NEURAL ASPECTS OF TEMPERATURE REGULATION , Editors JOHN P. HANNON ELEANOR VIERECK — ^^■^^ . 1 cr o -D - ^^^^^ 5 m Xi -D |^= □ — igs^sa o gi r^ g^^sf a — ^— m ^^= a □ ^^= sss^s ARCTIC AEROMEDICAL LABORATORY FORT WAINWRIGHT ALASKA 1961 NEURAL ASPECTS OF TEMPERATURE REGULATION /j/v^ _____ ■ \__ r :.; ^^^^^ " "»....-... == ■ ■ cr . Editors JOHN P. HANNON HOI ill III b37 ELEANOR VIERECK g— CD '^ ° CD™^^^*™ 1,^— m ^B° i o ^5 ARCTIC AEROMEDICAL LABORATORY FORT WAINWRIGHT ALASKA 1961 rs Acknowledgements Grateful acknowledgement is made to the following for permission to publish copyrighted figures: University of Chicago Press for figures reprinted from Physiological Zoology, the American Physio logical Society for figures from Journal of Applied Physiology and American Journal of Physiology, and to the editors of the Journal of Neurophys iology , Biological Bulletin , Journal of General Physio logy, Journal of Cellular and Comparative Physiology, and Science ARCTIC BIOLOGY AND MEDICINE Transactions of the First Symposium Neural Aspects of Temperature Regulation October 10, 11, and 12, 1960 Edited by Eleanor G. Viereck, Ph. D. Assistant Professor Biology Department University of Alaska Sponsored by the ARCTIC AEROMEDICAL LABORATORY FORT WAINWRIGHT ALASKA PROCEEDINGS SYMPOSIA ON ARCTIC BIOLOGY AND MEDICINE I. NEURAL ASPECTS OF TEMPERATURE REGULATION Symposium held at Arctic Aeromedical Laboratory Fort Wainwright, Alaska October 9, 10, 11, 1960 Conducted under the auspices of the Geophysical Institute University of Alaska in accordance with Air Force Contract AF 41(657) — 356 Proceedings edited by John P. Hannon Chief, Department of Physiology Arctic Aeromedical Laboratory Eleanor Viereck Assistant Professor of Biobgy University of Alaska TABLE OF CONTENTS Hemingway, Allan The neural control of the physiological responses to cold. Historical review 1 Hensel, H. Recent advances in thermoreceptor physiology 37 Lim, T.P.K. Central and peripheral mechanisms in temperature regulation. 79 Roddie, Ian C. The role of vasoconstrictor and vasodilator nerves in the control of the peripheral circulation 113 Clark, George Heat dissipation functions following experimental anterior hypothalamic lesions in cats 149 Kawamura, Yojiro Neuro- muscular organization of shivering 193 Chatonnet, J. Nervous pathways in the chemical regulation against cold and their central control 225 Freeman, Walter J . A critique of the doctrine of centers in temperature regulation 255 Stuart, Douglas G. Role of the prosencephalon in shivering 295 82383 AUTHORS J. Chatonnet Dr. J. Chatonnet received the Doctorate of Medicine and the License of Sciences from the University of Lyon in 1943-44. In 1952 he received the academic degree, Agregation. He is now "Professeur sans chaire" or Associate Professor of Physiology at "Faculte' Meciecine de I'Universite de Lyon". Whereas his earlier fields of research concerned the autonomic nervous system, his interest is now focused on the respective role of shivering and non- shivering thermogenesis and the central control of body temperature. George Clark Dr. George Clark received his Ph. D. from Northwestern University in 1939. At that time, and again in recent years, his research work has been devoted almost exclusively to the field of temperature regulation and various aspects of the neural control of body temperature. He is now research anatomist at the US Army Medical Research Laboratory, Fort Knox, Kentucky. Walter J. Freeman Dr. Walter J. Freeman obtained his M. D. at Yale, 1954; sub- sequently he has pursued post-graduate work at Yale in pathology, at Johns Hopkins in internal medicine, and at UCLA in neuro- physiology. His major research work has been on the anatomy, histopathology, and physiology of the brain. Presently he is a Fellow, U. S. P. H. S., and Foundation Fund for Research in Psychiatry. Allan Hemingway Dr. Allan Hemingway received his Ph. D. in 1929 from the University of Minnesota. His research has been in the physiology of body temperature regulation of man and warm-blooded animals and in respiration. His early work consisted of standardizing electro- therapeutic treatments such as heat treatments with diathermy. He is now studying the nervous control of shivering, the problem of blood flow through the lungs, and the duration of the effect of move- ments in respiration on blood flow through the lungs. Dr. Hemingway and a colleague in pharmacology, Dr. P. K. Smith of George Wash- ington University, discovered that the drug hyoscine prevents motion sickness. Dr. Hemingway is now professor of physiology at the UCLA Medical School. ;„ H. Hensel Dr. H. Hensel obtained the M. D. degree from the University of Heidelberg. The subject of his major research work has been tem- perature regulation, thermoreceptors, general sense physiology, and peripheral circulation. His presentpositionisProfessor of Physiol- ogy, Director, Dept. of Physiology, University of Marburg/Lahn, Germany. Yojiro Kawamura Dr. Yojiro Kawamura obtained his M.D. at Osaka University in 1945, whereupon he continued to complete the post doctoral course at Osaka University, obtaining a D. M. S. degree in 1952. His major research interest is neurophysiology. Currently he is Professor and Head, Department of Physiology, Dental School, Osaka University, Osaka, Japan, and also has the title, Lecturer in Physiology, School of Medicine, Osaka University. Thomas P. K. Lim Dr. Thomas P. K. Lim was graduated from the Severance Union Medical College, Seoul, Korea, in 1948 with a degree of M. D. At Northwestern University he was awarded M. S. and Ph. D. degrees in physiology in 1951 and 1953 respectively. He has been engaged in research in the fields of respiration, circulation, and body tem- perature regulation since this time. Currently Dr. Lim is the Director of the Cardiopulmonary Laboratory at the Tucson Medical Center, Tucson, Arizona. Ian C . Roddie Dr. Ian C. Roddie received the M.D. at the Queen's University of Belfast, North Ireland, in 1953. He has returned there, following a year of internship at the Royal Victoria Hospital, Belfast, to work as a Walter Dixon Memorial Scholar, Beit Memorial Research Fellow, and Lecturer in Physiology. Currently he has leave of absence from Belfast to work in the Department of Physiology and Biophysics at the University of Washington, Seattle, as Harkness Fellow of the Commonwealth Fund of New York. His main research interest has been in the regulation of the peripheral circulation in man. Douglas G. Stuart Douglas G. Stuart is receiving his Ph. D. degree from the University of California at Los Angeles in 1961. His major research interests are the neurogenesis of normal and abnormal somatomotor activity and the physiology of telencephalo-hypothalamic relations. Currently he is a Bank of America - Giannini Foundation Fellow at the Department of Anatomy, School of Medicine, UCLA, and also is employed at the Veterans Administration Hospital, Long Beach, California. DISCUSSANTS Mr. Thomas Adams University of Washington School of Medicine Department of Physiology and Biophysics Seattle 5, Washington Mr. Charles Eagan Arctic Aeromedical Laboratory Fort Wainwright, Alaska Dr. J. P. Hannon Arctic Aeromedical Laboratory Fort Wainwright, Alaska Dr. Kjell Johansen University of Washington Department of Zoology Seattle 5, Washington Capt. David Minard Thermal Stress Division Naval Medical Research Institute National Naval Medical Center Bethesda 14, Maryland Col. Joseph M. Quashnock Arctic Aeromedical Laboratory Fort Wainwright, Alaska THE NEURAL CONTROL OF THE PHYSIOLOGICAL RESPONSES TO COLD HISTORICAL REVIEW Allan Hemingway Professor of Physiology University of California School of Medici Los Angeles 24, California In the animal kingdom, only birds and mammals have the abil- ity to maintain rectal temperature within a narrow range in spite of wide changes of environmental temperature. This property of homeothermy is not possessed by the lower vertebrates or any of the invertebrates. The homeotherms have physiological responses to cold and warmth under control of the central nervous system which are absent in the poikilotherms. In analyzing the physiological responses to cold and heat which the homeotherms possess, the following physiological mechan- isms are observed: For protection against heat, the homeotherms can (1) dilate their cutaneous vessels and raise their skin tempera- ture to facilitate heat loss and (2) increase evaporative cooling by sweating or panting. For protection against cold, the homeotherms can increase their metabolic heat by (1) shivering and (2) nonshiv- ering thermogenesis and can (3) reduce heat loss to the environ- ment by cutaneous vasoconstriction. The control of body tempera- tures by cutaneous vasoconstriction and vasodilation is a physio- logical mechanism functiong continuously throughout the day and throughout life. It is this mechanism which maintains rectal temperature nearly constant in a comfortable environment. Shiv- ering, panting, and sweating are emergency functions which only become active during cold or heat stress. This system of physiological mechanisms which functions to maintain homeothermy is under the control of the central nervous system which functions with remarkable efficiency as a thermo- stat to activate, inhibit, and regulate the component functional physiological processes of homeothermy. A survey will now be given of the experimental data which in recent years has served to establish our current concepts of how the central nervous system functions to protect against cold. HEMINGWAY, A. For protection against cold, the homeothermic animal can increase its heat production by shivering to a value as much as four to five times the resting heat production. In a recent paper where shivering was studied on man, lampietro, Vaughan, Goldman, Masucci, and Bass (1960) found the maximal shivering heat pro- duction rate to be 442 Cal/hr, and with a resting nonshivering heat production rate of 80, the heat production rate was increased to 5.5 times the resting nonshivering value. This value is, however, maximal and usually the heat production rate observed during shivering is two to four times the nonshivering value. In fact, the regulation of shivering is so well adjusted to temperature regu- lation demands that the shivering can be mild, moderate, or intense as required. Nonshivering thermogenesis, the term applied to an elevation of heat production caused by cold without shivering, can elevate the oxygen consumption rate from 10 to 20 percent in the larger animals and possibly more in the small mammals and birds. The other physiological protection against cold, the thermal cutan- eous vasomotor response, is the nervous control of the diameter of the blood vessels in the skin. Contraction of the smooth muscle of the cutaneous arterioles causes vasoconstriction and reduces blood flow through the skin. The skin temperature is lowered and heat loss from the skin is reduced. Methods of Testing Physiological Responses to Cold Before the details of the neural control of the physiological responses to cold are described, the methods for testing these responses will be reviewed briefly. The following methods are those which have been used by investigators for testing the functional activity of the various components of the temperature regulating system . Cold exposure test (test for homeothermy). An animal is placed in a cold environment, usually a cold room having a temperature of from 0 C to 15 C, and its rectal temperature measured at HISTORICAL REVIEW intervals in a 1- to 4-hour test. The presence or absence of shiv- ering is determined visually or by palpation. A normal animal will maintain a rectal temperature within narrow limits of 0.5 to 1.0 C, whereas a poikilotherm will have a continuously falling rectal temperature. Experiments which impair temperature regu- lation will have a temperature-time record intermediate between these extremes. Measurement of oxygen consumption rate. This measurement is perhaps the most useful and quantitative method of measuring shivering and nonshivering thermogenesis. An animal tested for shivering is first placed in a comfortable environment of 25 to 30 C, and after a time interval sufficient to achieve a steady state, its oxygen consumption rate (designated as VO ) is meas- ured. The oxygen consumption rate is usually expressed as milli- liters O STPD per kilogram of body weight per minute. The animal is next placed in a cold environment of 0 to 10 C, and the oxygen consumption rate continuously measured; the oxygen con- sumption rate will rise and finally reach a fluctuating steady state value. The fluctuation is caused by the irregularity in shivering. The ratio VO (cold environment)/ VO (comfortable environment) is a usetul index of the intensity of shivering. Oxygen consumption rate can be measured by one of several methods. A convenient method used in our laboratory has been to measure continuously the O content and the CO content of the gas within a sealed box in which the animal is placed. Carbon dioxide and water are re- moved by circulation of the gas through absorbing columns. Mechanical record of shivering. Since a shivering animal pro- duces a well-defined tremor, the presence of shivering can be detected and can be roughly measured quantitatively by measuring the vibration of shivering. A resting animal is placed on a platform which is suspended by wires or supported on springs. The shiv- ering causes a vibration of the platform and can be measured with a piezo-electric crystal, a sensitive pressure transducer, a strain gauge, or more simply with a thread which has one end attached to the platform and the other end to a lever arranged for kymograph recording. With kymograph recording, however, the frequency characteristics of the system are likely to be poor. HEMINGWAY, A. Electromyogram. The shivering muscles, which are under- going contraction and relaxation, produce action potentials which can be amplified and recorded on kymograph paper. The electro- myograms as recorded with an AC amplifier consist of a suc- cession of biphasic spike potentials, irregular in frequency and amplitude. Using a suitable rectifier, these biphasic potentials can be converted into an integrated monophasic record whose ampli- tude is roughly proportional to the intensity of shivering. The thermal cutaneous vasomotor response. The ability of the cutaneous vessels to constrict in the cold and dilate in the heat can be measured conveniently in dogs by measuring their ear skin temperature. In a cool environment of 15 to 20 C, cutaneous vaso- constriction is indicatedby a low ear skin temperature of 2 to 3 C, above environmental temperature. Vasodilation, in this environment, caused by central or peripheral heating, is indicated by a rise in the ear temperature which will reach values as high as 34 C. This test is simply an indicator of cutaneous vasoconstriction and vasodilation, and is roughly quantitative. Other methods of measuring cutaneous blood flow are the (a) photoelectric plethysmograph (Hertzman, et al., 1946); (b) the impedance plethysmograph (Nyboer, 1960); (c) the venous occlu- sion plethysmograph (Abramson, 1944, and Freeman, 1945); and (d) the opening of a cutaneous vein and measuring the blood flow rate of the blood lost by hemorrhage. The photoelectric plethysmograph and the impedance plethysmograph cause negligible discomfort to the animal or human subject, but the measurements are only approx- imate. The venous occlusion plethysmograph, while it gives quant- itative values of blood flow rate, does not measure cutaneous blood flow alone, but the blood flow of the extremity. The assumption is required that the venous occlusion obstructs all of the venous out- flow from the extremity and does not interfere with arterial in- flow. The measurement of the cutaneous blood flow by hemorrhage is suitable only for animals under anesthesia. HISTORICAL REVIEW Apparatus Used for Testing Cold Responses of Animals In studying the responses of homeothermic animals to cold, the animals used for testing have been mainly men, dogs, cats, and rats. The advantages and limitations of using men are well known. Of the various animals available for stucfy, there are certain anatomical and physiological characteristics of dogs, cats, and rats which make them useful for special purposes. Dogs and cats have a well-dev- eloped shivering mechanism which is similar in all respects to that of man. The presently well-known control for starting, stopping, and regulating shivering in these animals is identical with that of man. This is not true for the physiological response to heat. Where metabolic, circulatory, and respiratory investigations are being made and where adequate samples of blood for analysis are needed, the dog is the animal of choice. In addition, the dog can be trained to lie motionless on a table or platform without restraint. This is a most important characteristic where studies are made of mild shivering, or in attempts to separate the metabolism of shivering from nonshivering thermogenesis. The cat is particularly valuable where studies of the nervous system are required. The neuroana- tomy and neurophysiology of the cat are probably as well, if not better, understood than that of man. This is a result of the popular usage of the cat for studies of the nervous system because of the uniformity of the shape of the head of the cat for stereotaxis studies, and the ease, convenience, and cost of the cat in handling and main- tenance. For such considerations, the monkey, although useful, is much less popular. The rat, due to its omnivorous diet, is use- ful for nutritional studies. It shivers well in the cold and is a fairly good homeotherm. Apparently, however, the rat differs in its response to cold from the larger animals, relying on nonshivering thermogenesis. The apparatus which has been found the most useful for stucfying shivering in the dog (Hemingway and Hathaway, 1941) is shown in Figure 1. The trained dog lies on his side in a double-walled box with either cold or hot water being circulated through the space between the walls. A glass window inthe cover permits observation of the animal. The shoulders and chest of the dog are supported by a fixed platform, while the hind legs rest on a platform suspended HEMINGWAY, A. ? ''i; fell'*; HISTORICAL REVIEW by steel wires but hinged to the fixed platform. When shivering starts, the movable platform vibrates, andthe vibration is recorded to give a mechanical record of the tremor of shivering. The head of the dog is sealed within a chamber through which gas of the closed respiratory system is circulated with an airtight pump (re- frigerator pump). Respiration and oxygen consumption rates are recorded from a spirometer of the closed circuit system. Thermo- couples are used to record skin and rectal temperatures. Electrodes sealed to the skin with collodion areusedfor recording of the elec- tromyogram. Trained animals will lie for as long as 3 hours in this apparatus and endure cold sufficient to produce shivering without attempting to move. A record of the start of shivering as revealed by the record of rectified action potentials is seen in Figure 2. Before shivering, there was slight background electrical activity due to respiration and the heart beat. When shivering started there was a sudden rise in the rectified action potentials which was coin- cident with the appearance of visible shivering. For this animal, there was no increase of electromyographic activity preceding visible shivering. The apparatus used in studying shivering in cats is shown in Figure 3. A cat is not capable of being trained to lie in a relaxed resting condition such as is a dog. A carefully selected cat will, however, sit with little voluntary movement for an hour or two in a confined box such as is shown in Figure 3. The cat sits on a platform supported by springs or rubber stoppers and shivering is recorded by a strain gauge attached to a movable platform. The box which contains the cat is sealed airtight and the gas within the box is circulated by means of a sealed pump. The box having a glass cover is placed within a refrigerator, also with a glass cover, so that the cat can be observed. The gas of the enclosure is circulated through cooling coils immersed in ice water in the refrigerator. Continuous analyses are made of the O and CO content of the enclosed gas. This apparatus permits a measurement of oxygen consumption rate, but not a measurement of respiration. Thermo- couples and electromyographic electrodes record temperatures and muscle action potentials respectively. HEMINGWAY, A. g S o ^ ? B H {J- S' § jtS- '^ 2 fT 5 3 S — n> 3 - a 2 s" ^ ■§• ? 3 ■§ s :;■ a ^ I £ ^"^ S- o g S 3 g e. H3 2- a 2. 3 I HISTORICAL REVIEW Current Concept of Nervous Control of Temperature Regulation The current concept of the nervous control of temperature regulation is that the control is effected through a modified reflex system, somewhat similar to the control of respiration. This modified reflex system consists of afferent pathways, a center, and efferent pathways to peripheral effector organs with, at least for shivering, a "feed-back" pathway for control of shivering rhythm. Afferent system. Afferent impulses from cutaneous thermal receptors, to be described later at this Symposium by Professor Hensel, travel via afferent nerves through the dorsal roots, the spinothalamic tracts, and the fifth cranial nerve to reach the thala- mus. From the thalamus, which functions as a distributory center for afferent impulses, the impulses are relayed to the cerebral cortex and hypothalamus. This theory of the thalamus functioning as a distributory center may not be strictly valid since animals with a larger part of the thalamus destroyed regulate their body temperatures efficiently. This is the so called "peripheral" control of body temperature, and requires further stucfy. There is considerable evidence from experiments on brain cooling and heating that the temperature regulating activities can be motivated or suppressed by temperature changes of the brain without any temperature change of the skin. This leads to the con- clusion that there is a "central" or "central thermostatic" control- ling system. Temperature regulating center. The term "center"' for such activities as respiration, vasomotor activity, or temperature regu- lation has been justly criticized as a term which has a vague meaning and uncertain anatomical localization. Nevertheless, the term serves a useful purpose in the original sense as used by Sherrington when applied to a region in the brain which serves to integrate, coordinate, and regulate the afferent influences for moti- vation, inhibition, and regulation of the motor effects. An anatomical center may subserve more than one physiological function. The term "center" will be used here for want of abetter substitute. HEMINGWAY, A. 10 HISTORICAL REVIEW glass refrigerator cooler t connpres.jI . (cooling coils I J not shown) window ^^ REFRIGERATOf EMG thermocouple recorder vibration recorder copper tubing ^ ra Figure 3. Apparatus for stucfying shivering in cats. 11 HEMINGWAY, A. All evidence indicates that the region of the brain most effec- tive and important for temperature regulation is in the hypothalamus, and that this part of the brain must be functional and connected through nervous pathways with the lower motor centers. However, temperature regulation can be influenced by activities of the cere- bellum, the cerebrum, and the midbrain. Efferent system. The efferent system for temperature regu- lation involves several pathways. Shivering pathways essential for effective shivering travel downward through the midbrain and medul- la tothe lower motor neurons. A possible pathway for these impulses will be described later. Panting involves pathways of the respiratory system. Cutaneous vasomotor activity and sweating are mediated by fibers of the automatic nervous system. Feedback. The riiythm of shivering is controlled by proprio- ceptive activity (Perkins, 1945; Lippold, Redfeam, and Vuco, 1958). The level in the central nervous system at which this feedback control is integrated is as yet undetermined. Experimental Evidence This concept of the neuralmechanisms controllingtemperature regulation is based on a large amount of experimental evidence accumulated mainly in the past 50 years. Some of the experimental data which have contributed to our understanding of how the nervous system controls temperature regulation, especially shivering, will now be reviewed. The experimental methods used by neurophysiologists in studying functional activity of the nervous system have consisted of the following four procedures: (1) Studying physiological response to cold after transections, (2) study ingles ions of the central nervous system, (3) studying the responses evoked by electrical stimulation of various points ofthe brain, and (4) recording of electrical activity associated with a particular physiological activity. Examples of these procedures with their role in temperature regulation will now be given. 12 HISTORICAL REVIEW Transections of the central nervous system. In locating centers within the central nervous system which are essential for function, one of the oldest procedures has been the central nervous system transection. Transections can be made at all levels in the central nervous system, excluding regions above or below the lesion from functional activity. Transections between the fourth cervical level and the lower medulla cannot be made without artificial respiration because of the failure of spontaneous respiration. In the classical work of Sherrington (1923-24), spinal cord transections were made in dogs and the animals were studied long after spinal shock had subsided. Sherrington found that shivering and cutaneous vasocon- striction failed in the cold in the structures innervated from the spinal cord below the lesion. When the cerebral cortex is removed by a brain transection just below the thalamus and the animals allowed to recover from the operation, there is no appreciable reduction in shivering. These observations, first made by Dusser de Barenne in 1920, have been confirmed by Pinkston, Bard, and Rioch (1934) and more recently by Birzis and Stuart at the University of Calif ornia. Bard and Rioch (1937) tound that ablation of the cerebral cortex resulted in a warm skin in a cool environment with more rapid onset of severe shiv- ering. This cutaneous vasodilation in a cool environment occurred after removal of the ansate cortex. In spite of a warm skin, the piloerection occurred. Whereas a brain transection removing the cerebral cortex from the brain stem has only a slight effect on temperature regu- lation against cold, a midbrain transection made between the dien- cephalon and mesencephalon (the decerebrate preparation) has a profound effect on temperature regulation. The decerebrate prep- aration is (for all practical purposes) poikilothermic, as shown by experiments of Bazett and Penfield (1922), Keller and Hare (1932), and Bard and Macht (1958). However, a few 'investigators jiave reported a sligjit residual tremor which can occur after transection in muscles innervated from levels of the central nervous system below the lesion. Dworkin (1930) found that shivering which was insufficient to prevent a fall in the rectal temperature of rabbits in the cold, occurred in medullary and midbrain preparations. Thauer and Peters (1937a and 1937b), using rabbits, found that after 13 HEMINGWAY, A. a 5-day recovery period from a trauma of surgery, there is an increase in the ability of animals to resist cold. These observations while somewhat limited in scope, and contradictory to the majority of the results of others, would lead to the rather important general- ization that a controlling mechanismfor defense against cold occurs in the brain stem below the hypothalamus. In order to investigate this question further and to extend the work to other animals, a study of chronic decerebrate cats has been made in our laboratory. In these animals, the entire forebrain above a plane extending from the superior colliculi to the mammillary bodies is isolated from the brain stem by transection followed by removal of a 2 mm wedge of midbrain tissue. Considerable nursing care is required for these animals for the preventionof hypo- andhyperthemiaand cage sores. Hand feeding is required. However, with this time-consuming effort, the animals can be maintained for periods of several days to several months. These animals have been studied carefully by the methods already described. It has been found that in two of nine animals, a tremor only remotely resembling shivering can be induced by cold exposure and stopped by warmth. However, this tremor is valueless in protecting an animal against cold. The tremor does not raise the oxygen consumption rate and does not prevent a fall in rectal temperature in the cold. Professor Philip Bard of Johns Hopkins University, who has made extensive studies of the chronic decere- brate cat and has been interested particularly in the temperature regulation impairment caused by decerebration, has also observed this cold-induced tremor in the chronic midbrain preparation (1958). The occurrence of the cold-induced tremor in the chronic decere- brate cat, while it serves no useful purpose in temperature regu- lation, does raise the question of the extent in the central nervous system of the controlling mechanism for body temperature regu- lation. Lesions of the central nervous system. The transection tech- nique for separating completely the central nervous system into two unequal parts with no intracerebral or intraspinal nervous connection between the two parts has served a useful purpose in anatomical localization of the level in the central nervous system for control 14 HISTORICAL REVIEW of a particular function. However, the site of injury is extensive and excessive loss of function occurs. In recent years, most of the investigative work seeking to understand the extent and activity of various regions of the central nervous system has consisted of making small discrete lesions and testing the animal before and after the operation to determine loss of function caused by the lesion. In the older work, the puncture technique was used. This con- sisted of first exposing the surface of the brain, inserting a probe or small scalpel, and destroying mechanically from the surface inward, a region adjacent to tne surface. This was the method used in the classical work of Isenschmid (1912, 1914), whose work first definitely localized within the hypothalamus, the control for temper- ature regulation. The anatomical site of lesions in the brains of rabbits produced by the puncture technique of Isenschmid is shown in Figure 4. The stippledareas of the brain cross-sections represent the regions destroyed by the puncture. The clear unstippled areas represent intact brain. Below each of the four sections of Figure 4, the interference with homeothermic function caused by the lesion is indicated. In the section designated "abolished", there was no temperature regulatory function - the animal was poikilothermic. In the section labeled "undisturbed", temperature regulation was normal. If one superimposes the "undisturbed" section in the lower left-hand corner over the "abolished" section of the upper right- hand corner (since both of these sections are at the same level), it will be found that a small region in a transverse plane at the caudal border of the mammillary bodies, dorsolateral to the mam- millary bodies and at a position laterally about one-fourth of the dis- tance from the wall of the third ventricle to the lateral surface, must be intact and have neural connections with more caudal portions of the central nervous system for maintenance of homeothermy. The role of this particular region in temperature regulation will be discussed later in the Symposium by Douglas Stuart. In more recent years, a similar technique has been used by the neurophy Bi- ologists of the Army Medical Research Laboratory at Fort Knox, particularly Keller and Batsel (1952) and Keller (1956). Using elec- trocoagulation, a region of the brain was destroyed which resulted in converting a dog to a poikilothermic state (or nearly poikilo- thermic), producing a preparation called the "poikilothermic dog'". The region destroyed was at the junction of the mesencephalon and diencephalon and included the region found earlier by Isenschmid to 15 HEMINGWAY, A. BODY TEMPERATURE REGULATION ALTERATIONS 2-10 da/s post probe lesions (Isenschmid - 1914) Habenula Jnfundibjlum Hirnschtnlijl raclus pluus UNDISTURBED I'entnkel CoW: Corpus mamilla MILDLY DISTURBED Figure 4. Effect on temperature regulation of lesions in the brain of a rabbit. The part of the brain destroyed by the puncture is indicated by the stippled area. The lesions in the two left-hand sections left temperature regulation undisturbed as determined by a cold exposure test. The lesion represented in the upper right- hand figure abolished temperature regulation, making the animal poikilothermic. Data of Isenschmid and Kiehl (1912), 16 HISTORICAL REVIEW be essential for homeothermy. However, the Fort Knox workers, in contrast with earlier workers, were able to keep their animals alive for many months and make homeothermy tests long after the traumatic shock of the operation had subsided. The most popular technique for producing localized lesions in the central nervous system has beenthe electrolytic lesion produced at the tip of electrodes placed in the brain of anesthetized animals from the Horsley- Clark head frame. With this procedure, it is possible to make a lesion of a desired size at any desired location within the brain. The cat is the animal of choice for these studies due to the uniformity of the size and shape of the head. The first study of impairment of temperature regulation caused by these dis- crete lesion was made by the Ranson school of investigators. The earlier work of this group has been reviewed by Ranson in his clas- sical 1940 paper. Examples of the type of lesion made by this method from work of Birzis and Hemingway are shown in Figure 5. In this figure, the regions of the brain destroyed are small areas in the lateral pons with the level of the section being shown in the upper half of the figure. This lesion abolished shivering in a cat. The advantage of this method of producing a lesion over the transection and puncture technique is that regions remote from the brain sur- face can be destroyed with little injury to surface structures. Using the Horsley- Clark method of producing bilateral lesions at different levels in the brain stem of the anesthetized cat, Birzis and Hemingway (1956) have found that a region of the brain stem essential for shivering comprises a long pathway extending from the posterior hypothalamus to the medulla. This pathway starts downward from a position just dorsal to the junction between the medial edge of the cerebral peduncle and the lateral border of the mammillary bodies. As it descendsthroughthe midbrain, it courses dorsolaterally to the red nucleus, and in the pons it deviates later- ally, reaching a superficial position immediately adjacent to the transverse pontine fibers. It retains its lateral position in the upper medulla. This pathway through the brain stem is shown in Figure 6. The pathway consists of the region within the elliptical figure drawn on the sections. It will be noted that this pathway does not involve the corticospinal tracts. Figure 7 shows the site of lesions which destroyed the corticospinal tracts and adjacent areas of the 17 HEMINGWAY, A. ^ mm f Figure 5. Electrolytic lesions which abolished shivering in the brain of a cat. The lesions are laterally located in the pons. The electrodes were inserted from a Horsley-Clark stereotaxic frame. Blrris and Hemingway (1956). 18 HISTORICAL REVIEW * ^^J^M Figure 6. Site of lesions at various levels in the brainstem of cats, extending from hypothalamus to medulla oblongata, which abolished shivering. The site of each lesion is encircled. HEMINGWAY. A. 5- S •9. g •--• 0) 1 I ~w S' Oi ro — < 1 3 1 1 1 m 1 '->! OJ o a» z 1 I s X 3 3J CO 1 z JO X 1 0) O. C 3 :x} z m a o -< > > m CO o -< CO o 20 HISTORICAL REVIEW medulla of six cats. Shivering persisted in these animals after the lesion. It would be an attractive hypothesis to propose that the path- way described, which was found to be intact for shivering to occur, was the efferent pathway from the hypothalamus to the lower motor center for control of shivering. However, this hypothesis must be proposed with caution and final decision awaited for confirmatory evidence, preferably using unanesthetized animals. Electrical stimulation. A method currently used in electro- physiology for determining the site of control of a particular func- tion is to stimulate a point within the central nervous system and to observe the physiological response to stimulation. This technique has been usedeffectively in studies of respiration to determine loca- tion of the inspiratory and expiratory centers in the central nervous system. If stimulation of a point in the medulla oblongata invokes inspiration, then the point stimulated can be assumed to be located in the inspiratory center. This technique has been used by a number of observers in studying shivering. Akert and Kesselring (1951) in reviewing the extensive diencephalic stimulation data from the laboratory of Hess found a number of instances where stimulation of regions in and adjacent to the hypothalamus produced a tremor resembling shivering. The tremor was produced by low frequency stimulation of eleven points in eight cats. Six of these points were in the septum pellucidum, three were in the caudate nucleus, one in the thalamus, and one in the posterior hypothalamus. Birzis and Hemingway (1957) were able to produce a tremor similar to shiv- ering when nineteen points in the midbrain of four cats were stimu- lated. These points were on the pathway described previously which was established by lesion experiments; the region dorsolateral to the red nucleus and a region in the lateral pons where stimulation evoked shivering is shown in Figure 8. From the record below the figure, the onset of shivering after stimulation is shown. This is a continuous record of muscle electrical activity (electromyogram)in two muscle groups. After stimulation of apoint near the red nucleus, there was a brief latent period followed by shivering. With cessation of the stimulus shivering subsided during a brief after-discharge period of approximately five seconds. Interest in the septum pellu- cidum as a controlling center for shivering was again aroused by the work of Andersson (1957). He, like Akert and Kesselring, was able to induce shivering by stimulation of points in the septum pellucidum of goats. The role of the septum pellucidum in shivering 21 HEMINGWAY. A. 3-' ^'^ S "8 I 01 g p p CO o S, o ?5 3 I O P 3 < oq c- e 5- §■- Kb 55. n =- ^ \ ^ : . s "5 - \ ;" - - - ' - - - - ^ _ - 4 ; . ^ \ ■■ - ^ r : - i [ : : '-- ^ < - - - ro 1 J : _ - ^ Ul -- \ ^ CO ^ ^ m o ; =" ': "■ ; - - - . ^ \ - : ; '-- I \ \ ; \ cc Z2 HISTORICAL REVIEW and its possible relation with the posterior hypothalamus will be discussed in the work of Stuart later in this Symposium, Shivering cannot only be induced by electrical stimulation of points in the brain but it can be suppressed, if it has started spontaneously by cold, by stimulation of other points. Kaada (1951) was able to inhibit shivering by electrical stimulation of points on the cerebral cortex while Hemingway, Forgrave, and Birz is (1954) found that spontaneous shivering could be stopped or attentuated by electrical stimulation of points widely distributed in the hypothalamus. A particularly active suppression region as determined by low stimulus threshhold was found in the pre-optic region. Electrical stimulationof various points in the brain has revealed that shivering can be started or, it preceding spontaneously, stopped by electrical stimulation of points within the brain. The points for shivering suppression are more widely distributed than the points for shivering stimulation. The most active point for suppression is found in the pre-optic region, while the most active point for stimu- lation, as Stuart will report later, isinthe posterior hypothalamus. Other points of higher threshold for both suppression and stimulation can be found throughout the brain. At the present time, only a tentative "working" hypothesis of the control of shivering in the cat can be made resulting from inter- pretation of the stimulation experiments. The active region in the posterior hypothalamus for stimulation may be a primary center for starting and control of shivering, but its function may be under sub- sidiary influence of secondary centers, particularly in, or rostral to, the feline septum . The widely distributed suppression effects may be interpreted as due to the existence of an inhibitory mechanism for shivering. Shivering, if not too intense, can be suppressed volun- tarily, which may involve the cortical suppression control of Kaada. Shivering is also inhibited by voluntary movement. When a voluntary movement is made by a limb, shivering ceases. It seems that in the control of the skeletal musculature, shivering is a function of oecon- dary importance to locomotion. This is reasonable in consideration of animal safety and maintenance that first priority should be as- signed to movement for flight, fight, and defense. When the skeletal musculature is needed for these life-saving functions, shivering can be temporarily suspended in favor of emergency action. This 23 HEMINGWAY, A. would serve to explain the existence of a widely distributed shiv- ering suppression mechanism which can be immediately initiated when needed. Recording of electrical activity. A new tool in electrophysiol- ogy which has been used rather extensively in the last 10 years is the use of microelectrodes and semi-microelectrodes for recor- ding of electrical activity of the cells of the central nervous system. These electrodes have a diameter of from 0.2/i to 50m and are, in diameter, of the same order of magnitude as nerve cells and their processes, i.e., 0.5iJ. to 20/i. When placed in contact with a nerve cell (or nerve fiber) which is actively functioning, the electrical of the cell is transmitted to the electrode (Freeman and Hemingway, 1959). The electrical potential of the active cell consists of a series of rapid transient potentials called "spike" potentials. These poten- tials are amplified and transmitted to a cathode ray oscilloscope and photographed. Birzis and Hemingway (1957) were the first to record well defined spike potentials associated with shivering. These potentials appeared with shivering and disappeared with cessation of shivering. Records of these potentials are shown in Figure 9. These potentials were recorded from electrodes placed on the "shivering pathway" whose boundaries were determined by lesion experiments as described earlier in this paper. A semi- transection made just caudal to the point from which the potentials were recorded did not prevent or change these spike potentials (see Figure 10). This is evidence that the potentials were action potentials of fibers carrying shivering impulses downward in the central nervous system to lower motor neurons, that is, they were part of the efferent fiber system controlling shivering. These spike potentials associated with shivering were extensively studied by Freeman and Hemingway (1959), who found that the action potentials could be recorded from the "shivering pathway" previously des- cribed and extended from the fields of Forel in the posterior hypo- thalamus to the olive in the medulla oblongata. By using a high amplification and rapid sweep of the cathode ray oscilloscope, a characteristic change of potential pattern was discovered in prece- ding from the fields of Forel to the olive, which permitted a tentative interpretation of these potentials (based on a theory proposed by Lorente de No) as spike potentials travelling downward in the brain. There were, however, two disturbing factors of this study. The impulses seemed to arise from the fields of Forel and travel down- 24 HISTORICAL REVIEW lO m — id * 1 CM Zi < s 1 3 £ .5 la ^ a H be c 1 1 i 1 § 2 ■a "S to ■S ;5 5 S B 5 > E a ^ -a ■g S :3 £ i •a ^ 25 HEMINGWAY, A. electrode: _ Figure 10. The site of theelectrodefrom which shivering spike potentials appeared. The pothentials were obtained after the hemi- section and saggital section shown bv the dotted line. 26 HISTORICAL REVIEW ward. If these impulses were the efferent action potentials con- trolling shivering which arose in the fields of Forel, then destruction of the fields of Forel should abolish shivering. In a study of small lesions made bilaterally to destroy the fields of Forel, shivering persisted (Stuart, Freeman, and Hemingway, 1959). Another dis- turbing feature was that the duration of the spike potentials was not exactly the same asthedurationof shivering. The potentials seemed to persist after shivering had ceased. These observations reveal that any interpretation that the large well defined spike potentials of the midbrain which come and go with shivering form the efferent con- trol of shivering must be proposed with caution. Further research is needed. This work will be discussed by Dr. Freeman. ACKNOWLEDGMENTS The history of the search for an understanding of the neural control of the responses of the hypothalamus to cold has revealed that considerable progress has been made since the days of Isen- schmid in 1912-1914. In acknowledging useful contributions to this field of study, the workof Sherrington, Bard, Keller, Ranson, Clark, Magoun, Thauer, and Hensel has advanced our knowledge over that known in 1914. For the studies conducted at the University of California, which have been briefly reviewed in the latter part of this report, the Arctic Aeromedical Laboratory in Alaska has played a major role by encouragement, scientific liason, and finan- cial support. It is a pleasure to acknowledge their cooperation and to commend the staff, particularly Colonel Quashnock, Lt. Colonel Herbert, Drs. Hannon and Eagan, for their contributions to the field of environmental physiology by arranging this Symposium. 27 HEMINGWAY, A. DISCUSSION DR. CLARK: I would rather not ask a question. I would like to get into this shive ring pathway because there is some of Dr. Keller's work that has been misinterpreted. If I could take a few minutes, I would like to go over this. If you make a complete transection throu^ the midbrain and then put the animal in the cold as indicated by one of Dr. Heming- way's diagrams, you are going to get a progressive loss in bocfy temperature with time. As you quite well know, when you are attempting to make lesions like this, youfailquite frequently. Now, say you leave a little bit of the cerebral peduncle. The temperature may go way down. These animals shiver; there is a little of the cerebral peduncle left. We do not know how the fibers go through there; we do not know where they come from; but we are attempting to get an anatomical basis for this finding. We have seen it in both cat and dog, but when the entire brain stem is transected except for a little of the pyramid, there is still shivering. DR. STUART: How long after transection were such preparations studied? DR. CLARK: I think the longest time has been seven or eight months. You probably would not see it until the animal is at least two months postoperative. DR. HEMINGWAY: How much of the other tissue is there, Dr. Clark, above the peduncle? None at all? DR. CLARK: There is only a part of the peduncle left, and the amount that is left determines the rate of cooling. The pathway for shivering is more medial than the pathway for panting; part of the shivering fibres are in the peduncle. I am not saying they are part of the pyramidal tract; I am simply saying that there is something resembling shivering after this transection, and that these physio- logical observations are confirmed by subsequent studies of sections. DR. STUART: If the animal is shivering, how does the body temperature fall? 28 HISTORICAL REVIEW DR. CLARK: Well, it is not a strong shiver, and you have the problem of defining shivering. If you have your hand on the hind quarter of such a dog, you can feel a trenoor which disappears when the animal is warm and shows up when he is cold. It is not enough to maintain normal temperature levels, but it is definitely there. DR. STUART; Have you ever observed shivering following complete destruction of the preparation's cerebral peduncle? DR. CLARK: Yes, but not as much as in the preparation with partial destruction. In the complete transection, if you lower his temperature to 29 G, quite frequently you can feel a tremor. Bard has found the same thing in his cats. DR. FREEMAN: This phenomenon of an increased thermogen- esis with some portions of the nervous system still intact would imply that there is some descending activity which potentiates shiv- ering that is set up in the spinal cord, or that there are parts of the brain stem affected by the external cooling. Have you attempted to stimulate electrically in various portions of the brain anterior to the sections to see if you can increase this effect? DR. CLARK: No, we have not. I would like to point out that we do not feel that these animals are truly poikilothermic. The fact that they still retain some indication of panting, some indication of shivering, indicates they are not poikilothermic. In addition, they show something in the way of vasomotor control. Thus, if you take a midbrain dog and lower his temperature into the cold and then bring it back into the warm, his skin temperature would jump up rather more rapidly than you would anticipate on the basis of simple rewarming. Consequently, we have fairly good evidence that, in the dog at least, there is some vasomotor control in the midbrain preparations. DR. HEMINGWAY: Would you care, Dr. Clark, to comment on Dr. Keller's statement that the descending pathways for shivering are in the corticospinal tracts. DR. CLARK: I think that this is one of his points which has been misinterpreted. He states that some of the fibers are in the 29 HEMINGWAY, A. corticospinal tract, not that the entire pathway is there. Some of the fibres are intermingled with corticospinal fibres. DR. HEMINGWAY: Dr. Birzis destroyed the corticospinal tract in the medulla in acute experiments. She tested them two or three hours after the lesions were made and observed no shivering. Some lesions were large, some were small; but whenever the corticospinal tracts were destroyed, the animal did not shiver. DR. CLARK: We do not know anything about the path below the midbrain, but it is quite definite that some of the fibres that mediate shivering are intermingled with corticospinal fibres in the midbrain. DR. FREEMAN: Both she and I looked for unit activity in the cerebral peduncles in the midbrain but were not able to find any. However, this may be, as you point out, a statistical distribution problem . DR. CLARK: With the number of fibres in this area, attempts to get unit activity from a particular type are like reaching into a bucket to pick out the one black marble. DR. FREEMAN: Well, in that case, it would be the Mexican jumping bean in the marbles. DR. HEMINGWAY: Dworkin observed shivering in decerebrate preparations, but others have failed to confirm this. Stuart and Bard have observed a tremor in decerebrate preparations, but this tremor is ineffective metabolically. DR. CLARK: From the work we have on the way the temperature falls, I would say it is not very effective, although it is still shiver- ing, DR. STUART: Rather than shivering, I would say "intermittent muscular activity" that has no thermoregulatory function. DR. CLARK: I would not say that because it appears and dis- appears according to the temperature. 30 HISTORICAL REVIEW DR. STUART: In our experiments the method of cooling has been critical. The animal might display a generalized avoidance response to a stimulus that is nocioceptive as well as cold. DR. CLARK: That is hard to accept under the conditions of our experiment. DR. STUART: It is easy to accept under the conditions of mine. DR. HEMINGWAY: Is it possible that a small amount of tissue is left just above the cerebral peduncle? DR. CLARK: What I am saying is that we found that these are usually cases where a complete transection was attempted, where the knife skipped a little bit over the bottom, so that only part of the peduncle is left and everything above it is gone. MR. ADAMS: I believe there is supportfor Dr. Clark's remark in studies that were conducted a long time ago where it was shown that vasomotor changes were associated with environmental temper- ature changes. DR. HEMINGWAY: But they are almost ineffective in preventing a fall in rectal temperature in the cold. MR. ADAMS: In this type of preparation one should not argue that the vasomotor changes are completely ineffective. DR. HEMINGWAY: There may be a tremor in a decerebrate preparation. Bard and Stuart have observed a tremor which appears in the cold and disappears with warmth, but this tremor is ineffec- tive in temperature regulation. DR. CLARK: I agree with you; it is not particularly useful because the animal's temperature falls markedly, but it is an indi- cation, primarily, that the pathway is quite diffuse. DR. HEMINGWAY: Yes, I would agree with that. 31 HEMINGWAY, A. DR. CLARK: Our preliminary, anatomical data, does not indi- cate that that is the case because in animals where we did a hemi- section of the internal capsule, in a whole sagittal section, I find two or three little fibres in the degenerating pyramid. We don't know where they come from. 32 HISTORICAL REVIEW REFERENCES 1. Abramson, D. I. Vascular responses in the extremities of man in health and disease. University of Chicago Press, 1944. 2. Akert, K. and F. Kesselring. Kaltzittern als zentralen Reiz- effect. Helv. Physiol. Acta. 9:290-295, 1951. 3. Andersson, B. Cold defense reactions elicited by electrical stimulation within the septal area ofthe brain in goats. Acta. Physiol. Scand. 41:90-100, 1957. 4. Bard, Philip and M. B. Macht. The behaviour of chronically decerebrate cats. Neurological Basis of Behaviour. Ciba Symposium, p. 55-75, 1958. 5. Bard, Philip and D. M. Rioch. A study of four cats deprived of neocortex and additional portions of the forebrain. Bull. Johns Hopkins Hosp. 60:73-147, 1937. 6. Bazett, H. C. and W. G. Penfield. A study of the Sherrington decerebrate animal in the chronic as well as the acute con- dition. Brain. 45:185-265, 1922. 7. Birzis, Lucy and Allan Hemingway. Descending brain stem con- nections controlling shivering in the cat. J. Neurophysiol. 19:37-43, 1956. 8. Birzis, Lucy and Allan Hemingway. Shivering as a result of brain stimulation. J. Neurophysiol. 20:91-99, 1957. a. 9. Birzis, Lucy and Allan Hemingway. Efferent brain discharge during shivering. J. Neurophysiol. 20:156-166, 1957. b. 33 HEMINGWAY, A. 10. Clark, George, H. W. Magoun, and S. W. Ranson. Hypothalamic regulation of body temperature. J. Neurophysiol. 2:61-80, 1939. 11. Clark, George, H. W. Magoun, and S. W. Ranson. Temperature regulation in cats with thalamic lesions. J. Neurophysiol. 2:202-207, 1939. 12. Dusser de Barenne, J. S. Recherches experimentelles sur les fonctions du systeme nerveaux centrale,faites enparticulier sur deux chats dont neopallium avaiteteenleve.Arch. neerl. de physiol. de I'homme et des animaux. 4:31-123, 1919-20. 13. Dworkin, Simon. Observations on the central control of shiv- ering and of heat regulation in the rabbit. Am. J. Physiol. 93:227-244, 1930. 14. Freeman, N. E. The effect of temperature on the rate of blood flow in the normal and in the sympathectomized hand. Am . J. Physiol. 113S84-398, 1935. 15. Freeman, W. J. and Allan Hemingway. Localization of unit po- tentials in brainstem associated with shivering. Arctic A ero- med. Lab. Report 1959. a. 16. Freeman, W. J. and Allan Hemingway. Manual for the con- struction, use and interpretation of records from stainless steel micro-electrodes. Arctic Aeromed. Lab. Report 1959. b. 17. Freeman, W. J., D. Stuart, and Allan Hemingway. Nervous con- trol of shivering. IX. Role of the fields of Forel in shivering. Arctic Aeromed. Lab. Report 1959. 18. Hemingway, Allan, Paul Forgrave, and Lucy Birzis. Shivering suppression by hypothalamic stimulation. J. Neurophysiol. 17:375-386, 1954. 19. Hemingway, Allan and S. R. Hathaway. An investigation of chem- ical temperature regulation. Am. J. Physiol. 134:596-602, 1941. 34 HISTORICAL REVIEW 20. Hertzman, A. B., W. C. Randall, andK. E. Jochim. The estim- ation of the cutaneous blood flow with the photoelectric plethysmograph. Am. J. Physiol. 145:716-726, 1946. 21. lampietro, P. F., J. A. Vaughan, R. F. Goldman, M. B. Kreider, F. Masuci, and D. E. Bass. Heat production from shivering, J. Applied Physiol. 15 :632-33, 1960. 22. Isenschmid, R. and L. Krehl. Ueber den Einfluss des Gehirns auf die Warmeregulation. Arch. f. Exptl. Path. u. Pharm. 70:109-134, 1912. 23. Isenschmid, R. and W. Schnitzler. Bectrag zur Lokalization des der Warmeregulation vorstehenden Zentralapparatusim Zwischenhirn. Arch. f. Exptl. Path. u. Pharm. 76:202-223, 1914. 24. Kaada, B. R. Somato-motor, autonomic and electrocortico- graphic response to electrical stimulation of "rhinenceph- alic" and other structures in primates, cat and dog. Acta. Physiol. Scand. 23 (Suppl. 83): 105,1951. 25. Keller, A. D. Hypothermia intheunanesthetizedpoikilothermic dog. Physiology of Induced Hypothermia. Nat. Acad. Sci. Publ. No. 451, p. 61-79, 1956. 26. Keller, A. D. and Henry Batsel. The ambulatory and semi- poikilothermic dog (poikilothermic against cold). Army Med. Res. Lab. Rept. 84, Project 6-6-12-028, 10 June 1952. 1952. 27. Keller, A. D. and W. K. Hare. Heat regulation in medullary and midbrain preparations. Proc. Soc. Exptl. Biol. Med. 29:1067-68, 1931-32. 28. Lippold, O. C. J., J. W. T. Redfearn, and J. Vuco. The fre- quency spectrum of the muscular activity of shivering in the anesthetized cat. J. Physiol. 140 :49P, 1958. 29. Nyboer, Jan. The impendance plethysmograph. Medical Physics, 2:736, 1960. 35 HEMINGWAY, A. 30. Perkins, J. F. The role of the proprioceptors in shivering. Am. J. Physiol. 145:264-271, 1945. 31. Pinkston, J. O., Philip Bard, and D. M. Rioch. The response to changes in environmental temperature after removal of por- tions of the forebrain. Am. J. Physiol. 109:515-531, 1934. 32. Ranson, S. W. Regulation of body temperature. Proc. Assoc. Research Nerv. and Mental Dis. 20:342, 1940. 33. Ranson, S. W., C. Fisher, and W. R. Ingram. Hypothalamic reg- ulation of temperature inthe monkey. Arch. Neurol. Psychiat. (Chicago) 38:445-466, 1937. 34. Sherrington, C. S. Notes on temperature after spinal transection with some observations on shivering. J. Physiol. 58:405-424, 1923-24. 35. Teague, R. S. and S. W. Ranson. The role of the anterior hypo- thalamus in temperature regulation. Am. J. Physiol. 117: 562-570, 1936. 36. Thauer, Rudolph and Georg Peters, Warmeregulation olive Hypothalamus. Abhandl. Deutsch. Ges.f. inn. Med. 49:188- 190, 1937.a. 37. Thauer, Rudolph and Georg Peters. Warmeregulation nach operativer Ausschalltung der Warmezentrums. Pfluegers Arch.239:483-514, 1937. b. 36 RECENT ADVANCES IN THERMORECEPTOR PHYSIOLOGfY H. Hensel Physiologlsches Institut der Universitat Marburg/Lahn Germany I am going to discuss some recent problems of thermoreceptors which give the organism information about the temperature con- ditions of the surroundings. The previous concept of the electro- physiology of thermoreceptors was based merely on investigations of the afferent impulses in the cat's tongue. Figure 1 shows a record from a fine preparation ofthe lingual nerve of the cat during mechanical and thermal stimulation of the tongue. Mechanical stimulation alone will elicit a large discharge of impulses. During cooling we see a smaller single fiber discharge with a partial adaptation. When cooling the tongue and touching it at the same time, we get both types of discharges. Thus it is quite easy to discriminate in the cat's tongue between specific cold receptors which respond only to cooling and specific mechano- receptors which are not excited by cooling. Unfortiinately, we did not stop the experiment at this time, but started in Marburg two years ago with investigations of the afferent discharges from the cat's external skin. In connection with the problems of thermoregulation, it is important to record not only from the tongue, which is a rather specialized organ, but also to get information about the impulse traffic in the cutaneous nerves. We were very surprised to find practically no specific cold fibers in the external skin. There is quite a number of A fibers showing very much the same behavior on cooling and warming as found in the lingual nerve, but the cold sensitive receptors in the cat's skin were also sensitive to mechanical stimulation, as shown in Figure 2. The discharge of this fiber during cooling and warming the skin is the same as has been found for the specific cold fibers in the tongue, but on touching the skin with a hair, a marked response can be seen. Neither could we record any A fiber potentials during warming the external skin of the cat. ^y HENSEL, H. «^.f ^ ^ ^1 ig ECi = Cb i 38 rrHERMORECEPTOR PHYSIOLOGY I O -Q O -O Hti Hid H-H HHri HtH 0) 3 i o ID (U 4) <» ^2 f^ E 39 HENSEL, H. Now, the question remained open whether specific cold recep- tors might be connected with non-myelinated C fibers; about 80 per cent of the afferent fibers in the cutaneous nerves are C fibers. Recently, we had the opportunity of recording afferent impulses from single C fibers in the external skin of the cat. This work has been done together with Dr. Iggo, Edinburgh, and Dr. Witt in our laboratory, and we found that many cold receptors as well as warm receptors are connected with non-myelinated fibers. Figure 3 illustrates the method. The saphenous nerve was dis- sected into very thin filaments in order to record impulses from single non-myelinated C fibers. The C fiber was identified by measuring its conduction velocity. This was accomplished by apply- ing electrical stimuli to the nerve dist ally from the recording point. The impulses were recorded by means of two oscillographs; one oscillograph was used for normal recording of the spike, whereas the sweep of the second oscillograph was triggered by the electric stimulus. The conduction velocity could be found by measuring the distance between the stimulating and the recording point and by measuring the time between stimulus and the onset of the C fiber impulse at the recording electrode (Fig. 4). By means of this method, the conduction velocity of the non-myelinated temperature fibers of the skin turned out to be in a range of 0.5 to 1.5 m/sec. Figure 5 shows the discharge of a single non-myelinated cold fiber when cooling the skin. After the onset of the discharge, a partial adaptation can be seen. Even cooling by several tenths of a degree centigrade is sufficient to cause a marked increase in the impulse frequency. The C fiber also shows a steacfy discharge at constant skin temperatures. The short inhibition of the discharge before the onset of cooling is due to a slight increase in temperature. Plotting the discharge frequency of a single C fiber against time, we get curves as shown in Figure 6. Sudden cooling causes a phasic increase in frequency. On rewarming, there is a transient inhibition and then the discharge reappears and finally reaches the initial level. Figure 7 shows the discharge frequency of a specific non- myelinated warm fiber. The plots of the impulse frequency against time of this warm fiber show a phasic increase in frequency during warming (Fig. 8). Even warming by several tenths of a degree centigrade is sufficient to cause a marked increase in frequency. 40 THERMORECEPTOR PHYSIOLOGY .:.Q| T^ - * 1 i ( / \ \ S i i i ii a s 5l bC o •3 g ?-8 II « o 41 HENSEL, H. ^l s s a A P IB : § s ' 3 cr » 2. fO o 3 " i e 2 >i a 5 (R o » "S H jr Q. "^ 3- '^ t= • Ann 3 3 I Po (B A 3 IB a p ^i ? & $ I Urn IS JlS. 3 p a s A S 5" p a' * S- o 2 ^ O o A *" " "^ X °- rs p -^ o 5 o — - AC: " » « 5 •-► . PA 1 ? S'S 5. a o 5 A r* O _ o a- ►«. 5" "I ^: s: S J7 I 5 D. £. i, 2 2 » cr 6 $ 3 ^ 3§^ <<> 2 y^ o (B O Si 5 g £5 ? * o. A * <» 5^ »> <» <* 5 2 i» O W P 73 O «• ^ ET ^° I i 3"g 2 S 3 :i 5 g £? 2. » * l>B° 2 3 g"! "> «. A o r 42 THERMORECEPTOR PHYSIOLOGY Temp. 40r' 30 20 1 sec IIIIIIIIIIIIIIIIIMII'I I I I I'll . M 1 I II lllimiMIIIMimMlill !)! ! I I 40r' 30-. 20- I I llllllllllllllllllllllllllllllll|lllllllllll!!,!IIMIiy 40p^ 30- 20-" I I IMII ll|llllllll|imillllll||l!lil!l'i II II I f I 40r ! I I 30- 20- iiiifiiiiiiiiiiiiiiiiiiMiiiiihMiiMiiiimi Figure 5. Afferent impulses of a single C fiber in the cat and skin temperature when cooling the skin (fiber No.^1). Conduction velocity 1.1 m/sec. A, Cooling from 29 C to 25.5 C; B, 29.3 C to 28° C; C, 28° C to 26.5° C; D, 24.5° C to 22.5 C. (Hensel, Iggo, and Witt. 1960). 43 HENSEL, H. Skin temperature (*'C); A -1. -» NJ fsj Impulses/sec ' .5 3 a- e H a s (B p 5' ri cr 5? ^ a n j; 1 s S 1 » 8 s. I l!| « w S{ ? 3 > (g 1 i m ^ E 2. B ^ 2} g ■?< u- o ^ p ? ^ Skin temperature ( C); 8. C 44 THERMORECEPTOR PHYSIOLOGY OQ ^ ' ' ' ' ' §e U vo ^ r^i ■ I ' I ' tT o ^o 45 HENSEL, H. ■C; 5 00 5" » • III s R a ^ « s C e « g c D 20 ■hi mmm 1 1 I M [pi||||ll'i||ii[Ni|.''i|V|''iil|l||!""i|iill|ii"i'|ii,'i|'|l|l|i|l|!''|ll'I nMHil oL F 'nmmmmmmm oL llt III i I I I III I H^ohlli ill it ^1 oL i^i„| ll li.li I iiSigit Figure 10. Afferent impulses in single "cold" fiber in a human subject (No. 18a) and temperature during cutaneous thermal stim- ulation. A, stert of cooUng from 34 C to 26 C.^B, start of re- warming. C, start of cooling from 38 ^ C to 35^ C. D, start of rewarming. E, start of cooUng from 24 C to 16 C. F, start of rewarming. G, 13 sec after rewarming has started. H, continued from record G. (Hensel and Boman, 1960). HENSEL, H. in the cat. At constant skin temperature, we see a constant discharge of a certain frequency — on cooling, an increase in frequency, and on rewarming, an inhibition. The discharge frequency as a function of time during cooling and rewarming is shown in Figure 11. We were not able to stucfy the steady discharge of this receptor at very low temperatures, but I think that the maximum is to be expected at a temperature of about 15 C (Fig. 12). During this work we also found some non-specific fibers which were excited by cooling as well as by mechanical stimulation (Fig. 13). A sligjit pressure of 13 grams causes a great increase in frequency and a partial adapta- tion. During cooling, the discharge frequency increases, whereas rewarming causes a transient inhibition. Mechanical stimulation causes a much higher increase in frequency than does rapid cooling. The table shows some figures found in four single non-specific fibers from human skin. The maximum discharge during light pres- sure goes up to 125 impulses/sec, but on rapid cooling only a maximum frequency of 17 impulses/sec is reached. Now, what is the sensation connected with these non-specific cutaneous fibers? I believe it is a mechanical sensation. This can be concluded from an observation made by Ernst Heinrich Weber some hundred years ago, that wei^ts put on the skin feel heavier when cold. This is a well established phenomenon and proved by several investigators. Any opposite experience, namely, that pressure will elicit a cold sensation, is not known. Thus the conclusion is justified that the functional significance of the non-specific receptors is a mechanical sensation. As yet, we have not found any A warm fiber in human skin. Even when recording from multi-fiber nerve preparations, the total dis- charge frequency always increased during cooling and decreased during rewarming (Fig. 14). Of course, it is possible that specific warm impulses are travelling in the C fiber group. I think it is very probable, but as yet we were not able to record from single C fibers in human subjects. The movie I am going to show now was made in connection with a television broadcast from our institute. The television company was kind enough to give us the whole film, from which we have cut a short film. It is concerned mostly with the method of recording afferent impulses from cutaneous nerve fibers in conscious human 50 THERMORECEPTOR PHYSIOLOGY Temperature (®C) Frequency (Imp/sec! 51 HENSEL, H. 15 u 0) Q. J 10 >» u c 01 cr I 5 Q. E - 1 1 1 1 - - ^« s. - - \ - - 1 1 1 \ 1 \« : 10 20 30 Temperature (°C) 40 Figure 12. Frequency of steady discharge in a single "cold" fiber in a human subject (No. 18a) plotted against constant skin temperature. (Hensel and Boman, I960). 52 THERMORECEPTOR PHYSIOLOGY 1 T I I I I 1 1 1 I I I ■ 1 J ;: 1 1 1 1 1 A = 1 _^> 1 — 1 — 1 — 1 — 1 — 1 — 1 1 1 1 1 1 , , CM £ a « 51 (3as/duj|) ^3uanbdJ^ ds|nduj| (3(^ajn)ejadui»i 53 HENSEL, H, Figure 14. ToUl frequency of discharge In multi-fiber prepar- ation In human subject containing at least six fibers (No. I2c) during cutaneous temperature changes. (Hensel and Boman, 1960). 54 THERMORECEPTOR PHYSIOLOGY subjects. We are greatly indebted to our medical students for their cooperation as subjects in these experiments. MOVIE: CUTANEOUS AFFERENT NERVES IN HUMAN SUBJECTS We use the superficial branch of the radial nerve which can be found easily by palpation. The course of the nerve is marked on the skin. Exposure of the nerve is made under a very short general anaesthesia lasting for about five minutes. The experiment proper is performed in the non-anesthetized subject. If the preparation is made carefully, the subject will feel quite comfortable. The nerve preparation is made under a stereo-microscope. The superficial branch of the radial nerve is quite a thick nerve, about four or five millimeters in diameter, but we use only a very small part of the nerve, about one tenth. The nerve consists of several bun- dles connected with loose connective tissue. Before separating a bundle, the intact nerve is tested by means of weak electrical shocks in order to find the part with the most suitable receptive field. Then this bundle, which is about 0.5 mm in diameter, is separated from the remaining part of the nerve and cut proxim ally. At this moment, the subject feels a burning sensation in the receptive field which dis- appears after a few seconds. The electrode is fixed in a micromanipulator and adjusted to the nerve strand. The nerve is put onablack plate for the final dissec- tion into thin filaments. These filaments, of course, do not always contain single fibers in an anatomical sense, but they are thin enough to get mostly a single fiber discharge. The afferent impulses are recorded by a cathode ray oscillograph, whereas galvanometers are used for recording the skin temperatures. Heating and cooling the skin is brought about by water-circulated thermodes which rest on the skin in the same position for the whole time. Four thermostats set at various temperatures are connected alternatively with the thermode by means of a switch. 55 HENSEL, H. Figure 15. Experimental arrangement for recording afferent Impulses in cutaneous nerves in human subjects (before sterile covering). 1. Arm of the subject. 2. Metal holder for the arm. 3. Hand support for nerve preparation. 4. Adjusting screw for hand support. 5. Metal plate. 6. Skin marked for incision. 7. Lamp attached to microscope. 8. Stereo-microscope. 9. Support for stereo-microscope. 10. Magnification selector. 11. Black plate for nerve dissection. 12. Holder for electrode. 13. Electrode. 14. Micro-manipulator. 15. Shielding cage. 16. Lieads from electrode to pre-amplifier. 17. Oscilloscope for observation. 18. Thermode. 19. Water inflow. 20. Water outflow. 21. Leads from thermocouple to galvanometer. 22. Small thermode for testing the receptive field. 56 THERMORECEPTOR PHYSIOLOGY 57 HENSEL, H. SUMMARY In the external skin of the cat, no specific cold or warm fibers seem to exist within the group of myelinated A fibers but only non- specific receptors excited by pressure and cooling. However, numerous specific cold and warm receptors have been found with non-myelinated C fibers, the conduction velocity of which was be- tween 0.5 and 1.5 m/sec. The quantitative sensitivity of these re- ceptors corresponded well with the threshold for subjective temperaure sensations in man. In human subjects, only one specific cold fiber has been found as yet in the A group, whereas several A fibers responded to pres- sure as well as to cooling. The latter probably cause a mechanical sensation. 58 THERMORECEPTOR PHYSIOLOGY DISCUSSION DR. FREEMAN: I am enormously impressed with your re- cording from the C fibers. This is a remarkable accomplishment. I notice that the wave form was initially upwards, then downwards again. What is its polarity? DR. HENSEL: I do not know exactly how it was in this experi- ment. It is not always the case that you have waves in both direc- tions. DR. FREEMAN: Your technique for measuring time intervals will show your latency was at least as long as known latency. Is that correct? DR. HENSEL: We are measuring the time between the onset of the stimulus and the onset of the impulses at a certain distance from the stimulated point. DR. FREEMAN: Do you measure the degree of onset that accurately? DR. HENSEL: No, the conduction velocity has been measured by applying well defined rectangular electric shocks. The distance between stimulating and recording point of the nerve was about 4 cm. DR. FREEMAN: How do you figure the onset of your cold stimulus? DR. HENSEL: This was measured by a thermocouple below the therm ode. DR. FREEMAN: But if your latency is amatter of milliseconds, what part of the onset of the cold do you take as your starting point for the stimulus threshold? 59 HENSEL, H. DR. HENSEL: We did not measure conduction velocities of cold fibers using cold stimulation but only be electrical stimuli. We can be absolutely sure that the impulse set up by the electrical stimulus is identical with that caused by cold in the respective single fiber. The fiber can be identified by the characteristic shape of the impulse and by the collision technique shown in Figure 4. DR. FREEMAN: How does the latency compare with A fibers in this technique? DR. HENSEL: You mean the latency between the stimulus and the onset of the impulses? DR. FREEMAN: No, using this technique for measuring the conduction velocity and the duration of the impulse and its height. How do your figures for these C fibers compare with your observa- tions of A fibers under the same circumstances? DR. HENSEL: Well, I would say that the latency of the A fibers is about 1/30 to 1/50 of that of the C fibers. DR. STUART: Are the frequency spectra for pressure and temperature change in a single C fiber always different? DR. HENSEL: I am not sure about that. Recently we found a non-specific fiber which was quite as sensitive to cooling as the specific cold fibers. DR. CLARK: I think in one of your figures you said something about on and off potentials. Is that "off "due to the thermode change, or was it a real on and off effect? DR. HENSEL: I think it is a real on and off effect. The over- shoot in frequency during the onset of cooling is the mirror image of the false start during onset of rewarming. In some records, the whole course of the false start is not seen, as the impulse frequency cannot drop below zero. DR. CLARK: That is on all your records except one? 60 THERMORECEPTOR PHYSIOLOGY DR. HENSEL: Yes, that is right, but the general shape of the curve is the same in all experiments. This can be proved by apply- ing standard cold stimuli during the transient stop of the discharge after rewarming. The longer the period of time after the cessation of the discharge, the more effective the cold stimulus. When plotting the effect of the standard cold stimulus against time, the resultant curve is the mirror image of the overshoot during cooling. But, of course, the impulse frequency cannot be less than zero. DR. MINARD: Is there a rate of temperature change so slow that you do not observe the phasic discharge? DR. HENSEL: Yes. I think it is not even necessary to make the change very slow because this partial adaptation occurs rather rapidly. A slow linear change of temperature with time throu^ the whole temperature range will give very much the same frequency diagram as that for the steacfy discharge shown in Figures 9 and 13. DR. MINARD: There have been some reports that individuals who have been exposed to gradual increase in cold, such as when they have been lying outside in a sleeping bag, have died, apparently without even waking up during the course of this cooling process; and the assumption was that the rate of change was so slow that this was not perceived as a cold stimulus. DR. HENSEL: I think this concept contains some truth because the maximum frequency obtainable by rapid cooling is 15 times higher than that obtained during slow cooling or at constant temper- ature. COL. QUASHNOCK: Dr. Hensel, would the subject report any residual changes in sensation either transient or permanent after recovery? DR. HENSEL: Well, most of the subjects reported a decrease in sensitivity in an area about 2 cm in diameter. It was no complete anaesthesia but paraesthesia or hypaesthesia. In one subject there was no change at all because of the overlapping.- COL. QUASHNOCK: How many subjects have been done? 61 HENSEL, H. DR. HENSEL: Eight as yet. COL. QUASHNOCK: Did you suture the nerves after cutting them? DR. HENSEL: No, because it was only a thin bundle 0.5 mm in diameter. We only cut this small part of the nerve, whereas the whole remaining part was undamaged. DR. LIM: And you did not observe any warm receptors in all these subjects? DR. HENSEL: No, we always saw an increase in frequency during cooling and a decrease during warming. But we have not studied the C fibers as yet. I am quite sure that warm receptors could be found in the C group. DR. HEMINGWAY: Have you any idea what end organs are involved? DR. HENSEL: I can say only that there are some areas in human skin which contain only so-called free nerve endings. From these areas, any kind of sensation can be elicited. This holds true also for the tip of the tongue in the cat. A thorough histological stucty of this area has been made by Dr. Kantner in Heidelberg (1957). He found only a network of free nerve endings in the tip of the tongue — except the taste buds, of course. As you can see in Figure 1, there are quite specific pressure and cold impulses, in spite of the absence of any encapsulated specialized nerve endings such as Krause's and Meissner's end organs. DR. IRVING: I notice, sir, that the records from many of your cold stimulations cease at certain cold temperatures. Ami correct about this? DR. HENSEL: Well, the lowesttemperature for maximum firing that we have found so far in C fibers was about 16 C. Some C fibers were still firing at a temperature of 5 C, but at a very low rate. DR. FREEMAN: How much vibration in the thermode do you get as a result of passage of water through it? 62 THERMORECEPTOR PHYSIOLOGY DR. HENSEL: Oh, the vibration was quite small, but a few very sensitive mechano receptors could sometimes be excited by the vibration. DR. LIM: During cooling, you have a curve coming down, but it never reached the zero line. At the end of the cooling period, what were the impulses per second? DR. HENSEL: The maximum frequency of a cold fiber at con- stant temperature is about 10 impulses/sec. DR. LIM: I noticed that your time scale is about a minute or so. If you prolong it for ten minutes, what is the effect? DR. HENSEL: I would say that the final value is reached after about one minute. Thereafter the impulse frequency will remain constant as long as we are able to record from a single fiber, at least for several hours. DR. LIM: But still above zero? DR. HENSEL: Yes, for the whole time. DR. FREEMAN: What was the relation between accommodation and sensitivity or the sensations of the subjects with regard to intensity? DR. HENSEL: Generally, the impulse discharge is more sens- itive than the conscious sensation. For example, a skin temperature of 34 C is absolutely indifferent for the subject, but there is a considerable discharge of the cold fibers. DR. FREEMAN: Does the peak of sensitivity coincide with the peak of discharge? DR. HENSEL: Yes, roughly. It is difficult tomeasure this very accurately in a subjective way. DR. FREEMAN: Do you ever get a paradoxical sensitivity of warmth associated with this accommodation process? 63 HENSEL, H. DR. HENSEL: No, we observed only a decrease of the cold sensation. MR. ADAMS: What was the maximum impulse frequency? DR. HENSEL: 150 impulses/sec in the cold fibers. MR. ADAMS: Was this with cooling to 13° C7 DR. HENSEL: We did not record the 150 impulses/sec in the hijman cold fiber I have shown. I think it was about 100 impulses/sec, but when cooling the most sensitive cold receptors from 34 C to 0 C, the frequency will rise to about 150 impulses/sec. It depends on the rate of cooling. The higher the cooling rate, the hi^er the maximum frequency. MR, EAGAN: There is an inference here that one would not feel cold if the receptors stayed below 5 C. DR. HENSEL: Yes, I think so. The cold sensation will disappear at very low skin temperatures. DR. RODDIE: Do you use any painful stimuli? DR. HENSEL: Not yet, but this is part of our next program. We only used temperatures from 15 C to 43 C. DR. HANNON: Could this decrease pain as the temperature o goes below 20 C?I believe you say there is a decrease in sensation. DR. HENSEL: Yes. DR. HANNON: How do you determine whether this is an accom- modation or just a simple temperature effect? DR. HENSEL: Well, accommodation means that a time factor is involved. If the temperature is lowered to 5 C and raised again, the discharge will reappear and after sometime reach the initial level. 64 THERMORECEPTOR PHYSIOLOGY DR. HANNON: Would you care to speculate how much of this is an Arrhenius effect, the temperature just slowing up the reaction? DR. HENSEL: This is very difficult to say because the whole curve does not look like an Arrhenius diagram. I think this curve might perhaps be due to the difference between two processes, each of which follows the Arrhenius law. DR. HANNON: Many rate reactions will show this phenomenon if the temperature range is wide enough. They usually describe these by two mathematical rate-processes. DR. HENSEL: This is just a formal description. We do not know anything about the processes involved in the thermoreceptors. It would be very interesting to know more about the mechanisms for transforming the temperature into the impulse discharge of the receptor. COL. QUASHNOCK: Is it possible to take the proximal portion of the nerve and stimulate it at the various frequencies to determine the threshold for sensation? DR. HENSEL: We have not tried this, but it could be done very easily. We use electrical stimulation of the nerve before cutting it. The subject reports the site of the sensation. DR. HEMINGWAY: You stated that slow cooling is not nearly as effective as fast cooling, and this brings up the point raised by Capt. Minard. During World War II we had some subjects who were cooled slowly, at the same time being very heavily clothed; I think they had clothing that built up to fourclos. In these subjects we ob- served that rectal temperature would drop two or three degrees without the appearance of shivering. DR. HENSEL: Yes, the same as observed in the measurement of CO . The rectal temperature will drop without any considerable sensation of cold. It might be a matter of the slow rate of cooling. The rate of cooling is much more important than the absolute temperature. 65 HENSEL, H. DR. KAWAMURA: Does each single cold or warm fiber have its own temperature range? DR. HENSEL: Yes, eachfiberhas its range of temperatures, but if you take the whole output of the nerve, then the maximum will be at a temperature of about 15 C. The whole output of the nerve is integrated as a whole spectrum of nerve fiber discharges. If you count the impulse traffic in a rather thick nerve preparation — it cannot be done very accurately — it will have a maximum at about 15 C And this corresponds quite well to the subjective cold sen- sation under constant skin temperature. DR. STUART: What is the response to mechanical stimulus of the C fiber at the thirty-five impulse per second response? DR. HENSEL: Excuse me, this was no C fiber. This was an A fiber, a non-specific A fiber. DR. STUART: What was the discharge frequency during mechan- ical stimulation? DR. HENSEL: It was hi^er than 35. I think 120 or 130. With strong pressure you will get even higher frequencies, up to 300 impulses /sec. I would like to emphasize that the temperature sensitivity of the non-specific fibers might be in the same range as that of the specific ones, but still the response to mechanical stimulation is much higher. DR. STUART: This migjit suggest the central nervous system decodes diverse modalities by these diverse frequencies, even along the same fiber. DR. HENSEL: Let us assume a non-specific fiber firing at 35 impulses /sec. How can you discriminate between light pressure or strong cooling, both of which cause the same impulse frequency? I think the sensation mediated by this fiber is a specific mechanical sensation. There are also some experiences in daily life; for ex- ample, when a very cold wind is blowing into your face. You feel not only cold but you also have a strange sensation of pressure. This migjit be due to the discharge of such non-specific fibers. 66 THERMORECEPTOR PHYSIOLOGY DR. FREEMAN: In the light of your experience, would you have any suggestions as to how best to get into a cold swimming pool? DR. HENSEL: You mean just jump in? It depends on what you like. If you come from a hot room, it is very pleasant to jump into the cold water. If the same cold jump starts from a very high tem- perature level, we get a lower discharge frequency than when star- ting from, say, 34 C. DR. MINARD: I was wondering to what extent the static dis- charge is related to the spatial temperature gradient in the skin. DR. HENSEL: This concept has been completely disproved. We are sure now that the spatial gradient does not play any role in the excitation of thermal receptors. We can cool the receptor layer from above or from below. There is no difference in the discharge. I think it is well established that the discharge is only dependent on temperature itself, not on the spatial pattern of temperature. 67 HENSEL, H. REFERENCES 1. Hensel, H. Physiologie der Thermoreception. Erg. Physiol. 47: 166-368, 1952. 2. Hensel, H. and K. K. A. Bom an. Afferent impulses in cutaneous sensory nerves in human subjects. J. Neurophysiol. 23:564- 578, 1960. 3. Hensel, H., A.Iggo,andI. Witt. A quantitative study of sensitive cutaneous thermoreceptors with C afferent fibres. J. Physiol. 153:113-126, 1960. 4. Kantner, M. Die Sensibilitat der Katzenzunge. Acta. Neuroveg. 15:223-234, 1957. 5. Witt, 1. and H. Hensel. Afferent Impulse aus der Extremitaten- haut der Katze bei thermischer und mechanischer Reizung. Pfliigers Arch. ges. Physiol. 268:582-596, 1959. 68 THERMORECEPTOR PHYSIOLOGY DR. HANNON: If there are no further questions, we can go on with Dr. Hensel, I believe; he has something possibly related to this. DR. HENSEL: I appreciate very much this opportunity to dis- cuss some data on metabolism of the no n- anesthetized cat during hypothalamic cooling. Figure 1 shows a device for the cooling. In most cases, two cooling units were implanted symmetrically into the anterior hypothalamus. The thermodes have an outer diameter of one milli- meter. The right part or the left part of the anterior hypothalamus can be cooled separately, or both parts at the same time. The brain temperature at a certain distance apart from the thermode is measured by means of a thermistor. The cooling water comes in centrally and leaves the thermode through a lateral tube. I must say I disagree a little with Dr. Hemingway, because we found that the cats are quite cooperative andean be trained to quite a high degree (Fig. 2). A heatedthermocouple unit was used for recording the cutaneous blood flow in the ear. For measuring the metabolism, the cat is put into a double-walled perspex box which can be circulated with water and set at any desired temperature by means of a thermostat. The measurement of O consumption and CO output is made in an open system. The chamber is perfused with a stream of air of constant velocity and the air coming out of the chamber is continuously analyzed for CO and O . We studied quantitatively the heat production under various external temperatures, and at the same time, under the standardized hypothalamic cooling. We found the correlation shown in Figure 3. The lowest temperature in the chamber was about 5 C or 6 C, the highest 40 C. The lower the external temperature, the higher the metabolic increase during the same standardized hypothalamic cooling by about 2 C.I think this is another example of the cor- relation as shown by Dr. Lira between the external stimulus and the hypothalamic cooling in the unanesthetizedandunrestrainedcat. 69 HENSEL, H. Figure 1. Device for cooling and heating the hypothalamus in the unanesthetized cat. T, thermode; I, Inflow; O, outflow of water; P, metal plate with screws for the skull; N, needle with thermistor; L, leads to the thermistor. 70 THERMORECEPTOR PHYSIOLOGY 0) c I a> O 3 ^ „ «> Tt -a -c •Sea « •§ g § ■o g o S (1) jC F hC £ 5 5 il "^ 71 HENSEL, H. i" •-0 5 • 1^ €• s? cm ^ o-« ^ o°^-t J? o— • o • o • J^ ^ o ^ o -• S o o -• o — • - § o— o • • * o — • ^ ^ «>l 1 1 1 1 1 1 1 1 1 b O «D JO JJ 72 THERMORECEPTOR PHYSIOLOGY MR. ADAMS: Dr. Hensel, have you measured the temperature profile around the tip of the thermode? DR. HENSEL: Yes, we did; the temperature is measured mostly at a distance of 1.5 millimeters apart from the thermode. It is quite a steep temperature gradient, and the end of the temperature field might be at a distance of about three or perhaps four millimeters, if you have a very slow cooling. But, normally, I would say about three millimeters around the thermode. DR HEMINGWAY: The increase in oxygen consumption rate was not very great, not nearly as much as you usually find in shiv- ering. I think your maximum increase was about fifty percent. DR. HENSEL: The cooling was actually not very deep, and we did not observe shivering in this experiment. DR. HEMINGWAY: No shivering? DR. HENSEL: No, I would not say it was shivering. Sometimes we observed some sort of behavioral regulation. The cats were curling up a little bit, but no strong shivering occurred in this experiment. DR. HEMINGWAY: Would this increase be due to some motion, some voluntary activity on the part of the cat? DR. HENSEL: Not in all cases. DR. HEMINGWAY: If he is curling up and moving his muscles, it does not take much movement. DR. HENSEL: No, some cats sat mostly quietly, and there was no visible external motion. DR. HEMINGWAY: The cats were very kind to you then! DR. HENSEL: Yes. 73 HENSEL, H. DR. HEMINGWAY: It does not take much of a voluntary move- ment, just sitting up will increase the oxygen consumption rate twenty-five to fifty per cent, just a very slight movement and that is the difference. You were speaking about the difference between dogs and cats. A well-trained dog will lie quietly and will not leave the room. DR. HENSEL: In many cases the cats were lying on the bottom of the chamber very quietly, just like a well-trained dog. DR. STUART: Thesethermodes were implanted into the anterior hypothalamus. Have you ever implanted any into the posterior hypothalamus? DR. HENSEL: Yes, we did, but we did not see any considerable effect. We saw it in the anterior hypothalamus, but I must say we did not make very many experiments as yet in the posterior hypo- thalamus. I would not draw a definite conclusion, but we failed in getting an effect as yet. DR. LIM: In terms of oxygen consumption and differential cooling of the body, we found that there are two things we have to furnish. One is the onset of shivering; and in the onset of shivering, the peripheral mechanism is very important in the initiation of shivering, but in the maintenance of shivering, a central temperature is almost three times more important than the skin temperature in terms of oxygen consumption. DR. HENSEL: Yes. DR. CLARK: When you say "anterior hypothalamus", just how far anterior? Would it be over the optic chiasma? DR. HENSEL: Yes, the position was above the optic chiasma. But you see, even if you have a very accurate position of the ther- mode, you have a temperature field around the thermode. DR. CLARK: It would not get back to the tuberal region? DR. HENSEL: No. 74 THERMORECEPTOR PHYSIOLOGY DR. FREEMAN: Dr. Hammel, when he places his thermode, merely has a plate which he attaches to the skull under anesthesia and allows the skin to grow over it. DR. HENSEL: Yes. DR. FREEMAN: Now, when he wants to do an experiment, he sterilizes the skin over this plate and forces the thermode down into the hypothalamus for the duration of the cooling and heating period, and when he is through with the animal for the day, he pulls it out and puts the animal away. DR. HENSEL: Yes. DR. FREEMAN: Now, he says that over a period of time there may be some damage to the hypothalamus. DR. HENSEL: I would not agree with him. We had the thermode in the cat for half a year, and we have one cat in which we did the implantation five times, each time for half a year. It is now three years and the cat is quite as happy as in the beginning. So I do not think that there would be a serious damage. DR. FREEMAN: You would leave yours in, though, for the full period of time? You do not take them out? DR. HENSEL: No. DR. FREEMAN: You leave it in once it is there? DR. HENSEL: It remains for six months in the cat, and you can repeat this several times and go on without any disturbance. DR. FREEMAN: One other technical question: how do you prevent over-heating with your device; that is, when you want to warm the animal, what level dod you choose for your inflow tem- perature as an adequate stimulus? DR. HENSEL: As yet, we studied only the cooling. We were just interested in this effect because the effects during heating are more 75 HENSEL, H. generally accepted than the effects during cooling, as you know. DR. FREEMAN: You never over-cool? DR. HENSEL: No. DR. HEMINGWAY: You say you use a one-millimeter thermode and put that in the hypothalamus five times without damage? DR. HENSEL: I would say, without any visible change in the cat's behavior. DR. HEMINGWAY: Because the anterior hypothalamic region is not very many millimeters in extent. DR. HENSEL: Of course, in other cats, we made a histological investigation of this region, but not in the cat where the implantation was made several times. This cat is still living. DR. FREEMAN: That is humane, of course! DR. HENSEL: But in the other cats where the implantation was made for only one time, there was very little damage and very little reaction. DR. FREEMAN: How far laterally to the midline are these inserted? DR. HENSEL: Three or four millimeters. DR. FREEMAN: Any closer than that, we found that you either lacerate the choroid plexuses or you may obstruct the pyramidal row. DR. STUART: Was this the same in the posterior hypothalamus, three or four millimeters from the midline? DR. HENSEL: Yes, we tried various distances, but we did not see an effect as yet, nor with the vasomotor response. 76 THERMORECEPTOR PHYSIOLOGY DR. LIM: In this last slide you showed, is cooling the brain bilateral or unilateral? DR. HENSEL: Bilateral. But we didnotseea very considerable difference. It was a little bit more, but we got quite a high increase with unilateral cooling. 77 CENTRAL AND PERIPHERAL MECHANISMS IN TEMPERATURE REGULATION Thomas P. K. Lim Lovelace Foundation Albuquerque, New Mexico One of the outstanding characteristics of the mammalian body is its chemical and physical control mechanisms by which a constancy of internal environment is maintained. The behaviors of hemodynamics, gas exchange, electrolyte balance, energy trans- fer, and hormonal interaction clearly indicate that their ultimate goals are in preserving an optimum intrinsic condition which allows harmonious and efficient performance of complex processes. The regulatory phenomena of body temperature are no exception from this general premise. The understanding of temperature control has been greatly facilitated by the establishment of two cardinal neural mechanisms, namely, the major control elements located centrally in the hypo- thalamus which Dr. Hemingway described this morning, and the temperature sensing elements found peripherally in the skin, which Dr. Hensel described this afternoon. It is agreed that the central control mechanism may be subdivided into a motor unit which primarily governs heat dissipation and another which controls heat conservation. Furthermore, it is accepted that the peripheral thermoreceptors consist of two anatomical or electrophysiological units, that is, the receptors for cold and those for warmth, in the sense of functional anatomical units. Although the existence of temperature-sensing elements in the hypothalamus has never been demonstrated, it may be assumed that such receptors exist in the vicinity of the motor units of heat dissipation and conservation. Therefore, an hypothesis is advanced allowing functional elements of temperature detection in the hypo- thalamus as they exist in the skin. Then, central control is believed to depend upon thermoreceptors located in the hypothalamus, so that the relevant "central temperature" is that of the hypothalamus 79 LIM, T. P. K. itself or of its perfusing blood. On the other hand, peripheral con- trol is believed to depend upon efferent impulses arising in cutaneous thermoreceptors, so that the relevant "peripheral temperature" is that of the skin. On the basis of this general premise, the relationships between the temperature control and detector mechanisms may be inves- tigated, the principal aim of the study being to evaluate the relative roles of central and peripheral temperatures in the initiation of thermal panting in the warmth and shivering in the cold. The roles of central and peripheral temperatures in the initi- ation of thermal panting or shivering have never been adequately defined. Some investigators stress the importance of central control, others of peripheral control, and some admit the possibility that perhaps both are involved. One fact which can be regarded as firmly established is that so called "pure central panting" can be produced by local heating of the hypothalamus in the anesthetized cat (Magoun et al., 1938). In addition, recent data indicate the existence of "pure central shivering" which can be produced by local cooling of the hypothalamus in the unanesthetized cat (Kundt, Bruck, and Hensel, 1957) and in unanesthetized dogs (Hammel and Hardy, 1959). Nevertheless, a great deal of uncertainty exists as to whether cen- tral and peripheral temperatures interact to initiate thermal panting or shivering. The reasons for such ambiguity are partly due to (1) the erroneous choice of rectal or colonic temperature as representing the central (brain) temperature, in particular, during transient states of induced hypothermia or hyperthermia, (2) the inherent difficulty of dissociating central and peripheral tempera- tures in the intact animal, and (3) the poorly defined criteria for the onset of thermal panting or shivering. Thus it appears that the key for the successful answer to the problem is to overcome these three shortcomings revealed in the previous studies. In other words, (1) the brain temperature has to be measured directly in the vicinity of the hypothalamus instead of measuring rectal or colonic temperature, (2) the central and peri- pheral body temperatures have to be dissociated to allow a clear distinction between the action of two regional thermal stimuli and (3) the onsets of thermal panting and shivering must be defined. 80 CENTRAL AND PERIPHERAL MECHANISMS METHODS To meet the necessary requirements which are described above, it was necessary to use the anesthetized preparation. The mongrel dogs were anesthetized with barbital sodium with or without morph- ine sulfate. In some series chloralose was also used. The use of anesthesia has its advantages and also its disadvantages. On the side of advantages, it eliminates one of the sources of ambiguity, namely, the panting or shivering responses due to so-called conditioned reflexes. Barbital sodium is one of the long- acting barbituates and our preliminary study indicates that a stable background of normal blood pressure and respiration can be maintained for more than 8 to 10 hours following a single dose of this reagent. On the side of disadvantages, of course, it is well known that the barbiturates de- press the thermoregulatory mechanisms. However, this disadvan- tage is minimized by using the minimum amount of anesthesia necessary for the surgery. Body temperature was measuredby means of copper-constantan thermocouples soldered into hypodermic needles. In the study of thermal panting, the hypothalamic thermocouple was inserted through a trephine opening in the parietal bone. The angle and depth of insertion required to reach the desired area were determined previously by measurements on the brains of dogs of similar size. We used the medium size of dogs. The position was checked at post-mortem in each experiment, and was found to be within the thermosensitive area described by Magoun and his co-workers (1938). This technique of parietal approach, however, is rather time consuming. Therefore, a simple method of orbital approach was used in the study of shivering. The latter technique consists of implanting a thermocouple through an orbital fissure or optic foramen, into the hypothalamic region. In dogs these cranial openings are rela- tively large and straight. The hypothalamic, esophageal, gastric, and rectal thermocouples served as core temperature sensing ele- ments. Four oher thermocouples were inserted subcutaneously in front and hind legs and in the anterior and lateral abdominal walls. 81 UM, T. P. K. In the experiments in which head and trunk temperatures were independently regulated, head temperature was controlled by warm- ing or cooling the carotid arterial blood. This was accomplished by inserting a glass coil in the course of arterial blood and then the glass coils were immersed in the water bath (Fig. 1). Figure 2 shows the glass coils (length = 1,5 m, i.d. =4 mm, o. d. = 6 mm): the proximal end of a caret id artery is connected to the inlet B and blood temperature is monitored by the thermocouple D. The three-way stopcock E served as a tap throu^ which air bubbles or clots may be removed. The temperature oi warmed or cooled arterial blood is recorded by the thermocouple C and the blood returns to the head of the animal through the outlet A which was cannulated to the distal end of a common carotid artery. In the study of thermal panting, the temperature of the animal's trunk was controlled by heating or cooling the trunk alone in a tem- perature cabinet which had heating and cooling units. On the other hand, in the studies of shivering, the regional cooling or warming of the trunk was accomplished by means of ice or warm water con- tained in a plastic sheet which covered the animals from the neck down to the thigh and extremities. The onset of thermal panting was defined arbitrarily when the respiratory rate reached lOO/min. To define the initiation of shiv- ering, the electromyogram was monitored from four muscle groups at the regionsof neck, thigh, lower, and upper legs. For this purpose the Gilson's electromyograph and integrator were used. The tracing A in Figures indicates the muscle potentials for 10- second duration. When any one of the afore- mentioned muscles begins to show a deflection of 3 mm (i.e., 6 to 9 mV in terms of the integrated po- tential) it is designated as the onset of shivering. 82 CENTRAL AND PERIPHERAL MECHANISMS CAROTID HEAT EXCHANGE SYSTEM I k 1 r 0000000000 '^ -- glass coils water bath o shell thermocouple • core thermocouple Figure 1: The method of regulating head temperatures by warming or cooling the carotid arterial blood. 83 LIM, T. P. K. Figure 2. The glass coils (length - 1. 5 m, i.d. - 4 mm, o.d. - 6 mm) which were immersed in a water bath. 84 CENTRAL AND PERIPHERAL MECHANISMS ^^^',f(H^v^y*^JIi(^^^ ONSET OF CENTRAL SHIVERING ^'r#^;:w::^v^v'' *v0 TT ro O- ^ 0 O 0 0 -N o 3 41 41 41 41 41 41 r^ - f'^ X to > M ^ I— I c vO 0) ;^ o > •H g. -H 3 0 .^ Pw en -o vO m •> rt , u o 0) « -H fn l- (M 0) r- 03 a CO • E 0.5) which implies that the shivering observed in whole body cooling is primarily of reflex origin. From these i-esults it is concluded that shivering can be produced in anesthetized dogs by peripheral thermal stimulus alone when the hypothalamic tempera- ture is maintained at the range of 38.0 C to 40.0 C. Series VIII - Repetitive Cooling and Teimination of Shivering by Central and Peripheral Warming. The data acquired in the above three Series in the cold (Table II) indicate that both central and peripheral shivermg exist and that shivering observed in whole body cooling is predominantly of peri- pheral origin. To supplement these findings, four animals were repeatedly cooled and rewarmed centrally as well as peripherally. All animals responded to such a repetitive procedure. To explore further the interrelationship between central and peripheral temperatures during shivering, central shivering was LIM, T. P. K. produced in three animals (Table III), the hypothalamic temperature being maintained between 34 C to 37 C. This was followed by ° 0 0 warming the animal's trunk (water temperature =40 C to 45 C) to examine whether such a thermal counteraction on the part of the periphery could suppress central shivering. All three animals ceased to shiver within 1 to 2 minutes of peripheral warming at a mean surface temperature of 34.9 C. Utilizingthe same preparation the procedure was reversed. This time, peripheral shivering was produced by surface cooling and then central warming was initiated to 1-ist whether the opposing temperature on the part of brain center could arrest peripheral shivering. All three animals ceased to shiver after 10 to 28 minutes of central heating at an average brain temperature of 41.5 C. From the results of these cooling experiments (Tables II and III), it is inferred that shivering can be produced by differential cooling of the head or the trunk alone as well as by whole body cooling, but the peripheral mechanism of shivering is very potent. o In the studies of thermal panting, peripheral heating up to 45 C (brain temperature, 36 G to 38 C) did not produce panting in dogs under either barbital sodium orchloralose anesthesia. In the studies o on shivering, however, the reductionof skintemperature from 33 C to 30 C while maintaining a brain temperature of 38 C invariably produced a shivering response. This distinctive difference in the roles of peripheral stimuli in shivering andthermal panting leads us to conclude that, in general, the the rm al inf lax from the peripheiy in the cold plays a more dominant role than in a wami environment. To elucidate further, the relative importance of central andperipheral thermal stimuli in the initiation of thermal panting and shivering, the hypothalamic and subcutaneous temperature thresholds for panting and shivering are plotted in Figure 6. The open circles indicate^ temperature thresholds for panting and the closed circles those for shivering in whole body as well as regional heating and cooling ex- periments. The trend of the two curves is shown by the freehand drawings (Fig. 7) which are designated as the iso-panting and the iso-shivering line, respectively. It appears that the slope of the iso- shivering line is much steeper than that of the iso-panting line, which suggests that at a fixed level of brain temperature, the alter- ations in the skin temperature play a more significant role at the onset of shivering than thermal panting. Conversely at a fixed skin CENTRAL AND PERIPHERAL MECHANISMS ro GO O O 00 CVI (M C^l t^ CD ^' ^ U-) 'J- c\i — : c GO O If) ro O ro "^^ 'J' ro d uS -a o ro to ro ro ro CJ c — •~' ^^ — ' — • o 3 ^ O -Q 3 OJ K UJ 00 d CO E .y ro ro rO ro JZ ro ro ro ro ro ^ o ■ — ^^ . . _- ■ — ^ o oJ Q. > < X ID r^ o in 00 U) 00 in iri 00 00 — ro ro ro ro ro ^ o» o> 1 c ? _c o 8 c > o o i •J ■^ ^ •£ in o ■^ CD i O 0) (J) "o Q. (^ c o ^ s 1 O o — o Q. o (/> w c c U) c 'k. y) o (U (U o 0) a> < ^ o ^ m o Q. o 97 LIM, T. P. K. a: 35- 30 ^ mv ^^ SHIVERING THRESHOLD 10 20 TIME (min) Figure 6. Integrated muscle potential during cooling. 98 CENTRAL AND PERIPHERAL MECHANISMS Do (3yniVa3dW3i DIWVlVHiOdXH) ^i 99 LIM, T. P. K. temperature the variations in the brain temperature have a greater influence in the initiation of panting than shivering. The relative roles and interactions of central and peripheral mechanisms at the onset of panting and shivering, as revealed in our stucfy, are certainly established in the animals under anesthesia. However, it is not certain whether such relationships between cen- tral and peripAieral mechanisms also exist in the unanesthetized animal. The elucidation of control mechanisms of bocfy temperature in the unanesthetized animal is quite difficult since shivering or panting may be produced by a conditioned reflex, and also a regional or steady state dissociation of body temperature is almost impos- sible in the intact animal. In addition tothis, due to the current un- certainty, particularly advocated by Dr. Benzinger, a careful study concerning the control mechanisms of body temperature in the unanesthetized animal is needed. Althougji our data are generally in accord with the duality hy- potheses, that is, that both central and peripheral mechanisms are involved in the control of body temperature, the mode of interaction of the two cardinal mechanisms remains unknown. And in addition, despite the classical works by Dr. Ranson and his associates, which provided a valuable evidence for locating the central control mechanisms of body temperature in the hypothalamus, the necessity for additonal work concerning the neuroanatomy of the central thermoregulatory area is apparent. Emplpyingmodernneurophysio- logical tools, it may be possible to demonstrate more clearly the locations and functions of heat dissipation and conservation in the forebrain as well as the midbrain. In conclusion then, perijAieral and central mechanisms for the initiation of thermal panting and shivering have been studied in the anesthetized dog employing a thermal dissociation technique. Our first conclusion is that thermal panting can be produced by central but not by peripheral heating alone. The hypothalamic temperature threshold for panting is higher in central heating alone than in whole body heating. Thermal panting established by whole body heating disappears whenever either central or peripheral temperature alone is lowered. Thus, it is concluded that both central and peripheral temperatures contribute to the initiation of panting, although the central temperature is more important. 100 CENTRAL AND PERIPHERAL MECHANISMS Our second conclusion is that shivering can e produced during differential cooling of the head or the trunk alone as well as in the course of whole body cooling. Either "peripheral" or "central" shivering can be produced repeatedly and also inhibited by elevation of skin temperature in "central shivering" or brain temperature in "peripheral shivering". From these data, it is inferred that the onset of shivering in the anesthetized animals depends upon both central and peripheral temperatures; the peripheral temperature in shiveringinthiscase is more potent than the central temperature. 101 LIM, T. P. K. DISCUSSION DR. HENSEL: May I make a general suggestion not to use the word "receptor" for the central mechanism. I think it is better to use "receptor" only for a defined mechanism. Of course, the central mechanism can be sensitive to temperature. It can be changed in the activity of some neural unit, but it must not be a receptor in the sense of receptor. DR. LIM: Not in the anatomical sense. DR. HENSEL; May I ask another question? I think it is very important to discriminate between the natural conditions and the artificial conditions in the experiments because if you cool an animal from outside, under natural conditions, you would get at the same time an increase in hypothalamic temperature. I think under natural conditions it very seldom occurs that hypothalamic temperature actually drops during external cooling. The drop in hypothalamic temperature can be brought about by artificial cooling, of course, but it does not occur when cooling a cat in a cold chamber. DR. LIM: Yes, we knew this. That was the one reason we added this rather artificial situation. The main purpose was to dissociate two areas, by means of which we can maintain amy level of steady state. DR. HENSEL: Yes, of course, the discrimination of the mechan- ism is very important. DR. CLARK: May I also make a comment: First of all, I think in allprobability your conclusions are fine. lam not willing to accept 102 CENTRAL AND PERIPHERAL MECHANISMS that your experiments prove them. I would be much happier if you had sectioned the fifth nerve bilaterally, for one thing. Another thing, I am not willing to accept a respiratory rate of one hundred as equal to panting. Much more important than rate as far as panting is con- cerned is amplitude, and I would be much more happy to accept a small amplitude respiration. DR. LIM; Then, what is your definition of panting? DR. CLARK: If I am defining panting, I do not use rate at all. The cat and the dog have to open their mouth, retract the angle of the mouth, and move the tongue. That is a matter of semantics. Then, you said skin and brain are essential elements. Dr. Blatteis of our group has evidence that there are possibly some temperature sensing elements in the iliac vessels, and I do not know of any work myself that rules out the possibility of temperature sensing elements in several other places. I doubt if they are there, but I think that the data are not in. As for your hypothalamic temperatures, Magoun, in his dia- grams, has only indicated a position with reference to a sagittal plane, but gives no data on whether it is medial or lateral. Again, however, I think you are probably right. DR. LIM: Our areas were within three to four millimeters of the center linewhichis very close to the medial line. Oir definitions of panting and shivering are arbitrary, depending on our different workers. I do not know whether we have a clear-cut definition to make everybody happy on this, but I doubt it very much. DR. IRVING: Should you not have some indication of the function of the panting, namely, that it was dissipating heat in an additional degree? I think that is implicit in what Dr. Clark is saying. DR. LIM: Usually the tongue is protruded and, of course, the surface of the tongue is very important in the dissipation of heat. In most of the animals we studied their tongues were protruding. And we think this definition of respiration rate of lOO/min is ade- quate for our purpose. 103 LIM, T. P. K. DR. CLARK: Well, the chief thing is, of course, that with an increase in amplitude you are primarily getting a ventilation of the dead space, which is implied in the term "panting". Actually, I have seen cats panting at temperatures below 38 C. DR. HEMINGWAY: I have studied panting in both anesthetized and unanesthetized dogs and I agree with Dr. Clark; there is an abrupt change in the respiratory pattern when an animal pants. It is quite different from the hyperventilation that you get, say, with the decerebrate animal. It is possible just by raising the tempera- ture of a cold-blooded animal and warm blooded animals that have been made poikilothermic by decerebration to get this hyperpnea or hyperventilation, but the normal animal without anesthesia has a distinct change in pattern. There is a drop in tidal volume which is very striking. I do not think you can really tell with an anesthe- tized animal whether heispantingornot; it is so different from that of the unanesthetized animal. In the former you can have hyperven- tilation or h)T5erpnea, but it is difficult to say when you have panting. With an unanesthetized dog, it is very easy. DR. JOHANSEN: Is it conceivable that the vascular changes in the brain produced by your cooling of the carotid blood could bring in another stimulus that could interfere with the temperature, per se? DR. LIM: I did not mention it in detail, but we did obtain some data on the threshold levels with whole body heating in intact animals, without any insertion of coils into the brain circulation. In other experiments with repetitive heating or cooling, we produced panting by simultaneous heating of the brain and the skin. There was no statistical difference between the results of these experi- ments as far as thresholds are concerned. This would indicate that the disturbance of the cerebral circulation by insertion of the coil may not be a major problem. DR. JOHANSEN: To the best of my knowledge, the internal carotid is the main supply to the tongue of the dog. So, although you warm the nostrils and the skin of the face, you might also have an input from the tongue. 104 CENTRAL AND PERIPHERAL MECHANISMS DR. LIM: That is very possible. DR. HENSEL: There is practically nothing. DR. MINARD: Dr. Lim mentioned Dr. Benzinger's work, and I would like to point out that Dr. Benzinger found it possible in the human subject to bring about this dissociation between skin and deep temperature. He also felt, asDr. Lim does, that this is important in determining the regulatory responses in the conscious unanesthe- tized human subject. It is possible to bring about such dissociation by having the man either at rest or working, in a hot or cold environ- ment. By selecting various environmental conditions and work rates, it is possible to obtain a wide variety of core and skin temperatures. DR. LIM: In that connection, I was very happy to learn recently of Dr. Benzinger's finding, that the skin temperature plays a very important role in the cold. This was exactly what I found. DR. MINARD: I would say, from what I know of Dr. Benzinger's work, that there is no disagreement whatever in your results and his regarding cold conditions. I think he has come to some interpre- tation of a mechanism acting centrally, but as far as the experi- mental findings go, there is absolutely no disagreement between his findings and yours. DR. HENSEL: I think it is necessary to emphasize that Dr. Benzinger's concept is applied only to sweating in human subjects, not to vasomotor responses, nor to any kind of cold responses. It is a concept that should be restricted to mechanisms. I think he would completely agree with your findings, and I must say that I am very happy about them too. They agree very well with our results in the unanesthetized cat, where we made the local hypothalamic temperature change by means of a chronically implanted the rmode. DR. LIM: In some of our experiments we studied the contribution of the vertebral arteries to the cerebral circulation. Of course, the brain circulation is mainly through the carotid arteries. Ligation of the vertebral arteries did not make much difference in our experi- mental results, so we think, at least in dogs, that the blood going to the brain through the vertebral arteries is rather small. 105 LIM, T. P. K. DR. HEMINGWAY: It is possible, though, to tie out both common carotids. I have seen this done in a dog, and the dog is perfectly happy for a while, except for some atrophy of the facial musculature. He will live for a few days without showing any effect, thus indicating the vertebral circulation is extensive enough to substitute for the carotid circulation. MR. ADAMS: I wonder if Drs. Clark and Hemingway would be willing to accept an index of panting based on respiratory, evapora- tive heat losses, rather than on respiratory volume or pattern? DR. HEMINGWAY: Like your water loss, it goes up very abruptly. If you will notice, when a dog starts to pant, he first takes a few deep breaths; then all of a sudden there is an abrupt rise in respiratory rate and a drop in tidal volume at the same time. If you measure the water outputthat goes up abruptly, too. It should also be noted that it is also possible to stop shivering without any change in skin temperature by just warming the anterior hypothalamus. A number of people have done this. No change at all in skin tempera- ture is observed. MR. EAGAN: Is it possible thatthere is a sufficient association of the heat loss center and the heat conservation center, that you could get, with a suitable choice of conditions, a pre-panting hyperp- nea at the same time as shivering? DR. LIM: I do not know, but I accept it for the sake of the dis- cussion. I do not know how clearly the two centers are defined. DR. CLARK: I would say that the center for heat conservation is fairly well defined andthat the center for heat dissipation is much less clearly defined than we would be led to believe from reading the literature. MR. EAGAN: I was thinking of the functional association. DR. CLARK: I would not expect simultaneous shivering and panting because where you have opposing centers, you are going to have interconnecting inhibitory and excitatory fibres. 106 CENTRAL AND PERIPHERAL MECHANISMS DR. LIM: The only experiment I have that would suggest such a phenomenon is one where we studied exercise in the unanesthetized dog. We put the animal on a treadmill which was immersed in a water tank, so that he could be cooled during exercise. It appeared that such animals shivered as well as panted. Here again this was observed in the unanesthetized dog and there may be a conditioned reflex involved. DR. HENSEL: I think one must be careful with the studies during work because there is evidence that work will set the thermo- stat to another level. It is not just a rise in the core temperature. DR. HEMINGWAY: Anesthetics such as morphine really disturb temperature regulation. For example it is possible to get shivering and what appeared to be panting at the same time with morphine anesthesia. This is such a disturbance caused by the anesthetic. I have always felt that temperature threshold values for skin temper- ature and rectal temperature and brain temperature under anesthesia are quite variable. You can get almost anything you want, depending upon the depth of the anesthesia. After giving phenobarbital anes- thesia, all these thresholds are depressed, then they come back. Shivering comes back first andthethermocutaneous vasomotor res- ponse comes back later. However, you are never sure just what the anesthetic level is or how much depression has been caused by the anesthesia. DR. LIM: That is certainly true in short- acting anesthesia, par- ticularly amytal, but barbital- sodium is long acting. We have based several studies on how the latter affects the blood pressure, res- piration, and body temperatures. They all stay fairly constant, and within normal range for eight hours or more after a single dose. Thus, we thought barbital- sodium was the best preparation for our studies because it maintains a steady background. DR. HENSEL: Concerning the vasoconstriction during hypothal- amic cooling, there is quite a high impairment in the anesthetized as compared to the non- anesthetized cat. It can abolish the whole mech- anism even during the light anesthesia. I do not know how it is with studies of shivering, but this is a difficult problem in circulatory studies. 107 LIM, T. P. K. DR. HEMINGWAY: Anesthetics cause the skin temperature to go up very quickly, and it stays at certain elevated levels for a long period of time. Thus, it would seem that the cutaneous vasomotor response is so sensitive to anesthetics that they would seriously interfere with skin temperature measurements. DR. FREEMAN: There is one thing about your work that bothers me a great deal and makes me wonder about its relevance to normal processes. This is the high temperatures that you have to use to activate whatever you are stimulating. This has been my experience also. In order to get panting, or, for that matter, vasodilation in an anesthetized cat, you have to heat above the level which we ordinarily consider to be harmful to the brain. That is, above 42 C, which is so high that the response will decay if you maintain it for any length of time. Furthermore, if you try to get it twice or three times, you may get it the second time, never the third. This is associated with a marked disturbance of the animal. His pulse rate will usually go way down, his blood pressure will go down, his nocioceptive reflexes will disappear, and his muscle tone disappears completely; he is completely placid, and there is a very slow recovery period there- after. I think what this thing represents is some form of brain damage which may be related to manifestations you see in sunstroke. DR. LIM: With our dosage of anesthesia, as shown in the first figure, the temperature threshold in both central and peripheral areas is somewhere around forty-one. I would think of forty-two or forty-three degrees as being high, but we were using forty-one degrees centigrade in the panting studies. DR. FREEMAN: There is an interesting history of this so-called central response. In the first clear distinction of central versus peripheral heating by Roche", somewhere back around 1898, he de- fined central panting more in terms of its overwhelming durability in the face of anoxia. He set 40.7 C as a specific level of an emer- gency reaction. When panting did occur at this level, it would per- sist in the face of anoxia or asphyxia for a period of two, three, or even four minutes, whereas panting induced by external heating in an intact animal would be interrupted in a matter of thirty seconds or so. Now, most people that have tried to get this central phenomenon find that it is dependent upon the type of anesthesia being used. For example, Magoun published a short statement some two years after 108 CENTRAL AND PERIPHERAL MECHANISMS his original paper saying that they had only been able to get it in animals under urethane or under ether. This was our experience also. We did not get it under phenobarbital except in one instance. The people in Sweden have also found it only under chloralose and urethane anesthesia. Hess reported that he only saw it in cases where he was making lesions by local heating, but never in the course of minimal diathermy. Dr. Hemingway, what is your ex- perience in conscious animals? How high do you have to heat to get this response? DR. HEMINGWAY: We were not able to produce panting by heating the anterior hypothalamus with diathermy, but we were able to stop shivering and cause cutaneous vasodilation. During such ex- periments we could not measure the hypothalamic temperature except very roughly — perhaps to one-half or one degree. This was due to the diathermy current interfering with the temperature measurements. All of our measurements, therefore, had to be de- termined after the diathermy current had been turned off, during which time the temperature was falling quite quickly. Typically, heating of the anterior hypothalamus cause a cutaneous vasodilation, a cessation of shivering, but no panting. DR. LIM: Was there any increase in respiratory rate, say, in terms of so-called pre-panting hyperpnea? DR. HEMINGWAY: No indication at all. DR. HENSEL: We saw the onset of panting in the unanesthetized cat, during local hypothalamic rewarming. DR. HEMINGWAY: Where were you heating it? DR. HENSEL: In the anterior hjrpothalamus. DR. LIM: Dr. Hammond, in Dr. Hardy's group, produced it in the unanesthetized dog. DR. HEMINGWAY: Dr. Magoun reported it in the anesthetized cat, but he was heating the anterior hypothalamus. I was using dogs, unanesthetized dogs. 109 LIM, T. P. K. DR. MINARD: I want to clarify one point. When I spoke of Dr. Benzinger's work being in no disagreement with Dr. Lim's, I was referring to his recent work on the cold side; but with reference to what he has observed in the control of sweating on the warm side, there is this difference: he regards the central temperature as the only controlling factor in that sweat rate is independent of skin temperature. DR. HENSEL: This is quite possible, but in the unanesthetized condition, the peripheral mechanism may be quite different. DR. RODDIE: I think that we have quite good evidence that there is a peripheral element looking after both heat and cold. People like Kirschlich and Cooper ha ve shown radiant heat on one extremity will cause a vasodilation in the opposite extremity within several sec- onds; and this was abolished if they cut the sympathetic nerves which were supplying the radiated area. With respect to cooling, I think most people, if they cool a subject rapidly, find that the shiv- ering commences very soon, and at a time when esophageal temper- ature, which is about as close to a hypothalamic as one can get in man, is actually elevated. DR. LIM: Can you get a peripheral mechanism? DR. RODDIE: The peripheral mechanism is quite strong. 110 CENTRAL AND PERIPHERAL MECHANISMS REFERENCES 1. Benzinger, T. H. On physical heat regulation and the sense of temperature in man. Proc. Nat. Acad. Sc. 45:645, 1949. 2. Birzis, L. and A. Hemingway. Efferent brain discharge during shivering. J. Neurophysiol. 20:156, 1957. 3. Hammel, H. T., J. D. Hardy, andM.M. Fusco. Thermoregula- tory responses to hypothalamic cooling in unanesthetized dogs. Am. J. Physiol. 198:481, 1960. 4. Kundt, H. W., K. Bruck, and H. Hensel. Das Verhalten der Hautdurchblutung bei Kuhlung des vorderen Hypothalamus. Naturwissenschaften 44:496, 1957. 5. Lim, P. K. and F. S. Grodins. Control of thermal panting. Am. J. Physiol. 180:445, 1955. 6. Lim, T. P. K. Central and peripheral control mechanisms of shivering and its effect on respiration. J. Appl. Physiol. 15: 567, 1960. 7. Magoun, H. W., F. Harrison, J. R. Brobeck, and S. W. Ranson. Activation of heat loss mechanisms by local heating of the brain. J. Neurophysiol. 1:101, 1938. 8. Perkins, J. F. The role of the proprioceptors in shivering. Am. J. Physiol. 145:264, 1945. ill THE ROLE OF VASOCONSTRICTOR AND VASODILATOR NERVES IN THE CONTROL OF THE PERIPHERAL CIRCULATION Ian C. Roddie Department of Physiology The Queen's University of Belfast Belfast, North Ireland Over the last few years, there has been a gradual development in our notions of the nervous control of peripheral blood vessels in man. Today I want to review the pattern which seems to be emerging. As you know, blood flow in the limbs is mainly distributed to skin and muscle and both these tissues appear to have a rich vasomotor innervation. All these fibres are not actively concerned in tempera- ture regulation but an understanding of their various functions is very helpful in the design and interpretation of experiments on the peripheral vascular responses to heat and cold. First of all, there- fore, I will deal with the vasomotor control of muscle vessels and then with that of skin vessels. VASOMOTOR NERVES TO MUSCLE Vasoconstrictor Fibres Until 1943 the available evidence si^gested that muscle blood vessels had little or novasoconstrictor innervation. When measured as soon as one week following cervical sympathectomy, forearm blood flow was not found to be increased (Stein, Harpuder, and Byer, 1948). Blocking the deep nerves to the muscles did not increase the temperature of the overlying skin (Wool lard and Phillips, 1932; Grant and Pearson, 1938). Stimuli which caused large changes in vasocon- strictor tone in the hands and feet produced little or no change in flow in the muscular parts of the limbs (Abr am son, 1944). The infer- ence that muscle had a negligible vasoconstrictor innervation 113 RODDIE, I. C. seemed justified. However, appropriate experiments showed that this was not the case. Using venous occlusion plethysmography to measure forearm blood flow, Barcroft, Bonnar, Edholm and Effron (1943) found that blockingthe motor nerves to the forearm with local anaesthetic increased flow two to three fold. This increase did not seem to be in skin since it was seen even when the cutaneous circu- lation was suppressed by iontophoresis of epinephrine into the skin. Blocking only the cutaneous nerves to the forearm did not increase flow. From these and other control observations it was concluded that skeletal muscle is normally subjected to appreciable vasocon- strictor tone and that the fibres concerned were distributed to the muscle in the motor nerves. It was not clear at thattime what role these fibres played in the normal regulation of the circulation. Reflex changes in forearm blood flow were known to occur during body heating (Wilkins and Eichna, 1941). Barcroft, Bonnar and Edholm (1947) found that attempted suppression of the cutaneous circulation in the forearm by the epinephrine iontophoresis technique failed to prevent the re- flex vasodilatation during body heating. They concluded that release of vasoconstrictor tone in muscle contributed to this vasodilatation. However, in 1952, McGirr found that the rate of clearance of Na from human muscle did not increase during body heating. This result could not be reconciled with those of the plethysmographic experiments unless it were postulated that vasodilatation occurred in 'non-metabolic' vessels. This was not a very satisfactory state of affairs and further work on the subject became necessary, Barcroft, Bock, Hensel, and Kitchin (1955) used heated thermo- couples to estimate muscle blood flow. They found that forearm flow did not increase during body heating; a fall was the usual finding. Roddie, Shepherd and Whelan (1956) used changes in the oxygen saturation of effluent venous bloodfrom skin and muscle to estimate simultaneously the contribution of these tissues to the vasodilatation. Though the oxygen saturation of skin blood rose to almost full satur- ation, there was no significant rise in that from muscle (Fig.l). Im- provement in the technique for iontophoresis of epinephrine into forearm skin allowed Edholm, Fox, and Macpherson (1956) to pro- duce more effective suppression of the skin circulation than had been obtained in earlier experiments (Barcroft, Bonnar, Edholm, and 114 CONTROL OF PERIPHERAL CIRCULATION Effron, 1943). With this degree of suppression forearm flow did not increase during bocfy heating. It was clear, therefore that vasocon- strictor fibres to muscle were not involved in the peripheral vaso- dilatation during body heating. Blair, Glover, and Roddie (1960) found that the fall in forearm blood flow during body cooling could be prevented by blocking the cutaneous nerves in the forearm, pro- viding further evidence that vasoconstrictor fibres to muscle do not take part in thermoregulatory reflexes. The normal function of these fibres remained obscure. Over the last few years, the reflex effects of passively raising the legs of a recumbent subject on peripheral blood vessels have been studied in considerable detail (Roddie and Shepherd, 1956; Roddie, Shepherd, and Whelan, 1957). This stimulus caused reflex vasodilatation in the forearm but there was no comparable change in the hand (Fig. 2). The oxygen saturation of muscle venous blood was increased but that of skin was not affected. These findings sug- gested that muscle rather than skin vessels were responsible for the vasodilatation. The changes were mediated through sympa- thetic vasomotor nerves since they were abolished by acute nerve- block or by cervical sympathectomy. Release of vasoconstrictor tone rather than vasodilator fibre activity was never greater than could be accounted forby full release of vasoconstrictor tone; it was not reduced by atropinizing the forearm but was abolished by intra arterial infusion of the sympatholytic agent bretylium tosylate. It was concluded that alterations in vasoconstrictor tone in muscle were responsible for the changes in forearm blood flow with change in posture. A wide variety of stimuli are now thought to affect the level of vasoconstrictor tone in muscle. Negative pressure breathing (Blair, Glover, and Kidd, 1959), squatting (Sharpey-Shafer, 1956) and intra- thoracic pressure transients (Sharpey- Shafer, 1953; Roddie, Shep- herd, and Whelan, 1958) are thought to cause reflex vasodilatation. Tilting a subject into the vertical position (Bridgen, Howarth, and Sharpey-Shafer, 1950), positive pressure breathing (Blair, Glover, and Kidd, 1959), the Valsalva manoeurve (Roddie, Shepherd, and Whelan, 1958; Sharpey-Shafer, 1955), exercise (Blair, Glover, and Roddie, 1961), radial acceleration (Howard and Garrow, 1958), and hypercapnia (McArdle, Roddie, Shepherd, and Whelan, 1957) are 115 is o RODDIE, I. C. O2 SAT. % ■ SUPERFICIAL • DEEP y^*^^ /y*^ 100 JjT 80 60 Y %/ =^?-Hy^*^-^^ 40 - O 20 40 60 MINUTES Figure 1. The effect of body heating on the oxygen saturation of deep and superficial venous blood and forearm blood flow. TTie black rectangle represents the period of body heating. Left forearm blood flow, o ; oxygen saturation of superficial venous blood (right forearm) • ; oxygen saturation of deep venous blood (rig^t forearm) ai . (After Roddie, Shepherd, and Whelan, J. Physiol. (Lond.). 134: 444, 1956). SSI''"" J \ I t=^ «4 J Figure 2. Changes in forearm blood flow with change in boc^y posture. The open rectangles representthe periods duringwhich the subjects' legs were raised from the horizontal to the vertical position. 116 CONTROL OF PERIPHERAL CIRCULATION thought to cause reflex vasoconstriction. There is evidence that al- terations in the activity of stretch or baroreceptors in the low pres- sure area of the intrathoracic vascular bed may be responsible for some of the changes in the vasoconstrictor tone (Roddie and Shep- herd, 1958), but this part of the problem is quite complex and there is not time to discuss it today. Vasodilator Fibres The first evidence that human skeletal muscle had a vasodilator innervation was provided by BarcroftandEdholm (1945). They found that vasodilatation occurred in the forearm durningfainting(Fig.3). The conclusion that the vasodilation in the normal forearm was actively excited was based on a comparison of the average levels of flow in nonnal and nerve-blocked forearms during the faint; the flow in the normal arm rose to a higher level than that in the nerve- blocked arm. Though this evidence is highly suggestive, it is not absolutely conclusive since the observations on the normal and nerve-blocked forearms were not made simultaneously in the same subject. However, evidence has since been found that vasodilation in muscle is a constant feature of the vasovagal syndrome (Anderson, Allen, Barcroft, Edholm, and Manning, 1946). It seemed unlikely that a potent vasodilator system should exist just to facilitate fainting so a vigorous search was made to find a stimulus which would excite these fibres under more normal phys- iological conditions. As mentioned before we could find no evidence that these fibres were involved in bare receptor reflexes. Another incentive to search for a stimulus was the abundant evidence for these fibres supplying muscle in other experimental animals. In the cat and dog, cholinergic vasodilator innervation of skeletal muscle is well established and the central connections and efferent distri- butions of these fibres has been extensively studied (Uvnas, 1954). It was found that they played no part in either chemoreceptor or baroreceptor reflexes but the exact nature of the physiological re- flex stimulus was not obtained. Abrahams and Hilton (1957) have recently shown that stimulation by electrodes implanted in those 117 RODDIE, I. C. places in the brain stem which discharge these fibres in anesthe- tized cats provoke the 'fight or flight' reaction in conscious cats. It seemed possible, therefore, that emotional stress might be the effective physiological stimulus. It has been known for a long time that stimuli such as mental arithmetic frequently cause an increase in forearm blood flow (Abramson and Ferris, 1940; Grant and Pearson, 1938). GoUehhoff en and Hildebrandt (1957), using Hensel's heated thermocouples to measure skin and muscle blood flow, found that the vasodilatation occurred mainly in muscle, but this response was usually attributed to epinephrine. Evidence that cholinergic vasodilator fibres con- tributed to vasodilatation in muscle during emotional stress was ob- tained in some experiments where we tried to provoke a 'fight or flight' reaction in our subjects (Blair, Glover, Greenfield, and Roddie, 1959). Medical students were frightened, worried, or em- barrassed by means judged likely to be most effective for each in- dividual (Figs. 4 and 5). In most experiments the subject had an indwelling needle in the brachial artery. The operators, by their whispered conversation and demeanour, led the subject to believe that they were alarmed because of blood loss at the arterial punc- ture site and the precarious state of the subject's health. About half of the subjects complained of pain in the arm and throbbing in the head. After a few minutes the real purpose of the hoax was ex- plained and anxiety was promptly relieved. Figure 6 shows the result of one such experiment. During the period of stress forearm blood flow rose to about 50 ml/100 ml/min, a level similar to that found immediately after severe exerciseof the forearm muscles and much greater than that usually achieved by full release of vasoconstrictor tone. The size and rate of recovery of the vasodilation were greater than that usually seen during epinephrine infusions. The changes in hand blood flow and arterial pressure also were not typ- ical of those seen with epinephrine infusions. It seemed unlikely that release of epinephrine from the suprarenal glands in response to stress could fully explain the response. The fact that forearm but not hand blood flow increased during stress suggested that muscle vessels might be responsible for the vasodilatation. This was supported by the finding that the oxygen saturation of muscle but not skin venous blood increased during stress (Fig.7). Sympathetic vasomotor fibres to muscle contributed to the vasodilatation since it was reduced by blocking the motor nerves to the forearm and by 118 i CONTROL OF PERIPHERAL CIRCULATION lO- — ^ NORMAL FOREARM e — ^ DEEP NERVES BLOCKED - i^- - _W!"-^ i 25 SO Minutes FAINTING. 9 NORMAL 6BLOCKED Figure 3. Changes in blood flow in the normal and nerve- blocked forearm during fainting. Fainting occurred at the time indicated by the vertical dashed line. (After Barcroft and Edholm, J. Physiol. (Lond.). 104:161. 1945). 5 lO MINUTES Figure 4. Changes in forearm blood flow during emotional stress. At the beginning of rectangle A, the subject, who was a medical student, was told thathe would be given an oral examination in physiology within a few minutes. During rectangle B, he was given an oral examination. During rectangle C, personal remarks were made about an acquaintance of the subject. Though he showed no outward signs of emotion, forearm blood flow increased consid- erably. 119 RODDIE, I. C. MINUTES Figure 5. Forearm blood flow duringemotionalstress. During rectangle A, the subject, a departmental head, was given mental arithmetic to do. During B, he was led to believe that boiling water was being poured on his hand, and during C, he was told that a fire had broken out in his office. O 3 6 9 MINUTES Figure 6. Effect of severe emotional stress on arterial pres- sure, heart rate, forearm blood flow, and hand blood flow. During the time represented by the rectangle, it was suggested to the sub- ject that he was sufferingfrom severe blood loss (After Blair et al., J. Physiol. (Lond.). 148:633, 1959). 120 CONTROL OF PERIPHERAL CIRCULATION cervical sympathectomy (Fig. 8). It was not affected by blocking only the cutaneous nerves to the forearm. The following observa- tions suggested that the vasomotor contribution was mediated through activity of cholinergic vasodilator fibres rather than release of vasoconstrictor tone. In some cases emotional stress produced a greater vasodilation in the normal forearm than in the contralateral nerve-blocked forearm. The vasodilation was reduced and oc- casionally abolished by atropinizing the test forearm (Fig. 9). Bretylium tosylate, in doses which abolished reflex vascular re- actions usually attributed to alterations in vasoconstrictor tone in muscle, did not affect the vasodilation during stress. It was con- cluded that cholinergic vasodilator fibres are distributed to human skeletal muscle and that they are activated during emotional stress. It has been suggested that generalized vasodilator activity at the beginning of exercise might help to adapt the circulation to the circulatory needs of exercise rapidly (Uvnas, 1954). In an attempt to determine the part played by vasomotor fibres in the limbs in the general circulatory response to exercise, blood flow was measured in the hands andforearms of recumbent subjects during leg exercise on a bicycle ergometer (Blair, Glover, and Roddie, 1961). Exercise resulted in a considerable increase in vasoconstrictor tone in muscle butthere was no evidence that discharge of vasodilator fibres was an integral part of thegeneral vasomotor response to exercise. In inexperienced subjects vasodilator discharge was occasionally seen during exercise but it is likely that this discharge was due to associated emotional stress. Emotional stress may also contribute to the muscle vasodilation described during fainting (Barcroft and Edholm, 1945). Vasomotor Nerves to Skin Because of the relative ease with which estimation of blood flow can be made in the hand or foot, the vascular innervation of the extremities has been extensively studied. However, it should be stressed that the vasculature of the hand and foot are very hi^ly specialized and adapted to serve temperature regulation. Many false 121 RODDIE, I. C. MINUTES Figure 7. Oxygen saturation of blood from superficial forearm veins H and deep forearm veinsC3: forearm blood flow in the opposite forearm • . During the time represented by the rectangle, it was suggested to the subject that he was suffering from severe blood loss. (After Blair et al., J. Physiol. (Lond.). 148:633, 1959). Figure 8. Effect of stress on bloodflowthrough a normal • and a sympathectomized o forearm. At A, the subject performed a Val- salva manoeuvre, and during B, a mental arithmetic problem was given. (After Blair et al., J. Physiol. (Lond.). 148:633, 1959). 122 CONTROL OF PERIPHERAL CIRCULATION impressions of vasomotor regulation in the body have been gained by considering vasomotor responses in the hand to be representative of those occurring in other tissues. It is now quite clear that the conclusions from experiments on hand skin cannot be extrapolated to forearm skin, let alone to the other tissues of the bocfy. For this reason it is convenient to consider the innervation of different skin areas separately from that of hand skin. Forearrn skin. Vasomotor innervation of forearm skin can be conveniently demonstrated by the type of experiment illustrated in Figure 10. Blood flow was measured by venous occlusion plethys- mography in both forearms of a comfortably warm subject. At the beginning of the experiment the level of blood flow was similar on the two sides. When the cutaneous nerves to one forearm were blocked, there was little change in the level of flow on that side relative to that on the control side. This indicated that the cutaneous vessels are not subjected to any appreciable vasoconstrictor or vasodilator tone when the subject is comfortably warm (Edholm, Fox and Macpherson, 1957). When the subject was then cooled, by direct- ing a fine spray of coldwater over his chest and abdomen, the blood flow on the innervated side fell to a lower level than that on the nerve-blocked side. This result indicated that cutaneous vessels are innervated with vasoconstrictor fibres which are active in temperature regulation (Roddie, Shepherd, and Whelan, 1957a). It has recently been shown (Blair, Glover, Kidd, and Roddie, 1960) that these fibres are adrenergic, since the vasoconstriction of fore- arm skin on cooling the body can be prevented by intra arterial infusion of bretylium tosylate. When a recumbent subject is comfortably warm, the rate of blood flow through the forearm is about 4-5 ml/lOOml/m in. Cooling the body causes flow to fall to about 1.5 to 2.5 ml/lOOml/min, and as indicated above, this reduction is due to activity of vasocon- strictor fibres. A reduction of the same order is produced by sup- pression of the skin circulation by iontophoresis of epinephrine into the skin (Cooper, Edholm, and Mottram, 1955). It is likely, therefore, that vasoconstrictor nerves to the forearm skin, like those to the hand skin, can stop the flow of blood through the skin tissue completely. 123 RODDIE. I. C. 40 30 O c 5^ 6 20 2 ^ 10 & A 1 'K ?6 d CO J_, ^ (U o 0 o 0 o 0 o CO >^ C! c c c a c C GO S CO 0) 0 o 0 o 0 0 0 >< c c c c fi c fi rt Oh L >> « »^ Final r spirato; rate O (N] vO (V] -* sO ^ CM ^ r- ,—1 ro rg CO rsj CO i-H G . CO C ■rH OJ o ^ LD in (M ■* o ^ fvi (vi r^ ^' ^' tvi ^ CO Pi *^ ^ ^ CO rt (D O^ ^^ ro ro (M CO rj CO ^ fM O ^H ^^ ,— ( r-l ,—i ,_l ,—i ^ '^ ^ ^ ^ •^ ^ ^ (U p. H a; 0 C ^ 0 u ^ < — *i— oo uo rsj »— H- I— ro n "^ ^? 2 3 ^ a '^ '+^ooo^oooo^oocrors' o oooooooooooo 3333P 3333333 OOOOO OOOOOOO )-( vt) 3 fc era fj 156 ANTERIOR HYPOTHALAMIC LESIONS > i I 1 § s -S yes yes yes a, " = ° ° : : : : £ \l t ... S ! 5 . 2 - ::^ ^- - " :;■ "• " tih 41.2 41. 1 41. 1 39.9 41. 1 40.5 41. 1 > a. V" i 1 'i E E i- E (2 " E E E E E E E CO ? S 1 1 = £ 1 " 1 " Z o- o o- o o 40.3 40.2 39.6 39.5 39.3 40,0 39.3 o i- ?/% 1 = s , , (U en ■" O ;f si E| 157 CLARK, G. temporary and reversible damage to the heat sensitive area probably as a result of edema which subsided in the first postoperative month. If this is the case then it would appear probable that the results in the animals with symmetrical lesions were not due to edema. For this reason it was assumed that a steady state was present in the animals with symmetrical lesions, or in other words, that the damage was solely due to loss of tissue and therefore permanent. The foregoing was thought to supply a localization of the dual centers for temperature regulation. In his text Ranson (1947) sum- marized this theory as follows (Figure 7): "On the basis of all the available evidence it now seems clear that a center controlling heat- loss functions such as panting and sweating is situated in the pre- optic region and that a pathway from this center runs backward through the lateral hypothalamus. The center for preventing heat loss by vasoconstriction and for increasing heat production by shivering is situated in the hypothalamus proper; and its descending pathway also runs backward through the lateral hypothalamus. Both descending pathways run close together dorsolateral to the mammillary body and enter the mesencephalic tegmentum. Bilateral lesions in the caudal part of the lateral hypothalamus interrupt both pathways and interfere withbothheat loss and heat conservation mechanisms. Bilateral lesions in the preoptic region destroy the heat-loss center leaving the heat conservation center intact; and as a result the body temperature either remains normal or may be temporarily elevated." It has been routine in studies of the effects of various lesions on the thermoregulatory ability of experimental animals to utilize only external heat or cold loads. There are, however, other con- ditions where the dual center theory appears to make definite pre- dictions. Upon assumption of a constant work load it is generally found that when a steady state is achieved the body temperature is at a new and higher level. The dual center theory implies that ani- mals with the anterior hypothalamus destroyed would never attain a new steady state unless the increased heat production were by chance balanced by the increased heat loss occasioned only by the increased body temperature. In most cases in such animals a fatal hyperthermia should rapidly ensue. As far as I know such experi- ments have never been made and yet they are badly needed. 158 ANTERIOR HYPOTHALAMIC LESIONS Similarly the readjustments to pyrogen administration have only been perfunctorially studied. In our early studies (Ranson, Clark and Magoun, 1939) we had found that in cats with medial lesions in the anterior hypothalamus (animals which had no disturbances in ability to withstand cold or heat) the fever curves were markedly altered and there were fatal and near fatal hypothermic responses to pyrogens. On the other hand an occasional cat with disturbed temperature regulation had essentially the same responses as a normal animal. This is difficult to explain within the confines of the dual center theory for presumably animals with damage to the cen- tral apparatus for promotion of heat loss should have fatal hyper- themias to dosages of pyrogens producing only a high fever in normal individuals. Most of the other studies on pyrogens such as Chambers, Koenig, Koenig, and Windle (1949), add little knowledge to the field although two recent reports are of interest. Thompson, Hammel and Hardy (1959) reported that four hypothalectomized dogs failed to develop a fever following injection of a pyrogen. This was confirmed by Bard and Woods (1959) who found that cats with the brain stem sectioned at various levels between the upper mesen- cephalon and rostral third of the pons fail to develop a fever follow- ing injection of an adequate amount of typhoid vaccine. The brain stem of an animal similar to these (Bard and Macht, 1958) is shown in Figure 8. Despite the failure to develop a fever, these animals do show the typical transient leucopenia and other blood changes observable in normal animals. Both reports then indicate the importance of the functional integrity of the hypothalamic mech- anisms which enable an animal to regulate against cold in the devel- opment of a fever. The problem is: how is a rise in temperature in fever terminated in the presence of damage to the mechanisms for regulation against heat? It is also difficult to understand the maintenance of normal bocfy temperatures by cats with hypothalamic lesions in the relatively optimum temperatures of the usual animal room. If a cat is placed postoperatively in an incubator of 32.2 C after most lesions the body temperature will be normal or slightly above normal on the morning after the operation. However a majority of the animals with lesions in the heat sensitive zone will have hyperthemias, in some cases even above 42.2 C. Whenremovedto the animal room, these hyperthemias subside over 2-3 days or so and thereafter the body temperature remains within the normal range. However in cats 159 CLARK. G. Anienor commissure Internal capsule Preoptic region Preoptic region Basis pedunculi Mammillary body y Heat4oss pathway ~ H eal-consenation pathway Figure 7: Diagrammatic representation of the mechanism for temperature regulation superimposed upon schematic drawings of three transverse sections through the preoptic region and hypothal- amus. Figure 8: Dorsal and ventral vle^s of truncated brain stem of Cat 4. 160 ANTERIOR HYPOTHALAMIC LESIONS with damage to the posterior hypothalamus or in animals with ex- tremely large lesions anywhere in the hypothalamus, hypothermias are routine and many of these last a week or more. This finding with posterior hypothalamic lesions fits the dual center theory fairly well but with more anterior lesions the fit is not so good. These hypothermias following large anterior lesions have also been repor- ted by McCrum {X953); however, Keller (1960) has stated that this is not the usual occurence in his series. The difference may lie in the type of lesion for Keller's have routinely been produced by suction or thermocoagulation while ours and those of McCrum (1953) were electrolytic. On the other hand, such hypothermias usually do not result from large purely unilateral electrolytic lesions. One might assume that the reversible hypothermias and hyperthemias might be due to some type of reversible damage to remnants of the dual centers, or, on the other hand, that these normal body tempera- tures are due solely to vasomotor responses which remain active but not fully integrated even in the chronic midbrain dog (Keller, 1960). However an explanation rather than an assumption is needed. These three questions — the effect of exercise, the course of fever, and the maintenance of normal body temperature at the usual ambient temperatures in the presence of damage to one or both of the dual centers — present fields for research rather than insur- mountable difficulties for the dual center hypothesis. However, neg- lect of these problems and continued overgeneralization on insuffi- cient evidence (such as was perpetuated by Strom (1960) in the new so-called Handbook of Physiology) will not lead to clarification of the problems of thermoregulation. These three questions were not considered in Ranson's review and there have been no answers in the past two decades. There has appeared, however, much that has been considered confirmatory evidence for the dual center theory and some work which may be contradictory. Some of the confirming studies will be briefly discussed, then some of the contradictory data, and finally some of my own recent work. In these intervening years there have appeared a number of papers which have purported to confirm the localization for the dual centers as outlined by Ranson. In particular there have appeared a number of reports from Scandinavian workers who have heated the 161 CLARK, G. brain with diathermy or have cooled or heated with thermodes. In one of the early papers of this series (Folkow, Strom and Uvnas, 1949) there was also confirmation of the caudal higher threshold portion of the heat sensitive area of Magoun. In most cases, however, extensive explorations were not made and in few have the anatomical checkups been complete or even reported. They have found it diffi- cult and in most cases impossible to repeat the observations of Magoun et aU on panting. In cats anesthetized with urethane and alpha chloralose circulatory responses were readily obtainable while respiratory responses were slight to non-existent. This diffi- culty, I, also can confirm for although I have been able to elicit panting by localized diathermic heating, more often there were no respiratory changes even to heating in the most sensitive area. Figures 9 and 10 summarize one of these papers. The area sensitive to local heating has a greater rostro- caudal extent and is much narrower ventrally than indicated by Magoun et al. (1938). Like Magoun et al.(1938)EliassonandStr6m (1950) used cats and inserted paired electrodes equidistant from the midline. As they state,the reactive area could very well extend further laterally than their figures indicate. It is worth noting that their use of two figures gives a good three-dimensional diagram of the reactive area. In the goat (Andersson, Grant and Larsson, 1956) it proved possible to delineate an area where mild electrical stimulation would elicit panting and vasodilation of ear vessels. As shown in Figure 11 the area is very similar to that found in the cat. Perhaps as they state the larger brain made possible slightly more exact localization. It is, of course, also true that spread of stimulus would be much less with electrical than with thermal stimulation. Figure 12 from a paper by Birzis and Hemingway (1957) would probably have been hailed by Hanson as convincing proof of the local- ization he assumed for the dual centers. Electrical stimulation (at points marked X) and in more caudal levels elicited shivering as demanded by the theory. It should be noted that after lesions con- fined to this particular medial region dogs have no more difficulty than normal dogs in combating a cold load (Keller, 1960). However, after much larger lesions, as shown in Figure 13, there is a severe damage in ability to withstand a cold load. This is shown in Figure 14. In this test a normal dog would have a normal or slightly ele- vated body temperature but in the two dogs, one of whose lesions you have just seen, the rectal temperature dropped precipitously 162 ANTERIOR HYPOTHALAMIC LESIONS 20 18 16 14 12 10 ^ 6 4 2 2 4 6 8 10 12 14 16 18 20 MM. G. SIGMOID ANT. S. CRUCIATUS G. SIGMOID POST G. SUPRASYLV. ANT G. LATERAL SUPRASYLV. MED BULB. OLFACT G. PROREUS G SIGMOID ANT. G. ORBIT . OLFACT. ANT TR. OLFACT. L. OLFACT POST CUTANEOUS VASOCONSTRICTION DUE TO ELECTRICAl. STIMULATION CUTANEOUS VASODILATATION DUE TO LOCAL HEATING Figure 9. Dorsal and ventral views of the frontal lobe and hypothalamus of the cat, showing the localization of cutaneous vasomotor regions. The anterior and lateral coordinates of the Horsley- Clarke system are given in the figure. The size of the brain corresponds with that of a cat of 3.0 kg. body weight. 34 30 26 22 HORIZONTAL CORPUS CALLOSUM FORNIC. CHIASMA OPT CORP MAMILL. INFUNDIBULUM A CUTANEOUS VASOCONSTRICTION DUE TO ELECTRICAL STIMULATION O CUTANEOUS VASODILATATION DUE TO LOCAL HEATING Figure 10. Median section of the frontal lobe and hypothalamus of the cat, showing the localization of cutaneous vasomotor regions. The anterior and vertical coordinates of the Horsley-Clarke system are given in the figure. The size of the brain is the same as in Figure 9. 163 CLARK, G. Figure 11. A diagram of a parasagittal section (left) through the preoptic area and the hypothalamus of the goat. The cross- hatched area marks the region where electrical stimulations caused polypnea, vasodilatation in the ears, and inhibition of shivering. A diagram ofa horizontal section (right) of the preoptic ^rea and the hypothalamus of the goat at a level about 2 mm ventral to the anterior commissure. Filled circles mark points where electrical stimulation caused polypnea, vasodilatation in the ears and in- hibition of shivering. Open circles mark the localization of adjacent points where electrical stimulation was ineffective in these resjaects. 164 ANTERIOR HYPOTHALAMIC LESIONS Figure 12. Cross-section through tuberal hypothalamus of Cat 34, showing stinnulation points (X) which produced shivering tremor. 165 CLARK, G. Figure 13. Transverse section through area of greatest extent of lesion in Dog 35. 166 ANTERIOR HYPOTHALAMIC LESIONS liLj HOURS ILY 0 I 2 3 0 5 6 NOTE DASHED LINE INDICATES SHIVERING -DOG 28 5 MONTHS AFTER OPERATION ,^nM;1\^xx..^sx.x.^ ^5 ENVIRONMENTAL TEMPERATURE "C .x, ^^^^\\\\\W DOGS 28 a 35s\\\\\\\\\\sS> 39j_, 38^ 3eS 35_i 32^ 30 29 2-8 27 26 25 Figure 14. Rectal temperature curves of dogs (28 and 35) in which all central regulation against a cool environment was elim- inated by a large lesion placed in the posterior hypothalamus. These dogs' temperature curves are contrasted with an unoperated dog's (46) temperature curve, when stimulated by the same cooling load. Stippling in the curve indicates the presence of shivering. 167 CLARJC, G. and there was no shivering. The results of these tests, which were performed months after the operation, are similar to those seen in chronic midcollicular preparations. Figure 15 illustrates the res- ponses of these same two dogs to a heat load. In one dog there was no deviation from the normal and in the other only a slightly higher temperature (Keller, 1950). These dogs, then, are the fourth type required by the dual center theory, that is, animals with much dis- turbed regulation against cold but with little or no disturbances against heat. This latter finding, though, does raise some serious difficulty with the localization of dual centers as proposed by Han- sen. The lateral limits ofthe lesion apparently include the postulated pathway from the anterior center for regulation against heat. Either there are descendingfibers further laterally than had been originally thought or some other lower center is concerned. Neither of these questions can be answered by the available data. There has appeared one paper (Sherwood, Massopust,McCruin, and Buchanan, 1954) on the localization in the hypothalamus of the rat of thermoregulatory integrating processes. In this species it was found that bilateral lesions involving any portion of the lined area shown in Figure 16, whether anterior, posterior, or tuberal, resulted in an inability to maintain normal body temperature when exposed to a cold load. There was no mention of usual body temper- atures of these rats in the animal room nor of any heat load tests. It is probable that they were tested entirely too soon after operation but the findings are not compatible with current dual center theory. It was mentioned earlier that Birzis and Hemingway had been able to elicit shivering in cats by electrical stimulation. In their experiments the active areas fit in very well with the dual center concept; however, there are two other reports of the electrical elicitation of shivering. The earliest of these is that of Akert and Kesselring (1951). This was a part of the findings in the long series of studies of the effects of stimulation of the brain in unanesthetized cats conducted in the laboratory of Hess. In this series there was one cat in which stimulation in a location similar to that reported by Birzis and Hemingway (1957) elicited shivering but in the other ten cases the active region was adjacent to the lateral ventricle and included six locations in the septum pellucidum, three in the caudate nucleus, and one in the anterior thalamus. There is no mention of how often stimulation inthese same locations did not elicit shivering. 168 ANTERIOR HYPOTHALAMIC LESIONS A/ct^^A -A A A ^^o NOTE DOTTED LINE INDICATES BURSTS OF PANTINC 45 44 43 42 4 I 40 39^ 3 8i 3 7? 36i 35a 33'- 32< 31^ 30°^ 29 28 Figure 15. Rectal temperature curves, of the same operated dogs shown in "a", duringthe six-hour exposure to an environmental temperature of 37° C to 38° C. These dogs' entirely normal regu- lation against a heavy heat-load is contrasted with the rectal temperature response of another dog (72-D) in which all heat dissipating abilities were absent. Stippling indicates the presence of panting. CLARK, G. Andersson has reported (1957) that by electrical stimulation of the septal region in unanesthetized goats shivering can be easily induced. Figure 17 is a diagram of the active points he found. It should be reiterated that removal of the septum does not in any way interfere with a dog's ability to withstand cold or heat and the body tempera- tures of such preparations are within normal range (Keller, 1960). The role of the septum in temperature regulation is very difficult to visualize but any role at all is not in accord with the dual center theory of Ranson. Freeman and Davis (1959) have recently reported some results obtained by heating or cooling the brain of the cat that are not com- patible with the localization of the dual centers as proposed by Ran- son. Their results from heatingare shown in Figure 18, which again is one of those unfortunate parasagittal projections. Heating in the chiasmal region would be expected to bring about a fall in body temperature but a fall in body temperature on heating the anterior pons and a rise in body temperature on heating the midbrain and caudal hypothalamus are not in accord with the dual center hypo- thesis. The results with cooling as seen in Figure 19 are even less explicable. Cooling should give no response in the chiasmal region yet both a rise in some cases and a fall in others was observed. In the caudal hypothalamus only falls in bo(fy temperature were ob- served to result from coolingandthe points extend well into the mid- brain while the only rises in body temperature from posterior cooling were in the anterior pons. None of these results are in accord with the localization that was proposed for the dual centers nor would they be predicted on the basis of results of extirpations. This work I have covered maybe summarized as follows: In the past two decades, that is, since the dual center theory was thought to be validated by exact localization, there has appeared some con- firmatory work, none of which was crucial, and there has also appeared a body of work not explicable in terms of the localization proposed. Furthermore, none of the primary questions neglected by Ranson have been answered. 170 ANTERIOR HYPOTHALAMIC LESIONS MAMMILLARY ftOOT Figure 16. The cross hatched area indicates the portion of the hypothalamus the integrity of which is suggested to be vital to temperature maintenance in the rat. It has been projected upon a mid-sagittal section of a rat brain (left) and represents the loci of the sites of bilaterally symmetrical lesions in twelve animals which could not regulate their body temperatures while in a cold environment. On the right is the same area shown as a projection upon a horizontal section of a rat brain. Figure 17. A diagram of a transverse section through the fore- brain of a goat slightly in front of the anterior commissure. Black triangles indicate points where shivering, peripheral vasocon- striction and inhibiton of polypneic panting were obtained as effects of electrical stimulation. Encircled triangles indicate points where in addition to the above mentioned effects piloerection and a huddling up of the animals were observed. At right is a diagram of a horizontal section through the septal area and the thalamus of the goat at a level slightly dorsal to the anterior commissure. 171 CLARK, G. Figure 18. Sites of reaction to heating, shown in parasagittal section 3 mm from the midline. A thermode is shown in scale; the probable effective range of stimulation is 1 mm from the surface. Open triangles represent a fall, solid triangles a rise in rectal temperature. Scale in mm. 172 ANTERIOR HYPOTHALAMIC LESIONS Figure 19. Sites of reaction to cooling, shown in parasagittal section 3 mm from the midline. A thermode is shown in scale; the probable effective range of stimulation is 1 mm from the surface. Scale in mm. Open triangles represent a rise, solid triangles a fall in rectal temperature. 173 CLARK, G. The next four figures, 20, 2l,22,and23, represent some of my own recent work*. These are animals with rather unusual lesions. Routinely, extensive lesions were placed unilaterally in the anterior hypothalamus and contralateral to these large lesions were placed in the posterior hypothalamus. It was anticipated and found that such preparations would not present extremely grave nursing problems. As seen in the tests performed one month after the operation these cats would all be considered to have rather severe disturbances of regulation against heat. In this respect these cats are identical with those on which Ranson based much of his validation and localization of the dual center theory. In some of these cats even four months after the operation there remained a severe deficit in regulation. In all cases, however, the only permanent result was an increased panting level. In Cats 1 and 2 (Fig. 19) with the longest survival it is evident that this is probably a steady state defect and that no further improvement would have occurred; presumably this was also true in the others as well. A reasonable assumption would be that some sort of a learning process had occurred. This is negated by the fact that the first four animals spent most of their survival period in an air conditioned room in which the temperature rarely exceeded 23. 8 C and then only for very short periods. The animals were given short heat load tests at widely separated intervals. No opportunity for learning occurred. What then can be the mechanism for this return of function? In our original work the assumption was made that the effects of edema, etc., had disappeared by one month after the oper- ation. This assumption was thought to be validated by the fact that in cats with asymmetrical lesions there was an early loss in ability to withstand heat but one month postoperative responses were within the normal level. The assumption still seems reasonable and is sup- ported by the studies of Prados, Strowger and Feindel (1945). No explanation is in sight but one wonders what sort of an explanation would have been devised originally if we had continued to study our animals over a prolonged period of time. Of course it is probable that if we had waited for four months or so before testing the animals we would have found nothing to explain. •The Initial phases of this were conducted under a contract with the USAF 33 (616)-5657. 174 ANTERIOR HYPOTHALAMIC LESIONS What, then, are the essential details on which a theory of the neural mechanisms of temperature regulation must be based? The most important of these is that there is a separation of integrating areas for heat conservation from those regions essential for heat dissipation. This has been shown by two different types of prepar- ations, in the dog, Keller (1960) has established that it is possible for an animal to have minimal ability to withstand cold while having normal responses when exposed to a heat load. Our own original work andmy own recent work indicates that the other type of prepar- ation is also possible and we have been able to produce such animals. However the return of function in these cats poses more problems than it solves. Wliile we have good evidence for the localization of heat maintenance activities, the location or locations of these areas where heat loss activities are integrated are obscure. Cer- tainly the anterior hypothalamus must play some role but what the role is remains for the future to determine. 175 CLARK. G. 1 month 3 months 5 months 7 months 8 months Cat #1 Final Cat #2 Final Temp. Resp. Rate Temp. Resp. Rate 41.0 27 41.4 41. 4 36 40. 7 41. 2 192P 41. 4 41. 3 160P 39. 5 41.-2 132P 40. 3 32 120P 190P 200P 200R 1 month 3 months 5 months 6 months Cat #3 Final Temp. Resp. Rate 41. 0 24 40. 8 160P 41.2 128P 41. 2 130P Cat #4 Final Temp. Resp. Rate 41. 0 68 41. 4 104P 41. 3 122P 41.0 150P Figures 20-23. Time and results of damage to heat loss mech- anisms. The only permanent effect is an increase in panting level. 176 ANTERIOR HYPOTHALAMIC LESIONS 1 week 1 month 4 months 6 months Cat #5 Final Cat #6 Final Temp. Rasp. Rate Temp. Resp. Rate 41.2 70 41.1 68 41.1 120 41.1 40 41.2 140 41.0 60 41.0 BOP 40.2 120P Cat #7 Cat #3 1 week 1 month 4 months 6 months Final Final Temp. Resp. Rate Temp. Resp. Rate 41. 1 41. 0 41. 0 40. 5 34 20 60 41. 0 24 41.0 40 41.0 50 40.4 130P Temperatures in degrees Centigrade, respiratory rates per minute. This work was begun under USAF contract AF 33(616) - 5657. 177 CLARK, G. DISCUSSION DR. STUART: Do you have the histology of the animals that were shown in the last two slides? DR. CLARK: I do not have them with me, but the lesions were virtually unilateral hem isections. They were good, big-sized lesions. DR. FREEMAN: Since this question of the compatibility of my data with the dual center hypothesis came up, I might say that we were trying to prove this hypothesis when we started off with this work in 1950, with the idea that there was a cold center posteriorly and heat center anteriorly. If you stimulated them with conductive thermal changes, you should get appropriate activation of a mech- anism for heat loss by heating the anterior hypothalamus, and nothing if you cool. You should get opposite effects with the poster- ior hypothalamus. What we found was, as you see, that sensitivity to both heat and cold could be demonstrated to exist anteriorly and that the inverse changes took place in the posterior hypothalamus. This sent me back to Hanson's original description of this hypoth- esis, and it struck me then — and it does so now — that there was a peculiar confusion or ambiguity in his expression of this thing. He described heat loss mechanisms or centers in the anterior hypo- thalamus and in the posterior hypothalamus and did not specify whether these were sensory or motor or both. He never said, as we thought he had said, that these were heat- sensitive and cold- sensitive mechanisms. He was simply describing his results in more general terms without specifying as to whether temperature sensitivity exists in either one of these areas. DR. CLARK: At that time, it was thought that the work with the diathermy had conclusively proved that the region beneath the an- terior commissure was a heat-sensing area, but we had no data whatsoever about a cold sensing area. The only thing we had was the fact that the caudal and lateral hypothalamus lesions produced ani- mals that could not withstand cold as well as normals; those lines he drew were entirely hypothetical. 178 ANTERIOR HYPOTHALAMIC LESIONS DR. FREEMAN: We developed our own interpretation of this which we still think is compatible with a dual center data. The anterior hypothalamus consists of heat-sensitive and cold-sensitive neurons, and these have an inhibitory effect upon neurons in the pos- terior hypothalamus which in turn are responsible for activation of heat- conservation and heat-production mechanisms. Heating anteri- orly will, in effect, increase this inhibition and depress the mech- anisms, whereas cooling anteriorly will do the reverse, cause a release from inhibition. An inhibition of inhibition, if you will, allows these mechanisms to spring up. Now, if you make a lesion anteriorly, you remove this inhibition, you produce an an animal with tendency to excessive heat production. This animal has no difficulty in main- taining himself in the face of cold stress, whereas an animal with a lesion posteriorly will tend to lose the capacity for heat production. When the temperature now goes up, there is no automatic mechan- ism; when the temperature goes down, there is no preventative mechanism left. We think that these are compatible if you express the dual center in terms different from those originally proposed. DR. CLARK: They are not compatible with the rest of the evidence. DR. FREEMAN: In what way? DR. CLARK: You can produce a cat with a large electrolytic lesion just caudal to the optic chiasma. These are large lesions, say from L four to R four and from zero down to the base; those cats will show low temperatures; at a room temperature of 22 C the temperatures will be so low you cannot read them on a clinical ther- mometer, and I have had temperatures as low as 25 C. Those are electrolytic lesions; if you make a lesion with a knife mounted in the stereotaxic frame at the same location, you do not get these hypo- thermias. The only trouble is that a knife wound like that is very hard to see in histological sections. You cannot be sure. You know that you went down and hit bottom, but in the sections, you cannot show where the edges of the lesion are. In fact, you may hardly be able to see the lesion at all. 179 CLARK, G. DR. FREEMAN: I am not sure that I see the basis of the incom- patibility. DR. CLARK: There is a tremendous difference in the effects of the electrolytic lesions as compared to either section or knife cut. DR. FREEMAN: That is something entirely new. DR. STUART: When you make the knife lesion, how much hypo- thalamic and subthalamic tissue is left intact caudal to the section? DR. CLARK: All of it. DR. STUART: And this animal can regulate against the cold? DR. CLARK: Yes. DR. STUART: I cannot see where this would disagree with the classical literature. DR. CLARK: I was thinking primarily of the hypothermia that you obtain with the electrolytic lesion. As far as the rest, it is not too bad, but I still do not see that area in the anterior pons. DR. FREEMAN: Is there sensitivity back there? DR. CLARK: Yes. DR. FREEMAN: Well, you had the old data, yourself. This was not the anterior pons, by the way. This was the anterior midbrain. You yourself showed this trailing off of the region of sensitivity. DR. CLARK: Yes, that region is way dorsal. DR. FREEMAN: Well, do not forget that the stimulus we apply here is largely applied dorsally. Those points on the diagram repre- sent the tip of the thermode, and this is the greatest extent of projection. DR. CLARK: You did not try just going down a little way? 180 ANTERIOR HYPOTHALAMIC LESIONS DR. FREEMAN: We lowered the thermode routinely by short steps and stimulated each stage until we got a response. If you push it all the way to bottom, there is maximal damage to the structure that one is generally trying to stimulate. We went by easy stages. DR. CLARK: But, you did not get anything until you got deeper? DR. FREEMAN: Well, the points represent the position of the thermode at the time of an effective response. DR. CLARK: Presumably the thermode was in the area that is sensitive to heating. You did not get anything because that area should be up around the aqueduct. DR. FREEMAN: That is true; however, your pathway swings somewhat laterally there, is that correct? DR. CLARK: Well, there are no data on that. As far as the pathway for panting, yes, because if you leave just lateral ex- tremities of the midbrain, you still get panting, and those dogs cannot regulate against cold. DR. FREEMAN: This is Keller's work you are describing? DR. CLARK: Yes. DR. FREEMAN: That, I think, is also quite compatible with the finding of this region of sensitivity, that is, the anterior midbrain. DR. CLARK: No, Ithinkit is just a question of whether the path- way extends that far laterally. In other words, I think it is rather a diffused pathway that extends through the entire mesencephalon. DR. STUART: I think in fairness to Ranson and Magoun, it should be stated that they thought that the posterolateral hypothalamus in- tegrated heat conservation rather than heat productive mechanisms. In a 1940 review (Ergebn. Physiol. 41:56, 1939) they mentioned the difficulty encountered in instigating shivering pre-operatively with the applied cooling load. Thus they did not wish to attribute the inte- gration of shivering to that particular region of the hj^jothalamus. 181 CLARK, G. DR. CLARK: The only thing I did was to take the cat's tempera- ture and put it in a box; after three hours, I would go back and pull the cat out and put him on my lap and take his temperature and feel if he was shivering, and that is not too satisfactory. DR. STUART: Birzis and Hemingway stimulated a locus in the tuberal hypothalamus to produce shivering. Didn't you say that a lesion that spares this region has no effect on shivering? DR. CLARK: No, Keller has some dogs where he has had just that medial part destroyed, and they regulate against cold quite well. It takes a tremendous lesion to get any long-term effects. DR. FREEMAN: There is another point about your data that has always interested me. This also applies to the data from Igor and Hanson on exposing these animals to high ambient temperatures. Their body temperatures will go up in their terms to 42 C and level off there despite continuance of the heat stress, and I notice in your data that almost all these animals have temperatures that range from 41.11 C to 41.18 C, a very narrow range. DR. CLARK: Well, that is very simple. I did not want to hurt the animal. DR. FREEMAN: You were watching for it then? o DR. CLARK: When the temperature reached 41.11 C, I stopped the test; but since I only took the temperatures at intervals, in some cases the temperature went above that. DR. FREEMAN: That is humanitarian, yes, but how about the data of Teague and Ranson's studies, showing that with continuation of the heat stress there is prevention of hypothermia in all their animals? DR. CLARK: Well, I do not think that is quite the way it went because they also stopped their tests that way. DR. FREEMAN: I can recall some of the heat stress. I remem- 182 ANTERIOR HYPOTHALAMIC LESIONS ber this particularly because I am very interested in the phenomena. Teagueand Hanson's graph shows rectal temperature against time; it shows a period of stress; the temperature will go up, and then it will level off. This is my question: how come? Why does it not go up to 43 C or 44 C? You see, if they have lost all ability to regu- late against heat, they should be unable to control at any level, but now they have this cut-off level and they survive. DR. CLARK: I know, but that cut-off level is also due to the temperature that they are exposed to; when their temperature gets above 41 C, they are going to be able to lose heat to the environ- ment. DR. FREEMAN: Then, they should show some kind of gradual approximation to this thing, but this is not what the curve shows, by and large. MR. ADAMS: How does this approach the peak? Is it as sharp as you have shown? DR. CLARK: I would have to check back and see the data. One time I was told to take a cat's temperature up until it panted, and I did; I took his temperature up to, I think it was 44 C, and that was rather an acute preparation. In order to get it up that high, 1 had to raise the box temperature progressively and so the box temperature was way above 40 C when the cat's temperature reached 44 C. MR. ADAMS: Dr. Clark may still have a point in this being a net thermal balance. If I remember Andersson's data correctly, during his experiments his goats reached a lower level of heat ex- change where there was established a steady state heat balance which had been approached sharply. It was not reached gradually as you might expect. DR. FREEMAN: There is a good deal of other evidence, how- ever, that this level of 107 F, or about 41 C, is a critical point. DuBois wrote on this issue. He pointed ouf that fevers above 41.11 C are quite rare, and postulated that there is some kind of 183 CLARK, G. emergency mechanism which will keep body temperature from going above that level and Roche, in 189 8, found that same thing; you could drive an animal's temperature up by considerable amounts, up to this level, but to go beyond this level you had to either exhaust the animal by a prolonged application or else really provide a severe stimulant. DR. CLARK: You are quite correct in that, and it is not the usual thing right after anterior hypothalamic lesion, to get a temper- ature above 41.1 C or 41.7 C. I have one cat that had a tempera- ture of 43.6 C the morning after the operation. MR. ADAMS: This point that Dr. Freeman brings out I think is true. It is also seen in bats which may be essentially poikilothermic at lower ambient temperatures at rest but which can be seen to regulate their temperatures at higher levels. Since you mentioned that the upper temperature level is dependent on ambient tempera- ture, 1 think that this might preclude accepting this explanation completely. DR. FREEMAN: We have done the same thing with cats given strychnine, where they are given serial convulsions, and you can control this by the amount of strychnine that you give. By pounding on the table at fairly regular intervals, you can induce a series of repetitive convulsions which will drive the animal's body temperature up to about this level, and then it will flatten off, even though convulsions continue. About this time, also, they will start to pant. Another way of doing it is to put them in a very hot environ- ment. This is an intact normal cat. Put him in an environment of o 48 C, which is quite uncomfortable, and he will make vigorous attempts to escape but will not pant. Now, their body temperature will rise rapidly during this period. I could not get continuous records because they weretooactive, but abruptly they will collapse and start to pant vigorously. They just lie down; if you take them out at that point, you will find that their body temperatures are 40.7 C. They go so high and then no further. So, there is something that goes on here, and this something, this mechanism — whatever it is — is still intact in the animals you described. MR. ADAMS: Dr. Clark, were those lesions you reported on this afternoon or this morning dual lateral lesions? 184 ANTERIOR HYPOTHALAMIC I^ESIONS DR. CLARK: They were unilateral at two levels. You see, I have a virtual hemisection anteriorly on one side, and caudally on the other side. The third ventricle is convenient. There are no com- missural fibers in it. DR. HENSEL: May I make a more general comment on the eval- uation of impairment of thermode regulation of the lesion, here? I think I had better explain it on the blackboard. Mostly, we are speaking of a loss of the ability of the regulation against cold. We choose the temperature for testing this rather at random. I think it would be quite useful to test the ability of regulation systematically and quantitatively over the whole temperature range, because if you consider the curve of the poikilothermic and homeothermic or- ganisms as you showed in Figure 1, then the loss of the ability to keep the temperature at a certain level might have two explanations: one explanation is just limitation in the quantitative output of your regulating center; and the other might be a disturbance of the cen- ter — the feedback mechanism itself. So, if you have just a quantitative disturbance, the curve might regulate in a smaller range of temperature, but in Figure 24, your mechanism in terms of cybernetics and feedback is perfectly all right. It might be as precise as this one. This is the case, for example, in the pre- mature infant, and there are many wrong conclusions because the infant can not keep its temperature at this point, but it can maintain a perfectly normal temperature in this smaller range. DR. CLARK: That goes back to Isenschm id. DR. HENSEL: Yes, and I think it would be useful if you would do this. DR. CLARK: I have done it with some cats, but I did not have a good enough controlled room to really do it. You would have to have a pretty accurate room in order to test your animals properly. DR. HEMINGWAY: About this dual theory which we have been hearing of since before my time: the problem, I think, is trying to interpret the data. I was listening to Dr. Clark, who gave a very stimulating discussion of this. I justwonder if we have to be careful about being too rigid in defining the location and the functions of these centers. That is what it amounts to. You have a center in one 185 CLARK, G. place which has certain properties and certain functions, and another center in another place. I wonder if there is not considerable over- lapping in these centers. For example, there are two problems that have come up with shivering and panting. It is possible to control shivering by stimulation in the septum, which is certainly not in the posterior hypothalamus, and it is possible to inhibit shivering by stimulation in the anterior hypothalamus, which is far from the posterior hypothalamus. The other thing is the problem of panting. That is certainly a controversial problem, whether panting can be produced by heating or stimulatir^ the anterior hypothalamus. There is not general agreement on that, is there? DR. CLARK: No, there is not. Magoun, Harrison, Brobeck, and Ranson (1938), when they did that, were getting it routinely. I saw many of their animals and remember their rectal temperatures; they would be panting with a body temperature as low as 35 C. Of course, the Scandinavian workers (Strom, 1960) state that they had a hard job, and I do not know whether they ever saw panting or not. Do you know. Dr. Hensel? DR. HENSEL: Do you mean Andersson? DR. CLARK: No, Idonot mean Andersson. I mean the von Euler group, using diathermy in cats. DR. HENSEL: I think so. I think he saw panting. The only thing he could not see was the vasoconstriction during cooling. As far as I remember, he could see panting. DR. FREEMAN: But it was very uncommon in his animals. DR. HENSEL: Yes. DR. LIM: And also in Harcfy's work? DR. CLARK: Hardy was getting panting in the dog quite readily, and I have been doing some heat experiments with cats and oc- casionally getting it, in animals under chloralose and animals under urethane and animals under both. I have had trouble getting 186 ANTERIOR HYPOTHALAMIC LESIONS it in unanesthetized preparations, probably because there I cut down the power output of my diathermy apparatus. But we have a lot to learn. As far as the localization of these centers is concerned, I think we know quite a bit about heat conservation because Keller has had several animals that could regulate quite adequately in the heat, but could not in the cold. DR. HEMINGWAY: Does that include shivering, heat conserva- tion, and heat production? DR. CLARK: Yes. These animals when put in the cold would show a drop in body temperature similar to that you might see in a chronic midbrain preparation. DR. HEMINGWAY: Does that include shivering and cutaneous vasoconstriction, both in that one term? DR. CLARK: Yes. Of course, even these will show some vaso- motor changes. You get those even in the chronic midbrain dog. DR. LIM: From your recent data, do you suggest that chronic preparation should be made at least six months after operation? DR. CLARK: Well, there I do not know, but I do know that you have to differentiate between acute and chronic effect of lesions, and I think that any report that is based on animals a week, two weeks, one month, or two months after the operation should be questioned. DR. STUART: If the function does not return? DR. CLARK: Yes. Now, of course, determining the things that an animal can do a week after the operation as well as a normal dog is important, but one may be misled concerning the loss of function, for a week or up to two months or maybfe three months. Of course, the questions are: Is it due to tissue that regains function? Is it due to some other center vicariously functioning? You have a lot of problems in there, but the important thing after any brain operation is what the animal can do that approaches normal, not what is lost. DR. HEMINGWAY: That raises a question which Dr. Freeman and I have talked about many times. He thinks that if there is des- 187 CLARK, G. truction of part of the brainnecessary for some function, then some other part of the central nervous system can take over that function. This is certainly true in your asymmetric lesions, that you can de- stroy one side and the other side will take over the functions of both sides. DR. CLARK: Yes, but with the asymmetric lesions, I think you still have really intact tissue; as for complete vicarious functioning of the new area, I think we have to see that before we buy it. DR. FREEMAN: Well, you can see it in some, say, lesions of a sort in the peripheral end of the nervous system. For example, recall Sperry's transplant in which he exchanges tendons to differ- ent muscles and then forces these monkeys to perform various tasks with the tendons reversed. There is indication there that some form of reorganization does take place. It is not basically a struc- tural modification. That is to say, from all available tests, the pattern of nerve organization, anatomically, has not changed, but the functional abilities of the animal to use this limb have changed. DR. CLARK: That is true, but of course you remember, too, in Sperry's spider monkeys that after he moved postcentral gyrus, they lost that ability. And his rats never did learn how to use the muscles right in their legs, but continued as long as he kept them to step harder on a tack. DR. FREEMAN: On the other hand, in any form of hi^er, more highly organized activity, it is possible to see this sort of thing. For instance, Fleuron showed that by reversing nerves supplied to the extensor and flexor muscles in the wing, one could alter a bird's innervation, but yet the bird could still fly. So some form of re- organization obviously takes place. DR. CLARK: In the salamander, complete reorganization takes place. DR. FREEMAN: Well, that is a different story. 188 ANTERIOR HYPOTHALAMIC LESIONS DR. STUART: The way to get around this is to use our ablation and lesion studies as a preliminary investigation to stucfying the intact brain. What is going on in the normal animal during the heat conservation and dissipation events? DR. CLARK: Well, it is obvious that work in the future should not be two men working here and two men working there looking at a slightly different picture. What we need is a rather large team working together on all of this. 189 CLARK, G. REFERENCES 1. Akert, K. andF. Kesselring. KaltezittemalszentralerReizef- fekt. Helv. Physiol. Acta. 9:290-295, 1951. 2. Andersson, B. Cold defense reactions elicited by electrical stimulation within the septal area of the brain in goats. Acta Physiol. Scand. 41:90-100, 1957. 3. Andersson, B., R. Grant and S. Larsson. Central control of heat loss mechanisms in the goat. Acta Physiol. Scand. 37: 261-280, 1956. 4. Bard, P. and M. B. Macht. The behavior of chronically decere- brate cats. Ciba Foundation Symposium, Neurological Basis of Behavior. Boston, Little, Brown and Co., p. 55, 1958. 5. Bard, P. and J. W. Woods. Central site of pyrogenic action of bacterial endotoxin. The Physiol. 2:5-6, 1959. 6. Birzis, L. and A. Hemingway. Shivering as a result of brain stimulation. J. Neurophysiol. 20:91-99, 1957. 7. Chambers, W. W., H. Koenig, R. Koenig, and W. F. Windle. Site of action in the central nervous system of a bacterial pyrogen. Am. J. Physiol. 159:209-216, 1949. 8. Clark, G., H. W. Magoun, and S. W. Ranson. Hypothalamic regulation of body temperature. J. Neurophysiol. 2:61-80, 1939. 9. Eliasson, S. and G. Strom. On the localization in the cat of hypothalamic and cortical structures influencing cutaneous blood flow. Acta Physiol. Scand. 20:(Suppl. 70) 111-118, • 1950. 10. Folkow, B., G. Strom, and B. Uvnas. Cutaneous vasodilatation elicited by local heating of the anterior hypothalamus in cats and dogs. Acta Physiol. Scand. 17:317-327, 1949. 190 ANTERIOR HYPOTHALAMIC LESIONS 11. Freeman, W. J. and D. D. Davis. Effects on cats of conductive hypothalamic cooling. Am. J. Physiol. 197:145-148, 1959. 12. Isenschmid, V. R. and W. Schnitzler. Beitrag zur Lokalization des Warmeregulation vorstehenden Zentralapparates im Zwischenhirn. Arch. f. exper. Path. u. Pharmacol. 76:202- 223, 1914. 13. Keller, A. D. The role of circulation in the physiology of heat regulation. Physical Ther. Rev. 30:1-8, 1950. 14. Keller, A. D. Personal communication, 1960. 15. Keller, A. D. and W. K. Hare. The hypothalamus and heat regu- lation. Proc. Soc. Exper. Biol, and Med. 29:1069-1070, 1932. 16. McCrum, W. R. A study of diencephalic mechanisms in tem- perature regulation. J. Comp. Neurol. 98:233-282, 1953. 17. Magoun, H. W., F. Harrison, J. R. Brobeck and S. W. Ranson. B Activation of heat loss mechanisms by local heating of the ' brain. J. Neurophysiol. 1:101-114, 1938. 18. Meyer, H. H. Theorie des Fiebers und seiner Behandlung. Verh. Kongr. inn. Med. 30:15-25, 1913. 19. Nikolaides, R. and S. Dontas. Warmezentrum und Warm- polypnea. Arch. f. anat. u. Physiol. Anat. Abt. 249, 1911. 20. Ott, Isaac. The relation of the nervous system to the tempera- ture of the body. J. Nerv. Ment. Dis. 11:141-152, 1884. 21. Ott, Isaac. The interbrain: its relations to thermotaxis, polypnea, vasodilation, and convulsive action. J. Nerv. Ment. Dis. New Series 16:433-436, 1891. • 22. Ott, Isaac. The thermogenic center in the tuber cinereum. Med. Bull. 17:246-248, 1895. 191 CLARK, G. 23. Prados, M.,B.Strowger, and W. H.Feindel. Studies on cerebral edema. I. Reaction of the brain to air exposure; pathologic changes; II. Physiologic changes. Arch. Neurol. Psychiat. Chicago, 54:163-174; 290-300, 1945. 24. Ranson, S. W. Regulation of bocfy temperature. Res. Pxibl. Ass. Nerv. Ment. Dis. 20:342-399, 1940. 25. Ranson, S. W. (revised by S. L. Clark) The Anatomy of the Nervous System, 8th Edition, Saunders, Philadelphia, 1947. 26. Ranson, S. W. Jr., G. Clark, and H. W. Magoun. The effect of hypothalamic lesions on fever induced by intravenous in- jection of typhoid-paratypAioid vaccine. J. Lab. and Clin. Med. 25:160-168, 1939. 27. Sherwood, C, L. C. Massopust, Jr., W. R.McCrum, and A. R. Buchanan. The effect of hypothalamic lesions upon body tem- perature maintenance in the albino rat while in cold environ- ments. J. Neuropath, and Exp. Neurol. 13:191-208, 1954. 2 8. Strom, G. Central nervous regulation of body temperature, in Handbook of Physiology Vol. II, Chapter 46:1173-1196, 1960. (Editors: J. Field, H. W. Magoun, and V. E. Hall.) 29. Thompson, R. H., H. T. Hammel, and J.D. Hardy. Calorimetric studies in temperature regulation: the influence of cold, neutral, and warm environments upon pyrogenic fever in normal and hypothalectomized dogs. Fed. Proc. 18:159, 1959. 192 NEURO- MUSCULAR ORGANIZATION OF SHIVERING Yojiro Kawamura Department of Physiology University of Osaka Osaka, Japan Shivering in response to a cold stress normally waxes and wanes in intensity. Studying patterns of muscular activity at different levels of such intensity is a method of investigating certain neural aspects of shivering. In this experiment jaw and limb muscles were analyzed electromyographically during shivering and the effects of some physiological variables on such activity were noted. EXPERIMENTAL PROCEDURE Thirty-five male and female adult dogs (9.0 - 11.5 kg.) were anesthetized with sodium isoamytal barbiturate (0.6 mg/kg I. V.) and electromyograms of jaw, fore and hind limb flexor and extensor muscles were recorded simultaneously while the animals were shivering in the waning stages of anesthesia. The animals were not restrained, but rather positioned in normal sleeping posture. The electromyograms were recorded with a concentric bipolar electrode, C-R coupledamplifier,cathode ray oscilloscope and electromagnetic oscillograph. Respiratory movements were recorded by strain gauge transduction of pressure variation of a chest tambour. Shiv- ering movements were additionally recorded by strain gauge trans- duction of limb vibrations. 193 KAWAMURA, Y. RESULTS The Pattern of EMG Recordings During Shivering Preceding visible shivering a generalized increase in muscle tone was evident. This increase was reflected in the bipolar elec- trode recording of single unit discharges from one motor unit (NMU) (Fig. 1). Such NMU discharges from a single fiber motor unit ranged from 6-26/sec and 100-2 00 mV. They were detected from the mylohyoid, external oblique, flexor, limb, or tail muscles. There was no clear relationship of the discharge of any given motor unit with that of any other motor unit of the same or other muscles. At first these discharges had no close relationship to the respiratory cycle, but, as shown in Figure 2, as the number of unit discharges in- creased they tended to become grouped into the inspiratory phase. When shivering became visible, though still feeble, the EMG record illustrated grouped discharges consisting of fused NMU activity, together with separated NMU discharges. These grouped discharges were concomitant with lung inflation (Fig. 3). The intervals between both NMU and grouped discharges were longer at the beginning and end of lung inflation that at peak inflation. Similarly, the amplitudes of grouped discharges were spindle shaped, concomitant with one respiratory cycle and illustrated the fusing of additional NMU's into each grouped discharge at peak lung inflation (Fig. 4). As shivering became more intense, the NMU discharge had a frequency of 10-12 /sec (Fig. 5). One grouped discharge corresponded to one cycle of limb movement discharge and one given NMU oc- curred per given grouped discharge. At this stage there was no re- lationship between grouped discharges and they had an amplitude of 1 mV or more. During intense shivering these grouped discharges had a duration double that during feeble shivering. Vigorous 194 NEURO-MUSCULAR ORGANIZATION OF SHIVERING 195 KAWAMURA, Y. 196 NEURO-MUSCULAR ORGANIZATION OF SHIVERING 197 KAWAMURA, Y. ? » » CQ _. o 1 1?- 3 CO S CO m o o o 198 NEURO-MUSCULAR ORGANIZATION OF SHIVERING 199 KAWAMURA, Y. shivering continued for four to five hours and in the waning stages of anesthesia was again concurrent with lung inflation. The duration of grouped discharges shortened thus reflecting a lesser number of NMU discharges fusing into the group discharge. Intervals between each elongated grouped discharge lengthened and at this stage visible limb movements ceased (Fig. 6). Differences Between Muscles During Shivering Grouped discharges and NMU discharges were observed in jaw, tongue, andneck muscles before limb and trunk muscles and as shiv- ering waned it disappeared earlier from jaw muscles than from others. The above pattern was not evident when the head or limbs were restrained by being tied to a board which usually resulted in inhibition of the specifically restrained body part. Muscles of the hind limb had a tendency to become active earlier than the fore limb. Extensor muscle groups became active before flexor muscle groups, agonist and antagonist discharges being con- current rather than reciprocal; there was one grouped discharge from any one muscle fiber per one full limb cycle (Fig. 7). Addition- ally there were a greater number of NMU discharges fused into the grouped extensor discharges than NMU discharges fused into grouped flexor discharges. Jaw muscle shivering and electrical activity was most predom- inant in the horizontal position of the lower jaw. The Relation of Shivering to the Level of Anesthesia Under deep anesthesia (light mydriasis - corneal reflex absent) none of the animals shivered. Three of the 10 animals shivered con- comitant with lung inflation when a weak corneal reflex was evident. At this level of anesthesia shivering was induced in the other 7 200 NEURO-MUSCULAR ORGANIZATION OF SfflVERING Minutes 2 9 35 74 104 203 111 116 I 5€C » ♦ . / o _ - ^ / V < / o - Q> / c = o y^ o < 1 ^x^ x: - o ^^^^ 0> - -0 r 1 S l\ o lO o ir> ro O ro 209 KAWAMURA, Y. E o o o o X o c o o O ro CVJ ro ro ro 1 O -7JO — J I L J \ L J L O ID O ro O 210 NEURO- MUSCULAR ORGANIZATION OF SHIVERING a weak corneal reflex. At much lighter anesthetic levels noxious stimuli tended to inhibit shivering but such inhibition was followed by rebound facilitation (Fig. 14), DISCUSSION These observations must be considered in terms of neural mechanisms which serve to activate, inhibit, and regulate the rhythm of shivering. Activating Considerations Magoun et al. (193 8), Strom (1950), Von Euler (1950), Hemingway et al.(1954), andAnderssonetal.(1956)have localized a primary in- hibitory region at the junction of the ante ro-lateral hypothalamus and the lateral pre-optic region. A primary shivering activating sys- tem has been shown by Stuart, Hemingway, and Kawamura (1960) to exist in the dorso-medial portion ofthe posterior hypothalamus. The results here reported suggest that: First: The shivering activating region may have a stronger re- sistance to barbiturate anesthesia than the shivering inhibitory region. Shivering, once initiated during light anesthesia, continues as the rectal temperature rises above 38 C. Second; The excitability of the activating region is possibly affected by blood hormone levels particularly the thyrotropic hormone or thyroid hormone levels because in winter the thyroid function is active and it is somewhat depressed i» the hot summer season. This is suggested by the fact that it is difficult to induce shivering in anesthetized cats in the summer even though the blood and skin temperatures are low. 211 KAWAMURA, Y. 212 NEURO-MUSCULAR ORGANIZATION OF SHIVERING Third: The shivering activating region and the reticular activa- ting system have a close relationship since (a) strong noxious stimuli which stimulate the reticular activating system have a tendency to induce shivering even during deep anesthetic conditions, (b) the ret- icular activating system is possibly adrenergic (Bonvalletetal, 1954; Bradley, 1958; and Rothballer, 1956) and as such may be facilitated by epinephrine and depressed by acetylcholine. These results indicate that acetylcholine inhibits and epine- phrine facilitates shivering. Some conflict exists in the literature on this latter point. Cassidy, Dworkin,and Finney (1926) reported that shivering, abolished by insulin injection in anesthetized animals, was restoredby large doses of epinephrine. Hall and Goldstein (1940) reported that I. V. injection of 80-150 mg/kg epinephrine causes a transitory facilitation in anesthetized animals. Smaller doses (SO- SO mg/kg) produced inhibition without temporary facilitation. Un- published data of Stuart, George, and Hemingway suggest that shiv- ering is mildly facilitated in unanesthetized cats following sub- cutaneous injection of 40 mg/kg. The facilitation here reported following subcutaneous injection of 30 mg/kg into anesthetized dogs followed previous ACH injection. Obviously the question cannot be fully resolved until a more extensive study is performed on both anesthetized and unanesthetized shivering animals in which various doses of epinephrine are applied by both intravenous and subcutan- eous routes, uncomplicated by previous drug inhibition of shivering. Fourth, There is a relationship betweenthe activating region and inspiration during shivering that could be organized in one of three ways: (a) Excitatory impulses to the shivering activating region from pharyngeal cold receptors that are stimulated during inspiration (Cort and McCance, 1953); (b) Excitatory impulses ascendingto the activating region and/or descending to the spinal cord from pulmon- ary stretch or from the inspiratory center receptors, maximally stimulated at peak inspiration; (c) A generalized state of increased medullary excitability occurring during inspiration and facilitating descending extrapyramidal activity (Kawamura and Fujimoto, 1958). For example, the jaw opening reflex is facilitated during inspiration. However, it must be stressed that there was no clear relation- ship between inspiration and the increase in muscle tonus that pre- 213 KAWAMURA, Y. cedes shivering. This may mean that at this stage of anesthesia in which an increase in muscle tone but no shivering is evident, the animal is too deeply anesthetized to permit respiratory facilitation of shivering. However, this lack of relation between shivering and muscle tonus may be due to the latter phenomenon having spinal, rather than central origin. Classical decerebration studies have illustrated the predominantly inhibitory action the rostral nervous system exerts on the spinal cord. Just as decerebration releases the spinal cord from inhibitory influences so it may well be that in the light anesthetic state, the supra-spinal inhibitory regions are more depressed than the spinal cord. Hence, the initial increase in muscle tone is not directly related to temperature regulation but represents a stage of anesthesia in which proprioceptive hyper- activity predominates just as decerebrate rigidity is due to a hyper- activity of myotactic origin. In addition to a central activating region being necessary to in- itiate shivering, there is evidence that shivering is facilitated by afferent impulses from proprioceptive nerve endings that may exert these excitatory effects centrally and/or peripherally. The evidence is as follows: First. During deep anesthesia the somatic muscles are com- pletely relaxed and no electrical activity as revealed by unit potentials is noted. As the animal recovers from the anesthesia and the muscles regain tonus, unit spike potentials begin to appear and these are followed by shivering. Present results in these obser- vations confirm the earlier work of Burton and Bronk (1937). The work here reported suggests that as the muscle tone increases, the resulting activation in integrative afferent proprioceptive input from annulo- spiral, flower spray, Golgi tendon organ, and deep joint pro- prioceptors is favorabletotheproductionof shivering. For example, it is well known that increased annulo-spiral discharge from any muscle fiber lowers the alpha motor neuron thresholds of that muscle. Kawamura et al (1958) have shown that there is an increase in the discharge of frequency of the trigeminal mesencephalic nucleus (which is considered to receive proprioceptive information from the jaw muscles) immediately prior to the onset of jaw muscle shivering. 214 NEURO-MUSCULAR ORGANIZATION OF SHIVERING Second. Shivering is more intense in jaw muscles when the head of the dog is in the normal horizontal position. In this position, the anti-gravity jaw muscles are subjected to more tension with a con- comitant increase in proprioceptive activity. Inhibitory Considerations These results suggest the following concepts relating to inhib- itory shivering mechanisms: First. Uprus et al (1935), using unanesthetizedhuman subjects, have shown that shivering ceases as the rectal temperature rises and reoccurs as the temperature falls, even in the presence of a hi^ initial temperature. The results on anesthetized animals would suggest that shivering can be both inhibited by radiant heating and instigated by cold air without any change in rectal temperature. Shivering was inhibited in low rectal temperature dogs when the backs of the animals were briefly heated. This phenomenon may be due to a sudden increase in excitatory impulses to the anterior hypo- thalamic inhibitory region from dorsal heat receptors. Such a brief and sudden input could be sufficient to inhibit shivering temporarily. Unfortunately, it is not known if such inhibition mig^t be only temporary in the face of consistent low blood temperature since the period of radiant heating was of short duration. At any rate, this phenomenon is a clear example of a peripheral stimulus antagonizing the physiological effects of a contrary central stimulus. Second. In deep anesthetic states a noxious stimulus can facili- tate shivering. But in light anesthetic states noxious stimuli tend to inhibit shivering, even when the blood temperature is below normal. This type of inhibition may be due to the noxious stimuli evoking avoidance behavior in the animal and as such voluntary avoidance movements would tend to inhibit shivering by sucpessful seizure of the motor neuron pool. Such a concept is supported by Stuart, Free- man and Hemingway's unpublished findings that restraint devices inhibit shivering in unanesthetized chilled cats. It is difficult to ac- count for the earlier appearance of shivering in tongue, jaw, and neck muscles than in limb muscles. At first glance it appears analogous to the development of tetanus and tetany. However, restraint and 215 KAWAMURA, Y. proprioceptive input is more prevalent in these experiments in the limbs on which the animal is lying than in the freely suspended tongue, jaw, and neck muscles. When inducing shivering by electrical stimulation of the central nervous system, Stuart and Kawamura have found shivering easier to induce in the limb muscles; but in these experiments the animals' heads were held in a stereotaxic frame. Reerulatory Considerations Perkins (1945) has suggested that the rhythm of shivering is controlled by proprioception but instigated and maintained by central hypothalamic activity. Such a concept followed his recording a change in the frequency-amplitude characteristics of hind limb shiv- ering following deafferation. The most characteristic EMG pattern of a shivering muscle is a grouping voltage occurring 10-12 times per second, the same frequency as the hind limb oscillations re- corded by Perkins. When we cut the dorsal roots the grouping voltages became randomized and shivering lost its characteristic rhythm. This finding is a confirmation of Perkins* original findings. However, as mentioned earlier it is suggested that proprioceptive input as well as regulating the rhythm of shivering also facilitates its initial occurrence. SUMMARY Following electromyographic analysis, the muscles of dogs at various stages of shivering intensity, rectaltemperature, and anes- thetic level, the following conclusions appear in order: 1. The order of appearance of shivering in muscles is (a) jaw and tongue, (b) neck, (c) hind limb, (d) forelimb, with extensor muscle shivering more intense and appearing earlier than flexor muscle shivering. 216 NEURO-MUSCULAR ORGANIZATION OF SHIVERING 2. The increased muscle tonus that occurs in anesthetized dogs prior to shivering is not necessarily a phenomenon of tempera- ture regulation origin. 3. Lung expansion facilitates shivering. 4. Proprioceptor activity may facilitate as well as coordinate shivering. 5. Acetylcholine inhibits and epinephrine facilitates shivering, and suggests a relationship of shivering to the level of excitability of the reticular activating system. ACKNOWLEDGMENT I wish to thank Dr. A. Hemingway, Professor of Physiology, UCLA School of Medicine, for his help and encouragement in pre- senting this paper, and I also wish to extend my appreciation to Mr. D. Stuart for editing this paper. 217 KAWAMURA, Y. DISCUSSION DR. MINARD: Is it not possible that theACH acts by cutaneous vasodilation, warming the skin and reducing the afferent input from the skin, whereas epinephrine might have the opposite effect? Is it possible that acetylcholine, instead of acting on the ascending retic- ular formation, might act peripherally by vasodilation in the skin, thereby reducing the thermoreceptor inflow to the center? DR. KAWAMURA: Yes, there may be two processes. One is the effect of the autonomic function on the vasomotor system; the other is the direct effect of these drugs on the central somatic system. I woidd guess that the effect is either on the autonomic or somatic central mechanisms and that they are more important in this case than the peripheral ones. DR. RODDIE: How was the acetylcholine given? DR. KAWAMURA: By injection into the femoral vein. DR. RODDIE: It is just that the action of acetylcholine in the bocfy is a very short one; it is destroyed very quickly. That was shown by the heart rate records. But the inhibition of shivering lasted for quite a long time, right until the end of the record. I wonder whether there was any other effect that acetylcholine might have. DR. STUART: I do not think he showed heart rate. He showed blood pressure dropping. DR. RODDIE: There was an EKG record, and it came back to normal whereas shivering was inhibited throughout the record. DR. FREEMAN: Also, acetylcholine passes very slowly across the blood-brain barrier so that unless one gave it by intrathecal in- jection into the purported centers one would find it hard to see how it could escape destruction by acetylcholinesterase in the blood. 218 NEURO-MUSCULAR ORGANIZATION OF SHIVERING DR. HENSEL: In the record, I saw that after administration of acetylcholine the blood pressure fell from about 120 to about 60; and I wonder whether there isany possibility that the blood supply of the brain mi^t have changed. We know that shivering is very much in- fluenced by changes in oxygen pressure and CO pressure. Might this not be the cause of your change 7 1 do not know, but how did you exclude this possibility? DR. STUART: To say that the particular activating system is adrenegic and that it would be facilitated by an injection of epineph- rine and inhibited by acetylcholine, is a point of great controversy today, I think, among neuropharmacologists. I did notthink that Dr. Kawamura meant to imply that this was necessarily so; but he is suggesting that this is one alternative. DR. CLARK: That difference in shivering intensity between sum- mer and winter is intriguing. Did you, by any chance, give any of your winter-summer dogs thyroxin? DR. KAWAMURA: No, I have never applied thyroxin to the dogs in this experiment. DR. CLARK: And was the difference between summer and winter shivering only observed under sodium amytal anesthesia as well? DR . KAWAMURA: Yes, I tried anesthesia with other barbiturates too. Pentobarbital, thiopental, and ether were used. The summer- winter differences in shivering were the same, independent of the anesthetic drug. Sometimes I could induce weak shivering even in summer in the dog, but it was very difficult to get and was short lasting. DR. CLARK: I operate under nembutal and routinely I see my animals shivering when they are coming out of the anesthesia re- gardless of the season. DR. JOHANSEN: In regard to the interrelations you have pro- posed between respiration and shivering, have you ever tried to manipulate with the respiratory phases in an anesthetized animal? 219 KAWAMURA, Y. DR. KAWAMURA: This would be an interesting problem. DR. JOHANSEN: I was thinking that you mi^t try artificial respiration and attempt to interfere with the phases of the respir- ation, in order to exclude, for instance, the stretch receptors. DR. KAWAMURA: I have made a study of the relationship of j artificial respiration and shivering. Through my experience I can tell you that shivering movement is always followed by the inspira- tory phase, under fast or slow rhythm of artificial respiration. DR. FREEMAN: We found the same relationship between shiv- ering and inspiration when we were recording unit activity in the brain stem associated with the shivering. These unit potentials come with shivering and go when shivering stops, but during the early part when shivering is phasic with inspiration these bursts are synchronous with inspiration. Also, your observation of a painful stimulus is true of these units as well. There are many units in the nervous system which can be drivenby painful stimuli. Their activ- ity can be either started or increased and there are relatively few that will be inhibited; but these shivering unit potentials are very strikingly shut off, and when the painful stimulus is removed, there is a striking rebound phenomenon. Their frequency increases abrupt- ly above the preceding level. DR. HENSEL: Have you any observation of the topography of shivering in the proximal and distal parts of the limb ? We found in our investigations in the cold chamber that shivering starts in the proximal parts of the limbs as measured by the EMG. After pro- longed cold exposure, then the increased muscular tone in the prox- imal parts will disappear, and increased tone in the distal parts appears; so there is a temporal shift in the topography of shivering. DR. KAWAMURA: Yes, I believe there is a problem, and I vis- ually recognized that usually the peripheral region begins shivering earlier than the proximal vibration appears, but no exact recordings were made on the proximal and distal distribution of shivering. DR. FREEMAN:If you asphyxiate an animal or impair his res- piration, then instead of having simply the intercostal muscles and 220 NEUROMUSCULAR ORGANIZATION OF SHIVERING diaphragm active, you get accessory muscles that are inspiratory and these are predominantly in the neck and jaw. I wonder if you made any attempt to correlate the pattern of onset of shivering in muscles with the pattern of onset of these accessory muscles as asphyxia is brought into play? DR. KAWAMURA.: This was not measured. It would be worth- while to investigate this to determine whether or not the phasic activity of the accessory muscles facilitates shivering. DR. LIM: There is one comment. Dr. Hensel tells me that the carbon dioxide inhalation, five to six percent, increases shivering which phenomenon may be in favor of this last possibility in raising the respiratory activity. DR. HENSEL: Yes. DR. STUART: One important aspect ofthis paper is the demon- stration of synchronous activity in antagonistic muscles during shiv- ering. This, together with the shivering tremor frequency of 10-12 cps, is distinctly different from the Parkinson tremor (4-7 cps) in which antagonistic muscle activity is alternating. Now, you showed us evidence that the flexor and extensor fire synchronously. DR. KAWAMURA: Yes, I did. DR. STUART: I h ink this is unique in the literature. Have you recorded the tension developed in antagonistic muscles during shiv- ering? Or do you know of any accounts of it? DR. KAWAMURA: No, I have never measured that. DR. HENSEL: Did you record hypothalamic temperature during your experiments with local cutaneous heating? DR. KAWAMURA: No, I did not. DR. HENSEL: That is the question, because we found in the cat that local cutaneous heating might cause quite considerable changes in hypothalamic temperature without any considerable change in rectal temperature. 221 KAWAMURA, Y. DR. HEMINGWAY: I think Dr. Kawamura heated the back, did you not? DR. KAWAMURA: Yes, I did. j I DR. HEMINGWAY: Are you thinking of the head and back or just the back? DR. HENSEL: I mean just the limb, not the back but the hind limb. The hypothalamic temperature changed veiy slowly following heating of the hind limb. 222 NEURO-MUSCULAR ORGANIZATION OF SHIVERING REFERENCES 1. Andersson, B., R. Grant, and S. Larsson. Central control of heat loss mechanisms in the goat. Acta. Physiol. Scand. 37:261-280, 1956. 2. Bonvallet, M., P. Dell, and G. Hiebel. Tonus sympathique et activite electrique corticole. EEG clin. Neurophysiol. 6:119- 144, 1954. 3. Bradley, P. B. Microelectrode approach to the Neuropharma- cology of the reticular Formation. Psychotropic Drugs. Elsevier Pub. Co., p. 205-216, 1958. 4. Burton, A. C. and D. W. Bronk. The motor mechanism of shiv- ering and of thermal muscular tone. Am. J. Physiol. 119: 284, 1937. 5. Cassidy, G. J., S. Dworkin, and W. H. Finney. The effect of various sugars (and of adrenalin and pituitrin) in restoring the shivering reflex. Am J. Physiol. 77:211-218, 1926. 6. Cort, J. H. and R. A. McCance. The neural control of shivering in the pig. J. Physiol. 120:115-121, 1953. 7. Hall, V. E. and P. B. Goldstone. The influence of epinephrine on shivering and on metabolism in the cold. J. Pharma. and Exp. Therap. 68:247-251, 1940. 8. Hemingway, A., P. Forgrave, and L. Birzis. Shivering suppres- sion by hypothalamic stimulation. J. Neurophysiol. 17:375- 386, 1954. ). Kawamura, Y. and J. Fujimoto. A study of the jaw opening reflex. Med. J. Osaka Univ. 9:377-387, 1958. 223 KAWAMURA, Y. 10. Kawamura, Y., M. Funakoshi, and S. Tsukamota. Brainstem representation of jaw muscle activities of the dog. Jap. J. Physiol. 8:292-304, 1958. 11. Magoun, H. W., F. Harrison, J. R. Brobeck,and S. W. Hanson. Activation of heat loss mechanisms by local heating of the brain. J. Neurophysiol. 1:101-114, 1938. 12. Perkins, J. F. The role of the Proprioceptors in Shivering. Am. J. Physiol. 145:264-71, 1945. 13 . Rothballer , A. B . Studies on the adrenaline- sensitive component of the reticular activating system. EEG and clin. Neurophys- iol. 8:603-621, 1956. 14. Strom, G. Influence of local thermal stimulation of the hypo- thalamus of the cat on cutaneous blood flow and respiratory rate. Acta physiol. Scand. 20:Suppl. 70, 1-118, 1950. 15. Stuart, D., Y. Kawamura, and A. Hemingway. The Physiologist 3(3):156, 1960. 16. Uprus, v., G. B. Gaylor, and E. A. Carmichael. Shivering: A Clinical study with especial references to the afferent and efferent pathway. Brain. 58:220-232, 1935. 17. Von Euler, C. Slow temperature potentials in the hypothalamus. J. Cell and Comp. Physiol. 36:333-350, 1950. 224 NERVOUS PATHWAYS IN THE CHEMICAL REGULATION AGAINST COLD AND THEIR CENTRAL CONTROL J. Chatonnet Faculte Mixte de Medecine et de Pharniacie Universlte de Lyon France The nervous system can activate heat production in various ways, at least theoretically. Three main processes are usually called upon. First, the nervous system can increase the specific ac- tivity of every innervated organ; mechanical activity of muscle is a good example of this. The heat production will increase according- ly, the useful work being very small in the case of muscular tonus of shivering. Second, there are nonspecific reactions of emergency function, which, though they are not directly adapted to the struggle against cold, spend energy and produce heat useful for this purpose. An example of this is the activation of the sympathetic system and epinephrine secretion. Finally, there exists a specific and indepen- dent system which activates metabolism and heat production by various organs, especially the liver and splanchnic viscera, without any other efficient activity. Although the roles of these various means have often been in- vestigated in the past, only few works are at the present directed in this line; however, no definite solution has been found yet. We will consider the problem from the standpoint of the quantitative impor- tance of each of these factors and the conditions under which they are put into action. Most investigators are concerned with the disturbances of the central temperature after excluding one of the effector mechanisms. Such a method cannot provide sufficient and quantitative information. Also, the estimation of the heat production itself is not of much use in many cases. Indeed, the caloric response to cold depends more or less on physical regulation, chiefly on peripheral vasomotor activity, and it would be preferable if it were suppressed before- hand. Besides, one must be sure that the intensity of the cold action is responsible for the development of the caloric response as it may be the result of an increase of heat loss. Finally, a 225 CHATONNET, J. difficulty may arise from the great potency of the mechanisms of struggle considered as a whole in some animal species. In the dog, for example, the ability to withstand the cold is very important; the physical regulation may spare the load on heat production and in the severe cold the heat production can increase to very high values without any inefficiency being detected. Only extreme cold can sur- pass the total capacity for defense in the normal animal. In mod- erate cold exposure, oneeffectormechanismmay easily be replaced by the others. It is therefore necessary to know the greatest capacity for thermogenesis and, if that requires excessive conditions of temperature, to remove part of these effector mechanisms and to determine then the maximum capacity for heat production. After- wards, every further reduction of that maximum can be ascribed to the subsequent operation. The comparison of these various levels of thermogenesis is only valuable if they are true maxima and if no substitutes are possible. CONTROL OF THE MUSCULAR ACTIVITY Voluntary muscular activity, shivering, and thermal muscle tone are effected by the same muscle but according to different kinds of stimulation. Of course some parts of the muscular apparatus are concerned more with one sort of activity, but not exclusively. In the spinal cord, there is some evidence of the existence of descending pathways especially controlling shivering. It is known that corticospinal tracts are not necessary for that control. More- over, it appears that different patterns of muscular activity can substitute for each other. Indeed, in the dog deprived of reactions of shivering by section of the anterior part of the cord, muscular activity may still increase in cold ambient temperatures, in the form of agitation. Authors such as Freund and Strasmann (1912) and Isenschmid (1926) tried to measure the importance of thermogenesis provided 226 NERVOUS PATHWAYS by muscular activity by cuttingmotor nerves to an extensive region. This procedure pointed out the persistence of a potent increase of heat production against cold. But the active portion of the muscula- ture can still be responsible for that result. Lesions of the spinal cord have been attempted also, though they are less specific in their meaning. According to Freund (1913), section of the thoracic cord in the rabbit abolishes the whole phys- ical regulation against cold, leaving the chemical regulation un- changed or even more active in compensation. Kayser (1929), experimenting on the pigeon, observed similar results after section of the thoracic cord. Hermann, (1938), using the dog, saw the per- sistence of an efficient regulation against cold after destruction of the cord below thoracic one level. In cold ambient temperature, there is a distinctly greater increase of heat production in the same animal species we have seen than in the dog before the lesion. The maximum metabolism is not yet reached at -15 C, and the curve of metabolism is still ascending at this temperature despite extensive paralysis and the lack of epinephrine secretion. In that preparation, of course, a part of the musculature remains active and seems to be able to compensate for exaggerated heat loss in the cold. Indeed, in absence of vasomotor regulation, the heat production has to balance more exactly the thermal demand of the environment and any deficit must appear unmasked. Issekutz (1937) found that cervical section of the dog reduces the thermogenetic responses in the cold. According to Freund (1913), rabbits then become poikilothermic. These authors explain these results by the assumption of the suppression in that section of important vegetative outflow leaving the cord at this low cervical level. Our results confirm that section or destruction of the cord reaching the inferior cervical segments seriously disturbs the chemical regulation in the dog. In the cold, below -10° C, a maximum metabolism level of 2 to 2.5 times the basal value is reached with fall in central temperature, but there is no complete poikilothermia. Besides, we think that the difference between cervical and thoracic lesions can be explained more properly by the motor outflow of the brachial plexus at this level. In order to test this explanation, we tried to ascertain the role of the cervical and brachial plexuses in the potent thermogenetic 227 CHATONNET, J. capacities of dogs whose spinal cords are severed at thoracic one level (Fig. 1). In 10 to 12 dogs operated on in this manner, bilateral section of the brachial plexus induced an obvious disturbance of the regulation, very similar to those resulting from cervical lesions; below 0° C, the regulation failed and the central temperature de- creased when the cold exposure continued. Under the same conditions the heat production could not surpass a given level; we obtained a true maximum. Therefore, the muscular activity in the narrow field controlled by the cervical and brachial plexuses is able to provide an important though unmeasured part of the regulatory thermogenesis (Figs. 2 and 3). Also, one can suppose that the regulatory power remaining after that operation depends on the activity of musculature in the neck and face, which is almost impossible to eliminate. Many other results point out the importance of the muscular factor in the heat production elicited by cold exposure . Hammel and Harcfy (1960) observed a very close parallelism between the bursts of shivering and the curve of heat production in the dog. During the intervals between these bursts the metabolism drops down to basal values. Carlson (1955) made similar observations in the rabbit. The old experiments on curarization are liable to objections because of the non-specific actions of drugs. Davis (1955), using curare, assigns the extra-motor reactions for 40 per cent of the heat production by the mouse in the cold. Werner and others (1956) found spontaneous rewarming of curarized dogs with blankets after anes- thesia in a room at 24 C, but the role of basal combustion is not taken into account. Cottle and Carlson (1954) noted a distinct in- crease of heat production of curarized rats exposed to cold, chiefly in adapted animals. 228 NERVOUS PATHWAYS Col /Kg /H DOG 700 9 -+ normal >•• spinal cord severed of Th1 level 1 /*.'• -30 -20 -10 0 10 ^ .20 30 40 Ambient Temp. Figure 1. Heat production of dogs plotted against ambient temperature. The normal dog and dog after spinal cord had been severed at thoracic one level. 229 Col/Kg/H. CHATONNET, J. DOG 501 ♦ nrormal spinal cord severed ■*arTh2 level >o brachial plexus severed -»V. -30 -20 -10 0 10 20 30 40 "C Ambient Temp, Figure 2. Heat production of dogs. The normal dog, the dog after spinal cord had been severed at thoracic one level, and the dog after severing the brachial plexus. 230 NERVOUS PATHWAYS Ccl/ Kq/H OOC 500 ■ • sprnal cord severed a f 2Th level ...cervical plexus severed brachial plexus severed -30 ■20 -10 10 20 Ambienr Temp 30 Figure ;j. Heat production of dogs. The normal dog, after the spinal cord had been severed a thoracic two level, after the cervical plexus had been severed, and after the brachial plexus had been severed. 231 CHATONNET, J. THE ROLE OF EPINEPHRINE SECRETION Can the calorigenic action of hormones of the adrenal medulla play an important part in the regulatory heat production? In 1941, Hemingway and Hathaway ascribed to this factor the slight increase of metabolism observed at the beginning of the cold exposure, be- fore the onset of shivering. Of course, extraneous injected epineph- rine can substitute for the heat required by the cold exposure. How- ever, Lundholm (1949) estimated a maximum augmentation of basal combustions by epinephrine injection of only 28 per cent. But, sensitization of that action can be provided by acclimation (Cottle and Carlson, 1956). In experiments on removal of the adrenal med- ulla in the rat, Thibault (1948) estimates that epinephrine contributes as much as 35 to 40 per cent to the total power of thermogenesis. In the dog, such a procedure is inefficient, even in very cold envir- onments, because of the importance of other substitutes. In order to avoid these difficulties, we have tried first to elim- inate an important part of the muscular innervation, leaving the spinal control of epinephrine secretion intact, by section of the brachial plexus, associated with section of the roots of the spinal cord below lumbar two level. A maximum metabolism is still reached. Then, one adrenal medulla is denervated and cleaned, and the other gland removed. A significant but rather mild lowering of the maximum metabolism level is obtained (50 to 100 per cent of the basal metabolism) (Fig. 4). Probably the main role of the sym- pathicoadrenal system in the defense against cold consists of the reduction of heat loss. ZVl NERVOUS PATHWAYS 233 CHATONNET, J. SPECIFIC CALORIFIC INNERVATION A specific control of heat production has sometimes been as- sumed. For the muscles, no recent experience can support the views of Freund (1914) and of Wenger (1933) that a nervous apparatus controls the metabolism without any other activity. Certainly, the muscle can present various kinds of activity according to the con- ditions of stimulation (tonus, shivering). During acclimation, motor and electric activity could decrease (Sellers, 1954), or become dif- ferent (Carlson, 1955) despite an increase in metabolism and the development of vascularization pointed out in these conditions by Heroux, and Saint-Pierre (1957). However, these reactions can be the result of hormonal control, and no special innervation is neces- sary for them. It may be questioned, also, whether there is any special in- nervation of abdominal viscera, chiefly of the liver, subjecting their heat producing activities to the thermoregulatory control. Some investigators stress the importance of these organs in the chemical regulation against cold (Fedorov and Shur, 1942; Jitariuet al., 1941, Donhoffer et al., 1959). But evaluating the role played by these or- gans is very difficult. A nervous control of their activity has also been considered; Freund (1913) first, and then von Issekutz (1937) explained the importance of the low cervical cord for chemical tem- perature regulation because nervous impulses leave the cord here, and cross over the stellate ganglia or possibly the vagi, and thus they are able to reach the liver. On three dogs, whose active muscular fields were reduced be- forehand by section of the spinal cord at fourth thoracic level and section of the brachial plexus bilaterally, and which exhibited a "plateau" of metabolism in the cold, we performed a section of both pneumogastric nerves in the thorax. After this operation the maxi- mum metabolism remains unaltered (Fig. 5). On two other dogs pre- pared in the same way for a limitation of muscular activity, bilateral stellectomy failed to induce any change of the plateau level (Fig. 6). These results confirm the above-mentioned interpretation of a muscular destination of the cervical outflow controlling chemical 2 34 NERVOUS PATHWAYS Cal/Kg/H DOG 1601 3 - 1 - B norma I spinal cord severed of C7 level bilateral thoracic vagotomy -30 -20 -10 10 20 30 40 Ambrent temp. Figure 5. Heat production of dogs. Ttie normal dog, after tlie spinal cord tiadbeen severed at C7 level, and after bilateral tlioracic vagotomy. 235 CHATONNET, J. 3 - Cal/Kg/H •\ DOG 1510 Q Q normal spinal cord and brach. • plexus severed bilateral ablation oF stellate ganglia — ,-• ^» "^ •'\ -r-s- I.NITI^L TEMP. y" 0 1 2 >-|3 -30 -20 10 20 30 40 C AMBIENT TEMP. Figure 6. Heat production of dogs. Normal dog, after severing the spinal cord and brachial plexus, and after bilateral ablation of the stellate ganglia. 236 NERVOUS PATHWAYS regulation. However, a control of that sort by splanchnic nerves remains possible, but it is not yet thoroughly investigated. Thus the control of the musculature in its motor activity seems to bethechief nervous factor in chemical regulation in the dog, since the remaining thermogenetic power after extensive muscular de- nervation is poor. Moreover, the part of the hormonal control appears very small and restricted to the effects of epinephrine secretion, at least ifone lays aside the slow processes of adaptation. All of our preceding results are expressed as a correlation between thermal production and ambient temperature. Now, by varying ambient temperature we try to submit the animal to various thermal demands. One can ask whether the correlation is with am- bient temperature or with peripheral skin temperature. In the latter case, the value of our results must be questioned. Indeed, there is often in those animals a great reduction of the thermal sensitive skin area by nervous section. It is then necessary to know the significance of the measured calorification and of its maximum, and what kind of reaction is reflected by it. Besides the ambient temperature, two other values can be inves- tigated, the mean skin temperature and the difference between skin temperature and ambient temperature. This last value can provide information on the thermal gradient, and hence on heat loss. We made such an estimation on some of our animals which were para- lyzed or which lacked nervous section, by measuring three or four skin temperatures in the still sensitive area at different ambient temperatures in steady conditions. The calorific production, P, measured during a period of 20 minutes, is plotted against super- ficial temperature t , or t - t where t is ambient temperature. The difference between superficial and arnbient temperature (t - t ) appears to give a good correlation, while there is no correlation with t (Fig. 7). The deficit in thermoregulation of the dog with extensive mus- cular denervation remains unchanged for months. Besides, it becomes apparent only in severe cold. In mild cold exposure, heat production develops in a progressive and adapted manner. There is a limitation of but not a disorder in the regulation. 237 CHATONNET, J. ^HL£KgVK DOG 700 Yd 0,6 60 T.5-Ta ^ m 0,039 10 15 20 25 30 35 <0 45«>C Figure 7. Heat production of dogs plotted against t - t and plotted against T where t is the skin temperature and t is the ambient temperature. 238 NERVOUS PATHWAYS However, the existence of a central sensitivity towards the cold is still in question. Numerous attempts to demonstrate this sensitivity are negative, according to the work of Strom (1950), Forster et al. (1952), and Brendel (1960);sucha central mechanism is sometimes judged useless since, on exposure to cold, the im- mediate reaction is an increase of the core temperature. Positive results of Barbour (1912) are criticized because of the tremendous lowering of braii;i temperature required for the experiment. Since 1954, we have made several approaches to this problem, in order to know whether chemical regulation and particularly muscular activity and shivering can be put into action by a central action of the cold. At the beginning of this work we tried to obtain, by several means, a shift in the normal temperature gradient of the bocfy. Thus, the internal bocfy temperature might decrease (in order to trigger any central sensitivity) but the superficial skin tempera- tures would remain constant or even increase (in order that super- ficial cold receptors would not be stimulated). Experiments were performed systematically onunanesthetized dogs bearing chronically implanted thermocouples in the brain; several subcutaneous temperatures were also registered. Shivering and muscular tone were registered by electromyograph, and some- times by respiratory exchanges. In a chronic spinal dog, or in a dog whose spinal cord is des- troyed up to thoracic one level, central cooling may be induced by wrapping ice bags around the posterior part of the bocfy which is deprived of sensitivity. In such a case, warm blankets and infrared lights induce a warmer temperature on the anterior part of the body and particularly on the face. In other cases, for reflex cold stimulation, the face could be refrigerated. The results may be summarized as follows on a graph (Fig. 8): on the ordinate is central temperature; on the abscissa is mean superficial temperature. It is possible to show the existence of a "shivering zone" above a threshold curve. Thus, during the same experiment, shivering may be induced by the decrease of super- ficial temperature (with a constant central temperature) or by a decrease of central temperature (with no change or even an increase 239 CHATONNET, J. DOG PINGOUIN 47 57 117 CENTRAL TEMPERATURE oc • NO SHIVERING > SHIVERING o NO SHIVERING (CHLORPROMAZINE) 40- 37- 34 43 SUB-CUTAN TEMPERATURE °C Figure 8. Thermal shift through ice water ingestion. In normal unanesthetized dogs, ingestion of one liter of ice water induces a rapid fall of central temperature with much shivering. 240 NERVOUS PATHWAYS of superficial temperature). The threshold curve changes with the animal's conditions and may go up during fever (spontaneous, or induced by intravenous injection of vaccine). In normal unanesthetized dogs, with intracerebral thermo- couples, ingestion of one liter of ice water induces a rapid fall of central temperatures with much shivering, even whenthe superficial temperature is kept hig^. Shivering disappears when the central temperature comes back to the normal level. Finally, cold exposure on the thermal equilibrium can be com- pensated for by heat production of the muscular work that is exper- imentally imposed on the organism. There are two possibilities in this case. The autonomic thermogenesis of the organism may be regulated by the superficial stimuli; this thermogenesis would thus add to the heat production of the work. If the regulation is made by the need of constancy of central temperature, which is almost accomplished, there is substitution. We have shown that, in spinal dogs, the muscular work induced in muscles of the hind legs by electric stimulation increases the oxygen consumption only in neutral ambient temperature. In the cold, this consumption replaces the need for oxygen which is induced by cold, but does not increase it. Direct cooling of the central nervous system We implanted aseptically and stereotaxically in the brains of eleven dogs a thermode producing cold by gaseous detent of propane. The central temperature near the thermode was registered with a thermistor at the same time as the rectal and three or four super- ficial temperatures. The electromyograph was recorded in the awake dog with an ink-writer, with or without integration (Figs. 9 and 10). Results show that there is a close correlation between the cen- tral temperature and the electromyograph. Every decrease of the cerebral temperature is associatedwith an increase of the muscular tone or even with a clear-cut shivering. The most striking results are obtained when the thermode is located at the middle hypothal- amus. In such a case, only a decrease of 1/10 degree centigrade is 241 CHATONNET, J. t -i- B 1 I I 1 I : t r ■ - t s 242 NERVOUS PATHWAYS sufficient to trigger a muscular response. These reactions are ob- tained even with a constant superficial temperature; nevertheless, the superficial temperature has some effect on the amplitude of the muscular reaction induced by cold central stimulation. These results agree with those of Hammel, Wyndham, and Hardy (1958), who obtained shivering by local cooling of the dog hypothalamus. Freeman and Davis (1959) and Kri^er et al. (1959) were also able to register vasomotor reactions in the cat to cold acting in this region. Donhoffer et al. (1957), cooling the brains of rats, observed an increase of heat production. Lim (1960), from experiments in cooling carotid blood in the dog, concluded that shivering can be elicited either peripherally or centrally. In this respect, negative attempts can re suit from the use of anesthesia, from the operative shock, or from defective localiza- tion of the thermode, in consideration of the strong convection in the brain. Exclusion of central sensitivity. Chlorpromazine does not de- crease the capacity of resistance against cold in a normal animal. A dog treated with chlorpromazine in a cold environment may shiver normally and increase its therm ogene sis, and may be used as a control animal. But the hypothermic dog (spinal dog overstrained by a cold environment) given chlorpromazine, does not react against cold and stays hypothermic if there is no superficial stimulation, that is, if superficial receptors are submitted to a high ambient temperature by heating. But such a dog may shiver and its rectal temperature may increase when it is surrounded by a cold ambient temperature. This paradoxical behavior can be explained by a selec- tive suppression of the central component of the thermal sensitivity to cold by chlorpromazine. We performed electrolytic destruction of the hypothalamus in nineteen dogs. In several of them we obtained thermoregulatory be- havior resembling that of chlorpromazine. Such animals recover some reactions against cold post-ope ratively, such as increase of thermogenesis when placed in a cold ambient temperature, but this reaction does not permit an exact adaptation of the heat production to the loss. The central temperature remains unsettled and there is 243 CHATONNET, J. no muscular reaction when the central temperature is selectively decreased (Fig. 11). Keller (1959), in coagulation experiments on the dog hypothal- amus, reports cases of defect in the regulation of central temper- ature against cold, although potent reactions of shivering, no doubt of reflex origin, take place. On the other hand, the same author succeeded in observing shivering after elimination of the whole peripheral thermal sensitivity by cutting the brain stem except for the pyramidal bundles. In the last case the good resistance in cold exposure could be ascribed to brain sensitivity. The existence of a central cold-sensitive device in the brain seems well proved; nevertheless, its functional significance remains to be evaluated. One may also ask whether the thermogenetic response is spe- cific for each kind of activation (central or reflex). According to Davis and Mayer (1955), the chemical regulation, in its restricted meaning (without muscular activity), depends only on the central temperature. On the other hand, shivering and muscular activity would be under a pure reflex control. Some of the preceding results are in conflict with this view. As for the control of epinephrine, no recent direct evidence has been obtained. 244 6 - NERVOUS PATHWAYS Col/Kg/- DOG 100 Brain srem electro - lyCic lesion 'initial Temp -,f2 ♦1 0 _1 -2 ■30 -20 -10 0 10 20 ao'c AMBIENT TEMP. Figure 11. Heat production of dog following brain stem elec- trolytic lesion. 245 CHATONNET, J. DISCUSSION MR. EAGAN: In your first series of experiments, did your metabolism figures at the ambient temperature represent the maxi- mum metabolic rate that you measured at that temperature? Was this the sum of metabolism? DR. CHATONNET: The metabolism figures represent the meta- bolic rate actually measured at each ambient temperature. In very cold ambient temperature and only after exclusion of muscular activity by denervation in an extensive field, the maximum is reached. By adrenal demeduUation I try to suppress a thermogen- etic factor and, in such a way, to evaluate its importance. Of course, the vasomotor effect of adrenal medulla hormones plays some role in heat loss, but in the dog which showed beforehand a maximum of metabolism in the cold, the increase of heat loss follow- ing this operation does not modify this metabolism level. It only reduces the range of external temperatures regulated. DR. CLARK: Since you apparently kept your animals alive a very long time, one wonders about sensitization to circulating hormones. Were they more sensitive to adrenalin than normals? DR. CHATONNET: Yes, the chronic spinal dog, the acute dog, is very sensitive to epinephrine, as far as the vascular action. DR. CLARK: I was thinking of action on skin vessels. DR. FREEMAN: Doesn't enervation take place after spinal sec- tion? DR. CLARK: In the cat it does. And how about pitressin? Did you try pitressin in any of them? DR. CHATONNET: I didnottry it.I do not believe that hypophy- sis plays an important role in regulation in spinal dogs, at least in the capacity of thermogenesis. 246 NERVOUS PATHWAYS DR. CLARK: I know the C7 spinal cat becomes exquisitely sensitive . DR. CHATONNET: I know the sensitivity against posterior pituitary extract. DR. FREEMAN: Have you discussed your results with Dell or Stutinsky or Bonvallet, who worked similarly? DR. CHATONNET: Yes. DR. FREEMAN: They came, I believe, to the conclusion that the hypophysis played a major role in regulation in spinal dogs. DR. CHATONNET: I do not beUeve that the hypophysis plays an important role in regulation in spinal dogs, at least in the capacity of thermogenesis that I studied. Bonvallet and Dell (1946) observed that hypophysectomy or the section of the supraoptico-hypophyseal tractus deprived the spinal- dog of the ability to maintain his central temperature for exposure to 20 C to 22 C. Injection of posterior pituitary extract again set up this ability. The authors explain the fact at least partly thus: impairment of water metabolism by hypophysectomy lessens the efficiency of the remaining thermo- genesis in that animal. But they have not measured the metabolism of their dogs. One can also suppose that the essential disturbance concerned heat loss regulation by vasomotor control. In any way, the thermogenetic capacity of the C7 spinal dog is poor and does not depend on any "per se" thermogenic hormonal control, but on residual muscular activity. However, such a long-acting vascular factor should produce slow adaptation to environmental change. In the Bonvallet and Dell experiments the reactions were very fast. Our chronic spinal dogs and the dogs whose spinal cords have been destroyedcontroltheirtemperature in a limited range of exter- nal temperatures according to the extent of innervation abolished, but in this range their regulation is immediate and the thermogene- sis involved is steady from day to day. DR. FREEMAN: Did you compare your results with theirs? 247 CHATONNET, J. DR. CHATONNET: I had no results of hypophysectomy. DR. HANNON: Have you ever studied what I prefer to call non- shivering thermogenesis rather than thermochemical thermogen- esis? Have you ever studiedthis in climatized animals or acclimated animals? DR. CHATONNET: No. DR. HANNON: This mi^t prove interesting since these animals have almost completely replaced the shivering thermogenesis with the non-shivering thermogenesis. DR. CHATONNET: In my experiments, dogs are kept at constant room temperature of 25 C, except for cold tests. In these conditions the non-shivering thermogenesis is not very important. And it is doubtful that this mechanism can develop by acclimation in such large animals. DR. HANNON: I would like to toss out one other question to Dr. Chatonnet or any of the other participants here. In these cold clima- tized animals, during the processof climatization,they at first have a shivering thermogenesis when they go into the cold. This declines gradually and reaches minimum values after a few weeks. Concomi- tant with this decline, non-shivering thermogenesis increases and replaces it. Now if you take this animal out of the cold, his thermo- genesis generally drops down to about what a control animal would be — a little higher, maybe, with essentially the same as a cold animal. You put him back in the cold and almost immediately he starts this non-shivering thermogenesis. What is the mechanism? MR. EAGAN: It is indeed a mystery, because curarized animals and animals that are adrenal demedullated still control this in- creased potential for cold- induced thermogenesis. 248 NERVOUS PATHWAYS DR. FREEMAN: As Bonvallet pointed out, this tendency was predominant in small animals; the smaller the animal, the greater his range of variance. I think he quoted figures indicating that a rat can vary it over a range of forty times, whereas, in man, it is greatest at about one or two times. That is a large mammal. This form of regulation is important in the smallest one with a high ratio of surface area to body weight. DR. HANNON: Yes, this is true. DR. FREEMAN: So that you would not expect it to be as pre- dominant in dogs, but nevertheless, he thought there existed a mechanism in dogs which was the same as that in other animals, but not as effective. DR. CLARK: You can demonstrate that in spinal cats. DR. FREEMAN: How? DR. CLARK: If you take a cat that has had the cord severed at, say, C6, C7, and you had him alive for months, if you put him in a temperature of 18 C, his temperature will drop way down and then over a course of a few days, it will come up to almost what it was before. You take him then and put him in 29 C, and his temperature goes up to 39 C or 40 C, and it takes about 3 to 5 days for his temperature to come back to normal level. 249 CHATONNET, J. REFERENCES 1. Barbour, H. G. Die Wirkung unmittlebarer Erwarmung und Abkuhlung der Warmezentren auf die Kopertemperatur. Arch. exp. Pathol., 70:1-26, 1912. 2. Beaton, L. E. and C. R. Leininger. Spinal distribution of ther- moregulatory pathways in the monkey. J. Neurophysiol. 6:37- 38, 1943. 3. Birzis, L. and A, Hemingway. Descending brain stem con- nections controlling shivering in cats. J. Neurophysiol. 19: 37-43, 1956. 4. Bonvallet, M. and P. Dell. La regulation therm ique du chien spinal sans hypophyse. C. R. Soc. Biol. 140:942-945, 1946. 5. Brendel, W. Die Bedeutung der Hirntemperatur fiir die Kalte- gegenregulation. Pflilgers Arch., 270:607-647, 1960. 6. Carlson, L. D. Heat production and cold adaptation. Fed. Proc. 14:26, 1955. 7. Chatonnetj J. Importance du frisson dans les capacites de thermogenese du chien a moelle epiniere detruite. C. R. Soc. Biol. 145:1538-1540, 1951. 8. Chatonnet, J. Sur I'origine de la chaleur libe'ree dans la regu- lation chimique de la temperature. J. Physiol. (Paris), 51: 319-378. 1959. 9. Chatonnet, J. et M. Tanche. Sur I'existence dans la moelle epiniere de voies nerveuses distinctes pour la commande du frisson. C. R. Acad. Sci. 249:412. 1956. 250 NERVOUS PATHWAYS 10. Chatonnet, J. et M. Tanche. Dissociation du frisson "central" et du frisson "reflexe" chez le chien a moelle detruite. J. Physiol. (Paris), 48:439-442, 1956. 11. Chatonnet, J. et M. Tanche. Chlorpromazineet regulation ther- mique - Zone de froid. Therapie 14:778-792, 1959. 12. Chatonnet, J. et M. Tanche. Influence d'unabaissementde tem- perature localise de la region dience'phalique sur I'activite musculaire thermoregulatrice chez le chien eve ille. J. Phys- iol. (Paris), 52:48-49, 1960. 13. Cottle, W. and L. D. Carlson. Adaptative changes in rats ex- posed to cold. Am. J. Physiol. 178:305-308, 1954. 14. Cottle, W. and L. D. Carlson. Chemical regulation of heat pro- duction in cold adapted rats. Fed. Proc. 15:41-42, 1956. 15. Davis, T. R. A. and J. Mayer. Demonstration and quantitative determination of the contribution of physical and chemical thermogenesis on acute exposure to cold. Am. J. Physiol. 181:675-678, 1955. 16. Donhoffer, S., M. Farkas, A. Haug-Laszlo, I. Jarai, and G. Szegvari. Das Verhalten der Warmeproduktion und derKor- pertemperature der Ratte bei Lokaler Erwarmung und Kiihlung des Gehirnes. Pfliigers Arch. 268:273-280, 1959. 17. Donhoffer, S., G. Szegvari, I. Varga-Nagy, and I. Jarai. Uber die Lokalisation der erhohten Warmeproduktion bei der chemischen Warm e regulation. Pfliigers Arch. 265:104-111, 1957. 18. Eiff, A. Von, K. Spohn, W. Gnutchel, and F. Hertle. Stoffwech- selundersuchungen am curarisierten nicht narkotisierten Menschen. Pfliigers Arch. 253:283-285, 1951. 19. Fedorov, N. A. and E.I. Shur. The role of the viscera in regu- lating the temperature of the bocfy of an animal under phys- iological and pathological conditions. Am. J. Physiol. 137: 30-38, 1942. 251 CHATONNET, J. 20. Forster, R. E. and T. B. Ferguson. Relationship between hypo- thalamic temperature and thermoregulatory effectors in unanesthetized cats. Am. J. Physiol. 169:255-259, 1952. 21. Freeman, W. J. and D. D. Davis. Effects on cats of conductive hypothalamic cooling. Am. J. Physiol. 197:145-148, 1959. 22. Freund, H. Uber die Bedeutung der Vagi fiir die Warmeregula- tion. Arch. Exp. Path. Pharm. 72:295-303, 1913. 23. Freund, H. und S. Janssens. liber den Sauerstoffverbrauchder Skeletmuskulatur und seine AbhangigkeitvonderWarmereg- ulation. Pflugers Arch. 100:96, 1923. 24. Freund, H. und E. Schlagintweit. Uber die Warme regulation curarisierter Tiere. Arch. exp. Path. Pharm. 77:258, 1914. 25. Freund, H. undR. Strasmann. ZurKenntnisdernervosen Mech- anismus der Warme regulation. Arch. Exp. Path. Pharm. 69: 12, 1912. 26. Hammel, H. T., J. D. Hardy, andM.M. Fusco. Thermoregula- tory responses to hypothalamic cooling in unanesthetized dogs. Am. J. Physiol. 198:481-486, 1960. 27. Hammel, H. T., C. H. Wyndham, and J^D. Harc^. Heat pro- duction and heat loss in the dog at 8 C to 36 C environ- mental temperature. Am. J. Physiol. 194:99-108, 1958. 28. Hemingway, A. and S. R. Hathaway. An investigation of chemi- cal temperature regulation. Am. J. Physiol. 134:596-602, 1941. 29. Hermann, H., G. Morin, and J. Vial. Efficacite de la thermo- regulation Chez le Chien sans moelle au cours du refroidis- sement. C. R. Soc. Biol. 127:1491-1493, 1938. 30. Heroux, O. and J. Saint-Pierre. Effect of cold acclimation on vascularization of ears, heart, liver and muscles of white rats. Am. J. Physiol. 188:163-168, 1957. 252 NERVOUS PATHWAYS 31. Issekutz, B. von. Die RoUe des Zentralnervensystems und der Schilddriise bei der Warmeregulation. Pfliigers Arch. 238: 787, 1937. 32. Isenschmid, R. Physiologie der Warmeregulation. Handbuch der norm, und path. Physiologie. 12:23, 1926. 33. Jitariu, P., A. Koch, and U. Otto. Uber Temperaturdifferenz und Durchblutungmessungen am Leber undihre Auswertbar- keit zur Beurteilung von Stoffwechselanderungen im Organ. Pflugers Arch. 245:317-341, 1941. 34. Kayser, Ch. Contribution ^ I'e^ude dumecanismenerveuxde la regulation thermique. Ann. de Physiologie, 5:131-223, 1929. 35. Keller, A. D. Neurologically induced physiological resistance to hypothermia. Ann. N. Y. Acad. Sci. 80:457-474, 1959. 36. Kriiger, F. J., H. W. Kundt, H. Hensel, K. Brlick. Das Verhalten der Hautdurchblutungbei Hypothalamuskuhlunganderwachen Katze. Pflugers Arch. 269:240-247, 1959. 37. Lim, T. P. K. Central and peripheral control mechanisms of shivering and its effects on respiration. J. appl. Physiol. 15: 567-574, 1960. 38. Lundholm, L. The effect of adrenaline on the oxygen consump- tion of resting animals. Acta. Physiol. Scand. 19 (suppl. 67): 1-133, 1949. 39. Morin, G. L'adrenaline, hormone du froid. Biol. Med. 37:196- 230, 1948. 40. Sellers, E. A., J. W. Scott, and N. Thomas. Electrical activity of skeletal muscle of normal and acclimatized rats on ex- posure to cold. Am. J. Physiol. 177:372-376, 1954. 41. Sherrington, C. S. Notes on temperature after spinal transection with some observations on shivering. J.Physiol. (Lond.),58: 405-424, 1924. 253 CHATONNET, J. 42. Sinelnikow, E. Uber die Wirkungweise des Warmezentrum im Gehirne. Arch. f. Anat. und Physiol., p. 279, 1910. 43. Strom, G. Effect of hypothalamic cooling on cutaneous blood flow in the unanesthetized dog. Acta. Physiol. Scand. 20 (suppl. 70):27l-277, 1950. 44. Tanche, M. et J. Chatonnet. Sur I'influence d'unapport calori- fique impose a I'organisme (travail musculaire) sur les reactions thermoregulatrices au froid. J. Physiol. (Paris). 50:534-537, 1958. 45. Thibault, O. Facteurs hormonaux de la thermoregulation chim- ique. Ann. Nutrition Paris 2:89-109, 1948. 46. Wenger, H. Nachweis der Warmebildenden Funktion des Sym- pathikus im Skelettmuskel. Z. Biol. 93:307-316, 1933. 47. Werner, A. Y., D. Dawson, and E. Hardenber^. Spontaneous rewarming of hypothermic curarizeddogs. Science 124:1145, 1956. 254 A CRITIQUE OF THE DOCTRINE OF "CENTERS' IN TEMPERATURE REGULATION Walter J. Freeman Department of Physiology University of California Berkeley, California The title of this presentation was taken from a remark by Dr. Hemingway at the start of this conference, in which he said that he would describe a "center" for shivering, even though he did not like the term. I agree that it is not a good term in some respects, but on the other hand I think that if properly used, it can serve us as well in the future as it has in the past. HISTORICAL SURVEY One of the remarkable features about temperature regulation is that not merely one organ is involved but all the organs in the body. Respiration, digestion, or reproduction predominantly involve one organ and its ramifying parts, but in temperature regulation one cannot simply consider the skin, or the muscles, or the endo- crines, but all of them in the patterns of activation imposed on them by the central nervous system. One can make balance sheets of energy exchange during different patterns, or one can study the way in which the central nervous system creates these patterns, by means of which the brain regulates its own operating temperature. Systematic analysis of this property was initiated about one hundred years ago with the studies by Samuel Goltz in 1874 on the physiological effects of heating and cooling carotid blood. This was a particularly propitious time to begin, for operative intervention in the brain had just become feasible with the introduction of aseptic surgery. Claude Bernard had just published his PhysiologieGenerale in which the doctrine of the constancy of the internal milieu was first set forth, and there was ample evidence for the regularity of 255 FREEMAN, W. J. the body temperature from studies of clinical thermometry instigated by Karl Wunderlich of patterns of temperature change in normal cycles and disease. The early workers were impressed by the stability of body temperature, and they introduced three concepts into temperature regulation which have persisted to the present day. One of these was the placing of a line on the clinical thermometer at 98.6 F (37 C), which implies a higher degree of constancy of body temperature than is actually the case. The second was the notion that each of the components of thermoregulatory responses was represented by a group of cells in the nervous system, to which was given the name "center." The third was that this collection of centers was under the control of a master center described as a thermostat. Initially, this thermostat was placed in the medulla, and only following work by Isenschmidt and Krehl and others at the turn of the century was it placed in the hypothalamus. Much of the history of temperature regulation since that time has consisted of the attempt to define the location and properties of these centers. In reality, this has only been part of a much broader application of the same principles to other parts of the brain, with particular attention devoted to parcelation of function in the cortex and the medulla. In some respects this doctrine seems to have been higjily successful, notably in the cases of the "respiratory centers" in the medulla and the "feeding centers" in the hypothalamus. In other instances such as the "speech center" in the parietal cortex, the "sleep center" in the thalamus, or the "vasomotor center" in the medulla, identification at first seemed adequate and now has been brought into question. In the case of temperature regulation, with the exception of the so-called "heat center" in the hypothalamus (which actually turned out to be a sensory mechanism), the best that can be said after nearly a century of hard work is that efforts continue. In view of the large amount of work based on this doctrine and the still continuing doubts as to its validity, one is led to ask how it came into existence in the first place and what the underlying assumptions for it have been. This doctrine represents in essence an attempt to subdivide the brain into its physiological components or parts. There were at least three other methods for subdividing the brain introduced in the last one hundred years, each to a large 256 TEMPERATURE REGULATION "CENTERS" extent the work of one man. J. Hughlings Jackson was led by clinical observations of neurological disease to subdivide the brain in a series of horizontal components, each of which maintained excitatory or inhibitory dominance over the segments below. Analysis of these segments was carried out by transections of the brain at various levels. I. P. Pavlov, on the basis of his conditional reflex studies, divided the brain into a series of vertical segments, or what he called motor and sensory analyzers. These consisted of chains of neurons extending from the sensory receptor through the spinal cord and basal ganglia to the cortex, or in the opposite direction from the cortex down to the effector without significant transverse interaction being postulated at intervening way stations. Another series of subdivisions was developed by C. Judson Herrick and others on the basis of caom pa rat ive behavioral studies and the comparative anatomy of the vertebrates. He based his sub- division largely on the phylogenetic differentiation of the parts of the brain, which ultimately led to recognition of the importance of various parts of the limbic lobe in behavior. Each of these views was carefully developed over a period of many years and was based on a systematic observation of total behavior, that is, the behavior of animals in the intact state. The doctrine of the centers had no direct relationship to any of these systems, nor, for that matter, to any other. Rather, it developed from observations of isolated "evoked" responses in non-behavioral contexts. It was not the work of any one man or any one school and has yet to be given a formal and logical description. Yet it has been as widely accepted and used as any of the other three. Historically, the term 'center' first came into general usage after 1870. Prior to that time it was rarely used in text books, but thereafter it occurred in almost all major text books of physiology and clinical neurology. Specifically, it seems to have followed the discoveries by Fritsch and Hitzig in 1870 (see Brazier, 1959) of the excitable motor cortex and by Claude Bernard of the hyperglycemia produced by stimulation of the medulla. Several kinds of influence seem to have facilitated its immediate acceptance as a method of analyzing or describing brain function. First among these was un- doubtedly the development of new techniques for histological stain- ing, which showed that the brain could be subdivided, or was 257 FREEMAN, W. J. organized into a series of nuclear masses, which provided the ana- tomical substrate for 'centers.' A second" important influence was the continuing wide-spread interest in phrenology, which was kept before the public in the more acceptable form of localization of mental processes by the writings of Herbert Spencer. This implied that particular functions, which could be observed holistically in behavioral context, could be ascribed to particular parts of the brain. A less direct, more subtle, and yet equally important influence was the developing science ofthermocfynamics. This has been intro- duced earlier by Johannes Miiller in his law of specific nerve ener- gies. Throughout this time nerve discharges were described in physical terms such as nerve force or nerve energy. For example, Spencer (1863) stated (quoted from Darwin,1872, p. 109) as an "unquestionable truth that, at any moment, the existing quantity of liberated nerve-force, which in an inscrutable way produces in us the state we call feeling must expend itself in some direction — must generate an equivalent manifestation of force somewhere." This is a statement of conservation of momentum. Darwin (1872) remarked: "This involuntary transmission of nerve force may or may not be accompanied by consciousness. Why the irritation of nerve-cells should generate or liberate nerve force is not known; but that this is the case seems to be the conclusion arrived at by all the greatest physiologists such as Miiller, Virchow and Bernard, and so on." (Darwin, 1872, p. 70). As originally stated by Miiller, this hypothesis proposed that packets of vital fluid travelled down nerves at immeasurable speed. This was subsequently disproven through studies by Helmholz, Hermann, and others on the action currents of peripheral nerve and was finally discredited by the clear distinction that came to be drawn between vital spirits and electricity. In its place was intro- duced the notion of a unit or quantum of electro-chemical energy, the action potential, as the sole vector of information in the nervous system, with the sole attributes of number and location in time and space. These units of energy were thought to summate algebraically, in space andtime, and when concentrated in a local volume of tissue, 258 TEMPERATURE REGULATION "CENTERS" an anatomical center, to constitute a focus of activity. This focus was manifested either by an evoked potential or (much later) by trains of EEG waves. Hypothetical centers were conceived as main- taining levels of such energy of activity, and these levels of activity could be raised or lowered by synaptic activation. Eventually, this localized concentration of energy became identified with the central excitatory state of Sherrington, and subsequently with pools of unit potentials in the spinal cord or other spatially distinct volumes of brain tissue. Sherrington apparently had a great deal to do with the informal application of 19th century thermodynamics to analysis of the brain, perhaps less in his scientific publications than in philo- sophical works such as Man on His Nature (1953), in which he de- scribes his more tenuous analogies. He wrote, for example, "A nerve centre is a place of junction of nerve-lines, and of departure for fresh ones.... Signals convergent via many lines may in the centres coalesce and reinforce.... It is at such junctions that inhib- itions occur. It canthere suppress action, or, no less important, can grade it by mode rating it. In the net work of conductors, it can switch off one line as another is switched on." (p. 164). This was clearly analogous to a telephonic relay system, which is not too different from the idea of a thermostat, a mechanical contrivance, which by analogy was also conceived to exist in the brain. This conception of the brain as a mosaic of centers, each with its characteristic levels of energy, involved assumptions that were particularly appropriate for the prevailing technique of testing. The electrode inherently has a focal relationship to the brain. The application of an electric current through the electrode was con- ceived to raise the energy level of a given center in the brain much as the release of energy of a peripheral nerve was increased by electrical stimulation. It is characteristic of the looseness of think- ing involving "centers" that the logical possibility that electrical stimulation might increase a central inhibitory state was seldom considered. In the early part of this century the electrode began to be used also for making circumscribed lesions, which were thought to diminish the level of energy available or producible by a center. This was conceived as an algebraic effect rather than in terms of altering a pattern of function in the brain. More recently, the elec- trode has been used to record the level of electrical activity at various points in the brain, and the assumption has been made that 259 FREEMAN, W. J. these changes represent fluctuations in the energy level of centers in which the electrode has been found to lie. In each case, the elec- trode has been used as a focal tool and the interpretation has been made in focal terms. This fact of technique is crucial. When it was first introduced, the term center was simply a label applied to a given part of the brain which could produce a discrete response on stimulation, such as the eye-opening center, tongue-twitching center, orfinger- moving center, and so forth, without any firm commitment on the part of the user. But its subsequent application to all other parts of the brain was due more to the technology of the electrode than to any rational development of the system. Despite many difficulties with this doctrine, it has been the bulwark of studies in temperature regulation and there is no cliche more firmly entrenched in the medical liter- ature than the term "temperature regulation centers." I would now like to illustrate some of these assumptions and difficulties by describing three experiments. These have one thing in common. They failed. They gave negative results, or, perhaps better said, "indeterminate" results; and I would like to point out what the assumptions of these experiments were and how at least in two cases the assumptions were wrong. AN INCONCLUSIVE EXPERIMENT INVOLVING RECORDING The first experiment was to record the electrical activity of the hypothalamus in association with changes in body temperature (Freeman, 1957). This was a prelude to hypothalamic heating and cooling by means of which to evoke electrical changes possibly correlated with thermoreceptor activity. As we conceived the prob- lem initially, the hypothalamus was composed of a series of nuclei, and each nucleus was presumed to have its own characteristic form or pattern of electrical activity; one of these nuclei or collections of cells, even if it was intermingled with others, was sensitive to 260 TEMPERATURE REGULATION "CENTERS" temperature changes; and as temperature changed either the amp- litude, the frequency, or the pattern of electrical activity of that site would change with temperature. So we placed our electrodes through- out the anterior hypothalamus and subsequently all through the basal forebrain in a search for electrical changes. We induced changes in temperature ranging from 28 C to 45 C and found no localized pattern of electrical change that we could relate to these temperature changes. Pathological changes we re, of course, frequently and widely manifest. Essentially, the result was zero. Something was wrongwith the experiment. We might have attributed it to anesthesia or to surgical trauma or to some other cause, but as it turned out, this was not the case. The fault lay in the fact that we had assumed that the cells under study were producing electrical activity that we could record with this method, and this turned out to be not so. Figure 1 shows a series of recordings taken from electrodes placed in the hypothalamus. The sets represent a horizontal section through the cat brain showing the lower pole of the caudate nucleus, cerebral peduncle and optic tract aroundthe anterior hypothalamus. There were five electrodes spaced 1.5 mmtoS.O mm apart, dorso- ventrally halfway between the optic chiasm and anterior comm issure . Each of these electrodes was recorded from with respect to one of two indifferent electrodes on the scalp, the top tracing showing the activity occurring between the reference points. There was a re- markable similarity between the records taken from these different points in the hypothalamus. We had anticipated that there might be some overlap of wave forms in different records from this region. We did not anticipate, however, the fact that these different records could be superimposed after appropriate adjustment of gain in the different channels, the only difference between them being that there was a gradient of amplitude from anterior to posterior. Only as far back as the level of the mammillary bodies (Fig. 1, C, 5G) or later- ally near the hippocampus was there introduction of other elements into the records, and there they were very difficult to see. But an- teriorly where we were particularly concerned with temperature changes, the entire region had one pattern of electrical activity. The recordings in Figure 1 were monopolar. We also tried bipolar recording to see whether or not the universal electrical 261 & FREEMAN, W. J. 6-6" z4'^X''''~^''^:^^h ^\K'^y'Y- •-■ - " -'-..V /■"%/-■', v'^A-^'Vi./v',' I 200/iV [so/iv 6- 6 ■ ^v^'•-^--,•~^^^- /'^ 16 V^M^^^fV-y^l^AJ^^^ 3 6 \Ai/"^'AAv-V^wvN' '\ - ^ '.^ 1 1 vi ^j-^^'^"v^V,^~n,KV AV\wv 56 vWA/-.vrw v''^K, |^--^'^^V^--Na-''^~'^--/-' v^--'-''- -'-/^'■ Figure 1. Monopolar recordings from basal forebrain showing gradients of amplitude. The schematics represent horizontal sections through the cerebrum of the cat's brain; the location of each electrode is indicated by a numbered point. The notation "G-G"' indicates the recording between two reference electrodes. Electrode "G" was placed on the ipsilateral scalp and "G"' on the contralateral scalp. For abbreviations, see text. Time, 1 sec. 262 TEMPERATURE REGULATION "CENTERS" activity was masking local changes which might best be picked up by two electrodes as close together as 0.5 mm. The amplitude was clearly much less (Fig, 2) than that from either with respect to a distant electrode, but there was no other difference. This was also true for tipseparationsof 1.5mm,3.0mm, and 6.0 mm, the records from which were nearly identical except in aunplitude. There was no disclosure of any other signal than that recorded with monopolar electrodes, implying that the hypothalmus had a uniform electrical activity underthe conditions of this experiment, which did not involve evoked transients or abnormal temporal synchrony. This identity of pattern was true not only of the hypothalamus proper, but also of much of the surrounding region, i.e., inferior pole of the caudate nucleus, the olfactory striatum, parts of the septum and the cerebral peduncle. We attempted by diathermic heating locally in the hypothalamus to elicit local changes in the hypothalamic electrical pattern. There were four electrodes in this experiment with monopolar recording from each; the diathermy current was passed between the outer two, there being a thermocouple centered between the middle pair of electrodes for control of the hypothalamic temperature. The middle pair permitted study of the electrical activity without having to use theheatingelectrodes, around which the tissue was destroyed. Figure 3B shows that after heating to a temperature of 50 C, which is quite high for hypothalamic temperature, not only was the elec- trical activity from the intervening electrodes substantially unal- tered but so also was the electrical activity recorded from genera- ting electrodes. There was some difference in wave form, which was of the type which was seento occur spontaneously during the course of an experiment and did not represent a significant change. o In Figure 3C, the interveningthermocouple was heated to 70 C, almost, one mig^t say, the boiling point of the brain, and there was substantially no change in subsequent records. Histologically at post-mortem there was coagulative necrosis of the hypothalamus, parts of the septum, the caudate nucleus, and the thalamus. We did find a way of abolishing this activity, and that was to make transec- tions of the brain ventrolateral to the hypothalamus. Transections could be made anteriorly, laterally, dorsally, and posteriorly (Fig. 3F) with little change as long as the medial forebrain bundle was 263 FREEMAN, W. J. •o :; S S g a § ro ^ w n I ? ° § a. o «| 2 V 3 3 p § - O a ■o 5 a -^ 0. §3 ET O I I? g S S N. 1 « ►.. *= O 5" CO ? o a CL s <-► a !l g 3 « S SL y 2. (B ^ (8 pT •o 5. 1 "^ 0 !5 » CO » E o ^ 1 3 3 PD o' 2: 2. r m 3 S " ^1 3| o S- S a 264 TEMPERATURE REGULATION "CENTERS' ^ } \ sr-^i 2 S 2 o « m WHj) fi ? ^ • . 0) in S ►< t! ta - °- •3 ^ * M c E s : E > 3 (h •iS o "" « iJ ^ •§ * H .2- —I • J3 E - s JH a, 01 2 s 1 o a o a as 265 FREEMAN, W. J. left intact. As soon as that was cut, hypothalam ically recorded elec- trical activity (before or after coagulation) was greatly diminished (Fig. 3, D, H). This implied that the tacit assumption we made, that the electrical changes were generated locally by the tissue in which the electrodes were located, was not tenable. This raised the question, then, where was this electrical activity coming from? By systematic mapping of these potentials in this volume of brain tissue, we found that the primary olfactory (pre- pyriform) cortex generated an electrical field accounting for them, and that the electromotive forces of the field lay in the molecular layer of the prepyriform cortex (Fig. 4). This was a dipole field, meaning that the surface and base of the cortex were at all times 180 out of phase (Freeman, 1959). Such a dipole field existing in a volume conductor has no outside limits. In effect, this field spread throughout the brain and when its amplitude was high enough, or, correspondingly, when the amplitude of electrical activity of other structures was low enough, it could be detected in such distant points as the medulla, cerebellum, and temporal muscles. It was indeed a very powerful electrical field. The iso-potentials shown in Figure 4 continued into the hypothalamus and indicated the existence of current flow from the cortex into this nuclear region. The spread of current is a passive process resulting from the occurrence of a difference in potential in the cortex with spread of currents in all directions, much as there is spread of sound from a source of noise throughout the limits of the confining space. These alternating cur- rents spread outwards in all directions, producing that gradient of potential seen in Figures 1 to 3. Examination of the lesions found to abolish activity recorded in the hypothalamus showed that the effective sections had severely damaged the prepyriform cortex in all cases. 266 TEMPERATURE REGULATION "CENTERS' Si a ■" o 2 !3 o ^ 8 i; E 267 FREEMAN, W. J. AN INCONCLUSIVE EXPERIMENT INVOLVING STIMULATION Let us now look at a second experiment which also failed and inquire as to the reason. This experiment was derived from the one preceding, in which the prepyriform electrical field was localized. We became curious as to what its relationship to behavior might be, and we found that there was a variety of correlations with behavior involving sensory stimulation, learning, motivation, and so forth. The particular aspect of interest here is the fact that the amplitude of this electrical activity was correlated with the rate of work done by cats in pursuit of a goal object, in this case, canned milk. The electrical activity recorded from the structure is shown in Figure 5. It consistedof a series of bursts synchronous with respir- ation, which is of interest in view of Dr. Kawamura's findings. We determined the frequency and spatial distribution and the amplitude of this signal in order to find out what its parameter of change might be in relation to behavior, and found that the dominant parameter is amplitude, irrespective of the prevailing frequency and spatial distributions. Average amplitude was determined by rectifying and filtering the signal and measuring the surface area under this re- sulting curve (Fig. 5) graphically, or electronically as a matter of convenience and accuracy. As far as behavior was concerned, the animal waited in a starting box until the door was opened and then approached the m Jk and started lapping. It was this work period in which we were particularly interested. The work was measured by harnessing the animal after attaching the harness to a rope which pulled on a strain gauge. This rope then passed around a winch which rotated continuously. Whenever the animal was resting, there was no tension on the rope, and there was no rope paid out. As soon as the animal pulled, the rope tightened up on the winch; the winch pulled it through a friction drag and the animal was permitted to pull to reach the goal. It did not have to pull very hard in order to get there at a constant rate (the rate being determined by the winch), but the cats never seemed to learn; they always pulled harder than they needed to, and the hungrier they got, the harder they pulled. 268 TEMPERATURE REGULATION "CENTERS' si ii o 0) o — 4) * td 3 3 5) * t £ ^ ^ X> 7i 4) J3 73 2 rt OJ (U ■3 E •" o o s: >. Q. o -3. bo OJ ^ 2i -a :S -o ^ & a) " S ^ o ^ I * n - 5--a 2 » ^ « •o S 269 FREEMAN, W. J. This force of pull was integrated, and divided by the duration of pulling to get the average force. We knew velocity of approach; force times velocity gave the power or rate of work which the animal did to reach food. We then correlated this value with the amplitude of electrical activity generated concomitantly by this cortex. In general, we found a three-to-one ration such that for every one per cent increase in amplitude there was a three per cent in- crease in rate of work (Fig. 6). There was a dip in the upper half of the curve which was of considerable theoretical interest, but on the whole, it was apparent thatthere was a positive relationship be- tween amplitude and the rate of work done by the animal. It was known from previous work that the amplitude of electrical activity could be increased artificially by direct electrical stimu- lation of the cortex. Since the amplitude was related to work, one would presume that the increase brought about by stimulation would increase the rate of work done by the animal. We therefore increased the amplitude of electrical activity by direct electrical stimulation and measured the rate of work that the animal did during stimulus periods as compared to alternating non-stimulus runs. We found that there was no increase in rate of work. Again, something went wrong. We thought about it for a while and decided that our m istake was to equate the increase in amplitude of this electrical activity with an increase in output of the cortex. We went back through our records to find if there were any circum- stances under which we could get either positive or negative effects and, in fact, it turned out that this was the case. When we had first started out with our naive animals, there was a positive effect. Stimulation did increase the amount of work, but only for the first four days. Subsequently, this positive effect was averaged out to zero by a negative effect through most of the subsequent three weeks of measurement. In order to get a good measure of these changes, we started off with a fresh series of cats and found that, again, during the first four days of stimulation, as long as we were care- ful not to stimulate too hard or too often, stimulation would increase the rate of work done by these animals by some five to ten per cent, but that after four or five days, this effect dwindled to zero. We then found rather unexpectedly that we could increase the stimulus 270 TEMPERATURE REGULATION "CENTERS' O O > > < o UJ E ^- o O + kVWV^WX in O cz^Z lO O IT) + IDL Q ID + — 9pn|!|duJV .^ o LO I o I ID o o CD O OH LO CO I -+- CO CO =3 O 00 -D O LO LO 1^" c UD8IA1 P % D CD ^S. CO J= es ^ - O I E i ^ 2 E H S •3 ^ « 5 6 5 S.2 o. a £ 2 271 FREEMAN, W. J. intensity momentarily to a very hi^ level some ten times threshold and get what is known techanically as an "orienting reflex", in which the animal perked up its ears, sniffed the air, and moved its head and eyes as if in search. When this occurred, a rather re- markable change in the electrical activity did also, and thereafter, for at least two days, the positive effect returned. We again found a potentiation of rate of work by stimulation. Subsequently, however, this effect disappeared and in its place there came still a third phase, in which the rate of work was decreased by stimulation be- low the non-stimulus periods. Something happened that impaired the response; the same stimulus which still doubled the electrical activity and amplitude was associated with a decrease in rate of work. Still more complicated was the fact that the animal could be trained to perceive this stimulus, so that unless the stimulus was present, it would not pull at all. We could also train it to press a lever to get milk or perform any other type of directed activity. Furthermore, if the stimulus intensity was increased andthe effect tested on work, above a certain level the higher the intensity the greater the deficit of work done by the animal, until a point was reached at which the animal stopped completely. If this stimulus was maintained, there developed a behavioral and electrical seizure. So, although the intensity of electrical stimulation was linearly related to the amplitude of electrical activity of the cortex, the behavioral effects of electrical stimulation were much more complicated. The assumption derived from the doctrine of centers, that a correlation could be made between electrical amplitude and cortical function, was not valid, and the experiment that was based on this turned out to be inconclusive. AN INCONCLUSIVE EXPERIMENT INVOLVING ABLATION I have given examples of two experiments, one involving elec- trical recording and the other involving electrical stimulation, and would like now to give a third example involving lesion-making. I 272 TEMPERATURE REGULATION "CENTERS" have chosen these three examples becausethe classical approach to centers involves these three techniques. This problem involved an extension of Dr. Hemingway's work with Dr. Birzis on electrical recording of unit activity from the brain stem of the cat in association with shivering (Birzis and Hem- ingway, 1957). You recall seeing yesterday Dr. Kawamura's work on the correspondence between the electromyographic potentials of shivering and inspiration or the onsetof pain. The same correspon- dence was found between the electromyogram (Fig. 7, a, lower line) during a brief painful stimulus and, concomitantly, the interruption of a unit discharge associated with shivering, recorded in the brain stem with a rebound phenomenon occurring before the end of the painful stimulus. In Figure 7, b-d, you see phasic variations in shivering during its onset, when the relationship with respiration was most apparent, the temporal association of the unit spikes to the electromyographic change being quite clear. These unit discharges were related to shivering in their fre- quency. When shivering was present, they were present; when it was suppressed by whole body warming, these units stopped firing (Fig. 8). When shivering was re -introduced by cooling the animal again, shivering and the unit potentials returned, with the unit po- tential coming 5 to 15 seconds after the beginning of shivering. In general, this was a pulse modulated system, meaningthat the strong- er the shivering, the higher the frequency of the discharge. The location of these recording sites is shown in Figure 9 and reflects the distribution of a pathway in the nervous system which subserves shivering. Of particular interest was the origin of these unit potentials in the nucleus of the field of Forel. The efferent pathway of this nucleus goes down through the central tegmental fasciculus, turns laterally and, passing dorsal and lateral to the superior olive, goes to or throu^ the inferior medulla. These po- tentials in general had the same distribution. As I>r. Clark suspects, they were rather widely spread through the brain stem, but they showed a concentration in this pathway. These sites also correspond to the anterior and posterior placement of the lesions that Birzis and Hemingway (1956) found would prevent shivering in anesthetized animals. The suspicion arose that if lesions could be placed bilater- ally within the nucleus of Forel or the central tegmental fasciculus, 273 FREEMAN, W. J. bk HH-+ m 11 II. HI 1 1 I 0 seconds 5 1 ... I I Figure 7 . Concomitant changes are shown between electromyo- grams during shivering (upper trace) and brainstem unit discharge (lower trace); (a) arrest of shivering during nocioceptive stimulus (bar); (b-d) phasicvariationsinsynchrony with respiration; (e) onset of unit discharge after start of shivering; (f) onset concomitant with shivering. 274 TEMPERATURE REGULATION "CENTERS" TIME mm I mull* liMwHiii MiMiMi^^ 20 min iiiWiiiiiiiiiii mm ze mmmmmmmmmmmmm min 45 mmtimmmiiimiiMtmtmm iiiiiiiBliiiiiiiliiiiiiiiiliiiMiH iiilSiM 55 min g^|^g^,|gi,^,,g^g||jg|,yigy^lgg|laM^^ Figure 8. These tracings illustrate the procedures used to establish the association of a unit potential with shivering. The upper tracing of each pair is a continuous record from a microelectrode; the lower tracing is a concomitant electromyogram. At the right are tracings of the unit discharge as observed by triggering the sweep drive on an oscilloscope with the positive phase o^ the spike (cf. Fig. 3). Only the relatively long-lasting negative wave of the spike was reproducible in the records at the left. At 0 min a unit discharge was located and emptying of the cold water bath was be- gun. At 6 min the changeover to warm water was complete. At 20 min both shivering and unit were disappearing. Changing the bath to cold water brought an almost immediate reappearance of both the spike and of shivering; re-immersionin warm water again suppres- sed both. The change to a third cold bath did not produce an immedi- ate effect, but 3 min after complete cold immersion both shivering and the unit reappeared, and thereafter both rapidly increased in intensity. The close similarity of the 3 wave forms at the right in- dicates that the same unit was re-activated with each cold immer- sion. Time: left, 1.0 sec; right, 1.0 msec 275 FREEMAN, W. J. Figure 9. These schematic drawings were made from tracings of slides of coronal sections of the brainstem. The dots represent positive points, the circles questionably positive points. The photo- graphic insets show the configuration of representative spikes recorded at the various points. Abbreviations: ABD, n. adducens; AM, ansa lenticularis; BC, brachlum conjunctlvum; CF, central tegmental fasciculus; CG, central grey; CP, cerebral peduncle; CUN, n. cuneatus; DBC, decussation of brachium conjunctlvum; DEN, n. dentatus; DH, dorsal hypothalamus; DVN, dorsal vestibular n.; F, fornix; H,fieldofForel;IC, inferior colliculus; INF, inferior olivary n.; IP, n. intrapeduncularis; LH, lateral hypothalamus; LL, lateral lemniscus; LRN, lateral reticular nucleus; LVN, lateral vestibular nucleus; MB, mammillary body; MFB, medial forebrain bundle; MG, medial geniculate; ML, medial lemniscus; MLF, medial longitudinal fasciculus; MP, mammillary peduncle; NF, n. field of Forel; OM, n. oculomotor; PH, posterior hypothalamus; RE, resti- form body; RN, red n.; RS, rubrospinal tr.; SN, substantia nigra; ST, spinothalamic tr.; SU, subthalamic n.; SUP, superior olivary n,: SVN, superior vestibular n.; TEG, tegmental reticular for- mation; TF, thalamic fasciculus; TN, trigeminal n.; TO, tecto- olivary tr.; TR, trochlear n.; TS, tectospinal tr.; VSC, ventral spinocerebellar tr.; Zl, zonaincerta; V, VI, VII, VIII, X, XII, cranial nerves and nuclei. 276 TEMPERATURE REGULATION "CENTERS' 277 FREEMAN, W. J. shivering could be abolished with minimal involvement of other activities of the animal. This was a simple logical deduction from our data and from the doctrine of centers. So, then, a series of lesions was made (Fig. 10) in these sites in 34 animals, 29 of which survived for at least three days. Ability to shiver was measured by rectal temperature or by oxygen consumption rate in response to a standard cold stress three, ten, and twenty-one days post-opera- tive ly. In eight cats there was a transient impairment, two of the cats having developed intercurrent respiratory infections (Fig. 10, "deficient"). A similar series of lesions in animals which never showed a deficit are shown on the other side. The location and size of the two sets of lesions was as close as can be expected with the various technical difficulties of placing bilaterally symmetrical lesions. So, then, this must be regarded as essentially an inconclusive experiment. There was some transient impairment of the ability to shiver in response to cold in a small proportion of the animals, approximately one-quarter of them. In the remainder, there was no deficit and yet these lesions corresponded rather closely to the areas in which it had been predicted they would produce a deficit of shivering. The question must be asked, why did this experiment fail? I do not think that one can blame technical factors. I think, rather, that one must look to the basic h5T3othesis to the doctrine of centers. I think that we assumed a degree of stability and of localization of function in the brain which does not exist. In the first place, we have underestimated the number of func- tions or the degree of functional complexity of this area. This structure must do many other things besides regulate shivering, if, indeed, it does that. This is reflected in the fact that most of our animals during the early post-operative period showed impair- ment of other functions as well, such as walking, grooming, eating, etc. I think, secondly, we have overlooked the plasticity of the brain. The basic phenomenon apparent in this set of studies is that the brain recovered the function lost immediately after operation. It 278 TEMPERATURE REGULATION "CENTERS" repaired itself, if you will. The brain apparently has the ability to change its functional organization and so bring about a return of essentially normal output despite the loss of some parts. I have no notion of what this reparative or regenerative process might be. All I can point out is that the doctrine of centers does not take into account this plasticity of the brain in its formulations of the hypo- thetical function of various centers. THE ANALYSIS OF BEHAVIOR INTO ITS ELEMENTS These difficulties I have described so far are largely concerned with manipulative problems of stimulation, ablation, and recording. There are, in addition, some major observational difficulties con- cerned with how behavior is to be subdivided, and what behavioral elements are to be ascribed to specific sites. The assumptions are, in the doctrine of centers, that each center produces an organized, recognizable entity as its output, that the part of the brain respon- sible for each response is at a high energy level, and the others are quiescent. It is a doctrine of mutual exclusion. Consider as an example of these complex assumptions the phenomenon of reflex shivering, which we have already considered at some length. Specif- ically, is there any one part of the brain which is solely concerned with shivering and with no other reflex phenomenon? What parts of the brain are silent when shivering is going on, but when the animal is otherwise inactive? On the other hand, what parts of the nervous system concerned with shivering are also concerned with other types of behavior, whether thermoregulatory or not? Beginning with the periphery, it is obvious that cold receptors are not specific for shivering or, for that matter, even for thermo- regulatory responses. They can elicit a wide variety of autonomic or somatic responses depending upon previous circumstances, as with our experiment in stimulation of the prepyriform cortex. Nor can there be said to be shivering effectors, since the muscles are used also in a variety of other activities. 279 FREEMAN, W. J. ° 1 15 ^ » a s i s 2 >. « E i cs r ■£ o o D. « I .2 ^ .2 ^ o . g- c 0.00 o-x: 5j3-s q) il i ■§ I ^ ^ "lis - * 3 ° el '3 2 3 « SJ o bc ti S o 0) E £-5s»£«E4)c 8^1°ioE§| ^fc,— "Sic*"-? *2g«-§5|S*u gEEo°E|;a» ^Ct,t,mi:3'4,E5 4) 9J ™ r, ■_ ^ ,i, I" 2 m 5 S » 2 SJ a "2 € 2 2 S * s * S fc ? c 5 o o o 280 TEMPERATURE REGULATION "CENTERS" D 281 FREEMAN, W. J. At the level of the spinal cord, there is a highly complex set of structures concerned with shivering, including the terminal fibers of sensory nerves, the interneurons and their dendritic systems, recurrent collaterals, the afferent fibers from sensory organs in muscle, and so forth. All of these structures comprising the basic apparatus of the segmental spinal reflexes are unquestionably in- volved in shivering. They are equally unquestionably involved in all other types of directed muscular activity, and yet they are not sufficient for either. The particular contributionof these structures appears to be the adaptation of an overall pattern of body activity to local conditions encountered in the trunk and extremities. Specif- ically, the frequency of tremor of any given muscle or muscle group is determined at the spinal level. The question of whether or not shivering will occur, and if so, with what severity, is decided else- where. There is a great deal that ought to be known about higher struc- tures in relation to shivering, which we do not yet know. Perhaps Dr. Stuart will tell us something along that line. Proprioceptive activity is associated with the ascending discharges in the spino- cerebellar tracts, terminal fibers of which are broadly distributed over the cerebellar cortex. There are indications that cutaneous sensory activity may also reach the cerebellum. The spino-thalamic tracts subserving thermal sensation pass laterally and ventrally through the medulla over the lateral reticular nucleus, and they give off collaterals to it. Virtually the entire efferent discharge of this nucleus is distributed to the cerebellum. It is now known that electrical potentials related toavariety of afferent stimuli including visual, tactile, and auditory stimuli reach the cerebellum, and it seems likely, although the direct evidence is lacking, that thermal stimuli can induce cerebellar responses. In this case, the cerebellum can be regarded as a site for the mixing of cutaneous and proprio- ceptive discharges to produce a new pattern which is specific for these two types of input. Dr. Kawamura has shown in this symposium that an increase in proprioceptive activity associated with increased muscle tonus is a prelude to shivering. Obviously, cutaneous cooling is an essential element. When proprioceptive and cold receptor activity are mixed, as they may well be in the cerebellum, the groundwork is laid for an appropriate response peculiar to this unique combination of sensory input. Clearly the cerebellum is not specific for shivering, but it plays a significant role, since shivering 282 TEMPERATURE REGULATION "CENTERS" becomes irregular and spasmodic when the cerebellum is removed. I might also mention the contribution of the reticular formation to shivering, since this is another region which is involved in shivering, but not specific for it; however, time does not permit to go into it in any detail. The anterior hypothalamus is another sen- sory mechanism, which Dr. Lim and Dr. Hensel have shown to be clearly implicated in the shivering process. It is probably not essential for it to occur, and certainly there are other things that the anterior hypothalamus regulates besides shivering. One might now ask, what aboutthenucleusof Forel?ls it involved in shivering, as suggested by unit activity, or is it not involved, as the results of the lesions would imply? During the process of shivering there is muscular vasodilation, perhaps active. There is an increase in cardiac output, an increase in rate and depth of respiration, all subserving an increase in oxygen consumption and transport. Does this discharge that we see, which is related to shivering, subserve merely the skeletal muscular contraction, or does it subserve the entire autonomic- somatic pattern associated with shivering? The nucleus of Forel receives a powerful descending projection from the corpora striata, so that a large proportion of the entire efferent outflow of the basal ganglia passes through this nucleus. It would be rash to assume that other patterns of activity related to other forms of behavior are not partially organized in the nucleus of Forel. Is thereapatternof cellular activity specific for shivering and another pattern specific for each other form of behavior, or is there a common pattern to all these activities? If this structure is stimulated electrically, will it reproduce the pattern for shivering, or will it give something else, some entity which is not related to any observable phenomenon in the nervous system? If stimulation is applied in a situation in which some other rhythmic process is occurring, such as chewing, rutting during sexual behavior, or panting, will shivering occur or some other rhythmic contractile process? These considerations are summarized in diagrammatic form in Figure 11, showing a sketch of a longitudinal section of the cat brain. It is proposed that each of the structures indicated is active during shivering, and that each contributes some analytical element 283 FREEMAN, W. J. 284 TEMPERATURE REGULATION "CENTERS" to the completed process, e.g. frequency, rhythm icity, and intensity. But it is also proposed that these same structures are active during all other kinds of behavior produced by the intact nervous system, and that the essential nature of the contribution of each would, in some sense, be the same as that during shivering. The precise pattern of output would be determinedsolely by the previous pattern of input. CONCLUSION It can be said finally that even where the available evidence indicates the existence of a center in the classical sense, such an achievement can no longer be regarded as a satisfactory end point. For, if it be demonstrated that stimulation of a given structure will produce shivering, or, for that matter, feeding, non-feeding, rage, sleep, sexual behavior, etc., it must be recognized that the stimulus evoking this is not normally present in the brain. Where then does it normally come from? What is the nature of the cellular activity that the electrical stimulus has produced; and how does this compare with the normal pattern of electrical or cellular activity? How is this normal pattern brought into being by afferent fibers? Is a transmitter substance involved? Are there chemical gradients of ions moving in response to endogenous electrical fields? What is the nature of the efferent activity? These are all questions which are now within the realm of analysis and I think should con- stitute the bulk of work in neurophysiology in future decades. The conclusion of this review is that the various parts of the brain may not be so organized as to subserve particular functions throu^ the increased activity of particular parts, the other parts then remaining quiescent. It is suggested, rather, that during most reflex activities, virtually all parts ofthe brain are active, that each part of the brain contributes some basic element to all types of reflex activity, and that distinctive spatial-temporal patterns of behavior are determined largely by the pattern of sensory stimula- tion, which is maintained before and duringthe activity. In this view 285 FREEMAN, W. J. further studies of temperature regulation would require the iden- tification of the functional parts of the brain, of the contribution of each part to behavior as a whole, and finally, of the neuronal mech- anism of that contribution. This is the direction to which we must now turn in orderto avoid perseverative failures in our experiments. 286 TEMPERATURE REGULATION "CENTERS" DISCUSSION DR. STUART: Is it true that the posterior hypothalamus is out- side the dipole field generated by the prepyriform cortex? DR. FREEMAN: Excuse me. It does extend into that area. DR. STUART: I thought you showed its most posterior aspect as anterior to the posterior hypothalamus. DR. FREEMAN: That is somewhat different. It is mixed there with another field. DR. STUART: There is no perceptible disturbance of tempera- ture regulation in cats with the prepyriform cortex ablated. Could it be that the activity you recorded in the hypothalamus with macro- electrodes was masking the unitary activity of small hypothalamic cells whose activity is related to temperature regulation? DR. FREEMAN: If such activity exists, yes, but one would have to prove that it exists. This is precisely what we were looking for by means of various types of recording. I did not find it. I do not know of anybody else who has. I know several people who looked but were unable to find anything definite. This applies not only to EEG waves, but also to unit activity, which we also inade a serious attempt to find, and could not. We could find action potentials, but they came from other structures such as the optic tracts. In other words, they stopped and started with the opening and closing of the eyes, and, therefore, were irrelevant to this area. As far as I could find to date, this structure is electrically silent. I would hope that Dr. Hensel would be able to modify his techniques for recording C fibers to record from small cells in that area, since there are cells there and they can be presumed to generate electrical activity; but before you can say it is being masked, you have to know whether it is there. I have not been able to see it; as far as I am concerned, we cannot yet say that it exists. 287 FREEMAN, W. J. DR. RODDIE; Have any intracellular recordings been made around this area? DR. FREEMAN: These cells are somewhat smaller than red blood cells; the average intracellular electrode is, let us say, one- tenth of one micron in diameter, and its irregular movement due to vibration in this circumstance is too great. The average cell which can be successfully recorded from is some hundred microns is diameter. DR. RODDIE: Even though it is difficult, until this type of re- cording is made, I do not think any conclusions can be drawn; the failure of external electrodes to show any independent electrical activity really does not mean very much. DR. FREEMAN: No, all it means is that the electrical activity cannot be recorded with our presently available technique; and what I am saying is that the electrical activity that can be recorded there is not generated by those cells. It comes from somewhere else, and this is the point that is important in applying doctrinaire thinking to these problems. I started off with the assumption that what I recorded there was generated by cells foundthere, and this was not so. This is what I would emphasize above all, that which is recorded in a given structure may not come from that structure. In the case of a reticular structure like this, it probably does not. I fully agree that there must be cellular changes there. I think they will be more likely found not with electrical recording, but with other types of recording such as optical density, impedance, or paramagnetic resonance. There are a number of such techniques which are just coming into possibility now, which I think will give far better ap- proachability to this structure. There are obviously changes oc- curring here which are related to temperature regulation. I do not think electrical recording is the way to analyze them. DR. STUART: It seems interesting that the^unit potentials re- corded from the nucleus field of Forel begin fifteen seconds or so after shivering has begun. One wonders whether the units are driving the shivering or whether the shivering isdrivingthe units, particu- larly since once shivering begins, and a characteristic rhythm is set up in a limb, there is a proprioceptive input to the cerebellum, 288 TEMPERATURE REGULATION "CENTERS" and there are anatomical connections from the cerebellum to the nucleus of the field of Forel. I cannot see how you can conclusively say that the units you recorded are related to the efferent aspect of shivering. DR. FREEMAN: One can relate them to the shivering process, since they occur in association with it. What you are thinking of is the initiation of shivering. I do not think that they do this, since they usually begin after shivering has started. They are associated with the build-up of shivering, if you will, or the maintenance of the shivering at a level of adequate intensity. This tremor may or may not be shivering. A tremor in response to cold can occur in a de- cerebrated animal. I would think that the whole mechanism of shiv- ering exists downstream and what this part does is to facilitate this mechanism in response to the combined central- peripheral limbs of the cold receptor system. This means that this discharge is not necessaiy for the initiation of it, but it may be essential for the maintenance of an adequate degree of intensity. DR. STUART: It would be interesting to ablate the cerebellum and allow any rostral connection to the cerebellum with this region to degenerate and then see if you can still record these units. DR. FREEMAN: Well, a rather small proportion of the discharge comes up there in any case. Also, as we have seen, lesions of that area do not abolish shivering. I suspect that the re may be other parts of the brain which can do just as well, which also receive afferents from the cerebellum. DR. STUART: The units that were shown in this work were re- corded when animals were shiveringunder pentobarbital anesthesia, whereas after the lesions were made, the animals were tested in a cold stress in the unanesthetized state. Since this work has been done, I have made similar lesions in six animals in this subthalamic region — animals that, pre-lesion, could shiver under pentobarbital. After lesions had been made, the unanesthetized animals could still shiver in response to a cold stress, but they did not shiver under pentobarbital anesthesia. That makes me wonder whether or not there is a problem related to the effect of the drug on the response. 289 FREEMAN, W. J. DR. FREEMAN: There is undoubtedly a problem related to the effect of the drug. I often suspected that pentobarbital had abnormal potentiating effects on shivering. I have induced shivering in an ani- mal, for example, with a relatively high rectal temperature, and I have seen shivering continue even when the brain temperature has passed 42 C or 43 C.Itisan abnormal condition, but if one wishes to study these unit potentials, it is better to have an abnormal condi- tion in which shivering can be obtained than one in which it cannot. DR. STUART: Still, after the lesions in this region, if the ani- mal is given pentobarbital, it does not shiver at an abnormally hi^ temperature. It is very difficult to get it to shiver even if the animal is cool. DR. FREEMAN: That is a valuable finding. I agree that the use of anesthesia introduces difficulties, although I would not say that it is to be deplored. DR. STUART: In the work that you did with unit recording, did you find any units that began before shivering began? DR. FREEMAN: Yes, we discarded those as sensory. DR. STUART: What regions were they? DR. FREEMAN: They were further laterally in the vicinity of the medial lemniscus and spino-thalamic tract. I did not stucfy those to any greatlengthbecause of limitations of time, but certainly the way in which to approach unit activity of this kind is to study every unit one can find that is correlated with anything. Other types of units we found were those which were initiated or increased in their discharge with this painful stimulus as opposed to the decrease we saw with shivering units. Those also would be of considerable interest. DR. STUART: Often in neurophysiology we think in terms of recording channels rather than brain activity. When using an oscil- loscope we are two-channel thinkers, i.e., the unit and the EMG. When using an EEG machine we are eight-channel thinkers. This imposes a "centristic" approach to the nervous system. 290 TEMPERATURE REGULATION "CENTERS" DR. FREEMAN: I guess if we could afford it, we would be a twenty-five-channel thinker. DR. CLARK: I have one comment. You brought up the question of plasticity. Now, plasticity to me is a good deal like the spread of current. One may use spread of current to explain why the other man got results different from yours. Plasticity is to me just a recognition of lack of knowledge. For example, if you take out the motor cortex in a cat, and wait about ei^t months, he behaves quite normally. You do the same thing with a chimpanzee and eight months later he will still have the same paralysis that he began with. If you do it with a spider- monkey, the paralysis remains. Now, if you explain the return of functioning in the cat by invoking the term "plasticity", why does it not occur in these other two species? DR. FREEMAN: I take it that this is a rhetorical question? DR. CLARK: It is. And the same is true of the posterior columns. You damage those and the damage is apparently permanent, but as all of you clinicians know, the operation for relief of pain is not uniformly successful. If you invoke plasticity to explain the return of pain, why does function not return in the posterior columns? I think as far as return of shivering in those animals is concerned, there is a question of the probability of hitting a cell or a fiber. Where there is a fair collection of theiii this is easy, but when they are scattered throughout a large area, the probability of hitting enou^ of them becomes much less. DR. FREEMAN: I quite agree. It is a term to describe essen- tially the recovery process, but it does no more than show that we do not know what it is. DR. CLARK: It is worse than that, because when you invoke the term plasticity, you assume then that there is no localization, no permanent localization. DR. FREEMAN: I agree it is a loose term that one could take to mean that, but I would not go that far. 291 FREEMAN, W. J. DR. HENSEL: I think that even the results of the acute experi- ments have a certain significance. Plasticity is no argument against this, because you can get the acute disturbance in one area while in another one, you do not, and this means something, even if there is afterwards a plasticity in a certain recovery. DR. FREEMAN: It may explain the difference then between the acute experiments and the results of the chronic experiments. DR. HEMINGWAY: I would just like to point out the need for further work on the unit potentials which occur in the midbrain and are associated with shivering. We still do not know just what part they play in the shivering process. They are there; they come on with shivering and they disappear when shivering stops. They are not quite in phase. I would like to ask if you removed the cerebellum and then looked for them . DR. FREEMAN: No, I did not do that. DR. HEMINGWAY: Dr. Stuart suggested that these midbrain unit potentials mi^t have come in from the cerebellum which is a possibility, but experiments should be conducted in which the cere- bellum is removed. DR. FREEMAN: They were seldom encountered in that part of the brain; that is to say, they were not in the brachium conjunctivum. DR. HEMINGWAY: And remember, if you make a hem isect ion below where you are recording, they are still there, so they are coming down from above but they migjit go a round-about way, by way of the cerebellum. DR. FREEMAN: I think that is a safe conclusion. DR. HENSEL: I have another question. I was quite interested in your finding of the relationship of arousal of behavioral activity to the potentials of the hypothalamus. We recorded recently the local blood flow in the hypothalamus by means of very thin implanted heated thermocouples in an unanesthetized cat, and we found that there is also very close correlation between the activity or arousal 292 TEMPERATURE REGULATION "CENTERS" of the cat and rate of blood flow. If you show a mouse to the cat, you get an enormous increase in the hypothalamic blood flow by about 100 per cent. When the mouse is gone, then the blood flow drops. And another observation is, if the cat is not able to catch the mouse after some time, say if it were in a glass container, hypothalamic blood flow would drop, which is a sign of resignation. 293 FREEMAN. W. J. REFERENCES 1. Birzis, L. and A. Hemingway. Descending brain stem con- nections controlling shivering in the cat. J. Neurophysiol. 19:37-43, 1956. 2. Birzis, L. and A. Hemingway. Efferent brain discharge during shivering. J. Neurophysiol. 20:156-166, 1957. 3. Brazier, M. A. B. The historical development of neurophysiol- ogy. Chapter I, in the Handbook of Physiology. Section I: Neurophysiology, Vol. I. Amer. Physiol. Soc, Washington, D. C, 1959. Ed. J. Field. 4. Darwin, C.The Expression of the Emotions in Man and Animals. London, John Murray, 1872. ch. xiv, pp. 374. 5. Freeman, W. J. Oscillating corticonuclear dipole in the basal forebrain of the cat. Science. 126:1343-1344. 1957. 6. Freeman, W. J. Distribution in time and space of prepyriform electrical activity. J. Neurophysiol. 22:644-665, 1959. 7. Sherrington, C. S. Man on His Nature. Doubleday and Co., Garden City, New York, 1953. 2nd ed., ch. xii, p. 316. 294 ROLE OF THE PROSENCEPHALON IN SHIVERING Douglas G. Stuart Department of Physiology University of California Medical Center Los Angeles, California The experiments reported in this paper were designed to localize control of the production of shivering to a specific region of the hypothalamus and then to study some of the telencephalic influences that by hypothalamic relay could instigate, facilitate, or suppress shivering. First it was necessary to recognize the results of workers who have claimed that decerebrate preparations can shiver (Barbour, 1939; Dworkin, 1930), even though such results conflict with those of the majority of investigators (Bard, 1961; Bardand Macht, 1958; Bazett and Penfield, 1922; Blair and Keller, 1946; Dusser de Barenne, 1920). If decerebrate preparations can shiver, it means that the primary control of shivering is in the midbrain or pons, since all the above investigators agree that shivering is abolished by transection of the caudal pons. However, over 60 years of evidence (Ott, 1895, to Keller, 1959) has implicated the hypothalamus as the primary "center" for the production of shivering. Since it was early shown that massive anterior hypothalamic destruction did not affect shivering (Bazett et al., 1922; Clark et al., 1939; Keller and Hare, 1932), it followed that the essential neurons were in the posterior hypothalamus. A more precise anatomical description invites controversy. Some investigators have studied this problem by noting the presence or absence of shivering after destruction of part of the posterior hypothalamus, but their results conflict in localizing the essential neurons to the dorsal posterior hypothalamus (Bazett et al., 1933; Blair and Keller, 1946), the ventro-medial mid hypothalamus (Frazier et al., 1936), the lateral posterior hypothalamus (Clark et al., 1939) and to the ventro-lateral posterior hypothalamus (Birzis and Hemingway, 1956). Other investigators have studied this problem by stimulating the intact brain to instigate shivering. In three laboratories 295 STUART, D. G. (Chatonnet, 1961; Hammel et al., 1960;Kundtetal., 1957), shivering has been evoked by thermal stimulation, but this technique does not permit a precise localization of the effective region. Others have demonstrated the production of shivering during application of an electrical stimulus to the forebrain's septum in anesthetized cats (Akert and Kesselring, 1951) and unanesthetized goats (Andersson, 1957) and to the midbrain and pons in anesthetized cats (Birzis and Hemingway, 1957). In the studies using cats, production of shivering by stimulation of the hypothalamus was mentioned. In each case only one hypothalamic locus was stimulated. In all three reports the investigators' efforts were devoted mainly to regions more rostral or more caudal than the hypothalamus. Since decorticate preparations with the septum destroyed can shiver (Aring, 1935; Bard, 1961; Dusser de Barenne, 1920; Pinkston et al., 1934), it would seem that the most rostral prosencephalic region whose stimulation consistently produces shivering is one whose ablation does not affect shivering. Anatomically the septal region of the forebrain, which is but a vestige in man (septum pellucidum), has been shown to have intimate connections with neocortical and rhinencephalic pathways to and from the thalamus, hypothalamus and midbrain (Fox, 1920; Pribram and Kruger, 1954). Thus, it mig^t well be, but is not proved, that the septum is involved in alterations in temperature regulation evoked by classical Pavlovian conditioning (Bykow, 1957), the production of shivering by hypnotic suggestion (Gessler and Hansen, 1927) and "psychological aspects" to cold tolerance mentioned in recent reports of human adaptation to cold stress (Adams and Covina, 1958; Scholander et al., 1958; and Wyndham and Morrison, 1958). However, if the septum is involved in the production of shiv- ering, its activity should be but secondary to hypothalamic activity in that it can be ablated without affecting shivering, and all known caudally projecting septal effe rents synapse in the hypothalamus to make direct connections with the midbrain (Nauta, 1958; Sprague, 1950). Neither these fibers nor those septal projections to the thalamus could produce shivering, since the thalamus and also the hippocampus, from which the fornical fibers arise, can be ablated without affecting the production of shivering. Anterior hypothalamic stimulation either thermally (Eliasson and Strom, 1950; Freeman and Davis, 19 59; Hemingway et al., 1940; 296 ROLE OF THE PROSENCEPHALON IN SHIVERING Magoun et al., 1938) or electrically (Anderssonetal., 1957; Anders- son and Persson, 1957; Andersson et al., 1960; Hemingway et al., 1954) is known to suppress shivering. There are septal projections to both anterior and posterior hypothalamic regions. Therefore, if septal stimulation can evoke shivering, it should also be capable of suppressing it. Such suppression has been reported during septal stimulation (Hemingway et al., 1954) and during stimulation of more rostral telencephalic structures, the orbito-frontal gyrus (Kaada, 1951) and the amygdala (McLean and Delgado, 1953), It is not known if the neurons activated to suppress shivering during stimulation of the orbito-frontal gyrus and the amygdala traverse the septum or relay within it. Some investigators have shown that on exposure to cold shiv- ering appears sooner and is more intense after decortication (Aring 1935; Bard, 1961; and Pinkston et al., 1934) and on this basis have claimed that the cerebral cortex is inhibitory with respect to shivering. However, evidence that the telencephalon can activate as well as suppress shivering is just as valid and suggests that the effects of decortication on shivering should be re-examined. From the above mentioned literature the following questions emerge: 1. Can decerebrate preparations shiver? 2. Is the shivering intensity of decorticate preparations equal to, greater than, or less than that of unoperated animals? 3. Is the septum or the posterior hypothalamus primarily involved in the production of shivering? 4. Can septal stimulation suppress shivering and if so, how is this related to septal activation of shivering and to anterior hypothalamic suppression of shivering? On the basis of both anatomical and physiological studies, it appeared that septal production and suppression of shivering should be secondary to hypothalamic production and suppression of shiv- ering. Secondly, if septal and midbrain electrical stimulation could produce shivering, so should electrical stimulation of the hypo- thalamus. It was felt that the use of electrical stimulation would 297 STUART, D. G. permit a localization of that region of the hypothalamus responsible for the production of shivering. Thirdly, it seemed that the net effect of decortication on shivering could be one of depression, facilitation, or nil, depending on the relative tonic preponderance of facilitating versus suppressive telencephalic influences on the critical hypothalamic region in the intact brain. In defense of these hypotheses the following experiments have been undertaken in Dr. Allan Hemingway's laboratory in the past two years: 1. Re -ex ami nation of the decerebrate cat's response to a cold stress. 2. Re -examination of the decorticate cat's response to a cold stress. 3. Localization of the hypothalamic region which when activated produces, and when destroyed abolishes, shivering. 4. Comparison of somatic effects evoked during stimulation of such a region and the septum. 5. Comparison of relative suppression evoked by septal, an- terior and posterior hypothalamic stimulation during shivering. 6. Effects of septal lesions on shivering. 298 ROLE OF THE PROSENCEPHALON IN SHIVERING EXPERIMENTS EFFECTS OF DECEREBRATION ON SHIVERING Methods Aseptic decerebration surgery was performed on seven cats under pentobarbital sodium anesthesia (35mg/kg I. P.). Care was taken to preserve the blood vessels on the ventral surface of the brain. Figure 1 shows the gross aspects of such surgery and illus- trates the fact that the only connection between the prosencephalon and the lower brain is by virtue of the meninges. By preserving the ventral diencephalon, the hypothalamic-hypophyseal system can function relatively effectively, as evidenced by the preparations' lack of poly urea after surgery. After surgery the preparations were poikilothermic, and body temperatures were maintained artificially at 36 C to 38 C by appropriate alterations of environmental temperature. Forthe first three days after surgery, the preparations were given intramuscular injections of a mixture of streptomycin, dihydrostreptomycing and sodium and procain penicillin (0.5 cc I. M., Dicrysticin Fortis, E. P. Squibb and Sons, New York, N. Y.). They were daily fed by stomach tube, 150 cc of a homogenized mixture of meat and milk, were kept dry and clean and by frequent alteration of their relatively immobile postures, their "bed sores'" were kept in check. One to nine days post surgery, the animals 'oxygen consumption rates (VO 's) were determined over 20-minute periods with rectal 299 STUART, D. G. DORSAL VIEW PARASAGITTAL VIEW Cms 1 2 3 4 5 6 7 , 8 1 1 1 1 Figure 1. Gross views ot decerebrate cat brain No. 7. 300 ROLE OF THE PROSENCEPHALON IN SHIVERING o temperatures maintained near 38 C. The body temperatures were then lowered 0.5 C to 9 C by immersing the cats in cool water for 10 minutes. Ten minutes after completion of this cooling stress the oxygen consumption rates (VO ) were redetermined over a 20- minute period. During the cooling stress and second determination of VO , the somatic responses of each animal were checked by in- dependent observation. Somatic responses included all forms of skeletal muscular activity and patterns of movement that were apparent on visual inspection. By independent observation it is meant that more than one investigator visually inspected each preparation and after such inspection the somatic responses were discussed. Comparisons were made of the somatic responses during such rapid cooling and during slower cooling (exposure to 2 5 C environ- mental temperature), rapid rewarming (immersion in 50 C water) and slower rewarming (immersion in 40 C water). After these tests an attempt was made to keep the animals alive for further tests at one month post surgery, but this met with failure. Of the 22 cats decerebrate surgery was performed on, 20 lived one day or more, 13 two days or more, 9 three days or more and five 5 days or more post surgery. After these tests the animals were sacrificed and their brains fixed in formalin for gross and sometimes histological inspection. Results Somatic observations. Table I summarizes the metabolic and somatic responses of seven decerebrate cats subjected to a rapid cooling test one to nine days after surgery. All the animals made somatic responses to the rapid cooling stress of immersion in cold water. Movements consisted of intermittent spasmodic twitches, jerks and large amplitude kicks and running movements of the somatic musculature. These somatic responses were accompanied by an increase in respiratory rate and depth. The violence and fre- quency of these intermittent movements were greater in the animals 301 STUART, D. G. >" S?-^ o ^^B O u z ™ t; 6 s £ ^ 302 a. S § S ? Q ° 2 ?s 0. U •" -S '^ H m s a 2 c-j£ ^ 2 |°|^^ s S s S I o"| C T3 " — J= ° a> g £ g ■" " ^ I s 1 1 i g "-g -a Is bo I; III o ROLE OF THE PROSENCEPHALON IN SHIVERING tested five and nine days after surgery than in those tested one and three days after surgery. Cat No. 7 gave evidence of a slow tremor interspersed between grosser arrhythmic twitches, and Cat No. 8 had a faster tremor that, by palpation, resembled shiv- ering. If the body temperature of the preparations were slowly lowered simply by decreasing the environmental temperature to 25 C, no somatic responses to cooling were evident even when the body temperatures fell to the same levels as those following im- mersion in cold water. When the animals were rewarmed by im- mersion in waim water (40 C), the movements disappeared when the body temperatures reached 36 C to 38 C. If the rewarming was rapid (immersion in 50 C water), the animal's somatic responses were similar to those seen during rapid cooling. This suggests that the response the decerebrate animal makes to rapid cooling and rewarming is a non-specific avoidance response to a nocioceptive stimulus. This concept is supported by the fact that similar somatic responses of briefer duration were evoked by other "unpleasant" stimuli such as rectal thermometer insertion, pinching the hindlimb or fast rotation of the animals. Metabolic and rectal temperature determinations. As shown in Table 1, the oxygen consumption rate before cooling for Cats 5, 6, 7, 8, and 10 lies within the normal range for unoperated cats of weight range 2 to 4 kg. Cats 17 and 22 had lower oxygen consumption rates in keeping with their enfeebled condition. There was little variation between the animals in somatic responses during cooling. However, the variation in somatic movements during the second VO determinations was quite marked. Cats 5 and 6 making few somatic responses, Cat 7 making intermittent gross kicks, Cat 8 maintaining the tremor evoked by the cooling stress, and Cats 10, 17, and 22 becom ing quite hypotonic after cooling stresses, although making violent movements during such stresses. The variations in oxygen consumption rate of the animals after cold exposure paral- leled the variations in their somatic responses. During the oxygen consumption rate measurements after cold exposure, the cat had cold wet skin and the rectal temperatures of Gats 5, 6, 7, and 8 continued to fall during the 20-minute period of the determinations. This was not the case with Cats 10, 17 and 22. The rectal tempera- ture of Cat 10 rose 0.6 G during the post-cold VO determination, but the rectal temperature of this cat was only 27 C at the beginning 303 STUART. D. G. of this determination. Cats 17 and 22 had rectal temperatures of 33.2 C and 32 C respectively at the beginning of this post- cold stress determination, similar to the beginning rectal temperature of 32.5 C of Cat 7. However, Cat 7 had a decline in rectal temper- o • ature of 3.0 Cduringthe 2 0-minute post-cold stress VO. determin- • 2 ation biit an elevation of VO of +32% over the pre-cooling stress level, in keeping with its intermittent somatic activity during this period. In contrast. Cats 17 and 22 were quite hypotonic after cold exposure, and their oxygen consumption rate determinations were less than 50 percent of that of Cat 7; yet the rectal temperature of o Cat 17 fell only 0.8 C and that of Cat 22 remained unaltered during this period. Comments Shivering was defined above as cold-induced rhythmic muscular activity resulting in a two to four-fold increase in oxygen consump- tion rate and limb tremor frequencies of 9 to 11 cycles/sec On the basis of this definition none of these decerebrate animals shivered in response to cooling. Even Cat 8, who displayed a fast tremor that appeared on cooling, disappeared on warming and by palpation and independent observation resembled shivering, could not be said capable of shivering in that there was no appreciable oxygen con- sumption elevation during the period of tremor. Additionally, Cat 8 and all the other cats displayed similar somatic responses during rapid cooling and rapid warming and no responses during slow cooling and slow warming. Thus, these limited observations suggest that the decerebrate cat is capable of making similar somatic responses to rapid cooling, warming, and other nocioceptive stimuli and also has the neurological '"substrate" that if activated produces tremulous motions that are not related to temperature regulation. This latter concept is based on the fact that tremulous activity was seen in two of the seven cats (No.'s 7 and 8). This may be explained by the resultsof Jenkner andWard(1953), Folkert and Spiegel (1953) and Wycis, Szekely and Spiegel (1957), who produced tremulous activity in anesthetized monkeys and cats by electrical stimulation of the reticular formation over a wide region of the midbrain, pons 304 ROLE OF THE PROSENCEPHALON IN SHIVERING and bulb. Although the production of tremulous activity probably involves a reticular formation activation from a more rostral structure than the midbrain (Carpenter, 1958), it is conceivable that in the decerebrate animal, there could be a nocioceptive stim- ulation of reticular neurons capable of producing tremor. This suggestion awaits experimental confirmation. The results listed in Table I confirm the work of Dusser de Barenne (1920), Bazett and Penfield (1922), Bazett, Alpers, and Erb (1933), Keller (1959), and Bard and Macht (1958). Bard and Macht kept animals alive for many months after decerebration and reported that several weeks after surgery there was a return of cold-induced fine rapid tremors in decerebrate cats that resembled shivering. However, in a later report, Bard (1961) stated that al- thou^ such cold- induced muscular activity increased in frequency and vigor during extreme falls in environmental and body tempera- ture, it did not affect the rate of body temperature decline. It would seem that Bard and Macht 's results support the above-mentioned concept of shivering being replaced after decerebration by somatic activity alternatively and intermittently rhythmic and arrhythmic and deficient in heat productive capacity. Unfortunately, all the above work refutes, yet cannot explain, Dworkin's (1930) report of shivering returning in chilled rabbits from 5 to 230 minutes after brainstem transection as low as the caudal pons. Moreover, the body temperatures of these preparations rose while "shivering." Barbour (1939) reported the return of shiv- ering in two cats one and two days after decerebration, but his account seems too fragmentary to support Dworkin and refute all previous and subsequent work. Although the majority of past and present evidence supports the concept of shivering production in- volving a more rostral neural structure than the midbrain, there is no satisfactory explanation in the literature of Dworkin's diverse results. However, on the basis of the metabolic data presented and the observations of Bard and Macht on animals more carefully nursed for longer periods oftime after surgery, the evidence seems to favour a lack of true shivering in decerebrate preparations. 305 STUART, D. G. EFFECTS OF DECORTICATION ON SHIVERING Methods . Removal ofthe cerebral cortex by aseptic surgery was performed on 21 cats all under pentobarbital sodium anesthesia (35 mg/kg I. P.). On the first three post-operative days these animals were given injections of Diciystin Fortis and were maintained in a 25 C environment. Figure 2 is a film positive of buffered thionine sections at various frontal levels of a decorticate cat brain. It demonstrates that in preserving the medial aspects of the anterior heads of the caudate nuclei, the septum is preserved, even thou^ the corpus callosum is removed (Fig. 2A). Additionally, althougji the fomices were removed, direct paleocortical- diencephalic connections still existed by virtue of the integrity of the ri^t amygdaloid nucleus (Fig. 2B). The thalamus, hypothalamus andmidbrain are undamaged by this surgery (Fig. 2C and 2D). Immediately prior to and three days after surgery, the resting and shivering oxygen consumption rates were determined. One cat was tested at3,28and470days after surgery. A mild cooling stress induced active and continuous shivering at a rectal temperature of 36 C to 37 C. This stress consisted of immersing each animal in o o 10 C water for two minutes and 40 C water for one minute. This was repeated once and finally the animal was immersed in 10 C water for one minute. The rectal temperature of each cat was mon- itored during the determination of VO . At the conclusion of these tests an attempt was made to keep the animals alive for 28 days, but, with the exception of one animal, this met with failure. Of the 21 cats on which the decortication was performed, 10 livedthree days or longer, 5 lived one week or longer, 2 lived two weeks or longer, and one animal was sacrificed 570 days after surgery. The most frequent cause of death was aspiration of vomitus. This was probably caused by undue traumatization of the vagus nerve during temporary clamping of the common carotid artery. If such traumatization permanently damages afferent fibers 306 ROLE OF THE PROSENCEPHALON IN SHIVERING A. Forebrain Level B. Preoptic- Supraoptic Level ■^^^: i. C. Tuberal Hypothalamic Level D. P. Hypothalamic- Midbrain Level Figure 2. Frontal sections of decorticate cat brain No. 14. 307 STUART, D. G. involved in vomiting reflexes, it could account for the animal's inability to vomit effectively. When the animals died their brains were fixed in formalin for gross and sometimes histological examination. The animals were not fed between surgery and the days of testing three days after surgery. Three intact cats were therefore tested in similar fashion before and after a three-day fast. Results . Tables II and III summarize the results for 10 cats tested before and after decortication, one cat further tested 28 and 570 days after decortication, and one hem idecorticate cat tested 42 days after surgery. Three days after decortication all the tested animals shivered in response to a mild cooling stress. It was evident, by independent observation, that the shivering of these animals lacked continuity as the tremor frequently ceased in the various somatic muscles. This observation was supported by VO determinations indicating a decreased VO associated with shivering after decortication. The resting oxygen consumption rates of the animals before and after decortication lay within the same range (Table III), but whereas the mean elevation in VO during shivering was X2.58 before decorti- cation, the mean elevation during shivering after decortication was XI. 6. The range of VO elevations during shivering was X2.1 to X3.8 when the animals were intact, but after decortication this range was more restricted, being Xl.4 to XI. 7. The standard de- viation of the ratio for the intact animals (mean value X2.58) was 0.48 while that for decorticate animals (mean value XI. 6) was 0.12. When tested after decortication, the animals we re hyperactive, with sham or undirected rage easily induced by such nocioceptive stimuli as pinching the pinna, rectal thermometer insertion or rapid body rotation. They tended to hyperventilate at intervals, to have violent fits of gross somatic activity (leaping, running) followed by periods of complete physical quiescence. They defecated, urinated, and vomited excessively in a discoordinated manner with adoption of faulty postures for these expulsions. The one animal tested 28 and 570 days post surgery became more subdued in the second week after decortication, assumed a crouched posture and moved slowly 308 ROLE OF THE PROSENCEPHALON IN SHIVERING Pi CO 0 > un Tt* in 00 r^ r^ t^ v^ vO r- CO CT^ ID M ON rg Tt^ vO r- O O >> h 0 •> cm' o (VJ ■* rg ■* - - - ro Pi 00 •-o O ■* rg r- o in o vO o' 00 r- r- 00 t^ 00 r- r- r- r^ ^ • > < 0) M >^ to n) 0 f^ P h4 ^ rg rg 00 m IT) in rg CO ^ in o ro II ii r-t "* *"* ^ Oj pc; ^ -«^ CO CO rg rg # 0 •> O •> in -* ro ^ o vO oo •H (\J -t in o 00 xO 00 o vO rg rg -* (M ro ^ ^ rg rg rg ro rg o rg 00 ^ CO rg O II CO rg — ^-~ U •> •> Cti » - CO 00 vO ro ^ rg ^ r- •* in o^ 00 o^ 0 ro (M rg rg rg rg rg rg rg rg r- II >^ •> rg Pd u 1^ .|o 6fi CO O O in rg sO •* 00 vO l> l> :3 O O in 00 o o o rg >— 1 ■* o C p; ■> (M (M --• ^ rj rg 0 o V oJ nj CJ O 0 Oi 0) (M ro in -' '-' o o •* ^ '-' u 0 u u 0 o W U •> o 00 o o ro 00 o o^ r- _i 4) (U T) XI +J bO O 00 o 00 rg ,-, 3! o^ r^ O in r^ o rg o O o^ o a. 0) ■* CM (M rg rg ro rg ro - in 11 tJ o o ^ C "^ i^ 0 rt -^ CO rg t) to .? §i OJ X) . M c o O rt o (U i:l o -^ 'i:! ll CO S ffi ^■^ (tJ ^ Ig^' U "^ ^Q >N a.f^ 0) &0 § f-* *H d m yj CO c CD Sd^ O C to 0) ^ CO >< ^Sco >< f^ g X) 0-13 ^^rr) h ■* CO ^ M -— CO 0 -d M m r:^ (U d 0)^ CO In 0. C (U OJ ^-^ > (U t3 c ^ ^ r~4- 3 *,.,^«— H. ■f-^ ■j r^^-^^^-^^^^lf^^^^^J^ 1600 uA/pulse Septal Stimulation-33 hrs. post medication 2 lfiii)ittpilM^it,milti^(hiji~MmiW''\>4'i''^^^ 4 ■ ■■ I ■ l-x .11 . ■.. I .11... ~ . -^ 5 6 800 uA/ pulse P. Hypothalamus Figure 3. Somatic responsef: evoked by septal and posterior hypothalamic stimulation at various times post alpha chloralose medication (Cat No. ST. 17). 317 STUART, D. G. chloralose medication. The posterior hypothalamic stimulation did, however, abolish bilateral EEG spindle bursts, whereas the spindling persisted to a lesser degree during the septal stimulation. Thirty- two hours after chloralose medication, stimulation of the posterior hypothalamic locus evoked an immediate "burst" of shivering fol- lowed by subsequent milder bursts after the cessation of the stim- ulus. This response was evoked with a stimulus intensity of 400 \xK/ pulse. However, a stimulus intensityof 800 ^t A/pulse was necessary to evoke less intense but more continuous shivering during septal stimulation 33 hours after beginning anesthesia. When this intensity of 800 A^ A/pulse was used in stimulating the contralateral posterior hypothalamus 10 hours earlier than this last septal stimulation, more intense and continuous shivering was maintained for the duration of the stimulus, as shown later in Figure 14. As shown later in Table VIII, shivering could be produced both by septal and posterior hypothalamic stimulation from 9 to 24 hours after induction of alpha chloralose anesthesia. Shivering could also be produced during stimulation of these structures in cats anes- thetized with pentobarbital sodium. This indicates that the response was not solely a characteristic of alpha chloralose anesthesia. Figure 4 illustrates the somatic responses evoked by stimulation of avarietyof hypothalamic loci in 6 cats. In 13 other cats not listed, stimulation of one or more of these intermediate loci evoked re- sponses similar to those presented. The schemata of stimulated loci are based on reconstruction of electrode tracts from the buffered thionine sections of the appropriate brains. Plane A is a frontal section 10 mm anterior to the interauricular line and Plane B, 9 mm anterior to this line. Loci 1, 2, and 3, and 10, 11 and 12 are 2.5 mm from the midline, loci 4, 5, and 6 and 13, 14 and 15 are 1.5 mm from the midline while loci 7, 8, and 9 and 16, 17 and 18 are 0.5 mm from the midline. Loci land 7 and 10 and 16 are 10 mm dorsal to the interauricular line and loci 4 and 13, 9 mm dorsal to this line. Each locus shown is 2 mm ventral to any other locus in the same vertical plane. As previously mentioned, loci were stim- ulated at 1 mm depth intervals. These loci stimulation responses are not shown or discussedfor the sakeof clarity and in recognition of the accuracy limitations involved in comparir^ the specific loci stimulated in any one brain to those of another brain. The plane of brain section varies from one experiment to another, as does the 318 ROLE OF THE PROSENCEPHALON IN SHIVERING 00 O g o 0 o I X X ffi -^ H w CO CO CO ^ O H O H O O O :£] O o O -1- o ^ K I X ^ H CO CO CO :^ S DC CO X CO O O H n (M o U H ^ H H m o O ^ ^ H H o o Oh O o H H H H ^ o o a- 0 H o 0 ►J 00 K w X CO X CO r- 0 O o O -o O O ^ o g ^ in H *- H ^ tf X •s Tf H H CO H 0 ro H « :S o ,N1 H H H O n ^ O h h d 2 H r- o 2 ^ ':o, i^ <; H H H H H H u CO W CO CO CO CO o (^ DO Xl Q in o 2! c s 3 CO ■il o o ^ « j: H H H CO O 319 STUART, D. G. alteration in brain volume and the alteration in brain size caused by formalin perfusion in brain fixation. Locus 14 in Figure 4 does, however, include responses evoked by stimulation of a locus 1 mm dorsal to it. Such loci stimulation responses were included under locus 14 rather than 13 becuase stimulation of a site 1 mm dorsal to the latter locus never evoked responses similar to those seen during stimulation of locus 13 and the intermediate locus. In Figure 4 the somatic responses presented are a generalized increase in muscle tone, or more precisely an increase in fore and hind limb rigidity, arrhythmic muscle twitching, alternating tremor and shivering. Alternating tremor is a4to7 cycle/sec limb tremor in which agonists relax when antagonists contract. In shivering the rate of tremor is faster (9 to 11 cycles/sec) and the muscles con- tract synergistically. Additionally, shivering characteristically waxes and wanes in intensity. Differences between these two tremors were observed both visually and electromyographically. Alpha cholralose anesthesia customarily evokes muscle twitching as the animal enters and emerges from the anesthetic state, but no stimuli were applied at such stages. All stimulation responses of the pre- sented loci were repeatable in the same preparation. The table in Figure 4 demonstrates that shivering was best and consistently producedby stimulation of a diencephalic regionbounded dorsally by a plane 9 mm dorsal to the interauricular line, ventrally by a plane 7 mm dorsal to the interauricular line, laterally by a plane 1.5 to 2.0 ram lateral to the midline and medially by the midline. Anatomically the region is bordered by ventral thalamic, dorsal hypothalamic, and medial subthalamic structures. Systematic exploration of a plane 1 mm rostral to Plane A was performed in 5 cats (No.'s ST 9, 10, 11, 13,andl7). Shivering was never evoked by stimulation of any such loci, but alternating tremor was evoked by stimulating loci 1 mm rostral to loci 1 and 2 . Shivering was evoked by stimulation of loci 1 mm caudal to loci 14 and 17 but not in any other loci of this plane in Cats No. ST 9 and 11. Loci 3.5 mm lateral to the midline in frontal planes corresponding to A and B were stimulated in Cats No. ST 14 and ST 33 , but shivering was not evoked by any such stimulation. 320 ROLE OF THE PROSENCEPHALON IN SHIVERING Within the region thus localized, it was not possible to demon- strate stimulation of a specific part as being more effective in producing shivering than any other part when the responses of one cat were compared to another. For example, Figure 5 illustrates the production of well defined shivering by stimulation of a locus 1.5 mm from the midline (arrow B) in Cat No. ST 10. In the same animal, stimulation of a locus 0.5 mm from the midline (arrow F) produced a much less defined shivering response. Stimulation of a locus 2.5 ram from the midline (arrow A) produced a larger and slower limb oscillation response that could not, by independent observation, be termed "shivering." In contrast, in other cats (e.g., ST 9) there was a better production of shivering by stimulation of a locus 0.5 mm from the midline with a more discontinuous production of less well defined shivering produced by stimulation of a locus 1.5 mm lateral to the midline. The parameters of stimulation were similar in each given experiment, but the intensity of stimulation necessary to evoke shivering varied from animal to animal, ranging from 200 ^ A/pulse to 1600 \xK/ pulse. The frequency of stimulation utilized ranged from 25 to 100 pulses/sec. In some experiments 25 pulses/sec seemed more effective in evoking shivering than 50 to 100 pulses/ sec, and in others 25 and 50 pulses/sec seemed more effective than 100 pulses/sec. With the exception of two experiments (Cat No. ST 17 and Cat No. ST 33 ) no attempt was made to explore systematically the frequency-intensity combination that would evoke shivering when repeatedly stimulating the same locus. Figure 6 illustrates an experiment in Cat No. ST 33 in which the minimum intensities necessary to evoke shivering at stimulus frequencies of 10, 25, 50, and 100 and 200 pulses/sec were determined. In all cases the stimulus pulse duration was 1 msec, the period of stim- ulation 30 seconds, and the locus of stimulation the same. An attempt was made to clarify the results by gauging shivering response as being "strong," "mild," or "dubious." This classification was based on a combination of independent observations and the pattern of recorded EMG's. This is shown schematically at the top of Figure 6. A response was classified as "strong" and coded as a filled- in circle if the electromyogram of at least one limb muscle recorded shivering for the durationof the stimulus. A response was classified as "mild" and coded with an open circle when visible shivering either began when the stimulus began but terminated before the 321 STUART, D. G. i 322 ROLE OF THE PROSENCEPHALON IN SHIVERING Observed response Code Strong EMG Recording (schematic) Duration nt stimulus — O: 18 T5 ^ STIMULUS FREQUENCY (pulses/ sec. ) i3o 200 Figure 6. Frequency dependent graded instigative responses. 323 STUART, D. G. cessation of stimulation or when visible shivering began at some latent period after the beginning of stimulation but ceased when the stimulus was teminated. A response was coded as "dubious" and coded with a small filled- in circle when neither visual inspection nor the pattern of EMG recording indicated whether or not the muscle activity evoked by the stimulus was truly shivering or merely an increase in muscle tone. Stimuli of 50 and 100 pulses/sec required less intensity to evoke shivering than stimuli of 10, 25 and 100 pulses/sec. The best shivering response evoked by a 10 pulse/sec stimulus was mild shivering lacking continuity. Con- versely, the best response evoked by stimulation at 200 pulses/sec was a tremor in which background muscle tone appeared excessive. In Cat No. ST 17, shivering could not be evoked by stimuli of 1600 /i A/ pulse intensity even though well defined shivering was evoked at 25 pulses/sec of 400 //A/pulse intensity. In this experiment 200 pulses/sec stimuli were not applied, but less intense stimuli were necessary at 25 pulses/sec than at 100 pulses/sec. In both experi- ments 50 pvilses/sec stimuli evoked well defined shivering at min- imal stimulus intensities. In early experiments (Cats No. ST 5 to ST 17) a stimulus pulse duration of 3 msec was utilized. In Cat No. ST 17, using a variety of frequencies, less intense stimuli were necessary to evoke shiv- ering with a pulse duration of 1 msec than with 3 msec. Following this experiment, a pulse duration of 1 msec was used, but no sys- tematic study was attempted of the intensity of stimulation needed to evoke shivering at a variety of pulse duration- frequency combin- ations. Comments. The dorsomedial region of the posterior hypothal- amus, which, when stimulated electrically, induces shivering, is dorsal and medial to the shivering pathway described by Birzis and Hemingway (1956, 1957). These investigators abolished shiv- erir^ by ventrolateral posterior hypothalamic destruction in anes- thetized cats. In our laboratory it is now the contention that there is a conflict in interpretation but not experimental results. For example, Figure 7 illustrates an experiment in which bipolar elec- trodes were inserted into bilateral ventrolateral posterior hypo- thalamic loci (the center of the lesions noted with insert A and B) and into the dorsomedial posterior hypothalamus (tip of electrode noted C). With the brain intact, 30- second stimulation of locus A 324 ROLE OF THE PROSENCEPHALON IN SHIVERING £ £ £ E V t ' u _ 0 7 ^■ o - o' £ E .\ I .. 325 STUART, D. G. with 400 ju A/pulse, 50 pulses/sec and 3 msec pulse duration sup- pressed shivering in the cat anesthetized with pentobarbital sodium (35 mg/kg I. ?.)• Shivering returned after 10 minutes and was again suppressed in the process of destroying tissue at locus A electro- lytically. Shivering was depressed for 10 minutes but was repro- duced by stimulating locus C at 1600 /; A/pulse with 25 1 msec/ pulse/sec. Following locus C stimulation, spontaneous shivering returned on the right hindlimb, but was suppressed by locus B stimulation with a 30- second stimulus of 200 m A/pulse and 50 3 msec pulses/sec. Shivering was then reproduced by stimulation of locus C (not shown in Figure 7) but suppressed by electrolytic destruction of locus B. Following bilateral ventrolateral posterior hypothalamic destruction, shivering did not reoccur while the animal was anesthetized, but thirty-one days after surgery it was immediately apparent on exposure of the cat to cold, the VO shiv- ering/resting being 3 .2 . The failure of shivering to occur immediate- ly after bilateral ventrolateral posterior hypothalamic destruction may have been due to an unmeasured cardio-vascular depression, in that Fuster and Weinberg (1960) have shown that stimulation of the regions destroyed in this particular experiment increases myocardial contractility. The main aspect of this experiment is the demonstration of Birzis and Hemingway's result in an experiment that also illustrates seemingly conflicting results. That is to say, shivering can be evoked by dorsomedial posterior hypothalamic stimulation and suppressed by ventrolateral posterior hypothalamic stimulation. It can also be suppressed by ventrolateral posterior hypothalamic destruction in an acutely observed anesthetized prep- aration. Following such destruction shivering does return in the unanesthetized, chronically maintained preparation. This does not mean that all Birzis and Hemingway's acute lesion experiments are subject to question. They have hitherto unreported confirmation of the validity of this shivering pathway in the midbrain, pons, and bulb in that they have demonstrated the lack of shivering in chronic animals with bilateral lesions in the region in which the tissue destroyed is similar to that destroyed in the acute experiments. Perhaps the reason for their diverse hypo- thalamic result is that the majority of their efforts were directed to the descending paths, rather than the central origin of impulses related to the production of shivering. 326 ROLE OF THE PROSENCEPHALON IN SHIVERING It is not here the purpose to claim that the dorsomedial region of the posterior hypothalamus is an exclusive "center" for shivering. Figures 3, 14, and 26, and Table VIII illustrate EEC, heart and respiratory changes during stimulation of this region. Rather it is to suggest that activation of certain neurons within this region pro- duces shivering along with other ergotropic effects. Shivering limbs have a tremor frequency of 9 to 11 cycles/sec. This frequency was not effective in producing shivering when applied to an electrical stimulus to the hypothalamus. Stimuli of frequency 25 to 100 pulses/sec were much more effective. This suggests that the rhythm of shivering is controlled peripherally, but shivering itself is instigated centrally. This concept will be elaborated at a later symposium. It helps to explain the paucity of information from Hess's laboratory (Akert and Kesselring, 1951, Hess, 1957) con- cerning the production of shivering during electrical stimulation of the prosencephalon. That is to say, when the brains of over 350 anesthetized and unanesthetized cats were stimulated at about 7,000 prosencephalic loci, shivering was observed but 11 times in 8 cats. The stimulus used in this laboratory consisted of a variable direct current that could be mechanically interrupted but 4 to 15 times/sec. If Hess had had the advantage of modern electronic stimulators per- mitting higher stimulation frequencies, he undoubtedly would have unmasked even more physiology of the diencephalon. Electrolytic Lesion Studies Methods. In these experiments bilateral electrolytic lesions were made in various hypothalamic regions of 29 cats. Each cat was anesthetized with pentobarbital sodium (35 mg/kgl. P.) and the head mounted in a stereotaxic frame. The scalp skin was reflected on appropriate holes burred through the calvarium to permit inser- tion of a stereotaxic ally oriented insulated monopolar electrode. Each electrode was 0.7 mm in diameter and without insulation approximately 1 mm from the tip. An electrolytic lesion was made by cathodal polarization of the electrode with a current of 1.5 to 2.5 mA passed for 1 to 2 minutes. A remote anodal connection was provided on the tongue. The scalp skin was repositioned with wound 327 STUART, D. G. clips. Great attention was directed to the postoperative care of each animal. The rectal temperature was kept at 36 C to 38 C by appropriate alterations in environmental temperature. For the first 2 to 3 weeks after surgery each animal was tube fed with frequent testing for the recoveryof licking and swallowing reflexes. Care was taken to keep each animal clean and dry and given daily periods of exercise. Twenty-one animals recovered from the above procedure, and once they had regained their pre-operative weight by spontaneous eating, their responses to cooling were studied. One animal died at 12 hours, one at 5 days, one at 8 days and one at 10 days after surgery. Three animals died during the coiirse of heat stress tests administered 4 days after surgery. The resting oxygen consumption rate of 18 of the 21 cats which recovered was determined 13 to 132 o days after surgery at an environmental temperature of 20 C to 25 C. They were then exposed to an environment of 0 C to 5 C temperature, and 15 minutes after such exposure their oxygen con- sumption rates were determined for a further twenty minutes of exposure. Rectal temperature was maintained throughout these determinations. Control experiments consisted of measuring oxygen con- sumption rates and rectal temperatures of unoperated cats in the warm (25 C) and cold (0 C to 5 C) environment. If the animals with hypothalamic lesions had oxygen consumption rates and rectal temperature responses that fell within the un- operated cat range of responses, they were sacrificed immediately after the test, their brains fixed in formalin and sectioned every 80 microns, and alternate sections stained with buffered thionine. If any of the animals did not have control unoperated responses they were tested at least twice more at varying lengths of time after the first test before being sacrificed. In subsequent figures illustrating the extent of various brain lesions, a schematic midsagittal diagram will show the rostrocaudal extents on each side of the brain. Such diagrams are based on re- construction of the extent of the lesions from the buffered thionine slides. Additionally two to three film negatives of buffered thionine slides are presented to illustrate the dorsoventral and medial lateral extents of tissue destruction. 328 ROLE OF THE PROSENCEPHALON IN SHIVERING Results . Table VI lists the oxygenconsumption rates and rectal temperatures of 9 intact cats while exposed to a temperature of 24.5 C to 29 C air and while shivering in a temperature of 0 C to 5 C air. The mean VO shivering/resting ratio was 2.7, range 2.1 to 3.8 and standard deviation 0.5. The mean drop in rectal temperature during the 20-minute determination of V0„ in 0 C to o o o o 2 5 C was 0.8 C, range 0.2 C to 1.3 C, and standard deviation 0.4 C. Expressed in other terms, the mean rate of rectal tempera- ture drop per unit time was 0.04 C/min, range 0.02 C to 0.07 C, and standard deviation 0.02 C/min. Table VI shows that on succes- sive days the ratio of VO shivering/resting was 2.4 and 2.6 for Cat No. 1 and 3.2 and 2.1 for Cat No. 2. In the former case the 0 o rectal temperature drops were 0.2 C and 1.3 C and in the latter case 0.6 C and 0.8 C. These figures would indicate that the in- crease associated with shivering could vary from a two to fourfold increase in both the same or a sequence of cats. Additionally, the range in rectal temperature drops that accompanied the period of shivering oxygen consumption rates determination could vary by over 1 C in the same cat and a sequence of cats and that this variance was not necessarily correlated with the VO . It was not considered necessary to run additional tests on the other 7 cats to illustrate this point. This was because, in a previous stucty, the oxygen consumption rates of 7 cats were determined on successive days while they were shivering for the duration of time necessary to elevate their rectal temperatures from 33 C to 37 C. The determined ratios were as follows: Cat A: 4.1, 2.4, 4.4 and 2.1 Cat B: 2.2, 2.8, and 2.5 Cat C: 3.1, 2.9, 2.2 and 2.8 Cat D: 2.1, 2.4, and 2.2 Cat E: Pre midbrain lesion: 2.8 and 3.1 Post midbrain lesion: 2.0 and 2.4 Cat F: Pre midbrain lesion: 2.0 and 2.3 Post midbrain lesion: 2.6 and 3.8 Cat G: Pre midbrain lesion: 2.1 and 2.5 Post midbrain lesion: 2.4 and 2.7 In the above determination there was no significant correlation between the ratio and the time taken by the cats to elevate their o o rectal temperatures from 33 C to 37 C. On the basis of this, and 329 STUART, D. G. "1 ^ « "7 i" * -o CM -i 00 r^ o (M o — oo r~ -. in N CM CO M N <^ m IM ro (M l^ ^ < .iS — ;3 1% l\ II as 1= p |sf asi ^^ :rd |t ^§ x§ s"a :&5i s|S < a; oJ oi a cu ci<' i oi 332 ROLE OF THE PROSENCEPHALON IN SHIVERING ^ 1 Figure 8. Extent of neural tissue destruction in Cat No. F 10. 333 STUART, D.G. 15 10 5 Mms. Anterior to Inferauriculor PLANE OF FRONTAL SECTION '^^ ^^ B f Figure 9. Extent of neural tissue destruction in Cat No. F 8. 334 ROLE OF THE PROSENCEPHALON IN SHIVERING 15 10 5 flnlenor to Interouriculor iw^^r -i®^^^"! f^ il . m^k Figure 10. Extent of neural tissue destruction in Cat No. ST. 25. 335 STUART, D. G. Cat No. H 13 - A Cat No. H 13 - B Cat No. H 13 - C Cat No. H 14 - A Cat No. H 14 - B Figure 11. Extent of neural tissue destruction in Cats No. H 13 and H 14. 336 ROLE OF THE PROSENCEPHALON IN SHIVERING cold test it sought a warm pad and heating lamp when they were moved around the laboratory. On the third day of testing, 40 days post surgery, the animal was placed in a cage with another larger cat renowned for aggressiveness. After a typical display of ag- gression by the second cat, H promptly escaped from the cage when the door was opened. During the cold test the warm pad and heating lamp 'were placed in the cage with the aggressive cat and at the conclusion of the cold test H , with a rectal temperature of 33 C and still not shivering, entereathis cage to stand on the warm pad despite the aggressive cat's auditory and somatic objections. Although the impairment of shivering in Cats No. F , ST 25 and H , is such as to implicate the integrity of the dorsomedial posterior hypothalamus in the production of shivering, the above results do not indicate how little of this tissue can remain in order to permit effective shivering. In Cat No. H an attempt was made to destroy all of the tissue of the posterior hypothalamus except the dorsomedial region. As Figure 11 shows, this attempt was only partially successful. At the tuberal level of the posterior hypothal^ amus in this cat, most of the dorsolateral ri^t side posterior hypothalamus was destroyed but the dorsomedial tissue was pre- served on the left side. However, on the left side most of the dorsal tissue was destroyed at the posterior hypothalamic level (Fig. 11- Gat No. H B), but the dorsomedial region was preserved on the rig^t side. Since this cat could shiver effectively, this would sug- gest the ipsilateral preservation of the dorsomedial posterior hypothalamus tissue was sufficient to permit effective shivering. This point is further illustrated in Cat No. ST 24, in which the pos- terior hypothalamic lesions were bilateral but only overlapped for a distance of 0.75 mm approximately. However, at the level or region of overlap there was widespread dorsomedial posterior hypothalamic destruction except in that the ventral tissue within this region was still intact. This animal could shiver quite effective- ly. Similarly, Cat No. H could shiverquite effectively with all the dorsomedial posterior hypothalamic region destroyed on the right and partially on the left. Again the more ventral tissue within these boundaries was intact on both sides. In summary, these limited observations on 18 cats would sug- gest that ipsilateral preservation of the dorsomedial posterior 337 STUART, D. G. hypothalamus is sufficient for an animal to shiver effectively in the cold. However, it is not possible on the basis of these experiments to designate a more precise or limited region as being responsible for the production of shivering. In Table VII cold test A was only used when the mechanical breakdown of the cold test B refrigerator system occurred. In such tests it was not possible to gain an indirect picture of an animal's heat retention capacity. However, in cold test B,by measuring both oxygen consumption and rectal temperature while exposed to 0 C to 5 C air, this was possible. As shown in Table VI, the range of rectal temperature drop /unit time in 0 C to 5 C air was 0.02 C to 0.07 C/min for intact cats. Table VII shows this figure was .13 C/min for Cat No. F , .13° C/min for Cat No. F , and .33 C/min for Cat No. H . Additionally, Cats No. F and H had figures of over .13° C/min and Cats No. F , F and F , ST 19 and ST 38 fig- ures of less than 0.07° C/min.'^ince Cats No. F and F and H and H had bilateral lesions that effectively destroyed the lateral tuberai and posterior hypothalamic regions and Cats No. H , F , F and F , ST 19 and ST 38 had lesions that destroyed only the ventrolateral tuberai and posterior hypothalamic tissue, it would appear that the effective retention of heat is a function of the dorso- lateral tuberai and hypothalamic tissues. Such a conclusion is further supported by destruction of this tissue being evident in Cat No. F (Fig. 9). It would appear that the dorsolateral anterior hypothalamus is not involved in heat retention in that in Cat No. H the dorsolateral anterior hypothalamic tissue was destroyed yet the animal both shivered effectively and had a rectal temperature rise/unit time in 0 C to 5 C air of .01 C/min. Although the animals with lateral and dorsolateral hypothalamic lesions appeared to have a diminution in heat retention capacity in 0 C to 5 C air, by indirect determination, this was not obvious in 25 C air. That is to say, these animals had, at this higher temper- ature, rectal temperatures and resting oxygen consumption rates within the normal range. No systematic attempt was made to study temperature regula- tion disturbances in the immediate postoperative period. Table VIII lists the responses of three cats to heat and cold stresses 4 to 46 days after surgery. Four days after surgery the three animals had 338 ROLE OF THE PROSENCEPHALON IN SHIVERING R. T. after Cat 8 hours of No. 25° E. T. 35.8 10 35.0 35.8 R. T. after 3 hours of Observations 45° E. T. 4 Days After Surgery No shivering 41,0 No shivering 43. 0 No shivering 41.0 Observations No panting Languid posture No panting Languid posture -Death- No panting Languid posture 46 Days After Surgery 10 39.2 39.2 Shivering X2.6 elev. VO2 Feeble shivering XI. 9 elev. in VOo 41. 0 41.0 Panting Languid posture Panting Languid posture Table VIII. The responses of 3 cats with hypothalamic lesions to heat and cold stresses. R. T. - rectal temperature in degrees Centigrade. E. T. - environmental temperature in degrees Centi- grade. VO - oxygen consumption rates. 339 STUART, D. G. low bocfy temperatures in a 25 C environmental temperature but could not shiver. When exposed to a 45 C environmental tempera- ture, the animals' rectal temperature rose to 41 C to 43 C, this rise being accompanied by assumption of a languid posture that would permit maximal conductive heat loss. However, panting was not evident. Thus at four days after surgery, these animals verged on poikilothermy rather than homeothermy. Forty-nine days after surgery the two animals that survived the initial tests had normal rectal temperatures in the 25 C environmental temperature, could shiver in response to cold stresses and pant in response to heat stresses. Two other animals, H and H , with massive hypothal- amic lesions, demonstrated similar results when tested five days after surgery. Both these animals expired during the heat stress tests, without assuming languid postures and without panting. These results are listed to illustrate the author's reluctance to expose animals with massive hypothalamic lesions to heating and cooling tests in the early postoperative period of recovery. In this somewhat enfeebled condition death often results and the somatic responses to both high and low body temperatures are usually im- paired above and beyond the impairment present after the animals have recovered relatively fully from the surgery. Comments. The results support the electrical stimulation data in implicating the dorsomedial posterior hypothalamus in the pro- duction of shivering. If it is also accepted that cutaneous vasocon- striction is activated by the dorsolateral posterior hypothalamic neurons, then some of the seemingly diverse results of other in- vestigators can be explained in rather simple terms. For example, Figure 12 shows typical lesions produced by three sets of inves- tigators whose results have been widely quoted. First, Isenschm id's work with Krehl (1912) and Schnitzler (1914) implicated the lateral hypothalamus in temperature regulation (Figure 12 - Isenschm id D) but not the medial hypothalamus (Figure 12 - Isenschm id C). However, in this work, the animals' rectal temperatures were taken over a range of environmental temperatures, and at very low en- vironmental temperatures the body temperature of the animal with lesion C began to fall. If it is assumed that all these animals were in a debilitated condition and that, at least in terms of functional and nonfunctional nervous tissue, "C" could not shiver but could vasoconstrict, then it is feasible that its body temperature would 340 ROLE OF THE PROSENCEPHALON IN SHIVERING KELLER (DOGS) These two dogs had subnormal temps. for several days but subsequently could maintain normal body temp, in an environmental temp, range of 5-35°C. A- Undisturbed Temp. B. Heat loss Deficit CLARK, MAGOUN, & RANSON (CATS) 341 STUART, D. G. A. Disturbed Temp. Reg. C. Undisturbed Temp. Reg. HirnschenKelfuss B. Undisturbed Temp. Reg. Hobenula Infundibulum irns Troctus opticus TuberCinareum N.m. D. Disturbed Temp. Reg. >- Corpus momillare ISENSCHMID (RABBITS) Figure 12. Lesions produced by various investigjrtors that failed to dis- turb and did disturb body temperature regulation. 342 ROLE OF THE PROSENCEPHALON IN SHIVERING be relatively normal until the environmental temperature plum- meted. Conversely, if the rabbit with lesion "D" could shiver but not vasoconstrict cutaneously, then it is conceivable that in an enfeebled condition its body temperature would fall despite the neural "drive" to produce heat by muscular shivering. That their animals, studied but for five days post- surgery, were enfeebled seems evident from their statement (1914) that "we have never observed that an originally disturbed heat regulation later became normal, but we have often witnessed the reverse." This completely contradicts the latter-day findings, exemplified by the work of Hess (1957), that functional deficits produced by hypothalamic lesions become attenuated the longer the preparation lives after surgery. Figure 12 secondly shows schema of four lesion preparations that were considered by Clark, Magoun, andRansonto implicate the anterolateral hypothalamus in body temperature regulation against heat stress and the posterolateral hypothalamus in regulation against cold stress. Admittedly in a later review Ranson and Magoun (1939) stated that their observations on shivering were fragmentary, but this work (which essentially confirms the work of Isenschmid and his co-workers but had the advantage of beingbased on observation of animals fully recovered from surgical trauma) is still widely quoted with respect to the neurogenesis of shivering. Their work does not conflict with our results if it is accepted that their lesion "C" does not include the dor somedial posterior hypothalamus and their lesion "D" affects both dorsomedial and dorsolateral hypo- thalamic tissue as well as the more pronounced ventrolateral destruction. Thirdly, two lesions from one of Keller's (1959) investigations are shown in Figure 12 . Their dogs, maintained for long periods after surgery, could shiver and maintain normal body temperature even when placed in a 5 C environmental temperature for several hours. The hypothalamic lesions spare only the dorsal posterior hypothalamus, thus suggesting that shivering involves the dorso- medial neurons and vasoconstriction the dorsolateral neurons. Time does not permit a more detailed comparison of our re- sults with divergent results in the literature. If it is accepted that shivering will not return in the early post-operative period after any hypothalamic lesion, and that cutaneous vasoconstriction 343 STUART, D. G. involves the dorsolateral posterior hypothalamus and will return in the early postoperative period after any hypothalamic lesion that spares this region, then many seemingly divergent results can be explained. COMPARISON OF SOMATIC EFFECTS EVOKED DURING SEPTAL AND POSTERIOR HYPOTHALAMIC STIMULATION Anesthetized Preparation Studies Methods . In 7 experiments on cats anesthetized with either alpha chloralose (40 to 60 mg/kg I. P.) or pentobarbital sodium (35 mg/kg I. P.)i the septum was systematically explored for sites which when stimulated evoked somatic responses. In each experiment the intensity of stimulation needed to evoke shivering by stimulation of a dorsomedial posterior hypothalamic site was noted. In some of these experiments comparisons were also made of alterations in heart rates and respiration rates during septal and posterior hypothalamic stimulation. In these experiments an attempt was made to maintain the rectal temperature of each cat between 37 C and 38 C by appropriate alterations of environmental temperature. The surgical, electronic, brain fixing, sectioning, staining, and electrode tract location techniques were similar to those described above. Results. In five experiments (Cats No. ST 5, ST 13, ST 14, STl5 and ST 17) the septal region was systematically explored for loci which when stimulated produced either an increase in muscle tone or shivering. The strongest somatic response evoked by stimulation of any given septal locus was com pared with the shivering response to posterior hypothalamic stimulation in the same cat. Comparisons were made of the latency and intensity of the response and the stimulus intensity necessary to evoke it. Additionally, in some 344 ROLE OF THE PROSENCEPHALON IN SHIVERING experiments a comparisonwas made ofthe alterations in respiration and heart rate produced by such stimulation. In two of these five experiments (Cats No. ST 13 and 17) limited posterior hypothalamic mapping was also undertaken. In two further experiments (Cats No. ST 18 and 24) the comparisons were not undertaken after extensive septal mapping, but rather the electrodes were oriented to septal loci when stimulation had produced somatic responses in previous experiments. Figure 13 lists the somatic responses evoked by septal stimula- tion in five cats. Planes A and B are schemata of typical cat brain frontal sections 16 mm (midseptum) and 14 mm (posterior septum) anterior to the interauricular line respectively. Loci 1, 2, 3 and 4 and 9, 10, 11 and 12 are 1.5 mm from the midline while loci 5, 6, 7 and 8 and 13, 14, 15 and 16 are 0.5 mm from the midline. Loci 1 and 9 are 15 mm dorsal to the interauricular line, and loci 5 and 13, 16 mm dorsal to this line. Loci ventral to these are at 2 mm depth intervals. As with the posterior hypothalamus, loci were stimulated at 1 mm depth intervals, but responses to stimulation of these loci are not listed for the previously mentioned reasons. Extensive mapping was undertaken at a frontal plane 15 mm rostral to the interauricular line, but responses to stimulation of these loci are not listed, since the responses at this plane were similar to those seen at Plane A. Only two somatic responses are listed, an increase in muscle tone and shivering. The table in Figure 13 indicates that shivering was observed in three of the five cats. The response was best evoked by stimulation of midseptal loci 1.5 mm lateral to the mid- line andfrom 13 mm to 9 mm dorsal to the interauricular line. Stim- ulation of loci 1 mm rostral to this region evoked an increase in muscle tone but not shivering. When shivering was observed it seemed of equal intensity and duration in limbs both ipsi and con- tralateral to the stimulated site. In these experiments this latter finding was based on both visual observation and EMG recordings. Table IX lists comparison of somatic, heart and respiration rate responses during stimulation of the most "active" septal loci and a posterior hypothalamic locus within the previously deter- mined region whose activation produces shivering. In the four cats (No.'s ST 5, 13, 18 and 24) in which shivering was not observed 345 STUART, D. G. o X 2 0 < t33 cn -un -IrT -77)- rn H H H H H n > ^- l_ ^ ^ U1 ^ (Jl ^ Oo p -1 o H H O O 0 cn o cn H M X O w o H H OJ u: o O ^ X £ 0 rf^ o 0 0 O Ui o 0 H o CT- o O o 0 0 o -J o o o o 00 o o H 0 H O xO n c CO o o H 0 0 ° 0 o H 0 o - s o H O o rj o o o H UJ u o o o O >f>. o o o 0 Ul ^ o o o O o^ 346 ROLE OF THE PROSENCEPHALON IN SHIVERING e .2 III ? '£ ^^'2: £1 "I fsj If |i lllj 3 1 |. S5 '5 ?3 IS 3-32 6 & 11. 3 C 5'^.. i 1 1 1 " i 1 1 1 " i 1 1 1 In 5 S P - 3 2 S. 1 ^ X 0,' 347 STUART, D. G. during septal stimulation, it was observed during posterior hypo- thalamic stimulation. In the three cats (No.'s ST 14, 15 and 17) in which shivering was observed during septal stimulation, a more intense and less latent shivering was produced by the same or less intense stimulation. In Cat No. ST 5, the comparison is hardly valid in that the septum was stimulated 3.5 hours after induction of alpha chloralose anesthesia and the posterior hjrpothalamus was stimulated 9 hours post medication. That is to say, the preparation may have been at too deep a level of anesthesia to shiver during septal stimulation. However, in all other cats the stimuli were ap- plied to the two loci within a short period of time of each other. Figure 14 illustrates the comparisons for Cat No. ST 17. In this figure there are schemata of septal loci labelled 1 and 2, 2 being 2 mm ventral to 1. The posterior hypothalamic loci 1 and 2 are within the previously determined dorsomedial region whose stimu- lation evokes shivering. The hypothalamic locus denoted 2 is 1 mm lateral and 1 mm ventral to the locus marked 1. Loci posterior hypothalamus 2 and septum 1 were stimulated for 30 seconds at 50 pulses/sec with a pulse duration of 3 msec. In both cases shiv- ering was produced, but the stimulus intensity necessary at the hypothalamic locus was only half that necessary at the septal locus (800 /iA/ pulses vs. 1600 /.lA/pulse). Additionally, as shown in records No. 4 and 5 for the loci in Figure 14, shivering was more intense and appeared with less latency during hypothalamic than during septal stimulation. Loci posterior hypothalamus 1 and sep- tum 2 were stimulated for 60 seconds using the same frequency and pulse duration. As shown in records 4 and 5 for these loci, shiv- ering was less intense during the septal stimulation and required a stimulus intensity double that used during hypothalamic stimu- lation. These loci were stimulated in the order posterior hypothal- amus 2, septum 1, septum 2 and posterior hypothalamus 1 at 22.3, 23.5, 23.6, and 24.5 hours after induction of alpha chloralose anes- thesia (60 mg/kg I. P.). The rectal temperature during stimulation of these loci was 37.8 C, 37.8 C and 38.8 C respectively. In all these cases the environmental temperature was 28 C . These results indicate that shivering can be induced by electrical stimulation of sensitive loci in both the hypothalamus and the septum, but the sensitivity of the hypothalamic loci is considerably greater than that of the septal loci as revealed by the necessary intensity of stimulating current and the EMG response. The EMG of the hind- 348 ROLE OF THE PROSENCEPHALON IN SHIVERING 1. Resp. Rate 2. EEG - L. Cortex 3. EEG - R. Cortex 4. EMG - L. Forelimb 5. EMG - L. Hindlimb Black line - Stimulus Duration All Stimuli 50 pulses/sec. Pulse Duration 3 msec. HYPOTHALAMUS Posterior Hypothalamus 2-800 uA/pulse Septum 1-1600 uA/pulse Posterior Hypothalamus 1-800 uA/pulse 2 ^Ml>|f4»4iW^wl*,i<|tBttH***^#*>I^M^^ u 1 \ I ml >i 1 11 % irti «m iiiw II — -i^ — -^ — -f — " 5 ■ -■ ■■- -' '•'■- Septum 2-1600 uA/pulse i.J»....,f».i||H». l»ll^»ll^)l>l>ll»4»W^^ll^l«l*|'^>tt^^»>^rv-^^^^^■^^ Figure 14. Comparison of somatic effects produced by septal and pos- terior hypothalamic stinnulation (Cat No. ST. 17). 349 STUART, D. G. limbs (record 5) for each stimulated locus was similar to that of the forelimbs (record 4) but weaker in intensity of response. As well as the production of greater somatic activity during posterior hypothalamic stimulation than during septal stimulation, respiration and heart rate increases were of greater magnitude in the former than in the latter case. The appropriate figures are listed in Table IX , but it must be pointed out that no systematic attempt has been made in this study to compare the alterations in these rates during stimulation of septal or posterior hypothalamic stimulation loci. Comments. These results tend to confirm neuroanatomical data that septal projections to the midbrain first relay in the hypothala- mus. It is not possible on the basis of the presented data to deduce the number of relays involved between septum and hypothalamus, but this was not the purpose of these experiments. Rather, it was to show that shivering can be induced during electrical stimulation with a greater stimulus intensity during septal stimulation. This suggests the possibility that the main control for the production of shivering is in the posterior hypothalamus and that facilitating influences can reach the hypothalamus via or originating in the septum. Akert and Kesselring (1951) first reported the production of shivering by stimulation of 10 septal loci and only one hypothalamic locus in 8 cats. Based on this work, they suggested that the septum was primarily implicated in the production of shivering, since Hess's results, from which their report was gleaned, contained but this one isolated example of hypothalamic stimulation producing shivering. However, as mentioned above, this was due more to the parameters of stimulus than of the locus of stimulation. Sim- ilarly, Andersson (1957), in reporting the consistent and repeatable evocation of shivering during septal stimulation in 3 unanesthetized goats, speculated that "it might be possible that an integrative action of all mechanisms concerned with heat conservation is exer- ted from this part of the forebrain." Possibly he would not have so speculated had he compared the shivering response during both septal and hypothalamic stimulation. Certainly the fact that shiv- ering occurs in animals with transection separation of the anterior from the posterior hypothalamus, first reported by Bazett, Alpers 350 ROLE OF THE PROSENCEPHALON IN SHIVERING and Erb (1933), would indicate that structures more rostral than the posterior hypothalamus could play but a secondary role in the production of shivering. The data listed in Figure 13 and those later to be discussed in Figure 21 indicate that the ventrolateral midseptum is facilitating and the ventromedial midseptum inhibitorywith respect to shivering. Jacobson, Craig and Squires (19 60) found that electrolytic destruction of the ventromedial midseptum suppresses the shivering of lightly anesthetized cats. This implied that the ventromedial midseptum was involved in the production of shivering, and as such conflicts with our concept. An experiment was then performed to illustrate that the results from the two laboratories do not conflict, only the interpretation. In this experiment, shown in Figure 15, bipolar electrodes were inserted bilaterally into the center of the lesion shown in the top lefthand corner of Figvire 13, i.e., the photograph of a frontal section through the septal region. The tracts of the electrode do not appear because the brain was sectioned at a dif- ferent angle to the paths of the electrode tracts. When these elec- trodes were inserted, the brain was intact; the lesions were produced later in the experiment. A bipolar electrode was also inserted into the posterior hypothalamus as indicated by the white stain in the lower photograph. The locus stimulated by this electrode was the lowermost tip of the white stain. (Unfortunately this electrode slipped medially several days after the experiment when it was being removed from the living cat. Hence the white stain shows a vertical and a slanted angle to the electrode tract.) This cat was anesthetized with pentobarbital sodium (35 mg/kg I. P.), and once spontaneous shivering began, the following sequence of stimuli was applied, as shown in the figure in which the upper record is the electromyogram of the forelimb and the lower record the duration of the stimulus: 1. 16 OOa^ A/pulse stimulation of the righ ventromedial mid- septum to suppress shivering for the 30- second duration of stimulus. 2. 800)Li A/pulse stimulation of the posterior hypothalamus to augment shivering for the 30-second duration of stimulus. 3. Right ventromedial midseptum destruction by cathodal 351 STUART, D. G. £§ E u5 iiiiiiii I iiHii 3 momtmmmm Locus B St ^ Figure 22. Suppression ot shivering during septal stimulation after bi- lateral fornical destruction (Cat No. ST. 29^). Stimuli 25 pulses/sec. , 800 uA/pulse, 1 msec, pulse duration. 366 ROLE OF THE PROSENCEPHALON IN SHIVERING Site A-Left Side Stimulation Site B-Left Side Stimulation " ^ * " * "^ * * "" ' m\f il \ .til Site A-Right Side Stimulation m m.<. ilttn^h < wa . M 1 I V» ' ' " Site B-Right Side Stimulation ; lin JW- ™« nir ». 1 ,tm||it 100 uV it 1. Resp. Rate J^'l Schema of Loci 2. EEG - R. Cortex 3. EMG - R. Forelimb 4. EMG - R. Forelimb 5. EMG - L. Hindlimb 6. EMG - R. Hindlimb 7. Stimulus Duration Figure 23. Suppression of shivering during septal stimuiation after ipsi lateral fornical destruction (Cat No. ST. 31). All stimuli 25 pulses/sec. , 400 uA/ pulse, 1 msec, pulse duration. 367 STUART, D. G. Q CAT NO. ST. 34 All stimuli delivered for 30 with frequency of 25 pulses/ s ^P Strong suppression ^^j Mild suppression — Septal stimulation A Hypothalamic stimulati }X': K NO. ST. 32" '\^ 1 1 1 1 1 1 3. 5 1 3 5 7 0. 5 PULSE DURATION (msec. ) 3 5 7 Figure 24. Pulse duration depend pendent graded suppressive responses 368 ROLE OF THE PROSENCEPHALON IN SHIVERING to suppress shivering was greater during septal than anterior hypothalamic stimulation. The anterior h5^othalamic locus stimu- lated was within the region found to be most effective in inhibiting shivering (Andersson, 19 57; Dworkin, 1930). The septal locus of stimulation was within the ventromedial region of the midseptum. Figure 25 is a schematic representation of the effects of alter- ations in stimulus frequency on the intensity of stimulation necessary to suppress shivering. The pulse duration was maintained at 3 msec and the septal and hypothalamic locus stimulated within the pre- 2 a viously described regions. In the three cats (ST 32 , 33 , and 34) less intense stimuli were necessary to suppress shivering at stim- ulus frequencies of 25, 50, and 100 pulses/sec than at 10 and 200 pulses/sec. In two of the three experiments, stimuli of 100 pulses/ sec were less effective than 50 and 25 pulses/sec frequencies. In Cat No. ST 33 both septal and anterior hypothalamic loci were stimulated with less intense stimuli needed to suppress shivering during anterior hypothalamic than during septal stimulation. During experiments on the inhibition of shivering the dorso- medial portion of the posterior hypothalamus was routinely stimu- lated to illustrate an augmented response. It was observed that ventrolateral posterior hypothalamic stimulation suppressed shiv- ering. Two experiments on Cat No. ST 36 and 37 were conducted to compare the relative stimulus intensities necessary to suppress shivering during anterior and posterior hypothalamic stimulation to the relative stimulus intensities necessary to suppress shivering during septal and anterior hypothalamic stimulation of Cats No. ST 33 and ST 34. In Cat No. ST 34 stimulation of a ventromedial mid septal locus suppressed shivering at a stimulus intensity of 400 ju A/pulse. Stimulation of a supraoptic locus suppressed it at a stimulus intensity of 200 /u A/pulse, but stimulation of a dorsomedial posterior hj^othalamic locus augmented shivering. In all cases loci were stimulated at 5- minute intervals and the frequency of stimu- lation of 25 pulses/sec with a 3-msec pulse duration. Figure 26 illustrates the experiment on Cat No. ST 37 that confirmed the results for the experiment on Cat No. ST 36. That is to say, stimulation of anterior and ventrolateral posterior hypothal- amic loci suppressed shivering with stimulus intensities of 200 and 150 ^A/pulse respectively, while stimulation of the dorsomedial 369 STUART, D. G. SEPTAL STIMULATION CAT NO. ST. 3? A. HYPOTHALAMIC STIMULATION CAT NO. ST. 34 •^.-" 10 Zb 50 100 200 CAT NO. ST. 33 25 50 100 200 10 25 50 100 20 STIMULUS FREQUENCY (pulses/ sec. ) Figure 25. Frequency and locus dependent graded suppressive responses. 370 ROLE OF THE PROSENCEPHALON IN SHIVERING 1. EEC - L. Cortex 2. Resp. Rate 3. EEG - R. Cortex ft. EMG - L. Forelimb 5. EMG - R. Forelimb 6. EMG - L. Hindi imb 7. EMG - R. Hindi irab 8. Stimulus Duration Locus All Stimuli 25 pulses/sec, pulse duration 1 msec Locus A Stimulation-200 uA/pulse -*-**►> +-> ■JJ- m r— 1 r- r-i vD v£> CO O CM ■rH (U . . . ^ M CO CM CO CO ,_l CM ro CO CO (M 0 • > >^ u (U M m !^ >^ 2 OJ ^ o O ^ ro (M O sO vO Q -- un Tt< CO CO CO lD CM CM tn O Pi. -i-> CO <=> -, ctJ 0 in , 1 ^ ro '-^ vO U 2 CO CO CO CO CD rO ^m Jh ^1 ^ u -^ o O (t o ^ •rH ^^ H CTJ 0) rt S^i^ S^ o H -M 0) en O en 5! o e o 0^ ^ ^ 1. oS 1 oa 1 =a 1 •l-l CO 0) 0) CO :^ CO 2 CO O ^ a g.S •> ■»-> i^ -d M G J-i 0) 'rt ra CO 4J Oi u o '^ CM O en s Z ^ > Pi c w bO o ^ W r! •r-l p< .' bJO Cti 0 .t; en •r-1 a ^ ^ "» ^ ^ ^ ^ o o W) •1-1 ^ -) H in ^ CO f\J IT) •i-' U 0 n ^ o O (tJ •r^ O •*-" & ^0 ^ ^ a o 0 en -t-> O > 0 o rt cn O +J 0 Pi tt ^ G nJ o cn J " (L) § 1 •iH (ti Q} j:5 n1 U •H >H j:^ ■u o cn "tj 0) C 1) (U C C (U - CO •rH ;:! •g^ (U tn -v a > :3 0 0) a o ^ 2 ^K '1 -^^ >H o en cti T3 XI 0) p. > 0) cn -rH 0) 1-1 'c3 B n1 en (t CU X) i3 • 1-1 >< (U (ti H 0) a 0) a 0) ■u 3 +j QJ . Cti CU 0) ^H ^ fl tn •-1 > rt ^ 2 5 d tti tt ^ x: M ^H ^ H CO a; (U 0 U •3«- Pi « P. PhO 377 STUART, D. G. CAT NO S 13 Mms Anterior to Interouriculi PLANE OF FRONTAL SECTION Figure 27. Extent of neural tissue destruction in Cat No. S 13. 378 ROLE OF THE PROSENCEPHALON IN SHIVERING 15 10 5 Anterior lo Interounculor PLANE OF FRONTAL SECTION Figure 28. Extent of neural tissue destruction in Cat No. S 11. 379 STUART, D. G. 15 10 5 Mms Anterior to Interauncular line PLANE Of FRONTAL SECTION yr B Figure 29. Extent of neural tissue destruction in Cat No. S 6. 380 ROLE OF THE PROSENCEPHALON IN SHIVERING studies in that electrical stimulation of a limbic structure, the septum, was shown to affect shivering, yet following ablation of the septum, there was no measureable alteration in shivering's meta- bolic effectiveness. The stimulation data indicated with respect to the modulation of shivering, the septum could be separated into medial and lateral regions. It was also mentioned that there is as yet no anatomical basis for assuming efferent lateral septal projections to the pos- terior hypothalamus and medial projections to the anterior hj^o- thalamus. Additionally, the similar intensity of shivering with medial vs. lateral septal lesions would suggest one of three alter- natives: 1. The electrical stimulation data are faulty in suggesting a topographic organization of the septum with respect to shivering. 2. The metabolic method of measuring shivering intensity is too crude to detect subtle differences evoked by septal destruction. 3 . The septal influence on the hypothalamus is tonically "silent" only coming into play when either hyperactivated by electrical stimulation or possibly during the formation of reflexes based on classical Pavlovian conditioning or hypnotic suggestions. My opinion is that the first alternative is acceptable for the posterior septum but not for the midseptum. The second alternative is of little value in this form of experimentation in that shivering intensity waxes and wanes, increasing the oxygen consumption from two to fourfold in the same or a sequence of cats from day to day. The third alternative is the most attractive, but is subject to exper- imental proof involving the recording of electrical activity of dif- ferent septal loci in unanesthetized animals during: 1. the nonshivering state. 2. the cold-induced shivering state. 3. the conditioned reflex shivering suppressed state. 4. the conditioned reflex induced shivering state. Since the last three have neverbeen attempted, the task, though formidable, appears of considerable value in elucidating the role of the telencephalon in such a primarily hypothalamic controlled phen- omenon such as shivering. 381 STUART, D. G. SUMMARY The results may be summarized as follows: (1) Decerebrate cats cannot shiver in the cold but can make tremulous spasmodic movements of limitedmetabolic effectiveness during rapid cooling. Such movements are more a response to a nocioceptive stimulus than a form of organized temperature regu- lating response. (2) The intensity of shivering in decorticate cats is depressed for a short time after surgery even when such animals are "auto- nomically" hyperactive. Within four post -operative weeks the autonomic hyperactivity ablates and shivering returns to its pre- operative intensity. It is thus concluded that the net telencephalic influence on shivering is not inhibitory but either a balance of inhibiting and facilitating influences or neither. (3) Normal shivering involves the integrity and activation of the dorsomedial posterior hypothalamus. Cutaneous vasoconstric- tion is controlled by neurons within the dorsolateral posterior hypothalamus. (4) Shivering can be both instigated and suppressed by septal stimulation with hi^er stimuli intensities than are necessary to evoke or suppress shivering during hypothalamic stimulation. (5) Shivering can be suppressed by ventrolateral posterior hypothalamic stimulation, and it seems that this suppression is mediated at a more caudal level than the hypothalamus. These results suggest that septal modulation of shivering is secondary to hypothalamic control of the function. There are prob- ably other secondary control systems within the telencephalon that can also facilitate and suppress shivering by modify inghypothalamic activity. This is suggested by the fact that not all telencephalic 382 ROLE OF THE PROSENCEPHALON IN SHIVERING projections to the hypothalamus traverse the septum. I would like to suggest that study of these secondary telencephalic control systems is as important as study of the primary control systems in that it may leadto an understanding of how man's attitudes, moods and emotions can modify temperature regulatory mechanisms. Obviously many experiments must be designed and executed before we can hope to understand the neurology of temperature regulatory responses evoked by hypnosis and classical Pavlovian conditioning. In such a context, the work here reported with respect to septal versus hypothalamic control of shivering is a minute aspect of the much broader problem of how the higher nervous system can modulate functions primarily controlled by the hypothalamus. That these results have implicated the dorsomedial region of the posterior hypothalamus in the production of shivering and the dorsolateral region in cutaneous vasoconstriction is of value in offering an explanation of the many seemingly diverse results in the literature. However, it tells us little more about the physiology of shivering. However, by localizing control of shivering toaspecific region of the brain, it should be possible to implant micro-electrodes and monitor the activity of single neurons whose function is related to the production of shivering. If we can next discover cells within the hypothalamus whose role is blood temperature detection and if we can further find cells in the thalamus and/or hypothalamus whose role is detection of skin tern peratur.;, then we can begin to study the adequate physiological stimulus necessary to evoke shivering. Obviously before one can study the neurogenesis of a function, the neural regions involved in affective and effective aspects of the function must be localized. It is in this context that these experi- ments reported today may be of somevalueto a future investigation of the neurogenesis of shivering. In conclusion I would like to thank the Arctic Aeromedical Lab- oratory forthe privilege of attending and address ingthis symposium. The aim of Dr. Hemingway's UCLA group has been to utilize a variety of neurophysiological techniques in studies of the physiology of body temperature regulation. It has been a privilege to work in this laboratory under Dr. Hemingway's guidance, which is deeply 383 STUART, D. G. appreciated. Dr. Walter Freeman's encouragement in the tran- section experiments is also appreciated, as are Dr. Yojiro Kawamura's most valuable contributions to the design and execution of the stimulation experiments. 384 ROLE OF THE PROSENCEPHALON IN SHIVERING DISCUSSION DR. FREEMAN: You are familiar with the experiments in which electrodes are implanted in the septal and hypothalamic areas whereby animals can be induced to stimulate their own brains, as power to avoid this self stimulation. I wonder, one often hears phrases as "trembling with joy "and "quaking with fear." How do you think such manifestations of tremor activity are related to the shivering process? DR. STUART: First, we must distinguish between pathological tremors that are 4 to 7 cycles/sec, in which the agonist- antagonist activity is alternating, and the 9 to 11 cycles/sec shivering tremor, in which the agonist-antagonist activity is synergistic. Physiological tremor (e.g., finger tremor) is also 9 to 11 cycles/sec in the adult but slower in youngsters, and to my knowledge there is no infor- mation as to whether or not it is an alternating or synergistic tremor. I do not think anyone has ever subjected "trembling with joy" or "shaking with fear" to neuromuscular examination, but I would think they are more closely allied to hypothalamically- induced shivering than to non- hypothalamically induced pathologic tremor. There is, of course, one difference in that shivering accomplishes something for the animal (i.e., an increase in heat production with- out an increase in external work), whereas the tremors you mention seem of little biological value. I would like to add that Dr. Kawamura and I have produced alternating tremor by stimulation of a prosen- cephalic locus that is anatomically different from those loci whose stimulation evoked shivering. Additionally, Dr. Hemingway, Dr. George, and I recently completed some experiments that demon- strated that reserpine blocks shivering but produces an alternating tremor, whereas atropine, which is known to block pathologically induced alternating tremor, has no effect on shivering. DR. FREEMAN: In other words, you draw a period of distinction between pathological tremors which are not related to shivering and other physiological tremors? 385 STUART, D. G. DR. STUART: Yes, but the differences between physiological tremor (microvibration) and shivering are more obscure. My personal feeling (subject to experimental confirmation) is that shivering is a cold- induced exaggeration of the amplitude of the neurological component of physiological tremor. Dr. Earl Eldred and 1 are particularly interested in patterns of alpha and gamma motorneuron and muscle receptor discharges during diverse tremors. The literature on some aspects of this will be reviewed shortly (Stuart, D. G., E. Eldred, and Y. Kawamura. Neural regu- lation of the rhythm of shivering. In "Temperature - Its Regulation and Control in Science and Industry. " C. M. Herzf eld (ed.) Washing- ton, Reinhold Publishing Corp. In press 1961). I feel your question is most pertinent but cannot be answered satisfactorily until more experimental evidence has accumulated. DR. CLARK: I would like to mention the figure that I showed of one of Keller's dogs. It has a lesion that fits in beautifully with what you have been saying because it went much further dorsal medially than it did laterally, so it would spare those Crosshatch areas you have outlined in your drawing but would have hit the medial ones. DR. STUART: With respect to the role of the hypothalamus in shivering, 1 believe that the data from our laboratory are in agree- ment with Keller's data. DR. CLARK: And that dog, of course, could pant but could not shiver. DR. MINARD: I would like first of all to express my great appreciation to the laboratory and to the speakers for a very illum- inating two and a half days. There are two statements I would like to make just to throw them out for possible criticism or disagree- ment; and the reason I am doing this is because there will be some studies reported from the Naval Medical Research Institute on humans in which these two statements are fairly basic in the inter- pretation of the results. The first of these statements is that the posterior hypothalamus is blind to temperature; that is, that it has not been possible to elicit responses by changing the temperature of the posterior hypothalamus. The second of the two statements is that in a shivering animal, heatingof the anterior hypothalamus will 386 ROLE OF THE PROSENCEPHALON IN SHIVERING inhibit shivering. Now, if these two statements are incorrect, I should like to know about it now, before I take this information back to the laboratory . DR. STUART: In reply toyour second statement, a report of the suppression of shivering by heating the anterior hypothalamus is not new. It was reported by Magoun and co-workers in 1938, by Hemingway and co-workers in 1940, and more recently by Strom and co-workers, and Freeman and Davis. Today Ihave reported the suppression of shivering during septal, anterior and posterior hy- pothalamic stimulation. The suppression during septal stimulation and the comparison of stimulus intensities at the three loci is new information, but the suppression during electrical stimulation of the anterior hypothalamus confirms the previous work of Hemingway, Forgrave and Birzis and Andersson, Grant and Larsson. Therefore, there is nothing in this work to conflict with your second statement. Your first statement that the posterior hypothalamus is blind to temperature has no direct relation to these experiments which are concerned with localization of neural regions relatedtothe efferent (motor) arm of shivering rather than the reception of temperatures. However, I would like to comment on your statement to the extent that it relates to a neurophysiolgical problem and involves neuro- physiological techniques of investigation. The experimental evidence on which you base your statement involves, I believe, gradient calorimetiy and very accurate measurements of temperature at various body sites. Such experiments have obviously been of great value, but before accepting the fact that the posterior hypothalamus is blind to temperature, I would expect reasonable experimental evidence showing that the firing pattern of single hypothalamic neurons is reversibly altered by cooling and warming, and evidence that such neurons are in the anteriorhypothalamus^and not the pos- terior hypothalamus. Since an animal with anterior hypothalamic lesions can shiver, it migjitwell be that third or second order skin temperature neurons impinge upon posterior hypothalamic neurons and that their discharge is capable of instigating and maintaining shivering. None of these things have ever been demonstrated and they will, I believe, involve microelectrode experimentation. DR. FREEMAN: The only evidence that I know that bears direct- ly on this is the series of attempts that Davis and I made to heat and 387 STUART, D. G. cool directly in the posterior hypothalamus as distinct from the anterior hypothalamus. The results we got were not as clear-cut by any means as results we got from the stimulation anteriorly. I believe one can sum the case up by saying that the term you used, "blind," is not adequately descriptive. One must think of sensitivity to thermal changes in two contexts: one in which the sensitivity is either specifically related to a sense organ or sensory receptor as Dr. Hensel described, and the other in which any physical event can alter neuro-function. Presumably, these nerve cells in the posterior hypothalamus are like any others. If youcoolthem off enough, their activity will be impaired. If you heat them up enough, it may be in- creased for a while, but eventually it will be impaired. Therefore, in the physiological range ofbodytemperature or brain temperatures of 35 C to 41 C, one would be willing to say, under most circum- stances, there is no thermosensitivity, but outside of this range, there may be critical thermosensitivity in the sense that gross im- pairment of function of these cells may take place. DR. MINARD: Thank you very much. DR. HEMINGWAY: I may add one thing to that. Many years ago as Dr. Stuart mentioned, I heated the anterior hypothalamus with surface electrodes, and alsotheposterior hypothalamus. Heating the anterior hj^othalamus would cause instantaneous cutaneous vaso- dilatation and shivering would stop, but there was no effect whatso- ever on temperature regulation when the posterior hypothalamus was heated. One interesting thing occurred, which I am not sure was significant: the animals went to sleep immediately upon heating the posterior hypothalamus, but there is no rectal temperature regulation. DR. HENSEL: We found the same result as did Dr. Freeman. As yet, we have seen only these reactions during cooling the an- terior hypothalamus, but I would agree completely with your state- ment that there is no tissue in the body which is completely blind to temperature, of course. It is a matter of qualitative activity and of the direction. Some are excited and some are inhibited. DR. MINARD: You might say compared to the anterior hypo- thalamus, it is relative. 388 ROLE OF THE PROSENCEPHALON IN SHIVERING DR. HENSEL: You can see it also in the receptors. I think there are more receptors reacting to temperature than blindto tempera- ture, but the question is the quantitative sensitivity. DR. HEMINGWAY: May I add one more thing? This work of Stuart's was designed to explain the findings of Andersson. He did find this interesting thing, that in the posterior hypothalamus there is a region which when stimulated produces shivering and it is much more sensitive than any other part of the hypothalamus. But in the septal region where Andersson was working, it is possible to find both inhibition and facilitation of shivering. And, an inter- esting thing, of course, you can take out this entire septal region with no effect on shivering. 389 7, STUART, D. G. REFERENCES 1. Adams, T. and B. C. Covina. Racial variations to cold stress. J. App. Physiol. 12:9, 1958. 2. Akert, K. and F. Kesselring. Kaltezittern als Zentrale Reiz- effect. Helv. physiol. Pharmacol. Acta 9:290, 1951. 3. Andersson, B. Cold defense reactions elicited by electrical stimulation within the septal area of the brain in goats. Acta physiol. scandinav. 41:90, 1957. 4. Andersson, B., R. Grant, and S. Larsson. Central control of heat loss mechanism in the goat. Acta Physiol. Scandinav. 37:261, 1957. 5. Andersson, B. and N. Persson. Pronounced hypothermia elicited by prolonged stimulation of the "heat loss centre" in unanes- thetized goats. Acta Physiol. Scandinav. 41:277, 1957. 6. Andersson, B., N. Persson and L. Strom. Post- stimulatory electrical activity in the pre-optic "heat loss centre" con- comitant with persistent thermoregulatory response. Acta Physiol. Scandinav. 50:54, 1960. Aring, C. D. Shivering and the cerebral cortex. Am. J. Physiol. 113:3, 1935. 8. Ban, T. and F. Omakai. Experimental studies in the fiber con- nection of the amygdaloid nuclei in the rabbit. J. Comp. Neurol. 113:245, 1959. 9. Barbour, H. G. Hypothalamic control of water movement in response to environmental temperatures. Ass. Res. Nerv. and Ment. Dis. 20:449, 1939. 390 ROLE OF THE PROSENCEPHALON IN SHIVERING 10. Bard, P. Temperature regulation in chronically decorticate cats and dogs and in decerebrate cats. In Temperature, American Institute of Physics, to be published in 1961. 11. Bard, P. and M. B. Macht. The behavior of chronically decere- brate cats. Chap. 4 in Neurological Basis of Behavior (edited by G. E. W. Wolstenhohne and C. M. O'Connor) Little, Brown, and Co., Boston, 1958. 12. Bazett, H. C. and W. G. Penfield. A stucfy of the Sherringtonian decerebrate animal in the chronic as well as the acute con- dition. Brain 45:185, 1922. 13. Bazett, H. C, B. J. Alpers, and W. H. Erb. Hypothalamus and temperature control. Arch. Neurol, and Psych. 30:72 8, 1933. 14. Benzinger, T. H. The sensory receptor organ and quantitative mechanism of human temperature control in warm environ- ment. Fed. Proc. 19, Suppl. 5:32, 1960. 15. Birzis, L. and A. Hemingway. Descending brain stem con- nections controlling shivering in cats. J. Neurophysiol. 19: 37, 1956. 16. Birzis, L. and A. Hemingway. Shivering as a result of brain stimulation. J. Neurophysiol. 20:91, 1957. 17. Birzis, L. and A. Hemingway. Efferent brain discharge during shivering. J. Neurophysiol. 20:156, 1957. 18. Blair, J. R. and A. D. Keller. Complete and permanent elim- ination of the hypothalamic thermogenic mechanism without affecting the adequacy of the heat loss mechanisms. J. Neuropath. 5:240, 1946. 19. Bond, D. D., T. Clark, M. D. Randt, G. Bidder, and V. Rowland. Posterior septal fomical and anteriorthalamic lesions in the cat. Arch. Neurol, and Psych. 78:143, 1953. 391 STUART, D. G. 20. Bykov, K. M, The Cerebral Cortex and the Internal Organs. (translated and edited by W. H. Gantt) Chemical Publishing Co., New York, 1957. 21. Carpenter, M. B. The neuro-anatom ical basis of dyskinesia. Chap. II, Pathogenesis and treatment of Parkinsonism, (edited by W. S. Fields), Chas. C.Thomas, Springfield, 1958. 22. Chatonnet, J. Nervous pathways in the chemical regulation against cold. This symposium. 23. Clark, G., H. W. Magoun, and S. W. Ranson. Hypothalamic regulation of body temperature. J. Neurophysiol. 2:61, 1939. 24. Dusser de Barenne, J. G. Recherches experimentales sur les fonctions du systeme nerveaux central faites en particulier sur deux chats dont le neopallium avait ete enleve. Arch, neerland. Physiol. 4:31, 1920. 25. Dworkin, S. Observations on the central control of shivering and heat regulation in the rabbit. Am. J. Physiol. 93:227, 1930. 26. Eliasson, S. and G. Strom. On the localization in the cat of hypothalamic and cortical structures influencing cutaneous blood flow. Acta Physiol. Scandinav. 20, Suppl. 70:113, 1950. 27. Folkerts, J. F. and E. A. Spiegel. Tremor on stimulation of the midbrain tegmentum. Neurol. Conf. 13:193, 1953. 28. Fox, C. A. Certain basal telencephalic centers in the cat. J. Comp. Neurol. 72:1, 1920. 29. Fox, C. A. The striaterminalis, longitudinal association bundle and precommissural fornix fibers in the cat. J. Comp. Neurol. 79-.277, 1943. 30. Frazier, C. H., B. J. Alpers, and F. W. Lewy. The anatomical localizations of the hypothalamic center for the regulation of temperature. Brain 59:122, 1936. 392 ROLE OF THE PROSENCEPHALON IN SHIVERING 31. Freeman, W. J. and D. D. Davis. Effect on cats of conductive hypothalamic cooling. Am. J. Physiol. 197:145, 1959. 32. Fuster,J. M. and S. J. Weinberg. Bioelectrical changes of the heart cycle induced by stimulation of diencepiialic regions. Exp. Neurol. 2:26, 1960. 33. Gessler, H. and K. Hansen. Uber die suggestive Beeinflussbar- keit der Warmeregulation in der Hypnose. Dtsch. Arch. f. Klin. Med. 156:352, 1927. 34. Gloor, P. Telencephalic influences on the hypothalamus. Chap. VI in Hypothalamic-Hypophysial Interrelationships (edited by W. S. Fields) Chas. C. Thomas, Springfield, 1956. 35. Hall, E. Efferent pathways of lateral and basal nuclei of the amygdala in the cat. Anat. Rec. 136:205, 1960. 36. Hammel, H. F., J. D. Harcfy, andM.M. Fusco. Thermoregula- tory responses to hypothalamic cooling in unanesthetized dogs. Am. J. Physiol. 198:481, 1960. 37. Heath, R. G. Studies in Schizophrenia. Harvard University Press, Cambridge, 1953. 38. Hemingway, A., P. Forgrave, and L. Birzis. Shivering sup- pression by hypothalamic stimulation. J. Neurophysiol. 17: 375, 1954. 39. Hemingway, A., T. Rasmussen, H. Wickoff, andA.T. Rasmus- sen. Effects of heatingthe hypothalamus of dogs by diathermy. J. Neurophysiol. 3:329, 1940. 40. Hensel, H. Recent advances in thermoreceptor liiysiology. This symposium. 41. Hess, W. R. The Functional Organization of the Diencephalon. Greene and Stratton, New York. 1957. 393 STUART, D. G. 42. Isenschmid, R. and L. Krehl. Uber den Einfluss des Gehims auf die W'arraeregulation. Arch. f. exper. path. 70:109, 1912. 43. Isenschmid, R. and W. Schnitzler. Beitragzur Lokalizationdes der Warmeregulation verstehenden Zentralapparatus in Zwischenchiim. Arch. exp. path, pharmak. 76:202, 1914. 44. Jacobson, F. H., P. H. Craig, and R. D. Squires. Cessation of shivering following forebrain lesions in anesthetized cats. Fed. Proc. 19:44, 1960. 45. Jenkner, F. L. and A. A. Ward, Jr. Bulbar reticular formation and tremor. Arch. Neurol, and Psych. 70:489, 1953. 46. Kaada, B. R. Somatomotor, autonomic and electrocortico- graphic responses toelectrical stimulation of "rhinencephal- ic" and other structures in primates, cat and dog. Acta Physiol. Scandinav. 24, Suppl. 83, 1951. 47. Keller, A. D. The separation of heat loss and heat production mechanisms in chronic preparations. Am. J. Physiol. 113: 78, 1935. 48. Keller, A. D. Neurologic ally induced physiological resistance to hypothermia. Ann. N. Y. Acad. Sc. 80:457, 1959. 49. Keller, A. D. and W. K. Hare. The hypothalamus and heat regulation. Proc. Soc. Exp. Biol. Med. 29:1069, 1932. 50. Koikegami, H., H. Hiroshi and A. Kimoto. Studies of the func- tions and fiber connections of the amygdaloid nuclei and periamygdaloid cortex. Experiments on gastrointestinal motility and bocty temperature in the cat. Fol. Psych, and Neurol. Japonica 6:76, 1952. 51. Kundt, H. W., K. Bruck and H. Hensel. Das Verhalten den Hautdurchbluting bei Kuhlung des Vorderen Hypothalamus. Naturwissenschaften 44:496, 1957. 394 ROLE OF THE PROSENCEPHALON IN SHIVERING 52. Lauer, E. W. The nuclear pattern and fiber connections of certain basal telencephalic centers in the Maccaque. J. Com p. Neurol. 82:215, 1945. 53. Liebermeister, K. PhysiologischeUntersuchenuberdiequanti- tativen Veranderungen der Warmeproduction. Arch, f . Anat. u. Physiol, p. 520, 1860. 54. Magoun, H. W., F. Harrison, J. R. Brobeck, and S. W. Ranson. Activation of heat loss mechanisms by local heating of the brain. J. Neurophysiol. 1:101, 1938. 55. McLean, P. D. The limbic system with respect to two basic life principles. The Central Nerv. System and Behavior, Transactions of the Second Conference (edited by M. A. B. Brazier), Josiah Macy Jr. Foundation, New York, 1959. 56. McLean, P. D. and J. M. R. Delgado. Electrical and chemical stimulation of fronto-temporal portion of limbic system in the waking animal. E. E. G. and Clin. Neurophysiol. 5:91, 1953. 57. Nauta, W. Hippocampal projection and related neural pathways to the mid-brain of the cat. Brain 81:207, 1958. 58. Ott, I. The thermogenic center in the tuber cinerum. Med. BuU. 17:246, 1895. 59. Pinkston, J. O., P. Bard, and D. Rioch. Responses to changes in environmental temperature after removal of portions of the fore-brain. Am. J. Physiol. 109:515, 1934. 60. Pribram, K. H. and L. Kruger. Functions of the "olfactory brain". Ann. N. Y. Acad. Sc. 58:109, 1954. 61. Ranson, S. W. andH. W. Magoun. The hypothalamus. Ergebn. Physiol. 41:56, 1939. 62. Richet, C. La frisson comme appareilde regulation thermique. Compte. Rend. Soc. de Biol. 44:896, 1892. 395 STUART, D. G. 63. Scholander, P. F., H. T. Hammel, J.S.Hart, D. H. Messurier, and J. Steen. Cold adaptation in Australian aborigines. J. Appl. Physiol. 13:211, 1958. 64. Sprague, J. D. and M. Meyer. An experimental study of the for- nix in the rabbit. J. Anat. 84:354, 19 50. 65. Wycis, H. T., E. G. Szekely and E. A. Spiegel. Tremor on stimulation of the mid-brain tegmentum after degeneration of the brachium conjunctivum. J. Neuropath, and Exp. Neurol. 16:79, 1957. 66. Wyndham, C. H. and J. F. Morrison. Adjustment to cold of bushmen in the Kalahai Desert . J. Appl . Physiol . 13 :2 19 , 19 58 . 396 ROLE OF THE PROSENCEPHALON IN SHIVERING EPILOGUE DR. HANNON: There is one final thing in our Symposium. We have been talking a lot about the maintenance of body temperature. Dr. Irving, from Arctic Health Research Laboratory, has brought out the term "peripheral heterothermy." Would you care to give us a few words, Dr. Irving? DR. IRVING: First, I want to express my thanks to Colonel Quashnock for the gracious manner in which he represented himself and the Air Command as our host, to Pat Hannon for his kindness in inviting me here for the extremely interesting program that has been presented, and to you, my friends and colleagues, I want to express my gratitude for having been introduced through your dis- cussions to subjects which were rather strange tome, for I usually look only at the periphery of animals. You have indicated some re- markable physiological complexities and some of the methods by which the warm-blooded animal system for communicating informa- tion is able to apply its very large capability for converting energy to the purpose of maintaining its individuality and specificity. I think that references to warm-blooded animals, as distinct from those that are cold-blooded, should emphasize the large order of the disposable conversion of energy that the warm-blooded ani- mals have in comparison with cold-blooded animals. I think we should also consider that from time to time, the central body tem- perature of the warm-blooded animals may rise three degrees or so, and that it regular ly falls half a degree or a degree during sleep. Schmidt- Nielsen's thirsty camels even elaborated this ability to modify the body's temperature to a cooling of some six degrees during the night time, making a total range of nine or ten degrees diumally in an undoubtedly homeothermous animal. Animals which are covered with fur and fat utilize their insu- lation. These materials are, however, inflexible insulators and animal producers of heat must utilize some variable insulator for the dissipation of heat. In cold climates this is effected by large variations in the temperature of the exposed extremities. One of the interesting examples of heterothermous tissues that John Krog and I observed was the foot of the seagull which, while 397 STUART, D. G. o its body temperature was around 41 C, maintained visibly active circulation in the thin web of the foot when it was kept on a cold plate so that the tissue temperature recorded by a thermocouple was close to freezing. We have a number of examples of that sort: the tail of the muskrat which John Krog and I observed and which Kjell Johansen plans to observe further, cools in ice water to near that temperature while the central part of the animal's body is 38 C warmer. There are other examples of this variation in tissue tempera- ture according to topographical anatomy as well as according to time. In recentworki studied a cold-acclimatized student who wears only a thin robe in Alaska, winter or summer. I was able to observe him last winter as he sat for an hour in a cold freezing room with only the li^t clothing that he had on when he was here yesterday. During much of that time, his toes and those of a colleague of his similarly lightly clothed remained at temperatures below ten degrees, and yet during that time they did not complain of pain. By their report of sensitivity to touch and their report of warming one toe as compared with another during the cyclic rewarming process, it was indicated that tolerance of cold was accompanied not by in- sensitivity but perhaps by even refined sensitivity and careful monitoring, both conscious and unconscious, of the thermal state in their tissues. What happens to the information system ope rating with therm o- labile components? I have recently been making observations on the detection of impact of small drops of mercury falling through a given distance and found that after a bit of practise I could get a regular threshold for the perception ofthe kinetic energy of the just detectable impact when my finger was at 35 C. But when the finger was cooled to about 22 C, the kinetic energy required to produce a detectable impact was elevated some five or six times. For some weeks in successive trials, the relation between temperature and threshold remained regular. We are thus presented with a problem incidental to thermal regulation which may provide a valuable clue to the nature of the communicating system by which animals maintain their integrity: how is it that the communication is effected through an extraordin- 398 ROLE OF THE PROSENCEPHALON IN SHIVERING axily labile system, and yet ends up with retention of central in- formation that is consistent with the steady existence of individuals and species. I mention these things by way of digression from the theme of your program because in a meeting like this, I suppose that we do not have to end up with a consensus of opinion. I believe that we have rather complete accord as to the value of many different opinions about physiology, and 1 find that pleasant, stimulating, and hopeful. 399 212-63