmeister’s Beitr., 1905, vi, 329. (9) Woopyratr: Journ. Amer. Med. Assoc., 1915, Ixv, 2067. (10) Davis anp Gorton: Journ. Urol., 1917, i, 136. (11) Zuntz: Arch. f. Anat. v. Phaaelt (Physiol. Abt.), 1898, 267; Biochem. Zeitschr., 1912, xliv, 290. (12) Wimmer: Zeitschr. f. Biol., 1911, lvii, 185. THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 43, NO. 4 FURTHER STUDIES ON THE EFFECT OF ADRENALIN | UPON MUSCULAR FATIGUE CHARLES M. GRUBER From the Laboratory of Physiology in the Albany Medical College Received for publication May 10, 1917 Muscular activity in conjunction with Addison’s disease is exten- sively discussed in the older literature. Before experimental work was begun on the adrenal glands it was well known that marked mus- cular weakness was characteristic of Addison’s disease. In 1892, Abelous, Langlois and Charrin (1) showed that persons suffering from. this malady were less capable of doing muscular work, as measured by the Mosso Ergograph, than were individuals suffering from tuber- culosis. When it was discovered that adrenalin excited the sympa- thetic nervous system, study of its effect upon the muscular system was for a time abandoned. Recently, however, the study of the relation of adrenalin to muscu- lar activity has been resumed. Three methods have been employed. They are, the older method, that of extirpating the glands, and the newer methods, those of splanchnic nerve stimulation causing the glands to secrete their adrenalin into the circulation and of injecting the adrenalin into the circulatory system or into the solution in which the muscle is contracting. In 1892 Albenese (2) removed the glands from frogs and rabbits. Upon cutaneous stimulation, he observed that a stronger current was necessary to elicit a response after extirpation than before. That this increase was not due to the operation was shown by subjecting animals — to more severe operations than decapsulation. He therefore concluded that the function of the adrenal glands was to destroy or to transform the toxic substances which, as a result of muscular or nervous work, are produced in the organism. His results were confirmed by Boinet (3) who reported that rats recently decapsulated were much more quickly exhausted in a revolving cage than were normal animals. A similar loss of muscular power was recorded by Biedl (4). ° 530 EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 531 Founding his belief upon experiments done alone and with Langlois (5), Abelous (6) maintained that the suprarenal capsules, in frogs and _ guinea pigs, can modify, neutralize or destroy poisons which are pro- duced in the course of muscular work or which accumulate in the or- ganism after destruction of the adrenal glands. Adrenalin acts, he believed, as an antitoxin to fatigue. He observed that if the blood of an animal dying from decapsulation is injected into a normal ani- _ mal true symptoms of fatigue result. Oliver and Schiffer (7) in 1895 showed that adrenal extract has a bettering effect on the contraction of an unfatigued muscle. After injecting the extract subcutaneously into a frog and then excising the gastrocnemius muscle, they obtained a curve of contraction which was 33 per cent higher and 66 per cent longer than that of the corre- sponding muscle not supplied with the extract. A similar prolongation of the muscle curve was observed after the extract was injected intra- -venously into a dog. Analogous experiments were performed in 1915 by Kuno (8) and in 1916 by Takayasu (9), who obtained opposite re- sults. Takayasu claims that adrenalin has a depressing rather than a bettering effect upon muscular contraction. He quotes Schafer as saying of his own early experiments that he believed the results he obtained were due not to the action of adrenalin but to some im- purities contained in the commercial extract which was employed. Schafer also makes note of this belief in his recent publication, The Endocrine Organs (10). These results would seem to dispose of Joteyko’s (11) arguments that adrenalin is a sarcoplasmic excitant. _ Betterments in fatigued muscles after injections of adrenalin were observed by, Boruttau (12) working on frogs and Dessy and Grandis (13) working on salamanders. Dessy and Grandis claim that the adrenal extract produces a beneficial effect on fatigued muscle either when injected subcutaneously or when added to the solution in which an isolated muscle is contracting. Cannon and Nice (14) were unable to confirm Dessy and Grandis’ observations in frogs’ muscles similarly exposed to the adrenal extract. In a case of neurasthenia, Pantanetti (15) recorded a marked better- ment in the total amount of work done after six subcutaneous injec- tions of adrenalin. The influence of fatigue upon the adrenalin content of the suprarenal glands in dogs was studied by Battelli‘and Boatta (16). In com- pletely exhausted animals the adrenal content of the suprarenal cap- sules was decreased to less than one-sixth the normal amount. These 532 CHARLES M. GRUBER investigators believe that during fatigue adrenalin is set free to keep up the blood pressure which has a tendency to lower itself by dilation of the vessels in the active muscles. That the chromaffin system is thus exhausted by prolonged muscular effort was later confirmed by Carl (17) on strychninized frogs. Bernard and Bigart (18) made examinations of the histological structure of the adrenal gland before and after muscular fatigue. It was discovered that after prolonged muscular activity the gland contained numerous vacuoles. Bardier and Boone (19) confirmed them. Carnot and Josserand (20) found that if 0.025 mgm. of adrenalin per kilo body weight was injected into the femoral vein of a normal dog there was an increase of 10 cm. of mercury in arterial pressure; if injected into the femoral artery (leg at rest) it produced an elevation in arterial pressure, of only 2 cm. of mercury. In one particular experi- ment upon a dog they injected 0.05 mgm. of adrenalin per kilo body weight into the femoral artery of a leg at rest and found that the blood pressure in the general circulation was increased 10 cm. of mercury. The opposite leg was then tetanized for fifteen minutes and at the end of that time 0.055 mgm. of adrenalin per kilo was injected into its femoral artery and the systemic blood pressure rose only 1.5 cm. of mercury. They claim that the adrenalin was neutralized or de- stroyed by the fatigue products. Panella (21) observed a marked improvement with the use of adrena- lin in the curve of contraction in heterothermic animals, frogs and toads, and also in homothermic animals, rabbits and guinea pigs, placed by special treatments (section of the bulb or profound narcosis), in a state of respiration, circulation and heat regulation somewhat like that of heterothermic animals. He believed that under normal mammalian conditions adrenalin has little effect because it is quickly oxidized or destroyed in the blood. Radwanska (22) found that when the gastrocnemius muscle of a frog was wholly fatigued it could be made to contract again by inject- ing adrenal extract subcutaneously.. He also found adrenin more effective in decapsulated than in normal frogs. Because the results were better when the muscle was stimulated through its nerve than when stimulated directly, be believed that adrenin acts upon the nerve endings. His belief was shared by Cannon and Nice (23) who thought the point of action of adrenin to be upon the nerve endings or neuro- muscular junctions. They found that adrenin injected in small doses or secreted during splanchnic stimulation caused a marked improve- ment in the activity of the fatigued muscle. ry EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 533 A decrease in the irritability of the nerve in the nerve muscle of a frog as a result_of the removal of adrenalin from the system by extir- pation of the adrenal glands and a restoration of irritability upon injecting adrenalin into the same animal were demonstrated by Czubal- ski (24). He has assumed that the first phase of the action current is due to dis-assimilation or katabolism and the second phase due to assimilation or anabolism. Working upon this assumption he obtained evidence that the quick exhaustion of the muscles of decapsulated ani- mals is due to slow and incomplete anabolism within the muscle. This process could be hastened by adrenalin. In a series of papers published a few years ago, I (25) showed that adrenalin increased the height of muscular contraction when injected intravenously in small doses (0.2 to 0.5 cc. of a 1:100,000 solution), and that the same dose administered in the same manner would bring back to normal in five minutes or less the increased threshold stimulus caused by fatigue of the nerve-muscle or muscle, as would rest of one to three hours. I also showed by means of a comparative study of the effects of adrenalin and amyl nitrite that this improvement after fatigue was not due to a bettering of the circulation. Invariably the effect produced by adrenalin was greater although the effect upon the general blood pressure was the same. In some experiments the - limb was perfused with warm Ringer’s solution. If adrenalin was injected into this perfusion fluid a betterment in muscular contraction but always a decrease in the flow of the fluid (vasoconstriction) from the venous cannula was observed. The theory held by Radwdnska that the action of adrenalin was on the nerve endings I disproved since I found that adrenalin has the same effect upon the muscles in which the nerve has been cut from nine to eighteen days as it has in nor- mal muscles. Recently Hoskins, Gunning and Berry (26) demonstrated that adrenin produces active vasodilation of muscle vessels and they be- lieved that the betterment in the height of muscular contraction, which I demonstrated, was due, in part at least, to the betterment in circu- lation. In my experiments there had been three factors tending to bring about extreme dilation of the vessels in the active muscle. First, the nerves were severed and the vasomotor tonicity from the vaso- motor center, therefore, inoperative; second, the rate of stimulation employed was favorable to dilation (27) and third, as Kaufmann (28) has shown, the fact that the muscles were active would presuppose actively dilated blood vessels. These circumstances led me to be- 534 CHARLES M. GRUBER lieve that in my former work any further dilation would have been so slight as to be negligible. Since I had not, however, separated the cutaneous circulation from the muscle circulation it seemed im- portant that I repeat my work with the view to determining the con- dition of the blood flow through the muscle alone. METHOD In the earlier experiments the animals (cats) were anaesthetized — with urethane (2 grams per kilo body weight by stomach), but later continuous ether anaesthesia was employed. By making a small slit through the skin on the outer side of either thigh, the anterior tibial nerve (peroneus communis) was isolated, cut and its distal end fastened in a Sherrington shielded electrode (29). The electrode was then held in place by fastening around it, with paper clips, the two flaps of skin. Through another slit in the skin in the same leg the tendon: of the tibialis anticus muscle was isolated from its insertion. It was then fastened to a muscle lever mounted upon a tripod base by a string pass- ing about two pulleys. These pulleys were arranged so that the muscle pulled in its normal direction. One loop of cord about the hock and another around the foot just below the fastening of the tendon ° bound the leg to the board and made a very satisfactory nerve muscle preparation. The strength of the stimulating current was 0.1 ampere in ‘the pri- mary circuit derived from a storage battery. The stimulating current in every case was a maximal break induction shock obtained from a glass knife blade key (30), which, propelled by a motor, made and broke the primary circuit. The rate ninety times per minute in these experiments was slow enough to produce vasodilation (27) in the ves- sels of the stimulated muscle. The muscle lever consisted of a piece of light straw 22 cm. in length from the axis to the tip of the parch- ment paper writing point. The tendon attached 3 cm. from the axis began at the moment of contraction to pull against an initial tension of 110 grams developed in a coiled spring. For each centimeter excur- sion of the muscle lever on the drum this was increased 5 grams. This spring was attached at the same position on the lever as was the mus- cle. Muscular contraction was, therefore, magnified about seven and one-third times. EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 535 The blood pressure, whenever recorded, was registered from one or the other carotid artery by means of a mercury manometer. A time marker which indicated intervals of five or of thirty seconds was placed at the atmospheric pressure line of the manometer. Thus, at any given muscular contraction, the height of blood pressure could be de- termined. Below the time marker was placed another signal magnet to indicate the time of the injection of adrenalin. The rate of blood flow through the muscle was recorded on the kymo- graph paper in the early experiments by a simple key and signal mag- net. Later an automatic drop recorder was substituted for the hand key. It consisted of a flat steel spring to which was soldered a plati- num disc in contact with a platinum point adjustable by a thumb screw: The platinum point and steel spring were connected by copper wires to binding posts on the vulcanite arm which supported them. Each drop falling upon the spring broke the contact between the platinum .dise and the point and the drop was recorded on the drum by the sig- nal magnet. The recorded blood flowed through a cannula placed in the femoral vein. All the branches to the vein were tied off except the deep anterior tibial vein which comes from the tibialis anticus muscle and muscles of that region. The cutaneous vessels were tied and the limb was either skinned or mass-ligated above the hock. In the early experiments a necropsy was made after each series of obser- vations to make sure that there had been a separation of the cutaneous and muscular circulations. _ Usually adrenalin chloride but in many cases crystalline adrenalin in solution was injected into a cannula placed in the external jugular vein. The dilution of adrenalin with mammalian Ringer was made in all cases just after the operation so that the solution had no appre- ciable time for deterioration. This solution was kept at body tem- perature for injections. ; Experiments were also performed in which the muscle was irri- gated. The medium for irrigation, a warm (38.5 to 39.5°C.) Ringer’s solution at a pressure of 60 to 70 cm. of water, ran through cannulae placed in the femoral artery and femoral vein after all the branches leading to and from these vessels were tied off except the two vessels of the deep anterior tibial region. The adrenalin, in most cases 1: 100,- 000 but in some cases 1:1,000,000 and 1:1000 solution was injected into the running fluid close to the arterial cannula. 536 CHARLES M. GRUBER RESULTS In no instance in my experiments did adrenalin in small doses (0.5 to 2 cc. of a 1: 100,000 or 1: 1,000,000 solution) produce dilation of the vessels in the limb in which the nerve was cut and stimulated at a rate favorable to dilation. In animals with the circulation undisturbed (except that the nerve was cut and stimulated) the vessels of the limb responded passively to adrenalin ‘(0.5 to 2 cc. of a 1: 100,000 solution) i.e., if the blood pressure decreased the rate of flow through the mus- eee decreased and vice versa. In the perfused limb with the nerve cut and stimulated vasoconstriction was obtained even with doses as small as 0.5 ec. of a 1:1,000,000 solution. The action of adrenalin in intact muscles Figures 1 and 2 are records made during one experiment and selected as typical of all the results obtained. The animal weighed 2.5 kgm. and was given 25 mgm. of hirudin intravenously as an anticoagulant just before the experiment and after the operation. In figure 1 the blood pressure was 54 mm. of mereury. Upon inject- ing intravenously 0.5 cc. adrenin (1:100,000 solution) the pressure — decreased to 44 mm. of mercury with a resultant decrease in the rate of blood flow through the muscle of 21 per cent and a concurrent in- erease in the height of muscular contraction of 55.5 per cent. In other animals upon injecting 0.5 cc. of a 1:100,000 solution of adrenin no measurable change in the blood pressure but always an increase in the height of muscular contraction, in some cases as much as 35 per cent, occurred. The rate of blood flow remained ‘unaffected. There is, of course, in all cases, a gradual decrease in rate and lower- ‘ing of pressure due to the loss of blood from the recording vein. Before figure 2 was recorded, 50 ec. of warm Ringer’s solution was added to the blood that was lost by hemorrhage and the mixture was perfused back into the animal. As a result of the transfusion the blood pressure increased slightly (20 to 52 mm. of mercury). At 1 in figure 2, 2 cc. of a 1:100,000 solution was injected intravenously. The blood pressure rose 88 per cent and the height of muscular con- traction was bettered 53 per cent. Since the same quantity of adrenin injected slowly at 2 produced no change in muscular contraction the betterment observed cannot be due to the injection of adrenin at 1 and 3. Both the dilation of the vessels of the muscles and the in- crease in the height of muscular contraction were due to the increase EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 537 Fig. 1. In this and the following figure the upper curve is a record of blood pressure with mercury manometer, below it the record of muscular contraction. The lowest record indicates the number drops flowing from the venous can- nula or limb circulation, above it is the signal record indicating the point of injection of adrenalin. Middle line time in thirty seconds. Hirudin was used as an anticoagulant. In this figure the time marker was placed 2 cm. below the atmospheric pressure level of the mercury manometer to allow room for muscu- lar contraction. At the point indicated in the record 0.5 cc. of adrenalin (1:100,- 000) was injected. ee Ee ee ees Fig. 2. Time marker 3.5 cm. below zero blood pressure. At the points indi- cated by the marker 2 cc. of adrenalin (1:100,000 solution) were injected. 538 CHARLES M. GRUBER in blood pressure. The increase in blood pressure is sufficient to account for the increased muscular efficiency (25). ' The results obtained upon animals into which the lost blood olus Ringer’s solution was perfused, were rarely satisfactory. The action of adrenalin in perfused muscles In the perfused muscles doses of adrenalin varying from 1 ce. of a 1:1,000,000 to 1 ec. of a 1:1000 solution produced a betterment in the height of muscular contraction with a decrease in the rate of flow of the perfusion fluid. Figure 3 is a record obtained from an animal which had been given hirudin., Adrenin, 0.5 ec. (of a 1: 100,000 solution) was injected at 1 and 2 cc. of the same solution was injected at 2. As a result of the first injection there was a betterment of 17 per cent in the height of muscular contraction and from the second injection there was a better- ment of 550 per cent with intermediary betterments, the whole last- ing for six minutes. Here, as in most instances, the increase and sub- sequent decrease were gradual. After the first injection the rate of flow. through the vessels decreased from seventy-four to thirty drops per thirty seconds and after the second injection the rate of flow de- creased from thirty-five to less than one drop per thirty seconds. It is observed in this experiment as in many others that the maximal increase in the height of muscular contraction occurs simultaneously with the maximal constriction of the muscle vessels. These condi- tions are probably coincident and not interdependent. ES When a muscle has ceased completely to respond to the original strength of stimulus it can be made to react again after an injection of adrenalin (see fig. 4). Here at 1, 0.5 ec. of a 1:100,000 solution injected into the perfusion fluid produced a rise in the muscle curve of 26.6 per cent with no noticeable change in the circulation.. At 2, 1 ec. produced a rise of 32.5 per cent in muscular contraction and a decrease of 5 per cent in the rate of flow,of the fluid through the muscle. At 3, 1 cc. of the solution produced no noticeable effect upon the muscle. From point 3 to point 4 the muscle curve dropped until no contraction occurred. At 4, 2 cc. of adrenalin were injected into the perfusion fluid. As soon as the adrenalin reached the blood vessels a marked vasoconstriction occurred, followed soon after by a slight muscular twitch which increased slowly until a contraction 6 mm. in height was obtained and then decreased to nil again. The length of time occupied by the entire recovery contractions was about six minutes. 539 SCULAR FATIGUE s MU ON ADRENALIN EFFECT OF ‘spuooes Aj1Iq} Ul OMIT, “B[huUBO 94} JO quourjsnipe ‘x {00 Z “Y 48 pue ‘00 [ ‘¢ 4B “00 T ‘g 9B “00 G0 ‘7 98 ‘poqooful SBM (MOTINIOS 000 OOT 1) Ul[eusIpy ‘“spuodes AY 103 poddo4ys sem UINIp ey} ‘A 4v ‘gjosnur posnjiod pondiyej Ajojo[du109 B Ul UOTPOBIYUOD IB[NOSNUT 94} JO AIOAONOI BUIMOYS OAINI VY “f “BLT : i rit i "a al " Fe Pith whi ‘opOsnuL oq} ysnory } ping uorsny1ed oyy jo MOY Jo sdoap Ul 07R1 OY} JO PAODOL B OUT] JSOMO] OY} PUB PLODOL OUUTY OY} FT MOTOR ‘pio0001 O[OSNUL OU} st eAano Joddn oy} onSy Burmoy[ojy oy} puve sry} UT “Nap oy} uo eoeds 1ojyowtjyu00 10d suleis 2 JO UOISUS} posBoIOUT Ue YIM SUIvIZ OPT SVM JUOUTTIOdXO STY} UL BuLIds OY} JO UOISUO} [VIPUL OUT, “SpUO9S OAY Ul CUNT], *‘popoolul a10M *09 Z iy ‘ping uorsnjsod oy} ojur pozooful svm (UOTINIOS YOO'OOT: 1) Ul[VUeIpw “99 g°0 ‘TT 4y ‘opognur posnjiog “¢g “3Iy Ea NIST SOS RINE SEW TP T SELL \ 540 CHARLES M. GRUBER Is adrenalin in large doses toxic? j | Takayasu, concluded jthat adrenalin in large doses was toxic in char- acter and its effect resembled that of increased potassium salt con- centration. Cannon and Gruber (31) remarked that in normal mus- cles adrenalin in large doses (1 ec. of a 1:10,000 solution intraven-- ously) sufficient to cause a marked constriction of the arterioles, results — in a lessened height of contraction and slowing or even a disappearance of ; the® wavelike variations observed in rhythmically’ &timulated mus- cles. “his muscular inefficiency they attributed to a deficient blood supply: The effect upon perfused muscles, of more concentrated solutions: than that employed by Cannon and Gruber, is not different from that produced by more dilute solutions. Here may be given the results of one typical experiment. Adrenalin 0.5 ec. (1:100,000) was injected into the perfusion fluid. There resulted from this’ injec-_ tion a betterment of 100 per cent in the height of muscular contrac- tion. After this injection 0.5 cc. adrenalin (1:1000) was injected and there was observed a-twenty-seven fold increase in the height of mus- cular contraction which required sixteen and one-half minutes for its development and return to the original height. Here‘as before marked vasoconstricton occurred. The effect of adrenalin upon the make shock. In heakad, the irritability was increased by adrenalin to such an extent that the make shock, which had been subminimal became submaximal at 7 and 2. In two instances in which a 1:10,000 solution was employed the make shock contraction became equal to the break shock contraction. In my experiments muscles exposed to the action of hirudin seemed to be more vigorous and to respond to adrenalin better than did nor- mal muscles. According to Tatum (32) adrenalin has no action upon hirudin blood and hirudin. plasma and vice versa. The question of the effect of anticoagulants upon muscular contragon is open for | further investigation. i $ Se How 1s the Specie effect of adrenalin in muscular fatigue shown? All the curves presented i in’ ‘this paper except figures 2 and 4, 3, show the specific effect ‘of adrenalin on muscular fatigue. This effect is noted whether the limb is intact or perfused. s Figures 2 and 4, 3, are given to show that at times adrenalin does not have this~effect. Cannon and Nice (23) shdw”a similar curve EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 541 (fig. 5, page 52) from a muscle of an animal undergoing splanchnic stimulation. In their experiment the blood pressure was kept at a fixed level by compression of the thorax. During that time, although there was undoubtedly a secretion of adrenalin caused by splanchine stimulation, the muscle remained unaffected. At the moment at which the thorax was released, the blood pressure rose and with it the height of muscular contraction rose 466 per cent. In an earlier article (33) I showed curves (figs. 6 and 7) in which adrenalin had the effect of lowering the blood pressure.and the mus- cle curve 18.7 per cent in figure 6 and 17.3 per cent infigure 7. It might be argued in these cases that adrenalin may have had its specific effect upon the muscle but that it was obscured by the fall in blood pressure and decreased blood flow. Such is probably not the case as seen from b in both figures. An equal decrease in the height of muscular con- traction and lowering of the blood pressure without the presence of adrenalin was brought about in figure 6, b, artificially by compressing the thorax. In figure 7 at b adrenalin was injected as at a but the blood pressure was maintained at a normal level by stimulating the splanchnic nerves with the adrenal glands tied off. The muscle curve was unaffected. In the above described cases the only effect of adrenalin on the muscle curve must be a passive one. By lowering or raising the blood pressure it decreases or improves the muscle irritability. The curves herein presented other than 2 and 4, 3, show the specific effect of adrenalin. In figure 1 this is especially great for animals in which the circulation is intact. The injection of adrenalin brought about a decrease in the rate of blood flow but an increase in the height of muscular contraction of 55 per cent. In perfused limbs the increase is so great as to be almost unbelievable. Figure 3 shows an improve- ment of 550 per cent. In other experiments improvements of 2730 per cent were observed. ‘How does adrenalin produce its effect? There has been much speculation as to the probable point of action of adrenalin. Although the question cannot be settled in this article it may not be amiss to discuss some of the possibilities. 1. Does it act upon Langley’s “receptive substance’? Langley (34) has demonstrated a hypothetical “receptive substance” between the nerve endings and the muscle. In another paper (35) I showed that 542 CHARLES M. GRUBER the threshold stimulus of a denervated muscle was unaffected by curare and that adrenalin had no effect upon the curare threshold in this same muscle. I found, however, that the threshold of a curarized muscle with the nerve intact was affected by fatigue and that adrenalin counteracts this fatigue. If then curare acts upon the “receptive substance,’ fatigue and adrenalin must act at another point nearer the muscle than the “receptive substance.” 2. Is adrenalin a sarcoplasmic excitant? It has been my observa- tion that adrenalin does not lower the threshold of a normal, unfatigued muscle (25). Were adrenalin a sarcoplasmic excitant as suggested by Joteyko (11) it would probably do this and if it were an excitant at all it would certainly not be necessary to fatigue the muscle before adrenalin produced a marked result. This conclusion is further sub- stantiated by Kuno and Takayasu. 3. Does it neutralize, destroy or transform fatigue products? One of the most reasonable explanations is this, that adrenalin has some effect upon lactic acid, rendering it less harmful to the active muscle. Various experimenters some of whom are Abelous and Langlois, Alba- - nese, Carnot and Josserand, Joteyko and Gruber, have thought this a feasible explanation. This leads to the fourth possibility. 4. Does adrenalin hasten the conversion of glycogen into sugar and does it assist in the reconversion of lactic acid into sugar? It has been found that in a normal muscle with the circulation intact the same dose of adrenalin injected repeatedly brings about approximately the same reaction while in a perfused muscle a second or third injec- tion has to be larger than the first to bring about the same response. This would seem to indicate that adrenalin acts upon some substance supplied through the blood or which is present in the muscle. Can- non and Nice (23) came to the conclusion that the increased muscular contraction cannot be due to hyperglycaemia because a typical rise could be obtained when the liver was removed. If an injection of adrenalin liberates sugar from the liver it does not seem improbable that the same injection could liberate sugar in the muscle. If that is the case, such liberation of sugar in the muscle might not be indi- cated in the blood and might along with the normal supply of sugar in the blood, supply the necessary energy for the increased muscular activity. This inference is in accordance with that of Cybalski (24) who believed that the decreased action current in the nerve muscles of decapsulated frogs and fatigued animals was due to the sluggish anabolic process and adrenalin is capable of accelerating this process, EFFECT OF ADRENALIN ON MUSCULAR FATIGUE 543. but it is not in accordance with that of Wilenco (36) who maintains that the ability of the organism to burn sugar is decreased by adrenalin. It must be admitted that none of these explanations is without its weak points and that none completely accounts for the various appar- rently contradictory results obtained. At present a chemical study is being undertaken to determine more definitely the point of action of adrenalin. SUMMARY In the fatigued unaltered nerve muscle adrenalin may increase the height of muscular contraction by a twofold action, by improvement of the blood supply (vasodilation) and by its chemical action upon some substances in the muscle. : In a muscle in which the nerve is cut and stimulated, adrenalin in small doses, however administered, does not better the circulation and must therefore produce its effect of increasing the height of mus- cular contraction by its chemical (specific) action alone. The following three processes which normally go on in the muscle may be greatly accelerated by adrenalin and it.is not improbable that ~ one or all of these will finally prove to be the way in which adrenalin produces its effects: sit 1. The conversion of glycogen into sugar. LR: 2. The reconversion of lactic acid into sugar (transformation of fatigue products). . zs 3. The oxidizing of lactic acid into carbon dioxide and water (de- struction of fatigue products). BIBLIOGRAPHY (1) ApeLous, LANcLoris anD Caarrin: Compt. rend. Soc. de Biol., 1892, xxiv, 623 and 721. (2) -Aupanese: Arch. Ital. de Biol., 1892, xvii, 239. (3) Borner: Compt. rend. Soc. de Biol., 1895, 273, 325, 498 and 646. (4) Brepu: Innere Sekretion, Berlin, 1910, 149. (5) ApeLous AND Lanetors: Arch. de Physiol., 1892, xxiv, 269 and 465. (6) Asexovus: Arch. de Physiol., 1893, xxv, 437 and 720; ibid, 1894, xxvi, 433. (7) Ourver AND Scuirer: Journ. Physiol., 1899, xviii, 230. (8) Kuno: Journ. Physiol., 1915, xlix, 139. (9) Taxayasu: Quart. Journ. Exper. Physiol., 1916, ix, 347. (10) Scu&reR: The endocrine organs, New York, 1916, 65. (11) Jornyxo: Journ. Med. de Bruxelles, 1903, viii, 421. (12) Borurrav: Arch. f. d. gesammt. Physiol., 1899, Ixxviii, 109. (13) Dessy anp Granpis: Arch. Ital. de Biol., 1904, xli, 225. any 544 CHARLES M. GRUBER (14) Cannon AanpD Nice: This Journal, 1913, xxxii, 56. (15) Panrane?tti: Arch. Ital. de Biol., 1895, xxii, 17. (16) BarreLyi AND Boarta: Compt. rend. Soc. de Biol., 1902, liv, 1203. (17) Cart: Deutsch. Med. Wochenschr., 1911, xxxvii, 1827-29. (18) BERNARD AND Bicart: Compt. rend. Soc. de Biol., 1902, liv, 1400. (19) BarpirER AND Bonne: Compt. rend. Soc. de Biol., 1903, 355. (20) Carnot anp JossrRAND: Compt. rend. Soc. de Biol., 1902, liv, 1472; ibid, 1903, lv, 51. (21) Panetua: Arch. Ital. de Biol., 1907, xlviii, 430. (22) Rapwinska: Anzeiger d. Akad., Krakau, 1910, 728; reviewed in Zentralbl. f. Biochem. u. Biophysik, 1911, xi, 467. ~ (23) Cannon AND Nice: This Journal, 1911, xxix, 24; ibid, loc. cit., 49. — (24) CzuBaALsk1: Anzeiger d. Akad., Krakau, 1912, v, 470; ibid, 1913, 184; re- viewed in Zentralbl. f. Biochem. u. Biophysik, 1912, xiv, 624; ibid, 1913, xv, 913. | (25) Gruser: This Journal, 1913, xxxii, 221 and 438; ibid, 1914, xxxiii, 335. (26) Hoskins, GUNNING AND Berry: This Journal, 1916, xli, 513. (27) BowpitcH AND WaRREN: Journ. Physiol., 1886, vii, 416. BrapFrorp: Ibid, 1889, x, 390. (28) Kaurmann: Arch. de Physiol., 1892, xxiv, 283. (29) SHeRRINGTON: Journ. Physiol., 1909, xxxviii, 382. (30) Martin: Measurement of induction shocks, New York, 1912. (31) Cannon AND GruBerR: This Journal, 1916, xlii, 40. (32) Tarom: Journ. Pharm. and Exper. Therap., 1912, iv, 155. (33) Grouper: Loc. cit., 228. (34) Lanetey: Proc. Roy. Soc. of London, 1906, Ixxviii, 181. Journ. Physiol., 1905, xxxiii, 374413. (35) GruperR: This Journal, 1914, xxxiv, 89. (36) WiLeNco: Biochem. Zeitschr., 1912, xlii, 49; Zentralbl. f. Physiol., 1913, xxvi, 1059. ‘THE EFFECT OF PHOSPHORUS POISONING ON THE : CATALASE CONTENT OF THE TISSUES W. E. BURGE From the Physiological Laboratory of the U niversity of Illinois Received for publication May 11, 1917 In phosphorus poisoning all the organs of the body are injured, the liver being the most affected. This organ first undergoes fatty de- generation with subsequent disintegration of the liver cells. It is supposed that phosphorus causes fatty degeneration by rendering oxidation deficient while the disintegration of the liver cells is supposed to be brought about by the increased autolysis. As a result of the investigations of Jacoby (1), Waldvogel (2), Welsch (3), Aberhalden and Bergell (4), Reiss (5), Wolgemuth (6) and others it is now generally accepted that the essential facts in phosphorus poisoning are an in- creased rate of autolysis and a defective oxidation in the tissues. It has been shown in this laboratory that the catalase content is an index to the amount of oxidation in a tissue, being greatest where oxidation is greatest and least where oxidation is least. Furthermore, it was found that when oxidation was decreased in a tissue, as was in- dicated by a decrease in the catalase content, the tendency of that tissue to undergo autolysis was correspondingly increased. In starvation, for example, it is known that the rate of autolysis is greatly increased in all the tissues of the body except the heart and central nervous sys- tem. We found in starving animals that catalase, and hence oxidation, was decreased in those tissues in which autolysis is known to be in- creased and remained normally high in the heart which is not. autolyzed during starvation. Furthermore, in thyroid feeding, it was found that the catalase content, and hence oxidation, was decreased in the fat, skeletal muscles and heart, with a corresponding increase in the rate of autolysis. In view of the fact that the autolyzing enzymes in common with all the ordinary enzymes are easily oxidized and de- stroyed and that autolysis can be increased in a tissue by decreasing oxidation, and vice versa, the assumption was made that there nor- mally exists a balance between the autolyzing and oxidizing enzyme, 545 * _ THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 43, No. 4 546 W. E. BURGE the idea being that when oxidation is increased in a tissue a larger amount of the autolyzing enzymes is oxidized and destroyed with re- sulting decrease in autolysis, and vice versa. It is known that autolysis is more intense in the liver than in any of the other organs of animals suffering from phosphorus poisoning. If, according to the preceding hypothesis, the amount of autolysis in an organ is controlled by oxidation, then oxidation should be decreased to a greater extent in the livers of animals suffering from phosphorus poisoning than in any of the other organs. The object’of this investi- gation was to determine if this is true. Cats were used in these experi- ments. Fifteen of these animals were placed in separate cages. Ten of them were fed daily 60 grams of salmon each, to which 2 mgm. of ordi- nary yellow phosphorus, previously dissolved in cod-liver oil, had been added; while the remaining five animals were fed the same amount of salmon without the addition of phosphorus. As a rule all of the ani- mals ate the salmon to which the phosphorus had been added, for the first two days. After this some of the animals would eat a part, others all of the material, while some refused to eat any of it. In the table after ‘cat fed phosphorus for three days’ are given the data from cats that refused to eat the phosphorus after the third day while after ‘‘cat fed phosphorus for six days” are given data from animals that had been fed phosphorus for that length of time. When the animals were etherized, approximately 25 cc. of blood were drawn and the blood vessels, by the use of large quantities of 0.9 per cent sodium chloride, were washed free of blood as was indi- cated by the fact that the wash water gave no test for catalase. The liver and heart were then removed and ground up separately in a hashing machine. Since the blood of the animals that were severely poisoned did not clot it was used without further treatment, while the clotted blood from the less severely poisoned and normal animals was pressed several times through several thicknesses of cheese cloth and ground up in a mortar. The catalase content of the heart was determined by adding 1 gram of the ground material to 45 ec. of hydro- gen peroxide, while 1 gram of the liver was added to 500 ce. of hydro- gen peroxide in a bottle. A greater amount of hydrogen peroxide was used for the liver because of the greater catalase content of this organ. As the oxygen gas was liberated by the heart muscle, it was conducted through a rubber tube to an inverted burette previously filled with water, and that liberated by the liver to a large, inverted, graduated cylinder. The amount of oxygen gas liberated by the heart and liver PHOSPHORUS POISONING AND CATALASE OF TISSUES 547 respectively was read off directly from the burette and cylinder where it had displaced the water. After this volume was reduced to stand- ard atmospheric pressure the resulting volume was taken as a measure of the amount of catalase in the ground material. In a similar manner the catalase content of the blood was determined by the addition of ten drops of blood to 500 cc. of hydrogen peroxide. A full description of the method may be found in a previous publication (7). It will be Afier liver, heart and blood, are given the number of cubic centimeters of oxygen os liberated from hydrogen peroxide in ten minutes by 1 gram of the heart, of the liver, and by ten drops of blood, respectively, of normal and phosphorus-poisoned cats P caT AVERAGE F AMOUNT OF 1 2 3 4 5 sie ce. Liver Normal gl SE a NES aia ee 755 | 620) 830] 712}; 610 705 Cat fed phosphorus three days...... 500 | 640| 602] 355} 600 539 Cat fed phosphorus six days.........| 250} 190} 320| 340/ 290 278 Heart Normal ony ieee 213 | 198| 266| 213} 220 222 Cat fed phosphorus three days...... 165 |" 165 | 197} 179} 209 183 Cat fed phosphorus six days.. .,....| 160 | 162} 165! 187) 140 162 j Blood CE a ee ee ee 575 | 640} 495 | 1070} 590 674 ‘Cat fed phosphorus three days......- 602 | 760| 600} 720} 680 672 Cat fed phosphorus six days......... 380 | 420! 580} 400} 630 482 - seen that the average amount of oxygen liberated by 1 gram of the liver of the normal animals in ten minutes from 500 cc. of hydrogen “peroxide was 705 cc.; that liberated by the liver of cats fed phosphorus three days, 539 cc., and that by cats fed phosphorus six days, 278 ce. of oxygen. The average amount of oxygen liberated by one gram of the heart of the normal cat in ten minutes from 45 cc. of hydrogen _ peroxide was 222 cc.; that by the heart of cats fed phosphorus for three days 183 cc., and that by cats fed phosphorus six days 162 ce. of oxy- gen. The average amount of oxygen liberated by ten drops of blood of the normal cats in ten minutes from 500 ce. of hydrogen peroxide was 674 cc.; that by ten drops of blood from the cats fed phosphorus three days, 672 cc., and that by cats fed phosphorus six days, 482 ce. of oxygen. 548 W. E. BURGE By comparing the data from the animals that had been fed phos- phorus with that from the normal animals it is seen that the catalase content of the livers of the animals that had eaten phosphorus for three days was decreased 23 per cent and 60 per cent in those that had eaten it six days; that the catalase content of the heart was de- creased 17 per cent in animals that had eaten phosphorus three days and 27 per cent in those that had eaten it six days; that there was practically no decrease in the catalase content of the blood of the ani- mals that had eaten phosphorus three days while there was 28 per cent decrease in those that had eaten it six days. Since catalase is an index to the amount of oxidation in a tissue, being greatest in amount where oxidation is greatest and least where oxidation is least, the de- crease in catalase in the liver, heart and blood of animals fed phos- phorus is interpreted to mean that the feeding of phosphorus decreased oxidation in these tissues. | : The livers of the animals that had been fed phosphorus three days presented the typical appearance of fatty degeneration with little or no indication of autolysis, while the livers of the cats that had been fed phosphorus six days showed extreme autolysis as well as fatty degeneration. The livers of these severely poisoned animals were literally in a state of falling to pieces as a result of autodigestion. From this it would seem that the amount of autolysis was inversely propor- tional to the amount of oxidation. CONCLUSIONS - 1. The catalase content of the liver, heart and blood is decreased in Z ; ‘ ae | phosphorus poisoning, the decrease being greatest in the liver. 2. The fact that there was a greater percentage decrease in cata- lase, and hence in oxidation in the liver, than in the heart, for example, and the fact that autolysis is greater in the liver than in any other organ of the body would seem to lend further support to our contention that oxidation controls the amount of autolysis. BIBLIOGRAPHY (1) Jacosy: Zeitschr. f. physiol. Chemie, 1900, xxx, 174. (2) WaupvoagE.: Arch. f. klin. Med., 1905, lxxxii, 437. (3) Wexscu: Arch. internat. de Pharm. et de Therap. 1905, xiv, 211. (4) ABDERHALDEN AND BerGELu: Zeitschr. f. physiol. Chemie, 1903, xxxix, 464. (5) Ress: Berl. klin. Wochenschr., 1905, Ixii, 44a, 54. (6) WoieemutH: Zeitschr. f. physiol. Chemie, 1905, xliv, 74. (7) Burges: This Journal, 1916, xli, 153. j THE NATURE AND PROPERTIES OF METATHROMBIN ARNOLD R. RICH From the Physiological Laboratory of the Johns Hopkins University Received for publication May 11, 1917 _ The literature concerned with metathrombin presents most widely divergent views as to the composition of the substance and contradic- tory experiments relating to its behavior. The belief of most later workers is that it is entirely absent from plasma but~ present in all sera. No one has succeeded in demonstrating the presence of meta- ‘thrombin in plasma. Fuld (1) expressed the belief that thrombin in solution does not remain unaltered but passes over into an inactive form which on treatment with alkali undergoes hydrolytic cleavage © with a new formation of thrombin. Morawitz believed at first that _ “8 proferment’’ or metathrombin must be produced either by the ealcium activation of prothrombin or at the time of fibrin formation, but later he (2) states simply in agreement with Fuld that the greater part of the thrombin formed during coagulation passes over into the inactive form of metathrombin. Weymouth (3) was led to believe that the substance is an antithrombin-thrombin compound. . Melanby (4) recognized the existence of an antithrombin-thrombin compound in serum but denied that alkali-activation of serum could split this ‘with the resulting liberation of free thrombin. He records his ex- periments which led him to the identification of metathrombin with thrombokinase. Collingwood and McMahon (5) support him in this view. During the progress of this work, Gasser (6) has recently offered further evidence in support of Weymouth’s theory that meta- thrombin is an antithrombin-thrombin compound. These experiments were undertaken at the suggestion of Dr. W. H. Howell to determine something of the composition and significance of metathrombin. The method of alkali activation used for the detection of metathrom- bin consisted in the following procedure: One-half cubic centimeter of the solution to be tested was incubated at room temperature with-an equal volume of 7 sodium hydroxide 549 550 ARNOLD R. RICH for fifteen minutes and then neutralized with 7 hydrochloric acid, Neutral Red being used as an indicator. Thrombin liberated by such activation was tested for by the addition of ten drops of a fibrinogen solution to the activated mixture. In all cases a control mixture was made which consisted of 0.5 ce. of the solution to be activated, to which was added ten drops of fibrinogen and a solution of 0.9 per cent sodium chloride, equal in quantity to that of the acid and alkali of the acti- vated specimen. Such a control would reveal any thrombin present before activation. In solutions in which free thrombin was Bispected (in sera, for exam- ple), the solution was heated before activation at 60° to destroy this thrombin. If the presence of prothrombin was suspected in the solu- tion to be tested, the solution was first oxalated in order to prevent any further liberation of {reel from the prothrombin, and then heated to 60°C. Repeated examinations of fresh and old oxalated plasmas (some of which were heated at 54° to remove the fibrinogen present and thus approach more closely the composition of serum) consistently failed to reveal any trace of metathrombin. It is regularly found in sera, however, as late as two weeks or more after coagulation. There are apparent, then, several possibilities as to the composition and time of formation of metathrombin. First. Metathrombin might be formed as a result of one of the three main processes concerned in coagulation: (1) the action of the thromboplastic substance; (2) the calcium activation of prothrombin; (3) the combination of thrombin with fibrinogen to form fibrin. As has been stated, the last two possibilities were suggested by Morawitz. Second. These processes leave in the plasma the following substances: antithrombin, prothrombin, free thrombin, calcium, thromboplastic substance in small amount and fibrin. If the fibrin be removed the remaining substances may be considered as the constituents of serum with which we are interested. In this serum there is present also metathrombin. It is well known that the thrombin of serum suffers a marked diminution within an hour after clot formation and then more gradually disappears, but the metathrombin persists for a much _longer time. Pure thrombin, however, may be kept for long periods in solution without losing perceptibly its power of acting on fibrino- gen. Alkaline, acid and neutral solutions of thrombin were kept for a week in the presence of calcium without the development of meta- thrombin. It is difficult to accept the suggestion that metathrombin : NATURE AND PROPERTIES OF METATHROMBIN 551 is formed by a change of some kind in the thrombin molecule. Experi- mental work indicates more strongly the occurrence of an interaction _ of substances. It is probable that the disappearance of thrombin from _ a@ serum may merely mark its combination with some other substance of the serum. From this viewpoint, metathrombin might be an anti- _thrombin-thrombin compound, a prothrombin-thrombin compound, a thrombin-calcium compound, or a thrombin-thromboplastic sub- ' stance compound. That metathrombin is not a prothrombin com- pound was determined by the alkali-activation of an oxalated serum that had been heated to 60°C. to destroy any free thrombin. Free thrombin was found after activation. Since this solution was calcium- free and thrombin-free to start with, the process of activation must have liberated thrombin rather than prothrombin. If the latter had been liberated it would have remained as prothrombin. It was de- termined also that alkali activation of isolated prothrombin does not yield free thrombin. The above mentioned possibilities in regard to the origin of metathrombin were each tested experimentally with the following results: 1. That the neutralization of antithrombin by thromboplastic substance is not attended by the formation of metathrombin was de- _ termined by incubating oxalated plasma (heated to 54°) with an equal amount of cephalin solution prepared after the method described by Howell (7). Alkali activation of the mixture gave no evidence of the presence of metathrombin. That neutralization of antithrombin by cephalin had taken place in the mixture was determined by anti- thrombin tests made on the mixture after incubation and at the same time onia control of plasma heated to 54°C., equally diluted with water and incubated for the same period. The cephalin plasma showed a marked decrease in antithrombin content in comparison with the water plasma. 2. The activation of prothrombin by calcium is neither attended by nor necessary for the production of metathrombin. This was de- termined directly by activating isolated prothrombin with a solution -of calcium chloride and testing the thrombic power of the mixture at intervals up to forty-eight hours, carrying parallel tests for meta- thrombin. The thrombin yield was plentiful and did not decrease perceptibly in the period of the experiment, but at no time was there any indication of the presence of metathrombin. bats More positive evidence of the fact that prothrombin activation by calcium is not necessary for the formation of metathrombin was ° 552 ARNOLD R. RICH later obtained in confirmation of results published by Gasser (6) by incubating active thrombin with calcium-free plasma that had been heated to 54°C. and filtered from the fibrinogen precipitate. This mixture on standing developed metathrombin with a corresponding loss of free thrombin as will be described further on. 3. That the actual process of coagulation with the formation of fibrin is not responsible or necessary for the formation of metathrombin was clearly demonstrated by such an experiment as the last mentioned in which metathrombin was developed in a solution in the entire absence of fibrin formation. Further proof that fibrin formation alone cannot . cause the production of metathrombin was obtained by clotting pure fibrinogen with pure thrombin; the serum from this coagulation showed no trace of metathrombin. Further, metathrombin was later found to be produced readily by the recalcification of fibrinogen-free plasma, which, of course, is parallel to the experiment of adding thrombin directly to fibrinogen-free plasma, since it has been seen that prothrom- bin activation plays no part in metathrombin formation. 4. That metathrombin is not a combination of thrombin left un- combined by fibrinogen, with the unactivated prothrombin of serum was determined by adding a solution of calcium-free thrombin to a solution of prothrombin. Activation of this mixture gave no evidence _ of metathrombin. 5. That the action of calcium on thrombin will not convert it into inactive metathrombin was determined by adding a solution of 0.5 per cent calcium chloride to a solution of pure thrombin. No meta- thrombin was developed. It was later found quite possible to produce metathrombin by addition of calcium-free thrombin to calcium-free plasma. 6. That metathrombin is not a ovorsblintecunt cine substance combination was determined by the addition of a strong cephalin solu- tion to a pure thrombin solution. Metathrombin was not developed. 7. As this work progressed, the results led steadily to the conclusion that metathrombin is a substance formed by the union of antithrombin and thrombin. Such a union might conceivably take place either after all of the fibrmogen has been satisfied by thrombin, i.e., after a clot has formed, or else the two processes might go on side by side—part of the thrombin liberated from the prothrombin being taken up by fibrinogen to form fibrin and part by the antithrombin present in plasma to form metathrombin, which then becomes detectable in the serum. Later experiments and general considerations of the significance of metathrombin have led to the latter conclusion. The negative re- NATURE AND PROPERTIES OF METATHROMBIN 553 sults obtained by experimentally testing other possibilities of the mode of formation of metathrombin served to confirm the theory of Wey- mouth that metathrombin is an antithrombin-thrombin combination. More direct evidence of the correctness of this theory was obtained by studying the mode of appearance of metathrombin in normal sera as well as by the experimental production of metathrombin in vitro by the recalcification of oxalated plasma and fibrinogen-free oxalated plasma; and finally by the direct addition of thrombin to the equiva- lent of an antithrombin solution. The attempt was first made to determine at. what point in the proc- ess of coagulation metathrombin’ becomes detectable. It has been the common experience of workers with metathrombin that the serum from normal coagulation apparently contains metathrombin in maxi- mum amounts from the first. Their serum activations seem to have been made between periods of twenty minutes and several hours, after coagulation. The attempt was made, therefore, to obtain a centri- fugalized serum as soon after clotting as possible in the hope of find- ing it free from metathrombin. A cat was anesthetized and bled through a cannula from the carotid artery into a clean glass vessel. The blood was whipped with a wire brush to accelerate clotting. As soon as the entire clot had formed, half of the defibrinated serum was imme- diately oxalated (one part oxalate to eight parts serum). At once the oxalated and unoxalated portions were poured into separate small glass centrifuge tubes which had been packed in the brass centrifuge cups with ice-salt mixture. These cups with their contained tubes had been kept in the freezing mixture during the operation. They were now centrifugalized for seven minutes at high speed and the clear serum was drawn off. Part of the oxalated serum was heated at 60° for one minute to destroy the free thrombin, and alkali activations were made immediately upon the heated oxalated specimen, the un- heated oxalated specimen and the unheated unoxalated specimen. Con- trols for each of these consisted in the same amount of serum diluted with a solution of sodium chloride, 0.7 per cent, in amounts equal to those used in the alkali activation, the same amount of fibrinogen being added in all cases. Activations were then made at intervals, using the unoxalated specimen which was allowed to stand at. room temperature. Small portions of this serum were oxalated (one to eight) and heated at 60° one minute just before each test to destroy free thrombin and prevent further calcium activation of the serum prothrombin during alkali activation. Experiments of this type showed that if an activation were made within twelve or thirteen minutes" 554 e ARNOLD R. RICH after clotting, no metathrombin could be detected. The table gives a record of such an experiment. ACTIVATED: UNACTIVATED: + FIBRINOGEN + FIBRINOGEN 1. Serum oxalated immedi- | No clot, 24 hours No clot, 24 hours ately after clotting, heated to 60° 13 min- utes after clotting 2. Serum oxalated and | Membranous clot, 6 | No clot, 24 hours. heated 60° 30 minutes hours after clotting 3. Serum oxalated and | Membranous clot, 20 | No clot, 24 hours heated 60° 5% hours minutes after clotting Gel, 40 minutes 4. Serum oxalated and | Firm clot, 11 minutes | No clot, 24 hours heated 60° 20 hours after clotting Serum oxalated immediately after clotting 5. Unheated 13 minutes | No clot, 7 hours Gel, 24 hours after clotting Gel, 24 hours 6. Unheated © 27. minutes | Noclot 1} hours Membranous, 1 hour, 30 after clotting Membranous clot, 3 minutes hours 7. Unheated 53 hours after | Good gel, 5 minutes, 10 | No clot, 13 hours clotting seconds Clot, 24 hours 8. Unheated 20 hours after | Good clot, 3 minutes,.| No clot, 50 minutes clotting 30 seconds Clot, 24 hours Unheated unozxalated serum 9. 20 minutes after clotting Membranous clot, 3 minutes 10. 53 hours after clotting Gel, 6 minutes Gel, 1 hour 11. 20 hours after clotting Gel, 4 minutes, 20 sec- | No clot, 50 minutes onds Clot, 24 hours Experiments of this type show four interesting facts: 1. Metathrombin is not detectable under the conditions of the ex- periment in cat serum immediately after clotting, (cf. 1 in table). 2. Metathrombin is not formed suddenly in maximum amounts, but grows gradually in the serum. The effect of chilling may be re- garded as merely the retarding of the normal process in which meta- thrombin is possibly formed more rapidly but nevertheless gradually. 3. The thrombin present in large amounts in the serum at first gradu- ally diminishes in amount. NATURE AND PROPERTIES OF METATHROMBIN 555 4. The formation of a clot after activation of a serum containing thrombin is no-eyidence of the presence of metathrombin. It has been stated that activation destroys any free thrombin present. There is no doubt that thrombin is weakened when subjected to alkali-acid activation but that it is by no means destroyed may be seen from a comparison of / and 4 in the above experiment. Both of these speci- mens were oxalated so that no further production of thrombin from _ prothrombin could be effected; / had its thrombin destroyed by heat and did not clot after activation; 5 unheated and containing free throm- bin did clot after activation. This latter result must be attributed to the presence of free thrombin in the serum and not to the activation _of metathrombin. The same type of experiment was carried out using oxalated plasma instead of whole blood. Centrifugalized oxalated plasma was recal- cified with a determined optimal amount of calcium, whipped, and fibrin removed and a portion immediately oxalated and heated as in the preceding experiment. From this cell-free serum it was possible to remove the fibrin without centrifugalizing and so the serum could be activated as early as three minutes after clotting. Even at this ‘point metathrombin was invariably detectable. A typical experi- ment will serve: ACTIVATED: + FIBRINOGEN UNACTIVATED: + FIBRINOGEN Recalcified serum oxalated imme- | Membranous clot | No clot, 24 hours diately after clot. Heated 60° 3 between 1 and 2 minutes after clotting hours -Recalcified serum oxalated and | Membranous clot, | No clot, 24 hours heated 60° 23 hours after clotting 37 minutes Recalcified serum oxalated and | Membranous clot, | Noclot, 24 hours heated 60° 20 hours after clotting 10 minutes Recalcified serum oxalated imme- | Gel, 53 minutes Clot, 3 minutes, 15 diately after clotting, unheated 6 minutes after clotting Unheated, recalcified serum, oxa- lated 23 hours after clotting Unheated, recalcified serum, oxa- lated 20 hours after clotting Gel, 5 minutes Membranous clot, 33 minutes seconds Membranous, 2 hours No clot, 1 hour Clot, 24 hours Unheated, unoxalated recalcified serum 7 minutes after clotting Unheated, unoxalated recalcified serum 23 hours after clotting Unheated, unoxalated recalcified serum 20 hours after clotting Gel, 4 minutes Gel, 3 minutes 10 seconds Membranous, 3 minutes Clot, 1 minute, 15 seconds Clot, 6 minutes Membranous, 20 minutes 556 ARNOLD R. RICH Metathrombin is then detectable in a serum obtained by recalci- fication of oxalated plasma earlier than in a serum resulting from normal coagulation. It is well known that an oxalated plasma recalci- fied with its.optimum. amount of calcium clots more rapidly than the same plasma allowed to coagulate unoxalated. The effect of recalci- fication is a more prompt activation of prothrombin with liberation - of a relatively larger amount of thrombin in a shorter interval of time than occurs normally. The bearing of such an accelerated thrombin — production upon a more speedy metathrombin formation will be seen from other experiments. A further experithent was carried out with the object of determining whether metathrombin can be detected in unoxalated plasma,: and of. studying the normal period and mode of appearance of metathrombin. Birds’ blood was used in this experiment. A large rooster was anesthet- ized and bled from the carotid through a paraffined cannula into paraf- fined tubes which were iced as in the preceding experiments. Part of this plasma was at once centrifugalized, the cell-free normal plasma drawn off and immediately activated, one specimen being heated to 60°C. and one unheated. Neither of these activated unoxalated plas- mas or the wnactivated controls gave clots during a period of twenty- four hours, indicating that metathrombin cannot be detected by alkali activation in circulating avian plasma. A second portion of this plasma was not centrifugalized but was allowed to. clot while cooled with its cell elements present. Portions of this clotting plasma were taken at intervals from the tube, oxalated, heated to 60° for one minute and activated. The table shows the results. Plasma taken from carotid of eyes oie 10.15 a.m. 10.15 a.m. 0.5 cc. plasma, oxalated, heated 60°, activated, 0.5 cc. fibrinogen added? - No clot, 24 hours. .10.35 a.m. 0.5 cc. plasma oxalated, heated 60°, activated, 0.5 cc. fibrinogen added. No clot, 24 hours. 10.50 a.m. Same procedure. 11.25 a.m. Fibrin threads present. Same procedure. Membranous clot, 4 hours. 11.25 a.m. 0.5 ce. plastale oxalated, heated 60°, unnoiivaked, using NaCl 0. 9 per cent to make equal dilution. 0.5 cc. fibrinogen added. No clot, 24 hours. NATURE AND PROPERTIES OF METATHROMBIN 557 = 11.35 a.m. Clot. 0.5 ce. serum, oxalated, heated 60°, activated, 0.5 ce. fibrinogen added. -Membranous, 12 minutes. Gel, i5 minutes. Same unactivated, using NaCl 0.9 per cent to make an equal dilution. No clot, 24 hours. 24 hours later. Same procedure (activating ‘fibrinogen used here one-half as strong as that used above). Gel, 13 minutes. Same procedure, unactivated, using NaCl 0.9 per cent to make equal dilu- tion. No clot, 24 hours. Clotting time: Unoxalated plasma, 1 hour, 20 minutes. 0.5 cc. oxalated plasma, 4 drops 0.5 per cent CaCl. solution, 30 minutes. 0.5 ec. unoxalated plasma, 4 drops 0.5 per cent CaCl, solution, 3 hours. 0.5 cc. unoxalated plasma, 4 drops H.O, 44 minutes. 0.5 cc. unoxalated plasma, 4 drops of Cephalin solution, 6 minutes. _ It is seen from such an experiment as this that metathrombin is- not detectable in birds’ circulating plasma. However, in the slowly — clotting blood there is a detectable trace of metathrombin before clot formation is completed. This is suggestive that the processes of meta- thrombin and fibrin formation go on side by side. It will also be noticed that the recalcified plasma clotted in a much shorter time than the normal plasma. _ Experiments such as the three just mentioned demonstrate the fact that metathrombin makes its appearance gradually and its develop- ment is attended by a gradual diminution of the previously strong throm- bie power of the serum. It is well known that if thrombin be added to serum it is readily inactivated. The possibility suggests itself therefore that this inactivation consists of a union of thrombin with antithrombin and that the gradual disappearance of thrombin from a serum marks the formation of just such a combination, the thrombin of which may be subsequently recovered by alkali activation—in other words, that metathrombin is a union of antithrombin and throm- bin formed whenever thrombin is inactivated by antithrombin. If metathrombin is such a combination of antithrombin and thrombin it might be expected that the union would begin to take place in clot- ting blood by the side of the fibrinogen-thrombin union as soon as thrombin is liberated from prothrombin. The above experiments offer a ¢ertain amount of suggestive information in this regard. In cats’ plasma allowed to clot unoxalated, and kept at a low temperature, metathrombin was detectable only some minutes after clotting had occurred. In the recalcified plasma under similar conditions it was detectable immediately after clot formation. Here, as stated above, the thrombin was more quickly liberated from prothrombin. 558 ARNOLD R. RICH If metathrombin is a compound of antithrombin and thrombin, formed whenever antithrombin inactivates thrombin, it should be possible to produce it experimentally by direct addition of thrombin to antithrombin. A difffeulty of this experiment lies in the fact that antithrombin has not been isolated from plasma. It has been shown, however, that there occurs no reaction between thrombin and either thromboplastic substance or prothrombin or fibrinogen which ends in metathrombin formation. The addition of thrombin to an oxalated plasma may be considered then, for experimental purposes, the equiva- lent of the addition of thrombin to an antithrombin solution, unless one assumes the presence in plasma of some as yet unknown substance with which thrombin combines. Further, it has been seen that coagu- lation is not necessary for the production of metathrombin. There- fore an oxalated plasma was heated to 54°, the fibrinogen precipitate filtered off and the plasma recalcified. This entailed the liberation of thrombin into the plasma which contained antithrombin. Tests made upon this mixture at first showed an absence of metathrombin and a high thrombin content, which relation after eighteen hours had be- come relatively reversed. As in the observations upon normal coagu- lation, there was seen here to be a gradual inactivation of the free thrombin with a corresponding appearance of metathrombin. Further experimental evidence that the inactivation of thrombin by antithrombin is always attended by the production of metathrom- bin was obtained by using an oxalated plasma heated at 60° for five. minutes. Such -treatment destroys the fibrinogen and though the antithrombin is considerably weakened, it is not destroyed. This plasma approximates more closely a free antithrombin solution. With this plasma an experiment was made similar to that described by Gasser (6). Thrombin prepared by Howell’s method (8) was dis- solved in this plasma and incubated twenty-four hours at 37°. Throm- bin and metathrombin tests showed a great decrease in free thrombin accompanied by the appearance of metathrombin. In two such ex- experiments, antithrombin tests were carried out with the object of detecting a decrease in antithrombin content which might be expected if part of the antithrombin is combined by the thrombin. The heat- ing of these plasmas at 60° for five minutes before the addition of throm- bin together with the reheating at 60° for two minutes to destroy any free thrombin before each antithrombin test weakened the antithrombin so markedly that the results, while indicating a loss of antithrombin, were not very striking. One such test may be given. NATURE AND PROPERTIES OF METATHROMBIN 559 Oxalated plasma heated at 60° for five minutes, incubated with dry thrombin. 0.5 ce. of this thrombin plasma added as soon as made to 0.5 ce. fibrinogen. ~¥irm clot, 1 minute, 30 seconds. * 0.5 cc. thrombin plasma 20 hours later, plus 0.5 ce. fibrinogen. No clot, 3 hours. ; 0.5 cc. thrombin plasma 20 hours later, heated 60° 1 minute, activated, plus 0.5 ce. fibrinogen. Flocculent fibrin, 30 minutes. 0.5 ce. thrombin plasma 20 hours later, heated 60° 1 minute, plus an equiva- lent amount of 0.7 per cent NaCl plus 0.5 ce. fibrinogen. No clot, 24 hours. Antithrombin: 1 drop thrombin plasma immediately after mixing, heated to 60° 2 minutes plus 2 drops thrombin (15 minutes) plus 10 drops fibrinogen. Clot, 6 minutes. Same, with 3 drops thrombin. Clot, 6 minutes. Same, with 4 drops thrombin. Clot, 6 minutes. 1 drop thrombin plasma 20 hours after mixing, heated 60° 2 minutes plus 2 drops thrombin (15 minutes) plus 10 drops fibrinogen. Clot, 5 minutes. Same, with 3 drops thrombin. Clot, 2 minutes, 30 seconds. _ Same, with 4 drops thrombin. Clot, 2 minutes, 30 seconds. Metathrombin, then, appears to be produced invariably in solutions that contain both free thrombin and free antithrombin. To test the possibility of its formation in the absence of antithrombin the follow- ing experiment was made. All of the known substances concerned ‘in coagulation with the exception of antithrombin were mixed in the following proportions: 4 cc. active prothrombin solution + 4 ce. fibrino- gen + 12 drops 0.5 per cent CaCl. + 20 drops cephalin solution. This mixture clotted and was allowed to stand twenty-four hours before removing the clot. The serum was tested for metathrombin. 0.5 cc. serum immediately after clotting plus 0.5 ce. fibrinogen. Membranous clot, 9 minutes. : 0.5 ec. serum 24 hours after clotting plus 0.5 cc. fibrinogen. Membranous clot, 10 minutés. 0.5 cc. serum 48 hours after clotting plus 0.5 cc. fibrinogen. Gel, 8 minutes. Activation of the serum at these intervals showed no trace of meta- thrombin. It is noticed that there is no diminution of free thrombin in such a mixture. As far as we know, this coagulation mixture con- tained every factor concerned in normal coagulation with the exception of antithrombin. It seems reasonable then to assume that the presence of antithrombin ‘is essential for the formation of metathrombin in any solution. Weymouth reached the same conclusion by weakening the antithrombin of oxalated plasma by dialysis and then recalcifying. The serum showed metathrombin in subnormal amounts. 560 ARNOLD R. RICH The above experiments indicate the following facts: 1. Metathrombin is not produced in solutions which lack either thrombin or antithrombin. 2. Metathrombin is readily produced in sola copbahien both thrombin and antithrombin. 3. In such solutions the free thrombin gradually decreases in amount. 4. There is some evidence that in such solutions the antithrombin also decreases in amount. It is evident that a combination of thrombin and antitheenea to form metathrombin might provide an efficient method, under certain conditions, for preserving the intravascular fluidity of the blood, as has been suggested recently by Gasser (6). It was shown by Davis (9) that large amounts of thrombin might be introduced into the circulation without causing thrombosis. Appar- ently the thrombin in these cases must have been combined with anti- thrombin and so rendered inactive. In the plasma after such an in- jection one would expect to find detectable amounts of metathrombin. Accordingly these experiments were repeated with that object in view. It is evident that thrombin may be introduced into the circulating blood either in pure solution or by the indirect method of injecting thromboplastic substance which should cause the activation of pro- thrombin and the liberation of thrombin in the blood. The latter method was first used. A cat weighing 2.6 kgm. was mhectiniilans the carotid artery of both sides and the femoral vein of the right side were cannulated. The femoral cannula was attached to a burette containing an active solution of cephalin in 0.9 per cent solution of sodium chloride. The activity of this cephalin was determined by its addition to recalcified oxalated plasma immediately before the injection. From the right carotid cannula 4 cc. of blood were run ate a grad- uate containing 0.5 cc. sodium oxalate, care being taken to make all measurements exact. The graduate’ was inverted several times to insure thorough mixing of blood and oxalate. Thirty-five cubic centi- meters of blood were then removed from the circulation and 15 cc. of cephalin solution were slowly run into the femoral vein from the burette, the entire injection consuming five minutes. Five minutes after the injection, 4 cc. of blood were received from the cannula in the left carotid into 0.5 ec. oxalate as before. The right carotid was then freshly cannulated and half an hour after the injection a third specimen of 4 cc. blood in 0.5 ce. oxalate was obtained. The recannulation of NATURE AND PROPERTIES OF METATHROMBIN 561 the left carotid and the withdrawal of a final similar specimen one hour and forty-five minutes after injection, completed the experiment so far as-the injection was concerned. The specimens of blood were centrifugalized, the clear plasma drawn off and tested as shown below. It was noted that no perceptible hemolysis had occurred. Plasma before injection of 15 ce. cephalin Plasma 5 minutes after injection - ; Metathrombin: Plasma 30 minutes after injection Plasma 1 hour, 45 minutes after injection A B Cc D Portions of A, B, C and D were heated to 54° and the fibrinogen precipitate filtered off. 0.5 cc. + alkali activation + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 ec. A + 20 drops 0.9 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 ec. A, unheated, + alkali activation + 0.5 cc. fibrinogen. No clot, 24 hours. The same series was carried using B, C and D with the same re- sults. After forty-eight hours these activation experiments were ‘repeated. In no case was there detectable a trace of metathrombin. The antithrombin tests gave the following results: Antithrombin INCUBATED 15 MINUTES THROMBIN sean pee t od FIBRINOGEN CLOTTED drops drops minutes 2 A 10 20 3 A 10 15 4 A 10 10 5 A 10 10 2 B 10 25 3 B 10 15 4 B 10 10 5 B 10 10 2 C 10 60 3 C 10 30 4 @ 10 25 5 Cc 10 20. 2 D 10 20 3 D 10 15 4 D 10 10 5 D 10 10 THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 43, No. 4 562 ARNOLD R. RICH A marked rise in antithrombin content is noted, followed by.a re- turn to normal. The rise appears to begin shortly after injection. In some instances the five minute specimen exhibited a fall in anti- thrombin content but in all cases the thirty minute specimen presented a marked rise. The return to normal occurs between thirty minutes and an hour and a half after injection. If the cephalin injected was — neutralized by antithrombin we should expect a fall in antithrombin content, and, as stated, this could be detected in some instances in the five minute specimen. _ It was believed that the failure to detect metathrombin in these experiments might have been due to a compensatory production of antithrombin which might thus render the cephalin valueless in aiding thrombin production from prothrombin. It has been shown by Dele- zenne (10), Nolf (11) and others that the liver is probably the seat of antithrombin production. Therefore a similar injection was made in an animal in which both the portal vein and coeliac axis had been ligated, with the object of removing the possibility of any antithrombin output from the liver during the experiment. Exactly the same methods were used in this case as were described in the preceding experi- ment. There was here a similar marked increase in antithrombin in both the five minute and the thirty minute specimens. Metathrom- bin was not detected. . Plasma before injection of 15 ce. cephalin = A Plasma 6 minutes after injection = B Plasma 26 minutes after injection = C Antithrombin THROMBIN pababgherciinay oder FIBRINOGEN CLOTTED drops drops minutes ee A 10 35 3 A 10 25 4 A 10 20 y B 10 45 3 B 10 35 4 B 10 25 2 C 10 1 hour, 30 minutes (imperfect clot) 3 ” 10 55 4 C 10 35 NATURE AND PROPERTIES OF METATHROMBIN 563 Apparently antithrombin was thrown into the circulation from some source other than the liver, for at autopsy the ligatures were found to be intact. It was decided then to establish a head-thorax circula- tion in an animal with the object of removing the influence of all organs below the diaphragm. A cat was anesthetized, tracheotomy performed and artificial respiration established. The thorax was opened and the aorta and inferior vena cava were ligated just above the diaphragm. Ten cubic centimeters of an active cephalin solution were slowly run into the right external jugular vein. Specimens of blood were collected before and after the injection as in the previous experiments. The antithrombin in this case showed a marked increase in amount in the thirty minute specimen. That the output of antithrombin began very soon after the injection may be inferred from the fact that there was no detectable diminution of antithrombin in the five minute specimen. Plasma before injection of 10 cc. cephalin =A Plasma 5 minutes after injection = B Plasma 30 minutes after injection = C Antithrombin THROMBIN Price t = oe FIBRINOGEN CLOTTED drops drops minutes 2 A 10 20 3 A 10 15 4 A 10 10 5 A 10 10 2 B 10 20 3 B 10 15 4 B 10 10 : > B 10 10 2 Cc 10 50 3 Cc 10 35 4 Cc 10 ; 25 5 Cc 10 15 Metathrombin was not detected. That this antithrombin reaction could not take place outside the body was demonstrated by incubating whole oxalated plasma with similar preparations of cephalin at 37°, testing the antithrombic power before and after incubation as in the injection experiments. The addi- 564 ARNOLD R. RICH tion of cephalin caused a diminution in antithrombin content’ which remained constant. There was no subsequent increase in antithrombin whatever. To determine whether the cephalin itself was the stimulus for this increased output of antithrombin, an injection of 20 cc. normal saline was made into the femoral vein of a cat weighing 2.7 kgm. Fifty cubic centimeters of blood were drawn before this experiment to approxi- mate the conditions of the cephalin experiments. Plasma before injection 20 cc. normal saline = A Plasma 5 minutes after injection = B Plasma 30 minutes after injection =C Antithrombin THROMBIN econ ¢ hed Bm dnd pie: FIBRINOGEN CLOTTED drops drops 3 A 10 No clot, 65 minutes 4 A 10 Clot, 50 minutes 3 B 10 Clot, 40 minutes 4 B 10 Clot, 20 minutes 3 C 10 No clot, 65 minutes 4 C 10 Clot, 55 minutes This blood was of a very high antithrombin content before injection and the sole effect of the saline seems to have been merely a primary decrease in antithrombin power, followed by a return to normal. ; Such experiments were unsatisfactory, however, from the stand- point of metathrombin, and it was decided to resort to the injection of pure thrombin. A very strong thrombin solution was prepared and dialyzed against water until the concentration in sodium chloride was about 1 per cent. An experiment was made upon an intact ani- mal in all details similar to the femoral injections of cephalin described above. Ten cubic centimeters thrombin solution were run into the femoral vein of a cat. Plasma before injection of thrombin =A Plasma 5 minutes after injection =B Plasma 30 minutes after injection = C Plasma 1 hour, 30 minutes after injection = D NATURE AND PROPERTIES OF METATHROMBIN 565 Antithrombin THROMBIN ee Besilv.vigay FIBRINOGEN CLOTTED drops tS a drops 2 A 10 No clot, 2 hours 3 A 10 No clot, 2 hours 2 B 10 No clot, 2 hours 3 B 10 No clot, 2 hours <2 C 10 Clot, 2 hours 3 Cc 10 Clot, 37 minutes 2 ; D 10 Clot, 1 hour, 10 minutes 3 D 10 Clot, 57 minutes . _ Metathrombin: : A, B, C, D heated to remove fibrinogen. 0.5 ee. A+ alkali activation +10 drops fibrinogen—no clot, 24 hours. 0.5 ec. A + 20 drops 0.7 per cent NaCl + 10 drops fibrinogen—no clot, 24 hours Similar tests were made on B, C and D with identical results. No metathrombin was detected. It is to be noted, however, that the thrombin injected must have been inactivated, for the autopsy re- vealed no thrombi and the oxalated plasmas remained fluid. Heating to 53° gave a good precipitate of fibrinogen in all cases, further indi- cating that the fibrinogen had not been combined by the thrombin injected. It will also be noted that this blood was of extraordinary antithrombin content, and the effect of the thrombin injection seems to have been merely the lessening of this antithrombin by combina- tion with thrombin. There remained then the possibility that metathrombin is gotten . ’ rid of in the blood as rapidly as it is formed. To test.such a possibility, a cat weighing 3 kgm. was used. The total volume of blood was roughly estimated at 150 cc.; one-third of the amount (50 cc.) was taken from the carotid artery, whipped with a wire brush to hasten coagulation, and the serum filtered off through gauze. This serum containing metathrombin was warmed to body temperature and injected back into the body twenty-five minutes after clot formation. 0.5 ec. serum before reinjection + 1 cc. 0.7 per cent NaCl Oxalated plasma 5 minutes after injection of serum A B Oxalated plasma 30 minutes after injection of serum Cc 566 ARNOLD R. RICH Metathrombin: 0.5 ec. A (heated 60° 1 minute and filtered) + alkali activation + 0.5 ce. fibrinogen. Found clotted, 1 hour, 15 minutes. 0.5 cc. A (heated 60° 1 minute and filtered) + 20 drops 0.7 per cent NaCl + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 ec. B (heated 54° 1 minute and filtered) + alkali activation + 0.5 ce. fibrinogen. Scant fibrin, ppt. 24 hours. 0.5 cc. B (heated 54° 1 minute and filtered) + 20 drops 0.7 percent NaCl + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 ce. C (heated 54° 1 minute and filtered) + alkali activation + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 ce. C (heated 54° 1 minute and filtered) + 20 drops 0.7 per cent NaCl + 0.5 cc. fibrinogen. No clot, 24 hours. Here then was an explanation of the fact that metathrombin could not be detected after cephalin and thrombin injections. Even if it were formed, it became quickly indetectable either by removal from the circulation or by the action of some factor existing normally in plasma. To exclude the possibility of body absorption of metathrom- bin from the circulation, a similar experiment was carried out using plasma outside the body. Fifty minutes after clotting had occurred, centrifugal serum was oxalated and heated to 60° for one minute to destroy free thrombin and prevent further activation of the serum prothrombin. The follow- ing mixtures were then made: A = 3 cc. whole oxalated blood + 3 cc. serum; incubated 30 minutes at 37°. B = 3 ce. centrifugalized oxalated plasma + 3 cc. serum incubated 30 minutes at 37°. C = 3 ce. centrifugalized exalaaa plasma heated to 54° and filtered + 3 cc. serum; incubated 30 minutes 37°. D = 3 ce. 0.9 per cent NaCl + 3 ec. serum; incubated 30 minutes 37°. Metathrombin tests made upon these mixtures after heating each to 54° and filtering are shown in the following table: . Metathrombin: 0.5cc. A+ alkali activation + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 cc. A + 20 drops 0.7 per cent NaCl + 0.5 ec. fibrinogen. No clot, 24 hours. 0.5cc. B+ alkali activation + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 cc. B+ 20 drop s0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 cc. C + alkali activation + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 ce. C + 20 drops 0.7 per cent NaCl + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 cc. D + alkali activation + 0.5 ce. fibrinogen. Firm clot, 3 hours. 0.5 cc. D + 29 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. NATURE AND PROPERTIES OF METATHROMBIN 567 Tt is seen that metathrombin apparently disappears when added to blood outside the body. Further the presence of blood elements plays no part_in this disappearance of metathrombin, for centrifugal- ized plasma exhibited the same result as whole blood. The clot formed after activation of D shows that the metathrombin of the serum was not destroyed by mere dilution. It was possible, of course, that plasma in some way destroyed metathrombin, but there is nothing in plasma so far as is known except fibrinogen which is not present also in serum, and it is clear that the fibrinogen-free plasma exerted the same influ- ence upon metathrombin as the other plasmas. A simpler explana- tion than that of a destruction of metathrombin by plasma suggests itself: We have reason to believe that in plasma there is a certain amount of free or uncombined antithrombin. In serum, on the contrary, some of the free antithrombin has been removed by combination with cephalin and the remainder is in ccmbination, loose or firm, with thrombin. When the serum is submitted to alkali activation the antithrombin of the metathrombin is destroyed and the liberated thrombin finds but little antithrombin to combine with and may there- fore be detected immediately after the activation. When the plasma is submitted to the same process, some of the free antithrombin is removed but enough remains to insure a rapid combination with the thrombin liberated by the activation and this combination occurs so rapidly as to obscure the detection of the thrombin. If this reasoning is correct, it would follow that a plasma-serum mixture might reveal the presence of metathrombin if the plasma were first heated sufficiently to weaken greatly its content of free antithrombin, or if the mixture were submitted to a reactivation by alkali, on the ground that the first activation would weaken the amount of free antithrombin. The following experiments were therefore made. Fifty minutes after clot formation, the centrifugalized serum was oxalated and heated at 60° for 14 minutes. 2 ce. oxalated plasma unheated + 2 cc. serum incubated 37° for 1 hour = A. 2 cc. oxalated plasma heated 56° and filtered + 2 cc. serum incubated 37° for 1 hour = B. 2 ce. oxalated plasma heated 60° 6 minutes and filtered + 2 cc. serum incu- bated 37° for 1 hour = C. 2 ce. oxalated plasma heated 70° five minutes and filfered + 2 cc. serum -in- cubated 37° for 1 hour = D. 2 ce. 0.9 per cent NaCl + 2 cc. serum incubated 37° for 1 hour = E. 568 ARNOLD R. RICH It is observed that the plasmas added to the metathrombin-contain- ing serum were heated to various temperatures up to 70°, the critical temperature of antithrombin. It is known that heating even to 60° weakens antithrombin. Metathrombin tests were made after each specimen was heated at 54° one minute and filtered. Metathrombin: 0.5cc. A+ alkali activation + 0.5 cc. fibrinogen. No clot, 24 hours. 0.5 ce. A + 20 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 cc. A activated, neutralized, reactivated, + 0.5 fibrinogen. No clot, 24 hours. 0.5 cc. B+ alkali activation + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 cc. B + 20 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5 cc. B activated, neutralized, reactivated + 0.5 cc. fibrinogen. Membra- nous clot, 24 hours. 0.5 cc. C + alkali activation + 0.5 ce. fibrinogen. Good gel, 3 hours. 0.5 cc. C + 20 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. 0.5cc. D + alkali activation + 0.5 cc. fibrinogen. Good gel, 3 hours. 0.5 cc, D + 29 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. Noclot, 24hours. 0.5 cc. E + alkali activation + 0.5 cc. fibrinogen. Membranous clot, 1 hour. Solid, 3 hours. 0.5 cc. E + 20 drops 0.7 per cent NaCl + 0.5 ce. fibrinogen. No clot, 24 hours. It is seen that the presence of metathrombin was detected in the mixtures the plasmas of which had been heated at 60° to 70° with a corresponding weakening of their antithrombin. Further, one acti- vation failed to reveal metathrombin in the 56° plasma mixture (B) whereas a second activation of the solution (then weaker in antithrom- bin from the effects of the first activation and by reason of some of the antithrombin having combined with the thrombin liberated by this activation) showed the presence of metathrombin. These experiments explain, possibly, the fact that metathrombin ~ has never been demonstrated in plasma, as well as the failure to detect metathrombin in plasma after injections of cephalin, thrombin or meta- thrombin itself. The demonstration is in agreement with the injec- tion experiments, as in these cases there was invariably a marked rise in the antithrombin content of the blood. The proof that metathrom- bin may be present in a solution and yet be “masked” by a high anti- thrombin content of the solution lends support to the view that meta- thrombin may be formed regularly in circulating blood as a protective mechanism as suggested above. The free antithrombin content of the plasma would render such small amounts indetectable by alkali NATURE AND PROPERTIES OF METATHROMBIN 569 activation. At all events it is clear that the failure to demonstrate metathrombin in the plasma by alkali activation cannot be accepted as an argument against such a view. CONCLUSIONS 1. Metathrombin is a thrombin-antithrombin compound. The fol- lowing facts may be offered in support of this view: _ a. The formation of metathrombin is not dependent directly upon any of the three essential processes of coagulation, the action of throm- boplastic substance, the calcium activation of prothrombin or the formation of fibrin. b. Metathrombin cannot be produced by the interaction of any known substances concerned in coagulation except thrombin and antithrombin. c. Metathrombin cannot be produced in any solution from which either thrombin or antithrombin is absent. d. Metathrombin is readily formed in solutions containing both antithrombin and thrombin. e. In such solutions the thrombin gradually diminishes in amount. jf. There is evidence that in such solutions the antithrombin also diminishes in amount. 2. Metathrombin added to blood inside or outside the body cannot be detected by the method of alkali activation. -The explanation offered for this fact is that the thrombin liberated by the activation is rapidly recombined by the free antithrombin of the blood. 3. On the basis of 2, it is suggested that metathrombin may be con- stantly forming in circulating blood although not detectable by the method of alkali activation; and that this process may serve to protect the blood from the coagulating effect of thrombin liberated within the circulation. 4. The injection of cephalin (tissue extract) into the external jugu- lar vein (cat) causes a marked increase in the antithrombin content of blood kept circulating through the head and thorax only. No in- crease in antithrombin content occurs in whole blood to which cephalin is added in vitro at body temperature. This is offered as evidence that the abdominal viscera cannot be regarded as the sole source of antithrombin. It is a great pleasure to me to thank Dr. Howell for his guidance in this work. 570 ARNOLD R. RICH BIBLIOGRAPHY (1) Fuxp: Centralbl. f. Physiol., 1903, xxvii, 533. (2) Morawitz: Ergebn. d. Physiol., 1905, iv, 367. (3) WrymoutH: This Journal, 1913, xxxii, 266. (4) Mexansy: Journ. Physiol., 1908, xxxviii, 28. (5) CoLttiInewoop anp McManon: Journ. Physiol., 1912, xlv, 119. (6) Gasser: This Journal, 1917, xlii, 378. (7) Howeu: This Journal, 1912, xxxi, 1. (8) Howretu: This Journal, 1913, xxxii, 264. (9) Davis: This Journal, 1911, xxix, 160. (10) DeLezENNE: Travaux de Physiol. (Univ. de Montpellier) 1898. (11) Nour: Arch. Internat. d. Physiol., 1910, ix, 407. THE CHANGES IN CLOTTING POWER OF AN OXALATED PLASMA-ON STANDING ARNOLD R. RICH From the Physiological Laboratory, Johns Hopkins University Received for publication May 11, 1917 Howell’ has shown that in testing blood to determine the presence of a hemophilic tendency, it is more satisfactory to obtain the clot- ting time of the oxalated and centrifugalized plasma after recalcifica- tion than to depend upon the time of coagulation of the whole blood, since in the latter case small variations in conditions may make large differences in the figures obtained. In connection with this procedure and also as a matter of general interest, it was thought desirable to ascertain to what extent the coagulating property of an oxalated plasma undergoes alteration upon keeping, and the effect upon this property of temperature and of sterile versus non-sterile conditions. The fol- lowing experiments were made at the suggestion of Dr. W. H. Howell with the object of testing these points. ; It is evident that determinations of the clotting time of blood kept over a period must be made upon oxalate or fluoride plasma. In these experiments, therefore, the following method was adopted: A cannula was introduced into one of the carotid arteries of an anes- thetized cat, and the blood allowed to flow into centrifuge tubes con- taining one part of 1 per cent sodium oxalate for every eight parts - of blood. The blood and oxalate were thoroughly mixed and then centrifugalized for twenty minutes. The cell-free plasma was pipetted off and divided into three parts, one of which was kept at 4°C. during the period of the experiment, another at room temperature and the third at 37°. The clotting time of these specimens was determined at intervals over a period of twenty-four hours by the method of recal- cification. It is well known that if calcium be added to a plasma kept fluid by oxalate precipitation of its ionizable calcium the plasma will readily coagulate, the rapidity of coagulation being determined by the 1 Arch. Int. Med., 1914, xiii, 76. 571 572 ARNOLD R. RICH amount of calcium added. There is for every oxalated plasma a certain ‘optimum amount” of calcium, the addition of which will cause coagulation in the shortest time possible for that plasma under given conditions. If calcium be added in amounts under this opti- mum, the clotting time is appreciably slower, and the same is true for calcium in amounts above the optimum. In order to eliminate the errors arising from recalcifying with uncertain proportions of calcium, a series was carried for each test consisting of five clotting tubes, each of which contained 0.5 ec. of the specimen to be tested. To these tubes were added respectively three, four, five, six and eight drops of a 0.5 per cent calcium chloride solution.2 The tube clotting first was -assumed to contain the optimum amount of calcium chloride. For the cat, using this method, the optimum amount of calcium chloride ranged between four and six drops, for different bloods. One such experiment upon human blood showed an optimum of four drops. The amount of calcium necessary to exert this optimum effect upon coagulation remained fairly constant during the period of each experi- ment, the slight variations falling well within the limits of experimental error. It is realized, when one considers the very minute quantities of calcium which affect coagulation, that even variations in the size of the drops due to temperature changes during the period of the ex- periment may influence the clotting time. The coagulation time of the recalcified plasma was found to en: _ markedly during a period of twenty-four hours. In most cases, the variations in the coagulation time of plasma kept at room temperature were negligible up to about four hours after the blood was drawn, when a marked lengthening occurred and persisted steadily, with the result that after twenty-four hours the clotting time was three to seven . times as long as it was at the beginning of the experiment. The temperature at which the plasma is kept was found to exert a definite effect upon the clotting time. The plasma kept at 4° exhibited a very much less loss of clotting power than did that kept at room temperature. Plasma kept at 37°, on the other hand,.showed a very much greater loss than that kept at room temperature (fig. 1). It was noticed in all cases that evidences of putrefaction were most prominent in the specimens kept at 37°—the ones which exhibited the most marked lengthening of the coagulation time during twenty- * The solutions of calcium chloride used for this purpose should be prepared from the crystal or ny Gee preparations rather than from the granulated or fused form. CLOTTING POWER OF OXALATED PLASMA 573 four hours. This was suggestive of the possibility that the loss of clotting power might be the result of bacterial action. Accordingly sterile plasma was collected in the following manner: _ The required amount of oxalate was put into small-necked centrifuge tubes which were plugged with cotton and autoclaved. A’ clean 24r No Clot in 24 24 hours } HEN a physician writes a prescription for a fluid extract, tincture, elixir, pill, tablet, or other form of medicament, what assurance has he that the agent he is prescribing is worthy of confidence? What guaranty has he that it possesses the requisite degree of activity? What warrant has he to expect a definite result from a definite dosage? These are important questions. One product may differ widely from another product identical with it in physical appearance and bearing the selfsame name upon its label. This lack of uniformity may arise from diversity in the process of manufacture. It may be due to the fact that the active constituents of the crude drugs entering into the medicaments vary from season to season and are modified by habitat, by climatic influence, or by differ- ing modes of curing and storage. Definite medicinal strength in a therapeutic agent is of the utmost importance. There is one method of assuring it. That method is by assay. That method is our method. We standardize our entire output of pharmaceutical and biological products—chemically or physiologically, according to the exigencies of circumstance. We were pioneers in standardization, putting forth the first assayed preparation nearly forty years ago. We championed standardization when it was ridiculed by “con- servative’ manufacturers throughout the length and breadth of the country, holding then, as we hold to-day, that the essential value of a therapeutic agent lies in the definiteness of its activity. The medicinal products of our manufacture are uniformly potent and reliable— the same to-day, to-morrow, next week, next year. a Parke, Davis & Co. ea Bee Pe aan BINDING SECT. MAR 4 1966 QP American Journa] of Physiology A5 NIG Biological & Medical | | | PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY a, We Be ITTLE VOvRenAL STORAG relate rete plas caf pita irae 3 etl Wy erry 14 uit ie Bae ay tn i nu i aed oH é pou ish anieaay ner w ye 4 i i Sana A : ati ana sites! Fp ane Ha erAtt ie i Ae i ke ier : Mt , Hee eid + bbe ; ied ity Nera if ih HUE hg eT Pat na erate hat te tt pried Ms reece i rf Pate te ee ik ‘ Ke ’ at | eal nas hae oe wieteeseyls } Ae iit fy ts pe} rh Pitan iy en De pera rae ov bh te why ‘ Leth gages pes : S Reposted y ey iv. > ape! Megat yt ea as end site eae aks nye ics Babee igtet rae re fi adecit eaten ee estat yin hes i ae weer sgisid : i it ue a = o a wt if an res ag, a ue barA See at iran as nop - ee ‘. vr Wasa EL As Pe Wh te We a a ts Boe aye bank donee Re Nd be