1. The so-called epilepsy which occurs spontaneously in dogs follows a distinct and plainly recognizable pattern. Various predisposing and exciting causes are discussed.2. The water intoxication discovered by Rowntree is characterized by precisely the same pattern of convulsions. It is readily distinguished from other forms of convulsions, such as occur in rabies, uremia or strychnine poisoning.3. Dogs with spontaneous "epilepsy" are invariably subject to convulsions with administration of smaller quantities of water than normal dogs. Certain other dogs, which remain free from spontaneous attacks during many months of observation, exhibit a similar subnormal tolerance of water, which may possibly represent a latent or sub-threshold stage of "epilepsy." These animals are free from obvious renal abnormalities, and with water administration they pass large volumes of pale urine, so that the sensitiveness is not due to urinary suppression such as Rowntree produced with pituitary extract.4. Chronic susceptibility to water intoxication and also to spontaneous epileptiform seizures can apparently be produced by suitable kidney resections together with repeated overdosage with water. It is still not certain whether this result can be obtained in all dogs or whether some undetectable individual predisposition plays a part. Although water elimination is known to be retarded after sufficient reductions of kidney tissue, the abundant polyuria in water tests seems to rule out this superficial explanation of the convulsions. The occult character of the relation with water balance or tissue hydration seems comparable with that in human epilepsy.5. The name of epilepsy, applied by veterinarians to the canine disorder, seems justified by a number of features, including the ability to elicit seizures by water administration on the same basis as used by some writers for the diagnosis of latent epilepsy in patients. Attention is called particularly to the neglected opportunities for experimental study of the disease in dogs.6. The apparent success in producing chronic idiopathic epilepsy for the first time experimentally is sufficiently important to deserve confirmation with a sufficient number of animals to test the influence of individual or accidental factors.Abstract Teaser