The Effects of Oxygen Lack on Peripheral Nerves Citation Wright, Ernest Bevier (1945) The Effects of Oxygen Lack on Peripheral Nerves. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/6kv4-r626. https://resolver.caltech.edu/CaltechTHESIS:06092025-160037685 Abstract No abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Biology) Degree Grantor: California Institute of Technology Division: Biology Major Option: Biology Thesis Availability: Public (worldwide access) Research Advisor(s): Van Harreveld, Anthonie (advisor) Wiersma, Cornelius A. G (advisor) Went, Frits W. (advisor) Thesis Committee: Unknown, Unknown Defense Date: 14 May 1945 Record Number: CaltechTHESIS:06092025-160037685 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:06092025-160037685 DOI: 10.7907/6kv4-r626 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17432 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 18 Jun 2025 17:30 Last Modified: 18 Jun 2025 17:34 Thesis Files PDF - Final Version See Usage Policy. 39MB Repository Staff Only: item control page

The Effects of Oxygen Lack on Peri pheral Nerves Thesis by Ernest Bevier Wri ght In Pa rtial Fulfillment of The Re quire ments For The De gree Of Doctor Of Philosophy California Institute of Technology Pas adena, California May 14, 1945 ACKNOWLEDGEMENT I wish to thank sincerely Dr. A. van Harreveld for suggesting the problem and Drs. C.A.G. Wiersma and van Harreveld for their advice and guidance throughout the work. I am also very grateful for the privilege of completing the minor requirement under the supervision of Dr. F. W. Went. I owe him many thanks for the facilities placed at my disposal by him. I am indebted to Dr. C.H. Ellis for criticizing the manuscript. INDEX Introduction- - 1 Peripheral Nerve Metabolism- - - - - - - - - 1 Peripheral Nerve Asphyxiation in Cold Blooded Animals- - - - - - - - - - - - - - - 2 Nerve Depolarization During Asphyxiation 6 Recovery - - - - - - - - - - - - - - - - - - - 8 Asphyxiation of Peripheral Nerve of Warm Blooded 12 Animals,--and Recovery of same Methods - - - - - - - - - - - - - 16 Nerve Asphyxiation Chambers- - - - - - - 16 Electrical Equipment - - - - - - - - 19 Temperature Control- - - - - - - - - - - 22 Gases- - - - - - - - 22 Nerve Preparations - 23 Results A, Mammalian Nerves- - - 23 Amphibian Nerves- - 26 Crustacean Nerves - 26 Asphyxiation- - - - - - - - - - - - - Survival Time- - - - 28 28 Rabbit Nerve- 28 Cat, Dog, Lobster Nerves- - - - - - 29 Frog, Crayfish Nerves - - - - - - - - - 30 Effect of Time in Solution before Asphyxiation on Survival Time - - - - - - - - - 31 Effect of Ionic Concentration on Survival Time - - - - - - - - - -- - - - - - - - - - 33 Temperature Effect on Survival Time- - - - 34 Stimulation and Fati g ue Effect on Survival Time- - - - - - - - - - - - - - - - 35 Excitability Threshold a nd Action Potential amp litude Chang es During Anoxia- - - - End-in Effect- - - - Reduced All- - 35 36 - - - - 41 Refractory Period- - - - 43 Chronaxie- 45 Me mbrane Dep ol a riz a tion Du ring An oxi a - - - - - - 46 Result s B, 49 Recovery- - - - - - - - - - - Mammali a n Nerve- - 49 Lob s ter Nerve- - 49 Frog Nerve - - - - - - - - - - - - - 50 Effect of Asp hyxi a tion Ti me on Recovery- 50 Effect of I mmersion in Solution on Recovery- 51 Effect of Temp era ture on Re covery- - - - - - 53 Fre quencyaf Sti mul a tion Effect on Recovery - - - 55 Recov e ry of Threshold of Ex citation and Amp litude of Action Pot e ntial- - - - - - - - - - - - - 55 End-in Effect- - - - - - - - 56 Refr a.ctory Peri od- - - - - - - - 57 Oxyg en 11 rest s 11 a nd Re p etitive Anoxia - - - - 58 De p ol a rization Recovery- - - - - - - - 58 To x ic Subst a nce- - - - - - - - - 60 - - - - pH Chang e of Sol ution During Asp hy x i a tion - 6 2 Te s t of Solution Surrounding Nerves Asp hyxi a ted Long Ti me s - - - - - - - - - - - - - - 63 64 DiscussionSumma ry - - - Bibliogr aphy- - - - - - - - - - - - - - - - - 77 - - - - - - - - - 79 INTRODUCTION Peripheral Nerve Metabolism The metabolic gas exchange of peripheral nerves was not scientiLically established until the beginning of the 20th century. Early attempts to measure the oxygen uptake in nerve were unsuccessful (Helmholtz 1868), the very small amount escaping detection by the methods used at that time. The first nerve asphyxiation experiments failed too (Ranke 1868, Ewald 1868) , 'b ecause no suitable physiological solution was to be had at that time, and nerves lost their excitability as quickly in oxygen as they did in hydrogen or nitrogen. However, the oxygen consumption of the central nervous system had been detected, so it became the concensus of· opinion that the nerve metabolism was confined to the nerve cell body. The axone activity was considered to be anaerobic, a purely physical phenomenon, and as would be expected from such a conception, it was indefatiguable. In 1904, the oxygen consumption of frog nerve was actually detected for the first time by Thunberg. He devised a micro-capillary-manometer in which small movements of a kerosene bubble recorded minute decreases in the oxygen volume of the system. This method was too crude for accurate quantitive determinations, but with some modifications (Fenn 1927-8-9) it proved capable of good volumetric measurements. The CO 2 emanating from a frog nerve was detected for the first time by Tashiro (1913-14-15). He described the formation of a cloudy p~ipitate of BaCO;t on the surface 2. of a droplet of BaCl2 solution sealed in an airtight chamber with the nerve. Although he claimed to be able to cal- culate by this method not only the actual amount of co 2 given off, but the change in this amount brought on by nervous activity, these data are questionable. Adoption of the Haas indicator method by Moore (1919) and Parker (1925-26) led to much more dependable results. This method was based on the color change of phenosulfaneph,lein caused by the production of carbonic acid. 1 Winterstein (1926) and Chang (193.,0') have also compiled data on the metabolism of peripheral nerve. The subject has been excellently reviewed by Gerard (1932). Peripheral Nerve Asphyxiation in Cold Blooded Animals Mathews, in 1901, was puzzled by the rapidity with which nervous activity disappeared in freshly killed animals. Em- ploying the string galvonometer to record the nerve impulse directly he discovered that if the nerves were removed immedi a tely upon the death of the animal and placed in the proper physiolog ical solution they were able to maintain their excitability for many hours. Delay in the removal impaired the survival of the excised nerve. This reaction suggested to Mathews the possibility of anoxia, so he re peated the experiments of Ranke and Ewald (Mathews 1902). A frog nerve was well washed in saline and then placed on electrodes in a small container which was sealed shut. Hydrogen, well-satu- rated with wa ter vapor, wa s made to fill the chrunber. was used in the control experiments. Oxygen After several hours the response from a nerve in hydrogen atmosphere disappeared, whereas nerves in oxygen were electrically as active as ever after the same length of time. Thus frog peripheral nerves were found to be asphyxiable. Nerve asphyxiation became the subject of a series of intensive investigations by the German school (vonBraeyer, Frohlich, Fillie, Thorner, Gottschalk) under the direction of Vervorn from 1900 to 1920. Further contributions in the field have been offered more recently by Cooper (1923), Koch (1926), Fenn (loo cit), Furusawa (1929), Heinbecker (1929-31), Gerard (1930), Cowan (1934), Thompson (1936), and Lehmann (1937). In all of the older work and in part of the new work the frog nerve was the preparation used. Gerard asphyxiated a few dog nerves, but collected most of his data from frog nerves. Koch and Lehmann worked exclu- sively with mammalian material, rabbit nerves and cat nerves respectively, while Furusawa and Cowan limited their study to crustacean nerves. Thompson observed anoxia in human nerves. Frog nerves were found to be able to maintain their activity for about an hour in an atmosphere of nitrogen or hydrogen, but then they began to lose their excitability slowly, finally becoming inactive after 3-5 hours (von Braeyer 1902, Frohlich 1903, Thorner 1911, Gottschalk 1914-18, Cooper, Gerard, and Heinbecker). The origi~al postulation by von Braeyer that the latency in the onset of the anoxia effect was due to an oxygen reserve in the nerve tissue was generally 4. agreed upon. From results obtained in experiments in which the nerve was stimulated at several places von Braeyer contended that the loss of excitability progressed centrally along the part of the nerve asphyxiated in the chamber. This contention was disputed by Frohlich who argued that the conduction along partially asphyxiated nerve became decremental near the chamber walls, an effect described later by Kato (1924) for narcotized nerve. Frohlich failed to find any directional loss of activity during asphyxia. Cooper's investigation of the least interval and recovery time of asphyxiated nerve also indicated conduction with decrement. 3 The least interval (Lucas 191)() was the least time between two stimuli, the second 8 times the strength of the first, to cause muscle summation. The recovery time was the least time between two stimuli, the second twice the size of the first, to cause the muscle to summate. The re- covery time was observed to remain constant, the least interval increased during the anoxia period. The constant recov- ery time seen by Cooper instigated an argument between her and Frohlich. Whereas Frohlich had concluded from experiments using stimuli of different frequencies on nerves in nitrogen that the refractory period of the nerve lengthened to as much as 0.1 sec during anoxia, Cooper claimed that the constant recovery time could only mean a constant refractory period throughout asphyxiation. She based her argument for decre- mental conduction on this claim. by both Heinbecker and Gerard. Cooper's results were refuted The latter repeated Frohlich's 5. experiments recording the integrated nerve action potentials rather than the muscle twitch and obtained results similar to those reported by Frohlich. Heinbecker could actually observe the broadening of the potential spike during nerve asphyxiation, and noted too the increase in the time interval between two identical stimuli necessary in order to produce two identical spikes. Thus the refractory period definitely varied during anoxia. The question of decrement although "tempting" as Gerard puts it, was not brought up again in connection with nerve asphyxiation. This was probably because of Kato•s thorough investigation of the s ubject using narcotics, the results of which failed to show a decrement in conduction in frog nerve single fibres. Working with whole nerves as Cooper had, the problem of decremental conduction in various numbers of fibres at various places in the chamber and at various times presents a difficulty which can only be eliminated with the elimination of all but one of the fibres. The blocking of nerve conduction by asphyxiation was reported by both von Braeyer and Frohlich. They distinguished between conduction and excitability by varying the position of the stimulating electrodes relative to the asphyxiated area. For conduction measurements this area was between the stimulus and the muscle. For examination of excitability the stimulus was applied directly in the asphyxiated area. The conduction was found to be normal for at least 2 to 3 hours after the nerve wa s in an oxygen free medium, then suddenly disappeared. It was very evident that most of the nerve 6. fibres were blocked all at once. This finding was not con- firmed by Gerard who contended that nerve fibres stopped conducting at varying times and at varying places. Frohlich was able to ascertain that the velocity of conduction was reduced by anoxia, the amoun.t of reduction depending on the length of nerve asphyxiated. lationship between the two was found. No definite re- Heinbecker describes the velocity reduction also, but do es not speak of the effect of the length of the asphyxiated part of the nerve. Temperature increase decreased the survival time of nerves in nitrogen or hydrogen (von Braeyer, Gerard) as was expected on the basis that such a change would tend t o increase the rate of the nerve metabolism. Fenn was able to measure the increase in Oxygen consumption of frog nerve with increasing temperature. Nerve Membrane Depoluriztion During Asphyxiation Koch, 1926, succeeded in developing a technique whereby certain rabbit nerves were made anemic by arterial occlusion. Before stopping the circulation, in the leg, however, electrodes were placed on the sciatic nerve. Care was t aken to first crush the nerve under one of the leads, or cut it, so that the demarcation potential was recorded. Then the cir- culation was cut off, and the variation of the demarcation potential was observed. At first there was a brief voltage increase, then a steady decrease to a very small constant value. With the resumption of circulation the potential re- covered rapidly to its original value, overshooting it in some cases. 7. The experiments of Koch were as important in the field of nerve anoxia as were those experiments of Mathews, for with these results it was possible to understand for the first time what was happening to the nerve deprived of oxygen. The nerve membrane was depolarized when the nerve was subjected to an oxygen free atmosphere. Koch simply assumed that the nerve became inactive and this drop in injury potential was a measure of the inactivation. He did not realize the significance of his results. Furusawa (1929) investigated.with a ballistic galvanometer the demarcation potential of the nerve from a crab claw noting the changes in this potential during anoxia. With the oxygen removed the voltage between injured and normal tissue (inside and outside of the membrane) slowly became smaller and disappeared. membrane was being depolarized. ll- Fur~sawa realized that the Gerard confirmed this dis- covery using frog nerves and Cowan observed the same with Maia nerves. Cowan suggested that the membrane lost its abil i ty to maintain the potassium ion .gradient across the membrane, its selective semi-permeability failing to function without oxygen so that potassium leaked out and the concentration gradient potential was reduced. V = RT log Cinside = 0.058 log E1_ NF Coutside Co Immediately the question of decrement was brought up again, but in a modified form termed reduced "all" by Gerard. Gerard did a rather pecuJiar thing as regards this problem. 8. He describes the assumption of linear blocking of nerve fibres, that is 11 a 11 fibres are blocked if 11 A11 length of nerve exposed and 11 2a 11 fibres block if 11 2A 11 length exposed, as unsafe, but adds that if it were true then the following calculation could be made. The reduction of the action poten- tial of the nerve at the end of the chamber to one half its normal value would mean non-conduction in 50% of the fibres at that point. Then only 2'5% would be non-conducting at the middle of the chamber so the potential should be 3/4 normal there. However he found the potential only 63% nor- mal in the center of the chamber instead of 75% indicating a slightly reduced potential in each fibre. He then con- cludes that the individual 11 all 11 potential of each fibre is reduced during anoxia as might be expected on the basis of the depolarization of the membrane without oxygen. As Gerard stated the method of arriving at the conclusion was not sound. A better method in which whole nerves are used is included in the results of this paper, but for such observations only single nerve fibres should be employed. According to Heinbecker the survival time varied directly with fibre size, the larger fibres lasting longer without oxygen. Gerard also came to this conclusion. Recovery The effect of anoxia was reversible with the substitution of air or oxygen for the asphyxiating gas, but the extent of the reversibility was cause for debate. Von Braeyer obtained complete recoveries in frog nerves after 3 - 10 9. minutes in oxygen. He used the muscle response to record the threshold of nerve activity so that· the actual spike potential could have been reduced either by partial fibre blocking or by a reduction of the ference being detectable. 11 all 11 without such a dif- Gerard also obtained complete recoveries after short periods of oxygen following asphyxiation. He observed this using the integrated nerve action potential as a measure of the nerve activity. At the start of recovery and also at the start of asphyxiation a period characterized by super normal action potentials was noted which he attributed to the broadening of the impulse. With increasing recovery the impulse shortened reducing the integrated action potentials. Frohlich observed that the threshold of many of the asphyxiated frog· nerves never returned to their normal value with oxygen alone. In fact washing with oxygen free saline had more effect on the reversing than did the gas, oxygen, by itself. Aerated physiological solution rarely failed to bring about complete recovery of a nerve washed in it long enough. The conclusion was drawn that during asphyxiation, nerve metabolic end products normally removed by oxidation collected, these metabolites eventually becoming so concentrated as to act toxically on the nerve itself. Oxygen a- lone was not able to remove them once they had accumulated in high concentrations. Evidently they could be carried off by washing in -the saline. These results and conclusions were confirmed by Fillie. Besides this he found that when submerged in saline the nerve survival time during anoxia was increased. Thorner, Gottschalk, and Cooper also observed the same effects described by Frohlich and Fillie. Heinbecker reported only partial recoveries with oxygen or air alone after asphyxiation and attempted to determine the postulated toxic substance. He investigated effects of high concentration of normal metabolites on recovery of asphyxiated nerve~-$.lbstituting CO 2 for nitrogen did not prevent the complete recovery of the action potentials with oxygen. The same results were obtained when high concentrations, compared to normal, of lactic acid were used. Thus neither of these common end-products of metabolism was the cause of the poisoning of the nerve. Gerard,having obtained complete recoveries, had discarded the toxic substance idea, but it was brought back again only a few years later by Cowan as an explanation for the partial recoveries of Maia nerves from asphyxi a tion. Cowan 1 s contention was that the potassium leaking out of the nerve fibre as suggested by the depolarization might reach a concentration outside the nerve which would depress the excitability. Potassium was known to depress nervous activity (Graham 1933). However, no direct tests for potas- sium were made by Cowan and no further attempts have been made since that time to find just what the cause of the inhibition of recovery is. Other nerve asphyxiation experiments of interest have 11. been reported and should be included here in a brief survey. Von Braeyer found that nerves could be asphyxiated by placing them in a brei of iron filings and NaCl solution. The oxygen was drawn from the nerve tissue to the iron to form iron oxide and the nerve lost its excitability. Complete recoveries were observed when the nerve was removed from this medium and oxygen administered to it. Frohlich found that nerves taken from frogs in the winter had a shorter survival time in nitrogen than those taken from frogs in the summer. He noted too that the survival time of nerves from starved frogs was always less than the survival time of nerves from well fed animals. Frohlich decided that in the winter the hibernating frog was in much the same condition as a starved frog, the lack of nutrition being the cause of the shorter time for asphyxia. Thorner reported fatigue in frog nerve becoming very prominent in the later stages of asphyxiation and early part of recovery. Gottschalk found that a 25 minute oxygen inter- val between anoxia periods was necessary to pring the nerve back to its original state as determined by the duplication of the original survival time. If the oxygen "rest" were e less than this optimum tb.«m a relationship between the time in oxygen and the following survival time was found to . hold. Oxygen (1) survival(l) = Oxygen(2) surviva1(2) etc. Gottschalk claimed that the nerve recovered its excitability in 5-10 minutes after complete asphyxiation, but that it took another 15-20 minutes for the reserve supply to reach 1 I 12. its normal level. The survival time of the repeated asphyx- iations was dependent on how full the reserve was. Asphyxiation of Peripheral Nerve of Warm Blooded Animals There have been only five previous investigations in which the mammalian nerve was the object for experimentation. The work of Koch, in 1926, has already been described. Yosomiya in 1927 asphyxiated rabbits, in toto, and found that the nerve sensitivity increased slightly at the start of asphyxiation but soon began to decrease, the reduction continuing until the death of the animal. The effect of oxy- gen lack on the nerve threshold became noticeable for peripheral nerves at oxygen tensions of 8-9%, for sympathetic nerves at 5-6%. Gerard found that the nerves from dogs were asphyxiable in less than an hour, a shorter survival than that exhibited by the frog nerve. These recordings were made at 23 degrees centigrade so that at normal body temperature it would be expected that the survival time would be still less. Asphyxiation of human nerves was accomplished (Thompson 1936) by applying a pressure of 160 mms Hg with a cuff around the upper arm. During the first 12 minutes there was a slow lowering of the threshold followed by a sudden rapid rise until no further response was evident. release of the pressure. Recovery was rapid with The threshold stimulating the ulnar and medial nerves was defined by the amount of alternating current necessary to be just felt by the subject. The most recent work limited exclusively to mammalian 13. peripheral nerve asphyxiation was carried out by Lehman (1937). It was stated by Lehman at the outset that results previously obtained with nerves in Ringers solution could not necessarily be considered reliable because the state of the nerve 'in ordinary physiological solution was not comparable to its state when in vivo. (see Grundfest and Gasser 1936). It was found that the hydrogen ion concentration of the solution had a strong effect on some of the electrical responses of the nerves. The threshold of activity and the after-potentials could be easily varied by small pH changes, but the spikes were found to be but little affected. The results of this preliminary survey showed that the threshold rose with increasing acidity and fell with increasing alkalinity even to the point where spontaneous discharges were observed. At pH 7.4 there was excellent correspondence be- tween nerve potentials recorded either from the intact animal or from excised material. In the same survey the CaCl2 and KCl balance in the solution was studied and it was found that removal of the CaC1 2 had the same effect as raising the pH, supporting the view that increased pH lowered the calcium ion concentration. Ample data had been gathered previously to do with K ion changes (Graham 1933, 1935 see discussion) and this was confirmed. The chrunber used by Lehman for the asphyxiation work was in the shape of a small cylinder. The nerve was sus- pended on Silver-Silver chloride electrodes placed in the chamber so that with the chamber half filled with solution 14. A 90 degree rotation the nerve was just above the surfac~. would completely immerse the nerve. Ordinary commercial nitrogen was used as the o. 11~ oxygen it contained was not enough to make any difference in asphyxiation time. Induction stimuli were used and all records were made with the cathode ray oscillograph. With the onset of asphyxia the threshold was seen to rise slowly after a preliminary drop so that cat nerves (phrenic, peroneal, tibial, and saphenous) were inexcitable in about 25 minutes. Oxygen restored the spike po- tential in only a few minutes, but the after potential became super sized at first before subsiding to normal. Most extraordinary, however, were the changes exhibited by the -/he. threshold in .e,a,e, course of recovery. In the first two min- utes it returned almost to normal only to rise again to 3 times its normal value and then slowly settle to its starting value, the whole procedure requiring 50 minutes. Importance was placed on the after-potential, i.e. the variation reported by Gasser et al as following the spike and only recordable with very high amplification. potential is a questionable phenomenon. This after- It is true that the spike whether monophasic or diphasic approaches its baseline only assymptotically since the membrane possesses a definite capacitance and reactance (Cole 1928), but for this trailing part of the deflection to show properties unique to itself seems doubtful. These findings of Lehman have been submitted as a possible explanation of the two muscle tones appearing in ani- 15. mals after spinal cord asphyxiation (van Harreveld 1943). The early threshold drop and then rise could cause the first tone which is very short-lived and then the second threshold drop brings on the second tone. However, since all other threshold recovery curves to be found in the literature are smooth, uniform expotential curves approaching a limit, it was decided that a comparative study of the subject of nerve asphyxiation should be undertaken to confirm or refute the claims of Lehman. The results of this study are described and discussed in the present thesis. 16. METHODS Nerve Asphyxiation Chambers The .technique used was similar in many respects to that of previous investigators in the field of nerve asphyxiation. It was found possible to improve upon the type of chamber used for asphyxiation and thus to incraase the durability of the nerve. This was done by constructing chambers of lucite a substance reported to have no noticeable toxic effects on nerves no matter how long the nerves are in contact with it (Marmont 1940). Also it is a very workable transparent ma- terial and an excellent insulator. It is non-soluble in water, gas tight, and is very rigid at low temperatures. It becomes pliable at about 100°c, which is well above any temperature to be used with living nerve tissue. Three chambers were made, one for vertebrate nerves, the other two for crustacean nerve preparations. The vertebrate chamber is diagramed in Figure 1, upper part. ..l_ It was made of 1/8 inch lucite, 2 1/.8 inches long, ' 1 1~ inches wide, and 2 inch deep. It was divided into 2 sec- tions by a I/8th inch partition (F). A narrow slot (J) was cut in the partition to allow the passage of the nerve (I). This slot could be sealed with vaseline when it was desired to use the chamber in two sections. Small glass pipettes served as gas inlets (B) through the side wall of the chamber and as gas outlets (C) through the lid (A) of the chamber. A ledge (G) ¼ inch high and½ inch wide was placed in the chamber to support the nerve. Six silver-silver-chloride A-----------lid B----------- gas inlet 0----------- gas outlet D u I E C WP E cB B He E D-----------electrodes E-----------lid screw F-----------partition A G-----------ledge 'H-----------fluid SIDE I-----------nerve J----------- partition slot I END A-----------electrode block B----------- plug C-----------electrodes D-----------holes for plugs (B) E-----------wire rod F----------- pre paration holder G-----------indent ure to fit limb A H--~--------propodite I-----------nerve -J-----------ishiop odite K-----------elastic band Fig.l. Diagrams of vertebrate nerve asphyxi a tion chamber (Upper), and crustacean nerve holder (lower). detailed descri ption. See text for electrodes (D) were set in grooves in this ledge so tha t they were flush with the surf a ce of the ledge. The nerve was thus supported along its whole length, not just by the electrodes. This means of holding the nerve was devised in order to preserve the nerve in its normal state the longest p ossible time. Two crust a cean nerve chambers were made. First, a smaller replica of the vertebrate chamber was constructed. It wa s made of 1/Bth inch lucite, 2½ inches long, 1 inch wide, and 3/4ths of an inch dee p . It was divided into 2 sections by a 1/Bth inch partition. 10 platinwn wire elec- trodes connecting with saline saturated filter paper strips (Furasawa 1929 ) and spaced about 1-2 mms apart were sealed into the side of the box, 4 on one side of the partition and 6 on the other. This chamber was made primarily for the re- cording of an action potential fro m a single fibre at various places on that fibre during anoxia. The only differ- ences from the vertebrate chamber were the platinwn electrodes and t he absence of the ledge. The platinum electrodes were used because silver is toxic to crustacean nerves (Marmont 1940). The ledg e wa s omitted because the filter paper stri ps provided ample support for the nerve. For really long anoxia periods it was necessary to leave the nerve attaarbed so that a third chamber was constructed l ar ge enough to include the whole limb of the animal if neeessary. 1 This chamber consisted of a large lucite box 52 in- ches long , 2½ inches wide, and 1-l inches high. A luci te holder adaptible to either the leg of a lobster or the claw of the crayfish was constructed as shown in the lower diagrams of figure 1. The wire connections (L) ran through small holes sealed with vaseline in the side of the large box. The gas inlets and outlets were the sarne as in the vertebrate chamber. parts. The claw or leg holder consisted of two The base section (F) was shaped (G) to fit the pro- podite (H) at least approximately. The limb was held firmly in place by elastics (K) looped over the propodite and the ishiopodite (J) to heavy wire rods (E) fixed in this section. The second part of the holder was made of a small lucite block (A) into which were sealed with parafin 4 platinum fine wire electrodes (C) connected with fine copper wire leads (L). The parafin served to cover the solder joints and the unlaquered copper so that no foreign ions came in contact with the solution. 2 prongs (B) fastened at the bottom of the "electrode holder" fit t ed into 2 holes (D) bored in the base section. It was found, as others had ob- served, (Furasawa 1929) that if small strips of filter paper (C) soaked in saline were placed on the platinum wire the nerves lasted longer than if in direct contact with platinum wire. Also in the case of small nerve bundles there was less leading off artifact than with direct wire contact. The mounting of the prep aration on this combined holdercont a iner is shortly to be described. A leg, or claw, was cut from the animal at the ischiopodite. The nerve was ex- posed (Wiersma and van Harreveld 1936) in the ischiop odite 19. and the limb fastened to the base of the holder with the elastics. The whole base with limb was held by a Jaw clamp to a ring stand and submerged in a dish of saline (sea water in the case of the lobster). The nerve fibre or bundle was then prepared {Wiersma and van Harreveld 1936), and when ready for placing on the electrodes the holder was rotated 90° clockwise so that the limb was on its side and the holes for the prongs of the electrode holder were facing up. With a needle the nerve was held away from the shell while the prongs of the electrode plug were inserted into the holes from above, the filter paper strips being slid in between • the nerve and the shell. The needle was then withdrawn and the holder rotated back through 90°. The weight of the nerve brought it slowly down onto the filter paper. The holder was then carefully raised until the preparation and holder were above the solution, whence the holder was removed from the clamp and placed quickly in the lucite box. The four leads were pushed through the vaseline in the holes in the side of the box and the lid put on so that the nerve would not dry out. Saline, about l/8th inch deep, covered the bottom of the box keeping the enclosed air saturated. With the connecting of the gas tubes to the chamber the experimental set up was complete. Electrical Equipment A square wave generator was used for stimulating the nerves. It consisted of a relaxation oscillator feeding into a peaking amplifier. The pulses from the latter tripped a self-blocking multivibrator circuit which swung one way and remained there for a length of time determined by the time constant of its circuit. The frequency of stimulation was governed by the time constant of the relaxation oscillator. The duration of the square wave was controlled by the time constant of the multivibrator. A potentiometer in the cathode circuit of an output triode determined the amplitude of the output linearly up to 22 volts. The wave shape from this instrwnent was extremely rectangular. The action potentials were recorded by a cathode ray osscillograph via a five stage amplifier. The potentials were led to the grids of a push-pull differential amplifier which to a large extent eliminated pick up of stray fields and reduced the stimulus artefact. Electric fields and stimulus artefact affect the input grids at the same time so that the output of the tubes would swing simultaneously in opposition to one another canceling any deflection, while the nerve potential causes the input grids to vary one at a time resulting in a faithful variation. The sweep of the osscillograph was linear when repetitive, but became partly logarithmic when single. Single sweeps were obtained by overbiasing the thyratron tube so that it could not fire unless a pulse was sent in to its grid from the outside. When repeatedly firing the condenser in the circuit discharges quickly to a certain voltage whence the tube cuts .off and the condenser recharges again. A very linear part of the discharge curve can thus be obtained because the condenser can only chaxge up to a value at which the tube will fire and this may be adjusted to any desirable amount. With the single sweep after the tube cuts off and the condensor starts to charge pulling the electron beam spot across the screen the condensor completes its charge bringing in the logarithmic part at the end of the sweep again because of the overbias. since the tube cannot fire Although the lack of linear- ilty might cause distortion of the potential wave shape the single sweep was advantageous in that it could be synchronized with the stimulator. This made it possible to record action potentials with only a few stimuli thus eliminating overstimulation of the preparations. A more detailed de- scription of this equipment is included in the thesis by G. Marmont (1940). For measuring nerve depolarization a slow moving galvanometer of 15000 ohms resistance with relatively low sensitivity was connected through a low-noise push pull amplifier to the preparation. This amplifier was designed by Mr. J.B. Hawes of the National Technical Labs and is diagrammed in Fig. 3. It consisted of 2 input tubes (932 1 s) operating at 21 volt plate voltage with a 15 volt balancing bias. Under these conditions the grid current was practically nil so the apparatus was essentially a true micro-voltmeter. The first stage fed into a 6C8G double triode output tube. The latter was operating at 66 plate volts and 1 volt cathode bias. The apparatus became quite stable after several hours of operation, this stability being insured at sensitivities , I 6C8G DOUBLE TRIOOE '\ 5000 I / OL/TPUT I 5000 I \ I / 66V _I__ -=i:..,_ _I__ 2 l/2V .....::r:.... _..I,_ __..:i:...__ 100 MEG ....:::c,_ 43V _.I,_ _.I,_ IOOMEG .....I... .....::r:.... .....::r:.... ....:t._ _I__ _..I,_ ....I.. .....:r_ ....I_ 14-IGV :_l_ \ I .,,, / / .....::I:,_ 2V ~ 14 INPUT Fig. 3. Diagram of D. c. Amplifer designed by J.B. Hawes of The National Technical Laboratories for depolarization of nerve experiments. GANO I I ' down to 1 millivolt per cm deflection. (Maximum sensitivity with the galvanometer used was 5 mms deflection per microvolt approximately.) Temperature Control The nerve chambers were placed in a large metal container which was submerged to its brim in a constant temperature water bath. A thick asbestos cover fitted on the container above the water surface keeping the heat from escaping into the room air. A hole was drilled in the cover to hold a thermometer. The bath was regulated by a thermostat control to½ degree Centigrade. Only for temperatures higher than room temperature was this control used. ice was melted slowly in the bath. At low temperatures The resulting cooling lasted sufficiently due to the large amount of water, so that experiments could be begun at 18 degrees, for example, and during a two hour period vary a degree to a degree and a half one way or the other. This provided sufficient ac- curacy, since most experiments with mammalian nerves required shorter times. Gases In all experiments nitrogen was used for asphyxiation. It was obtained from the Linde Air Products Co., of Los Angeles, and was rated by the company as containing slightly less than 0.3% oxygen. The nitrogen was led through a two foot metal tube containing heated copper gauze . to eliminate the small amount of oxygen. The oxygen free gas then passed through two wash bottles filled with water to saturate the gas before entering the chamber. At first mixtures were made using 5% CO2 to keep the pH in the chamber constant, (Lehman , loc cit) but it was soon found that this added detail made no difference on the nerve behavior during asphyxia or recovery therefore it was discontinued. It was also found that passing nitrogen through heated gauze had no noticeable effect on, the survival time, thus this procedure was discontinued. Nerve Preparations Mammalian Nerves Nerves from rabbits, cats, and a few from dogs were used, the procedure for their removal from the animal being the same in all cases. The animals were well narcotized with nembutal and the nerves carefully removed and placed in a beaker of physiological solution at room temperature. The tibial and peroneal of the sciatic nerve were separated and removed from the leg in lengths greater than 7 ems. The phrenic nerve also could be excised in great lengths, but this operation took longer and involved killing the animal by pneumothorax before the nerve was accessible. It was found that nerves removed even shortly after death showed a pronounced shortening of the survival time and were generally in worse condition than those taken from a live animal. Use of the phrenic nerve in most instances was avoided for this reason. The saphenous sensory nerve and the radial and ulnar nerves of the foreleg though shorter made excellent preparations. 24. To remove the tibial and peroneal nerves, the skin was cut and pulled back from the thigh and an incision made separating longitudinally the fibres of the biceps femoris. When this incision was spread apart the sciatic nerve was exposed resting on the adductor femoris. The nerve runs be- tween these two muscle layers mentioned. By moving the prong of a forceps gently along the nerve the sheath was strip_ped from it and the peroneal and tibial bundles fell away from each other. These were readily tied off and removed. The saphenous was taken from the medial thigh be removing the skin and very carefully separating the connective tissue underneath with a forceps. The nerve is attached a- long its length to the saphenous artery and the saphenous vein by connective tissue, and it must be teased away from both. The radial, medial and ulnar nerves are located in the upper foreleg lying between the biceps and epitrochlearis muscles. Pulling back the epitrochlearis muscle exposed these nerves which could be then teased away from connective tissue and blood vessels. The ulnar nerve lies adjacent to the axillaris artery, so again care must be taken in its removal to avoid rupture of this vessel. All nerves were placed in Tyrode 1 s solution at room temperature except in those cases described in the Results Section. This same solution was used for washing the inci- sion in the animal during the excision of the nerve whenever necessary. 25. To asphyxiate the nerve, it was taken from the beaker and placed gently in the trough on the ledge in the chamber. A few drops of Tyrode 1 s solution warmed to 38 degrees were placed in the bottom of the chamber and over the nerve. If it was desired to divide the chamber in two parts the vertical slit in the partition was sealed with vaseline before the cover was screwed down on the chamber. Glass pipettes connecting with the gas hoses were inserted in the holes in the wall of the chamber, whereupon the whole chamber was placed in a metal box- to which were fastened shielded cables from the stimulator and amplifier. The ends of the cables were soldered to jacks which could be inserted into sockets to which the Ag-AgCl electrodes were soldered. This way of electrical connection eliminated the moving around of cumbersome long heavy wires. The metal box was then lowered into the larger metal container submerged to its brim in the water bath and the asbestos cover placed on top. A thermo- meter was inserted through the hole in the top for temperature readings. A test was made by stimulating the nerve briefly and noting whether or not it responded with action potentials of the proper size and shape. The size of the potential disclosed whether the right amount of fluid was around the nerve and whether good ... electrical contact was being made. If the shape were peculiar, i.e. , monophasic in- stead of diphasic, something more serious was wrong and adjustments made. Usually oxygen was allowed to bubble into the chamber first, passing through the wash bottles i mmersed 26. in the bath. This helped speed up the warming of the in- terior of the nerve chamber to that of the bath. When the appropriate temperature was attained, the oxygen was replaced by nitrogen and asphyxiation begun. The normal nerve recordings, threshold, amplitude of potential and so forth were made during the first oxygen period when the right temperature was reached and just before the nitrogen was turned on. During a sphyxia recordings were repeated every 2, 3, or 5 minutes until the nerve no longer responded. The technique was the same during recovery except in many instances the records were noted every 30 seconds. Variations of these methods for special purposes are explained in the Results. Amphibian Nerves The sciatic nerves of Rana pipiens and Rana catesbiana were used. Both young and old frogs were used, 10 ems of nerve being easily removable from old and large bullfrogs. When removed from the animal these nerves were placed in Frog Ringers at room temperature. To be asphyxiated they were .placed in the chamber used for mammalian nerves and the procedure described above employed. It was not necessary except in special cases to use the temperature control bath. Crustacean Nerves Nerves taken from the rock lobster, Panulirus interruptus, and from the crayfish, 0runbarus clarkii, were used, the same technique applying to both. The leg (Panulirus) or claw (0ambarus) was removed from the animal by pinching the ischiopodite. The medial half of the chitin of the meropo- dite was chipped off with a pair of snippers and the tendon 27. of the flexor muscle was removed with forceps exposing the nerve lying along the extensor muscle. The extensor mus- cle was cut and pulled gently away leaving the nerve isolated. The limb was then fastened to the holder and sub- merged in sea water (Panulirus) or van Harreveld solution (Cambarus). Under a binocular dissecting microscrope the small nerve bundles were teased apart using fine needles (Wiersma and van Hareveld 1936). Although it has been re- ported that single fibres left attached to the muscle could last 6-8 hours (Marmont 1940) it was not possible to keep them that long in saturated gas surroundings on filter:i;:a.per supports. Long durability of single fibres depends on their being supplied with a fluid medium of the proper ionic concentration. On account of the large stimulation current escape especially with high thresholds during anoxia such a medium was impractical. However, much longer survival times could be obtained from small bundles of only a few fibres of which 4 or 5 functioned. Presumably the fluid between the fibres of such a preparation acts as a proper medium for the few active fibres. Although somewhat shortcircuited, the individual fibre potentials were s,till distinguishable and the stimulus artefact was not too great. tions lasted 5 hours or more. These prepara- When the nerve was prepared the electrode contacts were made as already described elsewhere. It was found that the lobster nerves remained in much better condition for greater periods if held at about 19 degrees. 28. RESULTS A Asphyxiation Survival Time The survival times of 28 rabbit nerves, 10 cat nerves, and 3 dog nerves are presented in Table I. Each of these nerves was placed in nitrogen immediately upon removal from the animal. Rabbit The average survival time of the rabbit nerves was 23.4 minutes. 3 rab bit nerves required a period of 35 minutes without oxygen before becoming completely inactive. One rab- bit nerve lasted only 15 minutes without oxygen before losing all signs of response to a stimulus. The survival times of the remaining 24 rabbit nerves are between these two extremes, 15 and 35 minutes. 50% of them are 20 minutes or less, 25% of them are 30 minutes or more. The specific nerve used is also listed in Tabler. It is clear that no specific survival time can be correlated with a specific nerve type. The saphenous nerve might be an exception (as reported for cat saphenous nerves by Lehman, see below) but since only 2 were used no conclusion is possible at present. All the nerves removed from one animal for experimentation were removed from that animal in one large operation and were placed in physiological solution all at about the same time. Thus only the first nerve to be asphyxiated of such a group could be considered fresh since the others remained in the Tyrode solution for 45 minutes or more while TABLE I The survival time of mammalian nerves asphyxiated i mmediately after removal from t he ani mal. Nerve Type Rabbit Nerve Survival Time Cat Nerve Survival Ti me 18 mins 16 30 20 28 20 16 45 mins 30 28 15 35 28 35 23 22 22 17 20 30 30 16 16 25 45 33 43 26 mins 35 18 16 35 35 23 15 20 Tibial fl II II II II II Peroneal II II II II II II II II II fl II II II Saphenous II Medial II Radial 25 Name not recorded II 32 35 ave 24 phrenic Dog Nerve Survival Time 18 35 ave 33 33 this asphyxiation was going on. Therefore only 1 fresh nerve per animal was asphyxiated so that specific survival times for different nerves from one animal cannot be correlated. Cat The survival times of 10 fresh cat nerves ranged between 18 and 45 minutes and averaged 32.9 minutes. ed an average of:-30 :ruinutes _for ca~ nerve. Lehmann reportHe also pointed out that the saphenous nerve survival time was always shorter than the survival times of the other nerves of the cat (16-18 min.). The 1 saphenous nerve used in the present work had a survival time of 35 minutes. Only 2 nerves lost their activ- ity in less than 20 minutes in nitrogen, while 7 of the 10 used were excitable for 30 minutes or over. Dog The survival times of the 3 fresh dog nerves listed averaged 33 minutes. Gerard gives 30 minutes as an average survival time for dog nerve. Lobster The survival times of 20 lobster nerves are illustrated graphically in Fig. 4. They ranged from 30 to 120 minutes with more than 50% having values between 60 and 90 minutes. 3 of these 20 nerves were not prepared beyond being exposed by removal of flexor and extensor . muscles in the meropodite. Each of these 3 nerves becrune completely inactive after 75 minutes in nitrogen. The average survival time of the other 17 nerves,which were prepared to fine bundles, was 77 minutes. LOBS'l'BR 6 AVBRAOI: OJ' ?.O DRVJ:S .. 77 Mins. 0 30 60 90 . 120 150 ,UPHYXU!IOll TIKI II IIIIIJ'l'IS Fig. 4. The survival time of Lobster nerves. Ordinate is number of nerves, abscissa is time (minutes) for complete extinction of the spike action potential of a single fibre. Frog Frog nerves could endure anoxia much longer. The nerves of Rana pipiens of which 3 fresh ones were used had survival times of about 2 hours. Table II. They are tabulated in Nerve 118 was bathed in ringers for 2 hours be- fore being asphyxiated and the resulting effect is described elsewhere. Only 2 fresh nerves of Rana catesbiana were asphyxiate~ numbers 155 and 186 in Table III. These maintained their activity in nitrogen for 2½ and 3½ hours respectively. The other nerves listed in this table were aged in physiological solution and are included in another discussion. Since al- most all the previous anoxia work on peripheral nerve has been done using frog nerves and since the results repo~ted are identical to those above it was deemed sufficient to limit the present work to the 5 experiments described. Crayfish Crayfish nerves subjected to a nitrogen atmosphere remained excitable to the end of the experiment in the majority of cases. The irritability of a prepared opener bundle was found to be as great after 5 hours in nitrogen as at the start. The activity of 3 closer bundles did not show the least sign of change in 100, 120, and 150 minutes anoxia. In control experiments both claws were taken from the same animal and placed one in nitrogen and the other in oxygen. After 5 hours the activity of both was identical. This was repeated with 2 more claws from another animal for 9 hours TABLE II Sciatic nerves of Rana pipiens (Leopard Frog) asphyxiated in nitrog en. Stimulation by direct current of 0.1 millisecond dur a tion. Nerve Time Temp. 63 0 2.3 Minutes in N2 Threshold Remarks 0 11 13 15 15 20 large, diphasic 15 37 65 95 65 0 23 0 5 10 15 20 30 50 60 119 0 24 2 hrs. 26 larg e, diphasic 18 22 22 monophasic 75 22 90 105 120 22 22 25 30 0 5 large, diphasic 45 14 14 14 18 85 100 25 monophasic, small X 10 118 11 13 14 14 becoming monophasic 1/3 orig inal size monophasic 0 30 60 75 90 105 120 135 150 160 180 190 200 210 5 5 5 5 5 6 8 10 12 15 18 at 2-5 shocks/sec. at 20-30 shocks/sec. not responding to every shock. large, diphasic less diphasic Wedensky block at 60 cycles 20 23 monophasic, very small, rapid fatigue TABLE III Sciatic nerves of Rana c a tesbiana (Bullfrog ) asphyxiated in nitroge n. Sti mul a tion with direct current of 0.1 millisecond duration . Nerve Time Temp . Einutes in •l'~T2 Threshold Re uarks 121 2 hrs. 25 0 25 75 100 125 155 185 213 245 275 6 6 large, diphasic 154 155 3 hrs. 1 ;:( 186 187 188 189 hr. 0 4 hrs. .~" hrs . W'± 48 hrs. 26 26 25 25 25 25 0 15 45 120 180 210 240 0 15 45 90 130 150 0 60 120 160 200 0 180 300 8 10 10 10 10 11 11 12 5 15 10 10 11 becoming monophasic 1/4 original size monophasic see fig. 2 1.2 normal size II tt 1.3 II II 1.0 tt tt 0.7 measurement no 0. 1 normal size large, diphasic 16 24 0.9 no1~mal size II monoph2,sic 0.8 II II 0.2 II tt 0.01 6 large, di:phasic 11 13 16 20 6 25 monopha,s ic almost gone large, diphasic 1/4 normal size small potential left, I mmersed throughout 0 60 180 240 330 11 large, diphasic 23 X 1/2 nonnal size gone 0 120 240 360 390 29 large, diphasic 54 monophasic almost g one gone with the same result. At the end of the 5 hour period the nerve action potentials were somewhat larger and both closer and bender muscles contracted, whereas only the bender contracted after 9 hourl:ffi. To check the apparatus for possible leaks etc., 6 legs from the same lobster were tested, 3 in oxygen and 3 in nitrogen for 3½ hours. All those in oxygen retained their ex- citability, but those in nitrogen were completely inactive after the 32·1 hour period. Thus the chamber is gas tight. The conclusion is therefore that it is not possible to sustain Cambarus nerves long enough to show any effect of anoxia. All mammalian nerves were asphyxiated at 38 degrees centigrade. Frog and crayfish nerves were held at room tem- perature which varied from 23-26 degrees. Lobster nerves were asphyxiated at 19 degrees. Effect of Time in Solution Before Asphyxiation on Survival Time It has been emphasized in the preceding paragraphs that the survival time determinations were made with fresh nerves, i.e. nerves just removed from the animal. This was done be- cause it has been reported that the metabolic rate of excised nerve in physiological solution changes ra~idly. The oxygen uptake rate ofmammalian nerve is quite reduced after bathing the nerve in saline for an hour or so. change occurs more slowly in frog nerve. The same Such a variation might affect the survival time of a nerve deprived of oxygen. A study of such a survival time effect is not to be found in the literature therefore it was undertaken in the present work. The results from mammalian nerves are collected in Table r:v, from frog nerves in Tables II and III. The sur- vival time of frog nerve, both species, increased with 2 to 8 hours washing in physiological solution. Since it takes approximately 2-4 hours to asphyxiate a frog nerve it is not surprising that frog nerves asphyxiated immersed in solution endure longer than those asphyxiated in gaseous media (Fillie et al) for the washing effect would become significant in that lenth of time. Washing before asphyx- iation ca.uses the survival time to rise toward a limit of about double the value for fresh nerve. In direct contra.st to this, the survival time of th9 mammalian nerve decreased with washing in physiological solution. This decrease is not continuous, for it approaches a limiting value approximately 70% of the value for fresh nerve. Effect of time in serum on survival time Gerard reported in 1932 that the metabolic rates of excised peripheral nerve remained more constant if the nerve was bathed in serum from the blood of the animal than when bathed in physiological solution. A brief study was there- fore made to compare the survival times of asphyxiated nerves in serum to those bathed in physiological solution. is recorded in Table V. The data TABLE IV Effect of a ging in physiological solution. The survival times are listed under the numoer of hours the nerve was left in saline before being asphyxiated. hours in s aline Rabbit nerves Cat nerves 1 2 3 30 mins 12 21 18 17 17 17 16 25 15 20 ave 19 4 15 mins 25 25 22 15 12 12 19 17 20 17 26 12 18 21 17 20 22 27 33 10 27 25 10 18 20 30 20 23 ave 23 Dog nerves 5 6 7 8 17 mins. 1811:i.n.13min s 18 17 14 25 Ts 10 TABLE V Asphyxiation times of nerves placed in saline compared with asphyxiation times of nerves placed in blood serum. Nerve Solution Asphyxiation Time 158 Rabbit serum radial 159 II saline 35 minutes 24 II 160 serum 32 II saline 20 II II 160 Remarks 158 and 159 from same animal. 30 minutes recovery in saline. 161 11 saline 22 II 30 min. recovery in serum. 161 II serum 20 II 160 and 161 from same animal. 166 Rabbit saline Medial 166 II II 26 II 1 hr. recovery in saline 16 II 2 hrs. recovery in saline. 166 II II 16 II 167 II serum 27 II 1 hr. recovery in serum. 167 II II 21 II 2 hrs. recovery in serum. 167 II II 22 II 166 and 167 from same animal. 168 II II 26 II 169 II saline 17 II washed in serum 3½ hrs. after removal from animal before experiment. washed in saline 3½ hrs. after removal from animal before experiment. 168 and 169 from same animal. 178 Cat serum Radial 179 II saline 36 II 180 30 II serum 21 II washed in serum ·l hr. after removal from a~imal. washed in saline 2 hr. after removal from animal. washed in serum 1 hr. after removal from animal. 178, 179, 180 from same cat. Nerve 158 was bathed in aerated serum 30 minutes before asphyxiation was begun, 159 was in aerated saline for about 1 hour and then asphyxiated. Nerve 160 was treated the same as 158 except that after being asphyxiated it was placed in aerated saline for 30 minutes and then asphyxiated a second time. This procedure was r .e versed with nerve 161. Nerves 166 and 167 were bathed only 5 minutes in serum and saline respectively and then asphyxiated. They were then placed in aerated serum and saline respectively for 1 hour and asphyxiated again. They were then given 2 hours washing in their solutions and asphyxiated a third time. The pro- cedures for nerves 168, 169, 178, 179, 180 are clearly outlined in the remarks column of the table. It is evident that the survival time suffers a lesser change with washing of the nerve in serum compared to washing of the nerve in saline. For example, the survival time of nerve 166 ,is re- duced 39% in saline whereas the survival time of nerve 167 is reduced 2.3% in serum in the same period of, time. The difference is demonstrated clearly by the survival times of nerves 158, 159 and 168, 169 and 178, 179. Effect of ionic concentration (Calcium) on survival time Chang and Gerar~ (1936) describe a depression of the oxygen uptake rate in peripheral nerves bathed in physiological solution containing excess calcium ion concentration. is reversed with calcium lack. This Therefore a study was made of the calcium (and potassium) ion concentration effect on the survival time of peripheral nerves in nitrogen. The re- sults are seen in Table VI. the asphyxiation number. The bracketed number indicate Thus nerve 35 was asphyxiated first in normal ringers requiring 33 minutes in nitrogen. It was then allowed to 11 rest 11 30 minutes in aerated normal ringers, placed in aerated calcium free ringers for a few minutes and then asphyxiated a second time requiring only 24 minutes. This procedure was reversed with nerve 36. The survival times were observed to be consistently longer than normal when the nerve was bathed in solution with excess calcium ions and shorter than normal when this ion was lacking from the solution. Temperature Effect on Survival Time 14 frog nerves were asphyxiated, 5 at room temperature (25 degrees) and 9 at temperatures of about 30 degrees. All 5 nerves asphyxiated at 25 degrees withstood 150 minutes in nitrogen before losing their activity completely, whereas those asphyxiated at 30 degrees or more survived only 70 to 135 minutes in nitrogen. The results are illustrated graph- ically in figure 5, upper half. The lower half of the figure shows the survival time data of 9 rabbit nerves asphyxiated while at various temperatures. 3 nerves were asphyxiated at 38 degrees, controls, the others at lower temperatures. The survival times at lower temperatures were longer than those at higher temperatures. It is known that the metabolic rate of peripheral nerve varies directly with temperature. TABLE VI The effects of variation of Calcium and Potassium ion concentrations upon asphyxiation times of nerves. Nerves Time in Normal Ringer Time in CA Free Ringer 35 Cat Saphenous 11 36 33 (1) 24 (2) 31 (2) 21 (1) 39 Cat Peroneal 24 1) 26 ~3) 18 (2) II 40 19 ~~~ 17 41 Rabbit Saphenous 25 (1) 43 Rabbit Medial 15 (1) 44 14 (1) II II 10 (1) 46 Rabbit Peroneal 36 35 ~~~ 45 Time in CA Excess Ringer Time in:K Excess Ringer 30 (2) 16 (2) 15 (2) 15 14 12 2) 3) 4) 1 13 (2) 20 (2) 47 II 20 (1) 48 II 28 (1) 23 (2) 49 II 17 (1) 8 50 Dog Saphenous 35 (1) 27 (2) 30 2) 35 ~3) (2) FROG 40 30 0€bde) 00 CID w C <C 0:: ~ I- (c O · 20 10 z: 1.1.J u ASPHYXIATION TIME 3 2 4 - e 1.1.J 0:: ::::, I- ~-40 w HOURS 5 6 RABBIT - 0 00 a.. 0 ~ 1.1.J 0 -~ 3-0 - - - ---· 0 0 - 2-0 0 - 10 0 Fig. 5. 0 10 20 30 -I 40 MINUTES 60 50 Graph showing change of surv:ival time' in nitrogen with change in temperature. Frog nerve (upper), rabbit nerve (lower). Stimulation and Fatigue Effect on Survival Time It has been shown that the metabolic rate of frog nerve increases with stimulation (Fenn 1928). It was expected that the survival time would be reduced by stimulation during anoxia compared to a similar period sans stimulus. This was investigated by using continuous faradic stimulation throughout the asphyxiation and by waiting until the nerve showed signs of losing its excitability and then beginning the faradic stimulation. Th~ survival 'times of nerves stim- ulated with frequencies of 0, 60, 180, and 240 shocks per second were compared. Rabbit nerve was used. The results showed no significant difference in the survival times. At first glance this was puzzling, but the met- abolic rate increase from continuous activity as measured by oxygen uptqke was less than 15% in frog nerve, any change in time the survival/of a rabbit nerve on that basis would be of the order of a minute which would not be significant. Fig. 18 shows a plot of two experiments on nerve 76, a rabbit peroneal. The decrease in amplitude of the action potential was actually more rapid with 15 shocks per second continuous stimuli than ,wi.th 60 shocks per second and since the 15 shocks per second test was the second asphyxiation of the nerve it is probable that the effect of aging smothered any small change due to the stimulation. Excitability Threshold and Action Potential Amplitude Changes During Anoxia When a peripheral nerve is subjected to an oxygen free atmosphere, two changes occur in the electrical activity of 36. the nerve. The threshold for electrical excitation rises and at the san1e time the amplitude of the action potential diminishes. Neither of these two changes, however, begins immediately upon commencing asphyxiation. There is always an initial period the length of which varies among the different nerves, during which normal excitability and electrical responses are maintained. Indeed, in some instances the threshold has been observed to decrease slightly during the very early stages of anoxia. Lehmann, Gerard~ Heinr• !~·. beaker have reported this super-normal stage, and i .. was observed in the course of the present work. In many experiments undertaken to investigate .J he thresh~:;;,- old and response changes in asphyxiated nerve, it w~s noted that a diphasic action potential became monophasic quite some time before disappearing altogether. This fipding was ~ .,,. .... . .. "' ,.: ....... important because it meant that the change in activity was ~' . '\ . :• - _ , not uniform along the nerve. is illustrated in Fig. 6. An example of the phenomenon The large negative spike was mostly off the cathode ray screen for the first 14 minutes of asphyxiation. In the same interval the positive phase disappeared altogether. Thus the region over the distal lead near the end of the nerve became inactive long before areas further in from the end lost their activity. This ef- fect was called, for convenience, the end-in effect. It was important to know whether this was merely a local occurrence, restricted to a small area around the cut end, or whether it progressed along the nerve. A frog nerve was asphyxiated Fig. 6. Positive phase of diphasic action p otential of cat phrenic nerve at various st ages of asphyxia . The p ositive pha se ha s d isapp ea red by 16 minutes, t he ne gative p hase l a sted 23 minutes . with 3 lead off electrodes, no.l contacting the middle of the nerve, no.3 contacting the end, and no.2 contacting the area betiveen the other tno. The frog nerve was used because the changes were slower and better records were obtainable. The results of this experiment are illustr2.ted in Fig. 7. The upper row of pictures, reading left to right, were taken 1.:csing leads 1 and 3 after 30, 60, and 90 minutes in nitrogen. The lower row were taken at the same times using leads 1 and 2. The ratio of the positive (down) phases at these times, leads 1 and 3/leads 1, 2, were 1.0, 0.6, 0.43, and 0.4. Thus lead 3 was becoming inactive more rapidly than lead 2. The rc\tio of positive to negative phase of the 1, 2 lead potential showed no significant change during the 90 minute asphyxiation. Thus the action potential in that area of the nerve was as diphasic at the end of the experiment as at the beginning whereas the action potential from leads 1, 3 became almost completely monophasic. It was concluded that the effect was limited to the region immediately surrounding the cut end of the nerve. Ex- periments by F.O. Schmidt (1939) have indicated a higher oxygen uptake rate at the end of the nerve. This conclusion sug 5 ested that the threshold rise too might be influenced by the end-in effect unless the stimulating electrodes were placed well in from the cut end of the nerve. The resultE of several experiments undertaken to test for this influence are collected in Table VII. The left hand column shows the change in threshold during anoxia with stimulating electrodes near the cut end of the nerve. The right .Fig . 7 . , Action potenti al of frog nerve c ombining common lead near middle of nerv e with one ne ar the end (uppe r), and one nearer t he middle (lowe r) . As asphyxia pro gresses (left to ri ght) note t hat the distal lead becomes monophasic, t he medial lead does not , showing the end-in effect fro m the cut end . TABLE VII Threshold chang e during anoxia with stimulating electrodes placed near the nerve end listed in columns on left; with stimulating electrodes placed well in, 1 to 2 ems, fro m nerve end listed in columns on ri ght. All nerves aged ½-1 hour in saline. Time in N2 Rabbit Saphenous 0 rnins 3 5 8 10 12 15 19 25 Dog Saphenous 0 mins 2 5 8 10 13 16 20 24 28 30 32 Threshold 6 6 8 9 Time in N2 Rabbit Perone al 10 10 11 23 infinite 2 2 2 2 3 3 3 5 7 9 10 infinite 0 mins 10 6P 10 lOP 10 15P 11 12 infinite 20P 25 Dog Tibial Threshold 0 mins 2 5P 3 10 3 17P 4 4 4 20 25P 30 P 4 35 infinite hand c ol umn shows t he t hreshold ch2.nge when the stimul a ting electrodes were pl a ced 1 to 2 ems away fro m the cut end. The threshold units are 1/200ths of a volt. It is clear that the t hreshold of excitability begins to change near the cut end of the nerve sooner than it does well away from this injured area. Photographs of action potentials of a rabb it nerve taken after 5, 10, 15, and 20 minutes in nitrog en are reproduced in Fig. Ba. The action potential of a dog nerve taken after 5, 17, 25 and 30 minutes in nitrogen are seen in Fig. Sb. In both cases the threshold remained very cons t ant (see table fi gures with suffix P for the thresold read ing when picture of action potential was t aken) while t he action potential decre a sed. The lead off electrodes were near the end of the nerve, the stimulating electrodes were well in from the end. Thus it can be concluded that nerve activity in general is more susceptible to asphyxiation near the cut end of the nerve than it is further in along the nerve on uninjured tissue. The end in effect was also ob- served with single nerve fibre preparations from lobster nerve. A p oor example of the effect is illustrated in Fig .9. The single action potential (middle photographs) is somewhat ., more diphasic after 20 minutes in nitrog en than after 115 minutes, but the distin9tion in this record is not a s clear as was observed with other preparations. It is not likely that t he effect would influence survival ti mes wery much, because surviv al ti me was determined uding strong enough stimuli to exc i te nerve tis s ue quite Fi g . Sa. Action potential of rabbit tibial nerve after 5, 10, 15 , and 20 minutes in nitrogen (left to ri ght). Note absence of diphasic ti p seen at 5 minutes in 10minute picture . See Table VII. , Fig . Sb . Action potenti al of dog tibial after 5, 17, 25, and 30 minutes in nitrogen. See Table VII. Fig. 9 . Potential of single fibre of lobster leg using four leads numbered 1 to 4 from middle toward cut end of nerve . Lead 4 was common and relatively inactive. After 115 mi nutes in nitrog en lead 1 is still full size, but 3 is definitely smaller showing end-in effect or decremental conduction(?). some dist a nce along t he ne rve fro m the stimulating electrodes. Howev er, it was clear t hat electrode position on the nerve was i mport ant for accurate t hreshold and potential studies on a s phyxiated nerve. Typical curves of thresh- old chang e during anoxia are presented in Fi g . lOA. Curves of a ction potential amplitude change during anoxia are in Fig . lOB. A curve for .the nerve of each kind of animal used is included. (See legend). These curves were made fro m data taken with the necessary precautions to eliminate the influence of the end-in effect. The action potential amplitude is measured in percent of initial value, the thresnold in 0 .02 volt units. The threshold of the rabb it nerve remained constant for about 10 minutes and then began to increase rather rapidly. The threshold of the ca t nerve remained constant for 13 minutes and then rose but more slowly at first than rabb it nerve t hreshold. The lobster nerve t hreshold remained constant for 50 minutes and then rose moderately fast, and fro g nerve threshold began a slow rise after 2 hours asphyxiation. The threshold of crayfish nerve did not change si gnificatly. It will be noted tha t the early drop in threshold Value is absent from these curves. This suggests that the drop was due to the adding of the demarcation potential from the end-in effect to the stimulus voltag e when the stimulating le ads were close to the nerve end. By placing t he electrodes well in fro m th e cut end, the demarcation p otential was ne gligible and the t hre shold did not decrease. 100 R C ,j L A C> _J 0 ::c: (/) Lu 0:: ::c: I- ~ HOURS F i,· F ". F ". Lu C> => I- :J Cl... ::::i: <( B F i g 10 . A. Curves showi ng t h reshold change in O. O2 vo l t · un i ts with t i me 1h ours , in n i t rog en f or r abb i t nerv e ( R), ca t nerve,c ), l ob st er sing l e f i ber ( L ), bullf rog ( F ), a nd crayfi sh( Cr ). B. Curv es showi ng runp lit ude chang e i n p erc ent maximum wit h t i me, h ours ,i n n it r og en . A dog ( D) nerve i s i ncluded , and three bullfro g n erv e s . The right ha nd curve repre s ents re s ul ts from a f r og nerve ag ed 3 h our s i n sa li ne before an ox i a . The other t wo c ur ves are fr om fresh fro s nerves one a t 35 deg r ee s , t h e ot h er a t 23 degr ees c en t ig r a de . The amplitude of t he nerve a ction potential reduced during anoxia. The curve of the reduction was similar to the thresh ol d rise curve for a nerve of the same ani mal. The action potentials of rabbit, cat, and dog nerves began to decrease in size after 5-10 minutes in nitrogen. The rabbit nerve p otenti al chang e beg ins s ooner than the potenti al chang es in cat and do g nerve. The potential change curves from 0 frog nerves are included in the fi gure B. These curves reading left to right, were made from data fro m a fresh nerve at 05 de gree s , a fresh nerve at 20 degre e s , and a nerve bathed 0 hours in saline at 20 degrees. The curve of the fresh nerve at 20 de grees corresponds closely to the t hreshold curve directly above it in the figure. The amplitude began to decrease after 1½ hours in ni- trogen. It was concluded fro m the threshold and amplitude data that rabbit nerve retained its normal activity for 5-10 minutes when asphyxiated; cat nerve and dog nerve sli ghtly longer, 10-15 minutes; lobster nerve for nearly an hour, and frog nerve more than an hour. The 2 other frog nerve curves are included in the figure to show how the effect of oxygen lack on excitability is influenced by te mperature and ag ing in s olution. It is seen that at hi gher temperatures the initial normal period is s hortened and the rate of change of a ctivity increased. The s olution effect causes a prolongation of the initial normal period and a reduction in the rate of change of excitability. Reduced All The decre a se in a ction p otential amplitude during anoxia could be caused in several ways. Ger ard (see introduc- tion) concluded t hat t he "all" res ponse of each fibre wa s re duced by anoxia. Two other possibilities are the block- ing of conduction in fibres a t various places along the nerve, and decrement a l conduction. In order to investigate Gerard's conclusion the following experiment was devised. A long frog sciatic was placed in the chamber a nd the partition sealed with vaseline. Nitrogen was made to enter one section, oxyg en the other. The part of the nerve in nitrogen contacted 2 pairs of electrodes. The distal pair wa s for stimulation, the proximal for leading off. There wa s contact by the part of the nerve in oxygen with another le ad-off pair of electrodes. Thus t he nerve was stimulated in nitrogen, an d t he a cti on p otential recorded first in nitrogen and then in oxygen. With all the fibres conducting past both lead off pairs of electrodes the r a tio of the 2 potential records would be 1. If the all of e ach fibre was reduced during ano x ia then the ratio of the potential size in nitrogen to the size in oxyg en would become less than 1. If the fi bres began to stop conducting in nitrogen then the ratio in nitrogen to oxygen would incre a se to greater than 1. Due to le ading off difficulties the initial ratio probably would v ary somewha t. The results from 2 frog nerves are given in Table VIII. In both c a ses the r a tio decreased for the first 2½ hours of TABLE VIII Nerve 2 29 Ra na c a tesbi ana 230 Rana ca te sbi a na 0 mi nutes s pi k e in N3 s p ike in a ir 1. 000 40 11 1.000 75 II .834 100 II . 6 80 130 II .640 150 II .550 165 II .562 1 80 11 .635 0 II .800 60 11 .7 83 90 II .70 5 120 11 .700 150 II .685 180 fl .70 5 210 II .750 Ti me in N3 Ra tio ( \ anoxia. Then a sli ght increase was recorded. Thus the re- s ponse in nitrog en in each fibre decre a ses until conduction blocking commences in the fibre. To observe the reduced all res ponse more clearly fine nerve bundles pre pared from lobster nerves were asphyxiated in a double chamber. Stimulation was applied to t he nerve in the aerated section. There were 4 lead off electrodes in the asphyxiation section which were numbered 1 to 4 progressing distally along the nerve. Lead 4 was com:-ion. Photo- graphs of the a ction potential of a single fibre recorded with lead combinations 1-4 then 3-4, after 20 minutes and 115 minutes of anoxia a re reproduced in Fig. 9. The 1-4 com- bination, top of fi gure, includes the artefact spikes from le ads 2 and 3 as would be expected. (Marmont 1940). Thus from this one picture it can be concluded tha t after 115 minutes of anoxia this particular fibre conducted a normal sized spike past electrodes 1 and 2, but a greatly reduced potential past electrode 3. The contrast in size of the s pike at lead 3 after 20 and 115 minutes of anoxia is clearly seen with leads 3-4, center p ictures. The lower pictures show si milar results from another fibre in a larger bundle. artefacts. The l a r ge bundle eliminated the The 1-4 le ad potential was much larger than the 3-4 s p ike after 70 minutes in nitrogen. After 90 mi nutes the 1-4 potential was the size of the 3-4 po tential in the fi gure. Thus the response of the single fibre definitely diminishes during anoxia but not uniformly along the fibre. Whether the conduction is considered decrement al, or whether it is considered irregularly reduced by the end-in effect becomes a question of definition. Referring back to t he fro g sciatic experiment, a careful me a surement of the 2 phases of the di pha sic s pike led off fro m the asphyxi a ted part of the nerve showed no change in the r atio of t hese potenti als during ano xi a . Thus the s pike size changed an equal amount at each electrode, wh ich suggests that decre mental conduction is not likely to occur in re g ions of the nerve far from the cut end. Refractory Period To measure the refractory period variations during anoxi a , 2 stimuli were necessary inste ad of 1. It wa s possible to cha nge the single rectangular stimulus to 2 shocks of short dura tion by connecting the output of the squc_re wave g enerator to a "separator" consisting of an audio transformer and 2 output diode tubes. The sudden potential ri s e at the st art of the s quare wave placed an induced voltage of short duration on the plate of one of the tubes so tha t it conducted momentarily. The sudden potential drop at the end of the square wave caused the other tube to conduct moment arily. These brief pas s ages of current were the 2 sti mulating i mpulses the separation of which was controled by the duration of the s quare wave. For a general study of the nerve refractory period during anoxia the second stimulus and the nerve impulse excited by it are made to occur in the rel ative refr a ctory st ate from the first nerve impulse. Thus t he sec ond nerve i mpulse could be reduced to only ½ or 2/3 the magn itude of the first. It was ass umed tha t the refr a ctory period would vary directly with nerve excitability and become shorter than normal if the nerve were in a _hyperexcit able state and longer if the nerve were hypoexcitable. This could be ea sily investigated by comparing the ratio between the amplitudes of the 2 nerve p otentials at v arious ti mes during anoxia. An increasing r a tio value would indicate a decre e.sing refr a ctory period and vice versa because an incre a sing ratio would mean the second nerve i mpulse was bec om ing larger which could only l'appen if the refr a ctory period shortened so tha t the impulse occured during a more excit able st a te. Table IX. Such r a tios have been gathered in Nerves 89 , 87, 85, and 83 show a steadily decreas- ing ratio so t here was no early shortening of the refr a ctory period in these c a ses, but in all the other nerves t here was a,t.increase of the r a tio value. Thi s increase . occured usual- ly during the first 6 minutes of anoxia. Let the difference between the first ratio at 0 minutes and t he other ratio v alues for ea ch nerve be defined as negative if t he other v alues a re smaller than t he first one and positive if they are large r than the first one. By dividing this posi tive or negative difference by the first ratio value it is converted to a percent age difference plus or minus. When these percentage differences a re plotted, a curve a s seen in Fi gure 11 is obtained. Thus for nerve 89 after 2 minutes in nitrogen the expression is 750-572 = 0.237, and '750 .800 .678 .305 .790 . 920 .70 0 .389 .740 .880 .700 .555 84 84 83 83 .906 .483 . 84 2 .830 .368 .545 .305 .825 .785 .305 .645 .718 .400 .770 .754 .727 85 86 . 294 .355 .625 .650 .400 .827 . 803 .7 8 7 .780 .780 .670 86 .2 20 , 256 .306 .46 6 87 . 286 ,317 .733 .7 33 88 .222 .704 .535 .625 .57 2 .750 89 .694 . .460 .375 ~ .250 .545 .334 .713 .148 .714 Time in Nitrogen in Minutes 8 10 12 6 ' 4 2 0 Nerve .214 .615 .48.3 ,167 .72 2 14 .136 .600 . 386 ,751 16 . .364 ,467 18 .310 20 Ra tios of t he sec ond i mpulse to t he fir s t when t he nerves were sti mul a ted with t wo s ti muli s o close together tha t t he second fell in the rel ative refr a ctory period of t he first i mpulse. TABLE IX 0 0 0 ' if) w ·U 2 ,z w i~ .2 4 Cl lo IP < a:: Cl 0 a: w 8 0 4 0.. • 89 >- a:: 0 I- u <: a::Q.2 LL. w ' a:: 0 • 0 2 a-- 0.4 0 7 i~ 11 . Gr~ph sh~vi ng t ~ e d iff erences i n s i ze r a tios of t wo nerve i mpulses at va ri ous t i mes du.ri n~ as.i)hyx i a t i on . The sec ond i mpulse is excited dur i ng the re l a t i ve refractory state caused by tlle first . Result s f rom r abb i t n e rve i n sol i d c i rcles , from ca t nerv e i n o)en circles . Bl a ck triang l es show refractory period a fter t i me ( rni nutes )i n ni tro gen i nd i cated by the nilll'lber ad j a cent to the tri ang l e . The pos it i v e slope represe nts the abso l ute per io d , the negat ive s l ope represents the relative ·::per i ad . since the 2 minute ratio value is smaller than the ratio value at 0 minutes the final figure is -0.237. The average of all t he rabbit nerve refractory periods has been plotted with solid dots, 1 cat nerve refractory period has been included, open circles. By reducing the time between maximal stimuli to the value at which the second shock can no longer excite an action potential, a measure of the change duri ng anoxia of the absolute refractory period for maximal stimulation was taken and by increasing the time between stimuli until the second nerve impulse was as big as the first, the total refractory period variation was measured. The results are pictured graphically as triangles in Fi gure 11. Each triangle was made from data taken at the time of asphyxiation indicated in minutes by the number adjacent to that triangle. .The average of results from 2 rabbit nerves was used, and it will be noted that both the absolute and the relative refractory periods first decreased and then incre~sed. Chronaxie The rheobases of 2 rabbit nerves were established with shocks of one second duration. The strength was then doubled and by varying the duration beginning at 0.01 sigma and increasing until the potential appeared, the chronaxies were determined. The nerves were then placed in nitrogen. The chronaxie of the first nerve was 0.12 sigma at the start of anoxia, dropping to 0.09 sigma after 9 minutes in nitrogen and rising abruptly to infinity aft er 22 minutes asphyxiation. 0.09.S- The second nerve had an initial chr onaxie o f ~ sigma, O•/ 0.13 .J...re" si gma after 5 minutes anoxia, ~ si gma after 10 min- utes, and 00 after 13 minutes. very short survival time. This nerve wa s one with a The chronaxie evidently changes very abruptly when the nerve loses its excitability from oxygen lack. Membrane Depolarization During Anoxia The cause of the reduced action potential in asphyxiated areas of a nerve could be membrane depolarization (Furasawa, Gerard, Koch, see introduction). ity was not difficult to test. This possibil- A nerve was placed in the double chamber, the partition of which was then sealed. Oxygen or air was allowed to enter one section and nitrogen the other. 1 electrode in each section led to the D.C. am- plifier previously described. galvanometer was observed. The swing of a slow moving If depolarization did take place during anoxia, as suspected, the electrode in the nitrogen section would become negative relative to the electrode in the a ir section of the chamber. The results of several such experiments with mammalian nerves are presented graphically in Fi gure 12. The slow rise during the first 10 minutes in several experiments was due to drift in the apparatus which could not be completely eliminated. 6 rabbit nerves were asphyxiated, 1 of them twice. These results are plotted with solid lines and small dots. The fi gures in brackets mark the first and second as phyxiations of the one nerve doubly tre ated. The results from 3 ca t nerves are plotted with dashed lines and 10 + 0 20 2 MINUT£S 10 30 Fig. 12. Resting potential change with t~me in nitrogen as re gistered with one lead on section of nerve being asphyxiated, other lead on section of same nerve in air. Curve (1) oxygen was used instead of air. Curve (2) same nerve re peated with air. Curve with large dots plotted with scale furthest left. Others with scale nearest ordinate axis. heavy line with larg e dots. The latter wa s gr aphed using the magnified millivolt scale on the very left of the figure. The nitrogen electrode did become relatively negative, the change beg inning after 5-10 minutes of anoxia, which is about the average ti me when the potential amplitude starts to decrease. The difference in potential between t he as- phyxiated and nonasphyxiated areas of the nerve was more pronounced when the nonas phyxiated area was in oxygen than when it was in air. This is recorded by curve (1) as com- pared to curve (2) both from the same nerve. For curve (1) part of the nerve was in oxyg en, for curve (2 ) the same part was in air. The potential difference when oxygen was used was about 15 millivolts, and with air varied from 5 to 10 millivolts. Theoretically the membrane po tential is about 60-70 mv so that the chang e observed was smaller than was orig inally expected. This smallness could be due to the fluid and connective tissue shortcircuiting the potential. It is important to mention here, though it will be discussed in detail below, tha t this potential change of mammalian nerves was reversible with the ~dmission of oxygen or air into the nitro~en section of the chamber. '-' 2 fro g nerves were as phyxi a ted at room temp erature using the same recording apparatus. A potential became evi- dent after about 00 minutes and incre a sed to 00 mv in one case and 45 mv in the other after 150 minutes. This result was not rev ersible with the admission of oxygen or air into 48. the chambe r. Again it was apparent that frog nerve dif- fered fro m mammalian nerve. It was concluded t hat the nerve membrane is depolarized during anoxia. 8 15 w N (/') .....J < I:z: w I- 4 0 a.. 0 6 7 8 9 MINUTES 10 11 Fig. 13. Graph showing relative spi ke size with ti me in nitrogen for r abbit nerv e wit h continuous stimulation at 60/sec. (s olid circles), and at 15/sec. (open circles). 12 I RESULTS B Recovery Mammalian Nerve When an as phyxi a ted mammalian nerve is placed in a fluid or gaseous medium cont a ining oxygen, the excit ability of the nerve recovers rapidly. The recovery times of 16 rabb it nerves, 15 cat nerves, 3 dog nerves, and 6 lobster nerves are presented in Table X. The left hand column of figures gives the time of the first signs of electrical activity after the beginning of oxygen treatment. This is expressed in seconds for mammalian nerves, minutes for lobster nerves. The right hand c olumn gives the time (in min- utes) which is required for the activity to reach its original value. All the mammalian nerves recovered in a like manner, the first signs of recovery appearing after 30-120 sec onds of oxygen, complete recovery being attained after 5-20 minutes. Lobster Nerve Lobster nerves recovered more slowly, the first si gns not being apparent for 5-20 minutes in oxygen. The orig- inal state of irrit ability was reached in 10 minutes by 1 nerve bundle, but the other 5 required 30 minutes or more. The recovery of a lobster nerve bundle is shown in detail in Fi gure Hf. In this case a small sensory bundle was as- phyxiated until only 1 fibre was still conducting. oxygen was turned on. Then After 5 minutes 1 fibre has recov- ered and its action potential has summated with the 1 present at the start of recovery. After 13 minutes the action TABLE X Recovery ti mes of irritability i n mammalian and lobster nerves . Fi gures in left column indic a te ti me after oxygen st a rts that rec overy first beg ins. Figures in ri ght colwnn indic a te ti me required for c omplete recovery. Rabb it Nerves Nerve Ti me start recovery Til:le complete recovery 90 secs 60 50 60 60 60 90 8 mins 8 Medial II II II II II II Pe rone al II II II II Tibial ti ti Saohenous 60 90 60 60 60 120 8 10 5 10 10 10 12 20 15 60 180 90 10 10 10 90 15 TABLE X CONT. Cat Nerves Ne rve Time s t art r ecovery Time complet e rec overy Per one a l 60 secs 60 90 60 60 10 rn in s 10 10 1.3 10 60 60 5 5 50 50 10 15 II II II II Phr enic II Saphenous II Radial 45 Tibi a l 60 5 Med i a l Cu t aneous 60 5 Dog Nerves Per oneal II Tibi a l 120 60 15 60 10 8 Lobs t er Nerves Whol e Nerve II I Cos er Bundle A 5 mins 5 1-l hours 6 10 10 '"' .:, 10 rn ins . Fig. 14. Recovery of lobster nerve in oxygen. At start only one fibre conducting, two in 5 minutes , 3 in 13 minutes, 4 in 45 minutes , and 5 in 60 minutes . This is an excellent example of the conduction velocity change, reduced by the nitrogen, increasing in oxygen. potential of a third fibre has appeared, but the conduction velocity is less than normal. ( see below). After 20 minutes the conduction velocity has incre a sed and summation with the other 2 spikes has begun. Summation is completed in 45 min- utes, and the impulse from a fourth fibre has appeared a gain with a smaller conduction velocity. This one sumr:iates in 60 minutes and a fifth fibre has started to function. Frog Frog nerves behaved differently in oxygen after asphyxiation. Of the 11 nerves listed in Tables III and II nerve 121 recovered completely and 186 about 3011a of normal at room temperature. The other 9 nerves never re gained any of the activity they lost in nitrogen. Heinbecker (1929) and Gott- schalk (1914) have reported many poor recoveries after frog nerve asphyxiation. To account for the poor recoveries the latter sugges ted the poisoning by high concentrations of metabolites collecting during anoxia. On the other hand Gerard and ot hers obtained complete recoveries using fro g nerves. Effect of Asphyxiation Time on Recovery The recovery times listed in Table X were fro m nerves g iven oxygen i mmediately upon the extinction of all nerve activity. Prolonging the anoxia beyond this period had a pronounced effect on the time required for the first si gns of recovery to appear when placed in oxygen. Once recovery had started it progressed at about the san1e rate as already described. Below are listed a few recovery results after p rolonging the asphyxiation with rabbit nerves. Prolongation period (minutes) Time in oxyg en to first sign of recovery (minutes) l 0 2 5 2 10 15 3 15 5 25 7 40 5 60 no recovery 70 II 55 " It was concluded that a) more than 40 minutes in nitrogen after complete asphyxiation destroyed some process neces sary for recovery; b) if no signs of recovery were observed after 8 minutes in oxygen there would be no recovery. Effect of I mmersion in Solution on Recovery An entirely different phenomenon was investi gated during recovery with immersion, namely, the possible washing away of poisonous metabolites as suggested by Gottschalk (1914). 4 frog nerves were asphyxiated for 3-~-, hours in nitrogen. si-, s-1- , and 6½ The first nerve showed no si gn of recov- ery after 30 minutes in oxygen, but did begin to recover after 30 minutes more, this time in aerated ringers. ery was complete after 10 hours. The recov- The second also failed to recover in 30 minutes of oxyg en alone, but began to recover a ft e r only 15 minutes in solution. Nerve 3 f a iled to show any respon s e a fter 1½ hour s in oxygen alone, but was 50% recovered a fter overni ght i mmersion in the saline. The f ourth nerve only recovered about 15% a fter overnight washing . These experiments confirmed Gottshcalk 1 s a nd Frohlich's cl a i ms. The e f fect of i mmer s ion was tried on mammalian nerves unable to recover after pr olong ed asphyxi a tion periods. A r a bbit p erone a l was a sphyx i a ted in a nitr og en at mos phere for 1 hour. Signs of recovery began to a ppe a r after 5 minutes in oxygen so the nitroge n was i mmedi a tely resumed for another hour. The surviva l ti me ori g inally was 20 minutes, but was only 7 minut e s the second ti me the nerve was a sphyxi a ted. After the s ec ond hour in nitr ogen oxygen f a iled t o revive activity i n t he nerve for 20 minutes. Th i s experiment was re peated with anot her r a bbit peroneal fro m th e s ame animal. The only difference was t hat the second nerve wa s immersed duri ng t he anox i a in s a line. The survival ti mes were about the same a s of t he first nerve. This i mmerse d nerve began to recov e r from the s ec ond hour in nitrog en after only 2 minutes in oxygen. Each of 2 r a bbit tibi als were subjected to 100 mi nutes in nitrog en, one submerged in s aline, the other in a gas a tmo~ere. minutes. Both nerves showed survival ti mes of 20 There was no res ponse fro m the nerve in gas after 20 minutes in oxygen, where a s t he submerged nerve was more t hat 25% norma l a fter only 4 minutes in oxygen. A cat p erone a l was submerged in s al ine and g iven 75 minutes nitrogen tre a t ment. This nerve recovered completely a fter only a fe w minut es in oxygen. Previously it was found t hat no cat nerve rec overed fro m mo re t han 55 minutes in nitrogen. It was concluded t hat whatever t he recovery inhibitor was, it could not be re moved by oxygen alone, but could be removed by oxyg en plus physiolog ical solution. Cowan has found t hat Mai a nerve recovers somewha t in oxygenles s sea water when no recovery was po ssible in oxygen alone. Effect of Temperature on Recovery Time 3 r abbit p erone als were a s phyxi a ted at different temper a t ures. Nerve 52 was asphyx i a ted fir s t a t 23 degrees. It began t o recover 60 sec onds after oxygen was applied and was completely recovered a ft e r 30 mi nutes in oxygen. The same nerve was a sphyxi ate d again this ti me at 11 de grees C. Recovery did not b eg in for al mo st 8 minutes and was only 50% c omplete af ter 30 minutes in oxygen. ted t he first ti me a t 23 degrees. Nerve 54 was asphyxia- Recovery, begun in 2 min- utes of oxygen, was complete in 20 minutes. The second an- oxi a was a t 35 degre e s and recovery began in 60 sec onds and was completed in 15 minutes. The nerve was ke pt in oxygen for 1 hour more, a t otal of 75 minutes. Then it was a sp hyx- i a ted a third ti me a t 24 degrees and did not begin recovery for 4 minutes in oxyg en, re quiring 32 minutes to c omplete the recovery. Ne rve 142 was a s phyxiate d a t 21 degrees and recovery did not st art for 3½ mi nutes in oxygen. A second anoxia a t 38 degrees was c arried out on t his nerve and recov- ery began in 30 seconds of oxygen, being completed in 8 minutes. Although the survival ti mes were increased by the lower temperatures which would affect the st arting time oxygen application, the increase was not enough alone to account for the above results. Therefore it was concluded tha t the recovery time of mammalian nerves was markedly increased by cooling. It was not practical to raise the temperature of mammalian nerves. However, this was tried with frog nerves in the hope of reducing the survival time and thus making recovery possible. The idea was based on the fact that mammalian nerves recovered poorly or not at all if subjected to long asphyxiation periods such as necessary to asphyxiate the frog nerve. C. 7 frog nerves were as phyxiated at 32.5 to 55 degrees The 2 nerves with the longest survival times at this tem- perature range, 130 and 210 minutes, failed to recover any activity. The 5 other nerves reacted as follows: Nerve 135, survival ti me 128 minutes, recovered 65% normal in 5 minutes. Nerve 120, survival time 120 minutes, recovered 80% normal in 15 minutes. Nerve 153, survival time 110 minutes, recovered completely in 20 minutes. Nerve 117, survival time 80 minutes, recovered completely in 30 minutes. Nerve 156, survival time 70 minutes, recovered BO% in 5 minutes. Nerv e s 117 and 120 were sciatics from R. p i piens which on the averag e have shorter survival ti 0 es than the nerves from R. ca tesbiana (135, 153, 156). It wa s concluded tha t the lack of recovery or p oor recovery in fro g nerve at ro om te mperature was due to t he l ong asphyxiation time, during which time recovery inhibitors collected. Temperature increas e decreased this time and recov- ery became possible. Frequency of Stimulus Effect on Recovery Fre quency of stimulus had no effect on recovery times. An example is g iven in Figure tf. A rabbit peroneal was sub- jected to continuous 60 cycle f ar adic stimulation causing maximal response, and another peroneal was sti mulated by 1 shock every fe w minutes. They recovered identically. Recovery of Threshold of Excitation and Amplitude of Action Potential The threshold recovery curves and action potential amplitude recovery curves are reproduced in Fig. ]§: The threshold recovery curve is mar ked T and is drawn through the da t a from 10 rabbit nerves (open small circles), and 1 cat ne rve (solid smal l circles). The a ct i on p otenti a l am- plitude recovery curve was plotted from the da t a of 3 rabb it nerves (large open circles). The data fro m a cat nerve (l a r g e solid circl es), and a dog nerve (half filled l arge circles) were a lso plotted. The threshold units ar e l/200ths of a volt (left hand ordinate), and the action potential amplitude is in perce nt of orig inal size (ri ght hand ordinate). w 0 :::, I_J a.. 0 .....J ' ~ <C 0 ::r: 50 (/) ~30 i- A z u w I I- a:: w CL T 0 0 0 10 100 I 4 Fig . 15. Graph showing recovery in oxygen (minutes) of threshold (T) with small open circles rabb it, solid circles cat; and recovery of amplitude (A) large circles rabb it, solid circles ca t, and half circles dog nerves. I The recovery of excitability progressed exponentially, appro a ching t he orig ina l v alue. The t hre shold of mammalian nerves decre a sed abruptly toward t he orig inal value, at taining it in less than 10 minutes. The action potential ampli- tude was slower, requiring 20-30 minutes to reach the maxi mal size. It was 80-90% of this size at the time when the thresh- old was normal. No such threshold fluctuation as described by Lehmann (1937) we re observed a t any ti me during recovery. The effect of prolonging a sphyxi a tion upon the threshold recovery curves is plotted in Fi g . J.6. The number near each line desi gna tes t he nwnber of minutes a ft er comple:te extinction of all nerve excitability that the nitrogen period was prolonged in that particular threshold recovery curve. There is a shift to the ri ght as t he ti me of prolongation incre as es, but the recovery curves maint a in a consistently similar shape. Thus the chief effect of prolongation, as said before, is to retard the initial appe arance of recovery. No effect upon the 4 recovery rate once it is started is evident. End-in Effect The end-in effect caused by anoxia was observed to reverse during recovery, Fi g . 19'. The photographs are of the recovery of an action potential of a frog nerve at 36 degrees. The ti me interv als a t which the pi ctures were recorded are ind icated in t he fi gure. The ratios of t he p ositive to ne ga- tive spikes were: at s tart--0.000, 5 mins--0.222, 10 mins-0.272, 20 mins- -0 .417, 30 mins-- 0 .462 . Thus at first there was only the negative spike, t han a very small positive one 0 80 60 0 __J 0 :r: ~40 a::: :r: f- 20 I 0 - --->-.L__- - - - - - - - i ~ 0 2 4 MINUTtS 6 Fi g . 16. Graph showing delay in recovery with i ncreasing time in nitrogen. The fi gures indica t e the time in minutes the nerve was held in nitrogen after disappearance of the action potential. Fig. 17. Recovery curve of frog nerve at 36 degrees centigrade in oxygen showing mono-to.diphasic change. appe ared at 5 minutes which increased in s ize r ap idly in the 10-20 minute interval of oxyg en. This effect was observed in the majority of experiments when the lead off electrodes were near the cut end of the nerve. It indicates that the tissue contacting the proximal lead became active before the tissue contacting the distal lead. Refractory Period The recovery of the refr a ctory period of the nerve is tabulated in Table XI. The manner of ca lculating was the s ame a s was used for the effect of anoxia on the refractory period, except here the last ratio of each nerve (10 minutes) was the value used for subtraction to obtain the ratio difference and for dividing to obtain ratio difference. Thus, taking nerve 87, for example, at ½ minute the exp res.ooo-.583 __ 1 000 sion is The graph, Fi g . 18, only . 583 • • s hows the first 5 minutes of recovery. The action potentials were recorded and measured for every 30 seconds during recovery. The refractory period recovery was very r ap id after the first minute until the third minute, following which the ratio decreased so that after five minutes the recovery was very slow. The aver age of 2 experiments in which just the absolute and total refra ctory period recoveries were made, showed that in 1 to 2 minutes a fter oxygen was st arted the absolute period was O. ?O sigma and the total refractory period was 2.84 sigma. At the start of asphyxia t hes e s ame 2 nerves averaged an absolute refra ctory period of 2.89 si gma, so within 2 minutes it TABLE XI Ratios of t he second nerve i mpulse to the first when the nerves were stimulated with t wo shocks following so close in succession t hat the se cond fell in t he rel a tive refr a ctory period fro m the i mpulse of t he first. Shows the rec overy i n oxygen of t he refr a ct ory state of t he mamn1ali an nerves, r abbit nerves. Nerve Time in Oxygen in Minutes 1.0 1.5 2.0 2.5 .ooo 0.143 .222 .ooo 0.000 .ooo .ooo . . ooo .156 .ooo .ooo .250 .ooo .ooo .143 .ooo .ooo .250 .ooo .ooo .ooo .308 .389 .435 . 428 .572 .167 .270 .214 .736 .375 .555 .600 .400 .430 .353 .100 .200 0 93 88 87 86 86 84 83 0 .5 .435 4.0 5. 0 10.0 .488 . 528 . 560 .367 .394 .583 .750 .286 .326 . 893 .315 .428 .364 .381 .740 .609 .667 2 0 M~UTES 3 4 s ;, 10 (/") l.&...J u z UJ a:: l.&...J LL. t....... 0 c5 0 250 I<( a:: I- z w u a:: w · a.. 100 w 0 ::::, 1_J a... ::::::!!: <( MINUTES 2_ _3_ _4_ _ _ 5 oc;...__ _ _ _ Fi g . 18. Refr actory peri od r e covery curve in oxygen (upper). Same recording syste m as used with a sphyxi a tio n , see te xt. Eff ect of no stimulation and 60 cycle continuous stimulation on recovery of the same nerve (lower). was possible for the refractory period to recover completely. This suggests that the limits of the refractory states of the nerve recover their original values rapidly, but the variations of nerve irritability within these limits caused by asphyxiation recover more slowly. In other words, if the curve representing the recovery of excitability in a nerve fibre after the passage of a nerve impulse will attain its 2 original limiting values before it will attain its original shape. No attempt was made to investigate this possibility. Oxygen 11 Rests 11 and Repetitive Anoxia 19 nerves (rabbit and cat) were asphyxiated 2 or more times with different oxyg en 11 rests 11 separating the anoxia periods. The results were as follows. 9 nerves were given an oxygen ~eriod of 20 minutes or more and showed the slightest decrease in the second survival time compared to the first. 10 nerves were given oxygen for less than 20 minutes and showed a decrease in the second survival time compared to the first of 15-201b. Those given only a 5 minute oxygen rest had an average second survival time 22% shorter than the first. Those given 10 minutes had average survival times 12% less than the first. All these nerves were first 11 aged 11 an hour or more in saline to eliminate the initial effect of ilaging 11 described in the a sphyxiation section. Thus an oxy- gen period of 20 minutes or longer is necessary for c omplete recovery. These results conform with Gottschalk's reports for frog nerve. Depolarization Recovery The results of depol arization recovery experiments are shown graphically in fi gure 19. The curves with the small circles represent the data fro m rabbit nerves, the large circles represent d at a from a cat nerve at higher amplification to be read with the millivolt scale to the right of the ordinate line. The open circles represent data fro m nerves as phyxiated 220 minutes and 90 minutes. The c at nerve was asphyxi a ted 90 minutes. The r abb it nerves for 30 minutes. about Re p olarization of the nerves began after/2 minutes in oxygen. All preparations did re polarize to the normal value, the large part of the process occurring during the first 2 and 5 minutes in oxygen. The belated start of the repolari- zation was probably due to the prolong ing of anoxia ~O minutes. Rabbit nerves are usually asphyxiated in 20 minutes. The appara tus wa s not desi gned or built so that action potential studies could be made si mult aneously with polarization studies, but it was only a matter of a few minutes to transport the chamber from this apparatus to the oscillo"'... gr aph. This wa s done in the c a ses of those preparations as- phyxi a ted for long periods of ti me (220, 90 minutes) which would not be expected to recover unless immersed in saline. None of these nerves s howed any a ction potenti al res ponse although the membr ane had been re pol arize d at least 5 mv during the 15 minute period in oxygen before the action potential measurements were attempted (small open circle s , large solid circles). Thus the action potentials were still absent after some 20 minutes of aeration, though the as phyxiated portion of t he nerve was well repolarized by then. The 2 rabb it V) f_j 0 > _j _j 10 MINUTES 15 20 Fig. 19. Curve showing reduction with ti me (minutes ) in oxygen of the resting potential between a le ad on t he nerve in nitrogen and one in air on the same nerve. Small circles are results fro m rabbit nerves, t he open circles indic a ting nerves a sphyxiated 220 minutes and 90 minutes. Large circle from cat nerve data using the fi gure s to ri ght of ordi nate. nerves were capable of pr oducing a 1/10 nor mal s pike after 2 hours w2, shing in s aline, and t he other nerve, cat, recovered its s pike completely a fter 3 hours washing . The p ossibility that the moisture entering the chamber with the ga s caused fluctuations in the apparatus which mi ght be interpreted erroneously as nerve polarity changes was dispelled by an experi ment using frog nerves. 2 frog nerves were de polarized with nitrogen as described in a previous section. Oxyg en wa s then admitted to the chamber. No visible chang e occurred for 15 minutes in 1 case and only a 5 mv restor ation after 105 minutes in the second case. This relatively very s mall recovery in the ca se of the frog nerve where a 65 mv depolarization was obtained with nitrogen, could easily have been due to slight depolarization of the section of the nerve held in air from ag ing. No action poten- tial could be obtained fro m either pre paration. The conclusion is that some change takes place in the membrane during oxyg en l a ck, besides the depolarization, which must also recover before the nerve can make use of the re polarized me mbr ane to conduct an impulse. Toxic Subst ance Experi ments by Heinbecker (1929) have proven tha t neither l a ctic acid nor CO2 is the metabolite causing the poor recoveries in nerves after long anoxia periods. Since then Cowan has suggested the collection of hi gh concentration s of potassium ions outside the nerve, leaking out fro m the inside, as being the c ause of the poisoning of the nerve. Several ex- periments were devised to test this supposition. First it was observed tha t 20 ti mes normal potassium ion concentration in the physiological solution extinguished rabbit nerve activity in about 1 hour confirming the findings of Graham (1933). Higher concentrations were poisonous in a shorter time, lower concentrations took longer to take effect. A rabbit peroneal was first placed in Tyrodes solution with 10 ti mes the normal potassium ion concentration and kept there for 1½ hours. It was then asphyxiated while sub- merged in the same . solution. 1 hour and 40 minutes. The anoxia was maintained for All activity in the nerve ceased after 20 minutes in nitrogen. The nerve began to recover, still submerged in the same solution, after 4 minutes of oxygenation. 2 rabbit tibials were asphyxiated simultaneously in a nitrogen atmos phere for 1½ hours. One was then placed in aerated physiological solution and the other was placed in aerated physiological solution containing 20 times normal potassium ion concentration. Both recovered identically. A rabbit peroneal was asphyxiated in the normal manner, becoming inactive in 15 minutes. It was g iven a 50 minute oxygen rest and then immersed in 20 times normal potassium ion concentration solution for 10 minutes. A second asphyx- iation period was then begun with the nerve still immersed. The spike disappeared again in about 15 minutes a nd after 2 minutes of oxygenating the solution, the nerve began to recover. Although the action potential never quite regained its original size, the threshold was normal after 15 minutes of oxygen. Another rabbit peroneal was kept in the 20 times normal potassium ion concentration solution for 40 minutes. The action potential was about 30% its normal size after this period. The nerve was then asphyxiated and had a survival time of only 5 minutes. However, the spike reappeared again after only 2½ minutes of oxygenation of the solution. Thus it was concluded that a nerve could recover from asphyxiation even if prolonged, as well in a 20 times normal concentration of potassium ions in the surrounding medium as in normal physiological solution. The 20 times normal potas- sium ion concentration was chosen because this is the concentration found inside the normal nerve, and it would be impossible to obtain a higher amount of this ion from leakage out of the nerve during asphyxiation. Therefore the nerve can recover from asphyxiation in a concentration of potassium ion higher than that possible biologically, and hence potassiwn cannot be the toxic substance. pH Change of Solution During Asphyxiation A group of several nerves were asphyxiated im~ersed in 1½ cc of solution for 2 hours. Others were simply placed in the same amount of fluid and allowed to remain there for 2 hours with oxygen bubbling through the medium instead of nitrogen. The pH decreased 0.15 units during anoxia compared to 0.10 units in oxygen. This difference of 0.05 units is not enough to have a poisoning effect. Test of Solution Surrounding Nerves Asphyxi a ted Long Times , 3 cat medial nerves immersed in 12 cc Tyrode solution were asphyxiated for 2 hours. utes. All activity ceased in 40 min- At the end of the 2 hour period these nerves were re- placed in this solution by a cat radial nerve. The latter was subjected to nitrogen and had a survival time of 20 minutes. Recovery was complete in 5 minutes after oxygen was admitted to the chamber, the nerve still being immersed in the fluid. Another radial nerve was placed in the same me- dium and asphyxiated. Its survival time was 40 minutes and it was completely recovered also 5 minutes after oxygenation of the solution. Thus 5 nerves asphyxiated for a total of 3 hours did not produce enough "toxic" substance to be biologically detectable in 11 cc of physiological solution. DISCUSSION When a nerve is placed in nitrogen it is able to maintain its original irritability for some time. This must be due to an energy reserve bound in the nerve. What the com- position of this reserve is, is not known. However, it is known that its manufacture is dependent upon oxygen. This is indicated by the results of varying the length of the oxygen intervals between repeated asphyxiations. With only short oxygen "rests" the second survival time is shorter than the first. The cause of this reduction of survival time is the abbreviation of the initial normal period. The rate at which the irritability disappears once this change begins is always the same. Thus after the recovery of the nerve activity in oxygen, which is very rapid, the reserve is formed and at least 20 minutes are required to complete the process. This reserve energy idea is not new. Gottschalk found that a 25 minute period in oxygen was nec~ssary to "fill" the reserve in frog nerve. Lehmann noted the decrease in succeeding survival times with repeated anoxia in mammalian nerve, but makes no mention of the correlation of the oxygen interval between anoxia treatments and this change. When the reserve is exhausted, the threshold for electrical stimulation risea and the action potential spike diminishes in size. Each of these changes when plotted forms an expo- nential curve (see Fig./0). An attempt to explain these changes on the basis of the membrane theory of conduction will be made. Let us c ons i der t h is t heory briefly. Normally a nerve maint a ins on it s membr ane a charge which i s an electric al double l ay er, negative on t he inside and positive on the outsi de. This can be recorded a s a de- marc a tion p ote nti al betwe en injured are a s, cont a cti ng the inside, and non-injured a r eas of t he nerve. Recently it has been meas ured directly a s t he p otential a cross t he membr ane betwe en a fine needle electr ode in s erted inside a single nerve fi bre and an electrode contacting the outside of t he fibre (Curtis and Cole 1940). This potenti al is due to t he uni que membr ane property of selective semi- perme~bility whereby concentr ation gr a dient s of d iff erent ions are set up across t he membrane. The po t a ssium ion concentra tion, for example, is 20 ti mes gre a ter inside a nerve fibre than in t he normal surrounding medium (Schmitt 1940 , Cowan 19.39 ). Calcium ions, on the other hand , cannot penetrate t he membr ane at all and remain outside of t he fibre. Thus the mem- bra ne potenti al is re ally a diffusion potential and its volt age i s g ive n by t he expression V =~log ~ derived by Nern s t (1 897), where R is t h ~ gas const an t, Tis t h e absolute te mperature, n is t he ion va lence, Fis t he f ar aday e quivalent, Ci is t he conce ntr ation of t he ions insid~ an d C0 t he c oncentr a tion of the outside of t he membr ane. If t he pot a s s ium ion concentra- ti on ratio inside t o outsi de i s substit uted for~ the exo expre ss ion becomes V = 0 .058 log 20, or V= 75 millivolts. oo. Cole and Curtis measured a membr ane p otenti a l of 40 mv, and a ction potenti als from single fibres have been recorded as hi gh a s 60 mv (MB.r mont 1 9 40). Allowing f or some shunting due t o t he c onductivity of the surrounding medium these results suggest tha t t he potassium ion gr adient a lone could be the c a u s e of t he membrane potential, although it i s possible t ha t other f a ctors enter in. During anox i a , besides those chang es already described, the de marc a tion p otenti al declines (Koch, Furasawa , and Ger ard) and if only part of t he nerve is as phyxi a ted, t his part becomes electronegative to the una s phyxiated part. Without oxygen, therefore, the nerve membrane becomes depolarized. This is probably due to the loss of the property of selective semi-permeability and hence the disappearance of the ion concentration gradients. Thus the membrane po- tential is extinguished an~ a s it is, the a ction potential declines. The exponential manner of this decline would be predictable from the diffusion potential equation. The velocity of conduction is reduced by asphyxiation as seen in the single fibre a ction potenti a ls (Fi g . 1~) and in a group of fibres from mamr1alian ne rve (Fig . 20). is probably due t o 2 f a ctors. This Firstly, the reduction in s i ze of t he spike p otenti al makes bre akdown of neighboring areas mo re difficult and hence not a s quick. Th e nerve is in a fefr a ctory s t a te. Lillie's ir on wire model provides a n example of this point. When an iron wire is pl a ced in 701b nitric a oid it develops an oxide coating on its surface. :'"'i .; ...(). .Po tent i a l of rablJi t :'t":di&l sr..o· ,ri n 0 .recovery o1 ~,o Grou~ s of fi bers , the VA l ocity of onA ,J reat ly re duce~ o;y as .:.)hyx i a tion , i n creas inB i n oxyGen Vlhen this coating is completed the wire is in a "pas s ive" st a te. If the surface of the coating is scratched off in a s mall a rea the whole co a ting disintegrates along t h e wire but at a very definite rate, not inst antane ously. The dis- integrating area moves along the wire in a manner. anal agous to the movement of the nerve i mpulse along the nerve. If during the for mation of the coating , however, before it is completed and t he wire is s a id to be "refra ctory", it is scr a tched, the disintegration occurs at a much slower rate. Whether it is a potential between the coated and non coated area or not which causes the breakdown of the oxide coating is not known. (Lillie 1919, 19_21). Secondly, and more like- ly, the rise of the local response in the nerve fibre around the negative stimulating electrode is slower and reaches the threshold value for set t ing off the propag ated s p ike potential at a l ater time (Hodgkin 1939). In other words, t he ti me constant k of the local potential (Hill 1935) is increased. The refractory period is lengthened by anoxia. This means t hat it t akes longer f or t he nerve t o restore its char ge than ori g inally. Without oxygen the power to move ions against a potential gr adient becomes more difficult. Frohlich, Heinbecker, and Gerard have all reported elongation of refr actory period, but Cooper ar gued against it. Since t h e recovery time remained constant and only the last period enl ar ged she claimed the refractory period remained const ant and that there was decremental conduction. In other words she observed that the muscle contractions did not summate unless the time between the 2 stimuli, the second 8 times the strength of the first, was increased. This meant, by the decrement idea, that the second impulse grew smaller so that it did not get through the chamber. Increasing the time interval allowed it to become longer at the start, and therefore proceed through the chamber. But, the very fact that the impulse decreased in size at the start means that the refractory state was becoming longer and smothering the response. Thus, Cooper's argument for decremental conduction demands a variable refractory period for its foundation. The rise in threshold can be explained by assuming an increase in the time constant k of the change in local response. Thus stronger stimuli became necessary to force this response over the threshold of the propagated spike. The accomodation factor probably changes too which makes the ca.use of the threshold change a more complicated mechanism. Graham (1933) found that the action potential was reduced, the threshold for excitation increased, the velocity of conduction reduced, and the refractory period increased by increasing the potassium ion concentration in the solution surrounding the nerve. present work. This was also observed in the Actually such a change in the outside medium simply reduces the potassium ion concentration gradient across the membrane and hence reduces the membrane potential. It is logical to suppose that a similar change in the potas- sium ion gradient takes place due to the ion's leaking out of the nerve durihg anoxia when the property of selective semi- permeability is lost (Cowan 1934). It is important to mention here the results of varying the calcium ion concentra tion around the nerve during anoxia. Excess of the ion caused the survival ti me to be prolonged, lack of it shortened the survival time. It is known that the permeability of a cell membrane varies inversely with the calcium ion concentration in the surrounding medium. Thus with excess calcium in the solution the nerve membrane is less permeable and it would require less energy to maintain the potassium ion concentration gradient. This is because the ions cannot easily leak out. With a low calcium concen- tration the reverse would be true. Therefore it would be expected that the reserve energy in the nerve would be used up quicker with the nerve in a solution void of calcium than it would in a solution with an over abundance of this ion, which is exactly what was observed. The end-in effect can be explained on the depolarization theory. To begin with, the area imrnediately around the cut end of the nerve is partially depolarized by the injury. Therefore, without oxygen, the remaining depolarization is completed in a shorter ti me than it takes to depolarize an area completely polarized at the start. The oxygen lack effect is different than that of narcosis. This was observed by Frohlich by adding ether to the nitrogen atmosphere surrounding a frog nerve. The irritabil- ity of the nerve was rapidly extinguished, but recovered in the nitrogen when the ether was removed. It has been noted (Wiersma and Keighley, by communication) that a "threshold" concentra tion of KON exists in crustacean nerves. If this concentration is increased by a very small amount the nerve activity disappears immediately. When the concentration is lowered again the activity reappears abruptly. Schmitt~narcotized a nerve with CO and obtained a sudden recovery of the action potential when a strong light was focused on the nerve. These very sudden changes could not occur if the membrane were depolarized, therefore the narcotic must attack some conduction process in the membrane without changing the electric double layer. Schmitt sug- gests that there is a photo sensitive hemin like enzyme which is poisoned by the CO. The toxicity from long anoxia periods may be due to a similar poisoning of a conduction process. Wiersma (1933) reports decremental conduction caused by KCN. Cooper (1923) claims that decremental conduction re- sults from asphyxiation (see note). Most of the results of the present work are against this possibility, except in the area about the cut end. The potential decline in this area is not a real conducti on with a decrement, but is due to the smaller degree of polarization of the membrane as pointed out above. However, the lower pictures in Fig. 'l ;J.,e suggest that there is a slow decline in the potential of a lobster single fibre as the spike progresses in the asphyxiation chamber. ""l'IJQ The pictures were interpreted, at first, as indicating that the end-in effect progressed along the whole asphyxiated length of a nerve. 7 Later experiments (Frog nerve, Fig. J:fr) showed that this was unlikely which leaves only decremental conduction to explain the phenomenon. Thus at present it is not possible to argue this point until further experiments have been carried out. During recovery in oxygen, the changes which occured in nerve irritability with anoxia, are reversed. The nerve mem- brane regains its selective semi-permeability, ion concentration gradients are reformed and the membrane repolarizes. This occurs rapidly. The threshold drops back to its normal value even before the membrane potential has reached its normal value. This is because, as we have seen during anox- ia, conduction can occur when the normally maximal action potential is not attainable. The velocity of conduction is slow and the refractory period long at this stage. They rapidly regain their normal values again. As pointed out above, prolonging the anoxia period produces a toxic effect on the nerve irritability. When the asphyxiated nerve is placed in oxygen, the membrane can repolarize but cannot conduct an action potential until the nerve has been washed in s~line. Even oxygen free saline can partially eliminate this blocking of conduction (Cowan 1934). CO2, lactic acid, and potassium are apparently not the causes of this inhibition. Thus, acidity changes are ruled out as the possible instigators of the phenomenon. A possibility not tested to date is NH3 (Harvey 1933) which if given off by nerve. The oxygen uptake rate of nerve has been found to have a pronounced effect upon the nerve behavior during anoxia. From data of other publications (Gerard 1932, Gerard and Chang 1936) together with data from the present work a very interesting relationship between oxygen uptake and survival time becomes evident. First, the oxygen uptake rate of a nerve increases with heating, decreases with cooling. survival time of a nerve does just the opposite. The Secondly, the oxygen uptake rate is reduced by high concentrations of calcium ion in the physiological solution, and is increased by the absence of the ion from the medium. vival time does just the reverse. Again, the sur- Thirdly, the oxygen up- take rate at the cut end of a nerve is higher than in normal uninjured areas, and the survival time in the cut end region is shorter than in the normal areas. Thus, it appears that the survival time varies inversely as the oxygen uptake rate. To investigate this problem, the oxygen uptake rates of rabbit, dog, lobster, and the two species of frog used were taken from the table in the review by Gerard. The average va1ue from two oxygen uptake rate experiments on cat nerves using the Warburg technique was included. below. They are listed Rabbit nerve 280 cc/gm/hr Cat nerve 200 II Dog nerve 190 II - Lobster nerve 80 II R. Pipiens nerve 40 II R. Cates nerve 30 11 If the above surmise is correct the ratio of the rabbit nerve oxygen upt ake rate to the cat nerve oxygen uptake rate should be equal to the ratio of the cat survival time to the rabbit survival time, the inverse ratio. Since the 2.3 minute surviv- al ti me of the rabbit nerve was taken fro m a mean of 28 nerves it is st atistically more accurate than any other values and will be used as the basic survival time. g ives 2-80 _ X '2DU - '2°3' X = 32.2 minutes. Setting up the ratio The mean survival time of the 10 cat nerves was about 3.3 minutes. (Lehman recorded a 30 minute survival for cat nerve.) Dog nerve irritability should survive about the saine length of time without oxygen as the cat nerve activity on the basis of their both having a similar oxygen consump tion r a te. This was true of the mean of 3 dog nerves. 3 nerves is not a large enough number for other than a very rough ap. 280 _ X proxi mation. From the expression T9TJ - ~ , X = 33. 9 minutes. This figure predicted by theory is not only in close agreement with the 33 minute survival time found experimentally with dog nerve, but bear s out t he postulation that the dog nerve should have a similar survival time in nitrogen as the cat nerve. The lobster nerve should survive much longer on the ratio 280 -- Z3 X ' X -- 80 mi·nutes' • From the expression . --su Actually the mean survival ti me of 6 nerves was 77 minutes, basis. the excitability of the 3 nerves left intact lasted 75 minutes each in nitrogen. The same equation gives theoretical survival times of 2½ hours for R. p i piens and 3½ hours for R. catesbiana . Ac- tua lly the nerves of R. p i piens had a survival ti me between 1J a.nd 3 hours V{hile the survival ti mes of the bullfrog nerves ranged between 2½ and 4 hours so again the order of magnitude of theoretical value and actual value are within comparable limits. Therefore the s urvival ti me does v ary inversly as the oxyg en upt 2..ke rate. This means also th2,t in all nerves the reserve supply is approximately the same, assuming that when anoxia is started the nerve draws on the reserve supply at the rate indicated by t he oxygen uptake rate. Examination of the r a tios of the initial normal per iods strengthens this argument. The ratio of rabbit nerve initial p eriod to lob- ster nerve initial per iod is 7/35 and the ratio of oxyge n upt ake is 10/35 for the same nerves. i mation. This is a go od approx- Similar r a tios of 1/10 to 1/9 hold between rabbit and bullfrog i ni tial periods and their respective oxygen uptB,k e rates. It is known that the nerve metabolic rate falls off during the fir s t hour or t wo in saline. In semm the fall is le s s rap id than it is when the nerve is in saline. It was found that the frog nerve followed the inverse law when placed in physiol og ical solution, but the mammalian nerves did not. The mammalian nerve survival time dec1·eased with time in saline until an equilibrium was at t ained after ½-1 hour. It must be concluded that when a mammalian nerve is placed in a physiologic al solution some metabolic process is changed which either does not exist in t he. fro g nerve or is not variable. Schmitt' l suggestion that the reduc- tion in the metabolic rate is due to a healing over of the cut end does not solve this problem, for the healing over would p robably occur in both types of nerve. The protein content of the serum mi ght b~ the cause of the difference between anoxic behavior (and metabolic behavior) of nerve in serum and in saline. However, the pro- tein content of fro g blood is about the s ame as that of mammali an blood, on the basis of simple acid and heating tests, so that the cause of the different reactions of these t wo types of nerves remains obscure. In conclusion it should be repeated that at no time was the threshold recovery curve published by Lehmann observed. Nor were peroneal and saphenous nerves found to differ in their respective behaviors during anoxia consistently, as claimed by hi m. Lehmann ha s suggested tha t~teration of the calcium ion bal ance is the caus e of the variations he observed in the irrit ability of mammalian nerves during anoxia. If this is true, then t he change in the calcium ion concentrations would have to be exceedingly large, becaus e in the pre s ent work nerves were immersed for 30 minutes in 250 ti mes normal calcium ion concentration and solutions containing no calcium whatsoever without initiating any tremendous variations of threshold or action potential as brought about by anoxia. A scheme is presented, next page, which pictures how oxygen lack may affect the metabolism and irritability of the nerve. The nerves of Cambarus clarkii do not fit into this picture at all. If there are any aerobic processes connected with excitability in crayfish nerves they are so small as to escape detection by the methods used here. It is of interest that the nerves of two related animals, lobster and crayfish, should react so differently to oxygen lack. It is worthy of more investigation to find just where the crayfish nerve respiration takes a turn strange to lobster, frog, and mammalian nerve. The crayfish is able to spend long times buried in mud where there is practically no oxygen at all. leads a more active life on the sea bottom. The lobster There are prob- ably others in the animal kingdom which also do not fit the scheme. More experi ments of a comparative nature are neces- · sary to lead to an understanding of the reasons for these variations and to a more complete scheme to include air types of nervous tissue. Oxyien___. Aerobic _ __.►• Energy - - - 1 - . Anae robic Energy - Stored ► Potential Reactions Reactions lleta'toli tea removed by aerobic reactions Membrane • Potential Reactton involving hemin-li:ke ~ : rvrembrane catali~er Conduction ( uses oxyg en?) Stored energ;y ----~>•Anaer0bic -----""'I~• Membrane used un reactions potential stl.11 reduces 1• Meta,boli tea Catali!:er inactivated accumulating---!~• by hi&h concentration of metabolites No Membrane 6<mductiem OXYGEN RECQVJi~~X Oxygen Aerebic Energy Anaerobic Membrane --~~~Energy--~~~ Stored--,>~Potentia1---••P0tential Reactions Reactions t i No Membrane Cataly~er still Conduction inhibited as oxygen removal process does not react with these bound(?)rnetabolites Newly formed metabolites r~mQved by aer~bic reactions / . t sa 1·1ne I mmers1on in removes me iaboli tes Cataly~er Reaction Membrane -------,~•conduction SUMMARY A comparative study was made of the effects of oxygen lack on the peripheral nerves of dog, cat, rabbit, frog (2 species, crayfish, and lobster. Dog and cat nerve activity survived about 30 minutes, rabbit nerve activity about 20 minutes, lobster nerve activity over an hour, Rana pipiens "I nerve activity l~-2 hours, Rana catesbiana nerve activity more than 2½ hours, a,nd crayfish nerve activity 9 hours without oxygen. It was found that the nerve membrane becomes depolarized during anoxia. The ratios of the survival times of nervous activity without oxygen were found to be inversely proportional to the ratios of the oxygen consumption rates of the same nerves (these rates taken from Gerard). The refractory period, relative and absolute, was prolonged by asphyxiation. The survival time of mammalian nerves varied inversely with temperature change and ti me washed in solution before asphyxiation. The survival time varied directly with calcium ion concentration in the physiological solution. The survival time of the mammalian nerve was not changed by i mmersion of the nerve in saline during· asphyxiation, was held longer at the initial value by bathing the nerve in serum and wa.s not changed by prolonged stimulation. Recovery of activity was rapid and complete in nammalian nerves with oxygen following asphyxiation of not too long duration was partial or not at all in fro g nerves unless the temperature was raised to shorten survival time, was complete but slower with lobster nerves. Prolonge d asphyxiation of mammalian nerves in vapor up to 45 minutes shifts the starting point of recovery. To recover fro m more than 45 minutes in nitrogen the marn.malian nerve had to be immersed during the asphyxia or ·washed after it in physiological solution, indicating possible accUJnulation of toxic metabolites. Potassium is not the metabolite which acts poisonously. No recovery of threshold pattern as Lehmann described was observed. Oxygen 11 rests 11 of 20-30 minutes were required to duplicate the original survival time. A possible scheme for the nerve metabolism is discussed. It is based on the assumption that there are two reaction groups, one aerobic, the other anaerobic. BIBLIOGRAPHY H. von B.-reyer, T.H. Chang, Zeits fur alge Physiol, Proo Soc Exp Biol and Med, Vol 2, 1902. vol 28(9), 954-956, 1931. T.H. Chang, R. Cole, Amer Jo urn Physiol, Journ Gen Physiol, R. Cole and H. Curtis, 147, vol 97, 511, vol 12, 44, 1931. 1928. Journ Comp and Cell Physiol, vol 15, 1940. J.B. Collip, Trans Roy Soc Canada, ser 3, sec 5, Biol Sci 21(1): 151, 1927. S. Co oper, Journ Physiol, vol 58, 41, S.L. Cowan, Proc Roy Soc, Bll5(793): 216-260, H. Davis, 1923. Amer Journ Physiol , v ol 86, 106, A.O. Downing and R. W. Gerard, 1934. 1 928. Proc Roy Soc Bi, 100, 223, 1926. Ewald, see von Braeyer , 1868. w.o. Fenn, Amer Journ Physiol, vol 80(2), 327-346, w.o. Fenn, ibid, vol 81(2), 476, 1927. w. o. Fenn, ibid, vol 85(2), 368, 1928. w.o. Fenn, ibid, vol 92, 349, 1930. Vi . O. Fenn , J ourn Gen Physiol, vol 10(5), 767-779, Vi . O. Fenn, ibid, vol 11( 2 ), 175-191, w. o. Fenn, Medicine, T. P. Feng, Journ Physiol, vol 77, 111-138, vol 7, 349, 1930. 1933. 1908. A. Forbes, 1929. Amer Journ Physi ol, vol 90, 119, Journ Physiol, vol 67, H.S. Gasser and J. Erlang er, Activity, 1927. 1927. H. Fillie, Zeits fur alge Physiol vol 8, 492, K. Furusawa, 1927. 1929. Electrical Si g ns of Nervous Un iv. of Pennsylvania Press. R. W. Gerard, Scand Arch Physiol, vol 49, 133-134, R. W. Gerard, Natur wissenschaft, vol 15(26), 538-539, R. W. Gerard, Journ Physiol, R. ~vv . Gerard, Amer Journ of Physiol, vol 90(2), 357-368, vol 63(3), 280-298, 1926. 1927. 1927 1929. R. Vl . Gerard, ibid, vol 92(2), 498-591, 1930. R. W. Gerard, ibid, vol 111(3), 681-710, 1935. R. 'vV. Gerard, Proc Soc Exp Biol & Med, vol 29(9), 1073-1076, 1930. R. \'i . Gerard, Physiol Rev, vol 12(4), 469-592, 1932. R. W. Gerard, Biochem Zeitschr, vol 191, 125, 1927. R. W. Gerard, Am Rev Biochem, vol VI, 1937. A. Gottschalk, Zeits fur allg. Physiol, vol 16, 513, A. Gottschalk, ibid, vol 18, 341, 1920. H.T. Graham, Am Journ Physiol, vol 104, 216, R. Guttman, Biol Bull, vol 69(3), 356-360, s. Harashima, 1914. 1933. 1935. Journ Cell Comp Physiol, vol 3(4), 419-424, 1933. A. van Harreveld, Amer Journ Physiol, vol 141, 97, A. v a n Harreveld and G. Marmont, 101-111, Journ Neurophys, vol II, 1939. A. van Harreveld and C.A.G. Wiersma, 78, K. Halter, 1944. Journ Physiol, vol 88 , 1936. Biochem Zeitschr, vol 257, 313-324, 1933. p. Heinbecker, Amer Journ Physiol, vol 89 , 53, P. Heinbecker, ibid, vol 96, 613, P. Heinbecker, Proo Soc Exp Biol Med, vol 27, 497, 1930. Journ Pharm Exp Therap, vol 29, 161-165, 1926. A.V. Hill, 1929. 1931. A.v. Hill, Proc Roy Soc, vol Bii5, 200-216, 1934. E.G. Holmes, Ann Rev of Biochem, vol III, IV, 1934, 1935. S.N. Kaganovskaia, Compt Rend Acad Sci U.S.S.R., 1, 261, 256- 1935. G. Kato, The Microphysiology of Nerve. E. Koch, Zeits f.d.gas.exp.Med, vol 1, 238, E. Koch, Pflugers Arch, vol 216, 100, 1926. 1927. J.E. Lehmann, Amer Journ Physiol, vol 119, 111-120, J.E. Lehmann, ibid, vol 118, 1936. R. Lillie, Science, vol 51, 259, 416, 1919. R. Lillie, Journ Gen Physiol, vol 3, 107, K. Lucas, Journ Physiol, vol 46, 470, 1921. 1913. E.J. Lund, Journ Exp Zool, vol 51, 265, E.J. Lund, ibid, vol 51, 327, 1937. 1928. 1928. H. Macheles and R.W. Gerard, Amer Journ Physiol, vol 93., 318., 1930. G. Marmont, Amer Journ rhysiol, vol 130, 392, A.P. Mathews, Science, vol 40, 492, 1902. A.R. Moore, Journ Gen Physiol., vol 1, 613, G.H. Parker, ibid, vol 12, 419-425 1 G.H. Parker, ibid, vol 6, 1925. G.H. Parker, ibid, vol 7, 1926. 1940. 1919. 1929. C.L. Prosser, Journ Cell Comp Physiol, vol 14, 287-297., Ranke, see von Braeyer, 1939. 1868. L.H. Rice and H. Davis, Amer Journ Physiol., vol 87, 73, L.K. Riggs, Journ Biol Chem, vol 39, 385, 1919. w.s. Root, Proc Soc Exp Biol Med, vol 29, 144, Y. Ryuzo Tohoku, Journ Exp Med, vol 9, 338, 1931. 1927. 1928. 82. F.O. Schmitt, Amer Journ Physiol, vol 90,506, F.O. Schmitt, ibid, vol 95, 650-661, 1930. F.O. Schmitt, ibid, vol 97, 302-314, 1931. F.O. Schmitt, ibid, vol 104, 303-319, 1933. F.O. Schmitt, ibid, vol 104, 320-330, 1933. 1929. F.O. Schmitt, Biochem Zeitschr, vol 213, 443, F.O. Schmitt, Science, vol 72, 583, 1929. 1930. F.O. Schmitt, Cold Spring Harbor Symposium on Quantitative Biology, vol rJ, 188-193, 1936. II. Shapiro, Journ Cell Comp Physiol, vol 9"", 381-396, II. Shapiro, Biol Bull, vol 77, 305, 1939. M.A.F. Sherif, Journ Pharm Exp Therap, vol 38, 11-29, M.A.F. Sherif, Biochem Journ, vol 24, 400-401, W. Solandt, Proc Roy Soc, B119, 1937. 1930. 1930. 1935. H.C. Syz, Journ Pharm Exp Therap, Vol 30, 1-19, s. Tashiro, Am.er Journ ihysiol, vol 32, 107, s. Tashiro, ibid, vol 34, 405, 1914. s. Tashiro, ibid, vol 36, 368 1 1915. 1926. 1913. I.M.Thompson and W.Kimball, Proc Soc Exp Biol Med, vol 34 1 601-603, 1936. W.Thorner, Pflugers Arch, vol 156, 1913. T.Thunberg, Zentralb f.Physiol, vol 18 1 553, 1904. H.Winterstein, Naturwiss, vol 19, 247-250, 1931. II.Winterstein, Pflugers Arch, vol 224, 749-759, 1930. II.Winterstein, Science, vol 71, 641-643, 1930. II.Winterstein, B1ochem Zeitschr, vol 232, 196, 1931. NOTE Adrian (1922.,1923)* has suggested that a narcotized nere.re conducts with a decrement.This suggestion was based on the results of a series of experiments in which varying lengths of nerve were subjected to a narcotic.An impulse excited by a strong stimulus was observed to traverse a longer narcotized length of nerve than the impulse from a weaker stimulus.The impulse excited by the strong stimulus was larger at its origin than the one initiated by the weak stimulus.It was concluded that the impulse amplitude became gradually less as it traversed t he narcotized region,and that due to its larger size at the start.,the impulse from the strong stimulus was still existant thou gh very small when it arrived at . the normal region beyond the narcotized area.The weak impulse died out completely somewhere in the narcotized area. Cooper based her claim for decremental conduction in asphyxiated nerve on this same principle.The original size of the impulse excited by a stimulus following the least interval became smaller as the interval increased with asphyxiation of the nerve.In traversing the asphyxiated region,this small impulse died out.on the other hand,the impulse stimulated after the recovery time remained large enough at the start to be able to travel through the asphyxiated region.However,since the least interval,and eventually the recovery time,did increase during anoxia.,it must be concluded that the refractory period of the nerve increased during asphyxiation.This is contrary to the claim of Cooper. * E . D.Adrian and A.Forbes, Journ Physiol, vol 56, 1922. E . D.Adrian, ibid, vol 57, 1923.