There has been reported investigation on 118 cases of this disease in the previous paper, which suggested marked discrepancy of its clinical symptoms between the juvenile form and adult one. Moreover, it has been pointed out that serial bilateral carotid and vertebral angiography should be absolutely necessary to distinguish those cases from other occlusive diseases of the internal carotid, and suggested that this disease would be a particular type of cerebral vascular abnormality of the Japanese. Postmortem examination on five autopsy cases has been done up to the present time, but the pathological findings brought no contribution to the pathogenesis of the disease. Therefore animal experiment has been attempted to elucidate the pathogenesis of the disease. Complete or incomplete occlusion has been performed bilaterally on the dog brain at the site of carotid siphon (C(1)) immediately distal portion to the origin of the posterior communicating artery. The arteries were successfully clipped in 8 dogs. Bilateral carotid angiographies was carried out with the intervals (from 3 months to 12 months) after the surgery on each dogs, but abnormal vascular network was never observed in all cases. The solid molding specimen of the circle of Willis also demonstrated no abnormal vascular network at the site of the occlusion in these experimental dogs. Histological examination revieled a minimal thickning of the wall of the arteries at the distal part of the occlusion and dilatation of artrioles which belong to the middle cerebral arteries. It could be denied by this experiment that the acquired occlusion of the circle of Willis will bring forth the abnormal vascular network at the base of the brain. The pathogenesis of the disease is still obscure but it is in my opinion that the previous clinical investigation as well as this animal experiment suggested congenital vascular malformation of the circle of Willis.