P la te 4 F ig u r e 5. Museum on second floor of south building.F ig u r e 6. Typical work room in south building. P la te 5 F ig u r e 7. R eading room on first floor of library.F ig u r e 8. Corner of physiological laboratory. P la te 6 F ig u r e 9. Chemical laboratory.F ig u r e 10. Y ard between north and south buildings showing reservoirs for sea w ater on left, outside tan k s and circulation bench on right, and end of E aster Course building in distance on left. P late 7F ig u r e 11. General view of aquarium . F ig u r e 12. Anemone ta n k in aquarium . P late 8 F ig u r e 13. R.V. Sabella.

THAT emotional stress may be the occasion for a fall in the rate of urine-secretion, even to the degree of total suppression, has been recognised since the reference by Bernard [1859] to certain cases of vesico-vaginal fistula operated upon by Jobert and in which "par suite de l'emotion, l'ecoulement de l'urine avait ete suspendu pendant toute la duree de l'operation, et quelquefois meme bien au dela." Bernard makes this reference in connection with the darkening of the blood in the renal veins which he observed to accompany the inhibition of urine-flow during severe abdominal operations on the dog and rabbit: the suggestion is implicit, therefore, that the inhibition following emotional stress has likewise a vasomotor cause. The fact, however, that an effective stress may be of apparently small degree [Mackeith, Pembrey, Spurrel, Warner, and Westlake, 1923;Theobald, 1934] and the consideration that a reduced blood-flow through the kidney forms, on teleological grounds, an unattractive accompaniment, have been partly responsible for recent attempts to discover the incidents of the phenomenon of inhibition and to disclose their accidental or essential relation with it. The demonstration of the persistence of the phenomenon in the dog after section of the renal nerves, and indeed of all organic connection between the animal and its kidney [Theobald and Verney, 1935], has now excluded these nerves from functional participation, and in so doing has testified to intervention through humoral channels. Reasons, moreover, have been advanced [Theobald and Verney, 1935] against this intervention's being in the nature of a change in the partition of water between blood and tissues: a humorally transmitted agent has therefore been postulated, and its identity with adrenaline made unlikely by the dissimilarity between the course of inhibition due to emotion and that due to an intravenous injection of this drug.The view that the agent is identical with the antidiuretic substance of post-pituitary extract, however [Theobald and Verney, 1935], is supported on the one hand by the superficial resemblance of the courses of inhibition from emotional stress and the intravenous 1 Fellow of the Rockefeller Foundation.

No abstract

The object of this investigation has been to define the site of the osmoreceptors, a term that has been applied to those hypothetical sensory elements that respond to changes in the osmotic pressure of their vascular environment, and through which the release of antidiuretic hormone from the neurohypophysis is physiologically regulated. Confirmation is given to the cephalic localization of these receptors, and an attempt has been made to discover where, within the substance of the brain, they reside. The test of their activation has been the inhibition of urine flow by intracarotid infusions of hypertonic solutions during established water-diuresis. By surgical means the cephalic field of distribution of the carotid arteries in the dog has been restricted to defined regions of the brain, and information has thereby been acquired on the osmoreceptive status of these regions. In this way it has been contrived that the evidence for the localization of the osmoreceptors should rest solely on the use of a blood-borne physiological stimulus appropriate to the sensory elements in question. The earlier sections of the paper present the results of the broader studies that this work has necessitated; first, an anatomical investigation of the arterial connexions of the circle of Willis and of the detailed vascular architecture of the diencephalon and hypophysis; and secondly, the devising of a method to trace the distribution of arterial blood and the application of this method to the demarcation of the cerebral fields of carotid and vertebral arteries. The middle sections of the paper describe the surgical and experimental techniques employed, and the manner in which the posterior lobe of the pituitary gland, the posterior brain stem and the cerebral hemisphere have been excluded from being the site of receptors. Also described are experiments in which the permanent unilateral suppression of responses following intradural ligation of an internal carotid artery has placed beyond cavil the assertion that the receptors he within the substance of the brain. The final sections present the evidence for the hypothalamic localization of the osmoreceptors and for the suggestion of a possible involvement of the thalamic paraventricular nucleus in the osmoreceptive process. , To trace the distribution of arterial blood two coloured suspensions have been employed. The suspensions are freely miscible with blood, cause no circulatory disturbance in the living animal during the period of their infusion, and tolerate histological procedures. The suspensions are infused into appropriate arteries at a terminal experiment in which the animal is killed before the suspensions have had time to recirculate. In the dog the brain is normally supplied with blood from both carotid and vertebral arteries. The telencephalic field of the carotid arteries is that part of the hemisphere supplied by the anterior and middle cerebral arteries and includes the striate body; the vertebral arteries supply the hippocampus and posterior cerebral artery field, the midbrain, cerebellum and medulla. However, this strict partition is defied by two seemingly normal occurrences: common carotid blood may pass via the occipito-vertebral anastomosis to join the basilar stream, and vertebral blood may pass forward from the posterior communicating artery to mix with the carotid cerebral supply. The thalamus is supplied almost exclusively by vertebral blood that reaches it via branches of the posterior cerebral artery and the thalamic branch of the posterior communicating artery. The hypothalamus is divided in the sources of its supply: carotid blood irrigates the anterior nuclei via arterioles arising directly from the internal carotid and the immediate vicinity of its trifurcation; vertebral blood supplies the posterior nuclei via arteriolar branches of the posterior communicating artery, but, in addition, this blood may stream forward to supplement the carotid supply to the pre-infundibular nuclei. Situated near that part of the circle where the carotid and vertebral streams meet are the main nuclear groups of the neurohypophysis, the supraoptic nuclei. Particular attention has been paid to the volume partition of these nuclei according to the origins of their blood supply, in order to gain an index of blood distribution in the anterior hypothalamus. In the normal animal between 10 and 30 % of the total supraoptic nuclear material is supplied with blood of vertebral origin. The naturally occurring asymmetry of the arterial supply to the posterior lobe of the pituitary gland has led to the elimination of this structure as a site of osmoreceptors. In each of a total of eight animals it was found that the posterior lobe was supplied with blood originating exclusively from one carotid, yet an osmotic release of antidiuretic hormone had been obtained from infusions of hypertonic solutions into the other carotid. The exclusion of carotid blood from the posterior brain stem has been achieved by ( the ligation of the two occipital arteries, and, on one side, by ( b ) the ligation of the occipital and posterior communicating arteries. Osmotic responses to carotid infusions were retained after these procedures and thus made secure the assignment of the receptors to the prosencephalon. Of prosencephalic structures the greater part of the telencephalon has been eliminated from being the receptor site by demonstrating the ipsilateral retention of responses after total hemispherectomy. This conclusion was supported by evidence from animals in which the carotid bloods, while evoking osmotic responses, were asymmetrically distributed within the telencephalon. The heavy degenerative cell loss in all thalamic nuclei of the hemispherectomized animal, excluding some cell groups of the midline, substantiated earlier indications that the receptors were not in the dorsal diencephalon. Alternative and equally compelling evidence for this conclusion was obtained from experiments in which carotid blood was directly excluded from the thalamus. Such exclusion was attained by tying the posterior communicating and occipital arteries of one side together with the middle cerebral or anterior cerebral artery. It was also attained in experiments in which the internal carotid of one side was ligated intradurally, this resulting in a redistribution of blood from the circle such that the thalamus of one side was deprived of any carotid flow. These same experiments afforded, too, unequivocal evidence for the exclusion of the posterior nuclear groups of the hypothalamus as a possible site for the receptors. The collation of the several evidences summarized above has led to the inference that the osmoreceptors are situated somewhere in the anterior hypothalamus or preoptic areas, that is, in the region comprised by the medial and lateral preoptic areas, the suprachiasmatic nucleus, the nucleus supraopticus diffusus, the anterior hypothalamic area, the paraventricular and supraoptic nuclei, the dorsomedial and ventromedial nuclei, and the dorsal and lateral hypothalamic areas. This region has always received carotid blood when there have been osmotic responses to intracarotid infusions. The restriction of the cephalic distribution of the carotid of one side of this receptive zone was successfully achieved by ligature of the anterior cerebral, middle cerebral and posterior communicating arteries just beyond the carotid trifurcation, together with occlusion of the ipsilateral occipital artery. The sequel, however, was disappointing in that osmotic responses were lost on that side; but the experiment provoked conjecture upon the possible neuronal organization of the osmoreceptive apparatus, since it was found that in this animal the operation had caused cystic destruction of the thalamic paraventricular nucleus together with all other anterior and medial thalamic nuclei. In all previous animals, including the hemispherectomized one, in which osmotic responses had been retained, the thalamic paraventricular nucleus and its connexions with the hypothalamus had remained intact. The conclusion is drawn that the osmoreceptors are situated in the anterior hypothalamus. There are indications that they are not of unvarying sensitivity, and that their functioning may be dependent upon the integrity of nervous connexions with the thalamic paraventricular nucleus.

1. Hypertension has been produced in dogs by obstructing the blood‐supply to their kidneys by a modification of Goldblatt's technique. Measurement of the arterial pressure was made by the application of a specially designed cuff to the van Leersum loop, the ipselateral carotid sinus having been denervated. 2. The time‐course of the blood‐pressure changes, and its correlation with structural alterations in the kidneys have been investigated. Results obtained from (a), obstruction of one renal artery only, followed later by the removal of the ischaemic kidney, (b) from complete nephrectomy, and (c) from the production of renal insufficiency in the absence of ischaemia, have led to the conclusion that the presence of ischaemic renal tissue is an indispensable condition for the development of hypertension. 3. The time‐course of development and disappearance of hypertension have been measured by a method which disposes of interference by surgical procedures under general anæsthesia. 4. In confirmation of the work of others, the ischæmic kidney was found to produce hypertension even if completely denervated. It undoubtedly did so, therefore, by the formation of some substance which escaped into the general circulation. 5. Sympathectomised dogs reacted like normal dogs to obstruction of their renal arteries. Any action of the hypertensive substance involving stimulation of the sympathetic chains or of the central connections of the sympathetic system was thus excluded. 6. Experiments in which an isolated kidney and a loop of small intestine were each perfused by a heart‐lung preparation have shown that the perfused kidney liberates a substance which produces vasoconstriction in the gut. It is assumed as a working hypothesis that this phenomenon is related to the hypertension observed in the whole animal under the conditions of renal ischaemia. 7. The sensitivity of hypertensive dogs to injected adrenaline was occasionally, that to tyramine more regularly increased, whereas decrease in the normal sensitivity to tyramine accompanied the hypertension produced by a constant infusion of tyramine. This and other evidence favour the view that the hypertension which accompanies renal ischsemia is not produced by excess tyramine in the circulation. 8. The reactivity of the carotid sinuses was unaltered in the hypertensive dogs. 9. Removal of one kidney after earlier obstruction of both renal arteries resulted in a further rise of pressure. The same held if, after obstruction of one renal artery only, the sound kidney was removed. Increased load, therefore, raised the hypertension level, even if the amount of ischaemic tissue remained the same or was diminished. 10. Increased dietary loads elicited reversible rises in the bloodpressure level of hypertensive dogs. In the dosage given, urea and meat were found equally potent; sodium chloride was still more effective. Excess of sodium chloride produced cedema in one dog, and convulsive “uraemia” in another. 11. Whein interference with the renal blood‐supply was very severe, a syndro...