That diabetes insipidus in man is caused by lesions in the environs of the hypothalamus is well established by human clinical and pathologic material. It is indeed pathognomonic of a lesion of some type in this general region. However, the precise limits of the neurogenic center have not been established. For example, the relationship of the stalk of the hypophysis (infundibulum), the supra-optic and paraventricular nuclei, the tuber cinereum and the component parts of the hypophysis itself still remain in dispute. Human pathologic material, particularly tumors, although the safest source of interpretation as applied to man, has nevertheless been of little value in sharply localizing this center because, with few exceptions, the lesions have been so large that more than one of the disputed areas have been involved. The animal experiments on this problem are open to even more criticism, because the component parts of the so-called hypothalamichypophysial ensemble are