Three groups of workers (Heller, Hasan & Saifi, 1968; Vorherr, Bradbury, Hoghoughi & Kleeman, 1968; Unger & Schwarzberg, 1970) have found that stimuli which are known to release endogenous vasopressin (vagal stimulation, haemorrhage, barbiturate anaesthesia) raised the antidiuretic activity (ADA) in the cerebrospinal fluid (CSF) of experimental animals (rabbits, dogs). The nature of the stimuli, the fact that the ADA was abolished by thioglycollate and the recent finding of neurophysin in the CSF (Robinson & Zimmerman, 1973) indicate strongly that the ADA was exerted by vasopressin. However, it is not clear whether the hormone is directly secreted into the CSF by juxtaependymal processes of neurosecretory neurones as suggested by morphological evidence (for references see Heller et al. 1968; Heller & Zaidi, 1974) or whether it reaches the CSF from the blood. Heller et al. (1968) and Vorherr et al. (1968) have tried to solve this problem by administering vasopressin intravenously