1. How shall we account for O’Donnell’s transient paralysis? We might invoke brain tumor, alcoholic pseudoparesis, or some form of neurosyphilis. The diagnosis of brain tumor seems quite untenable in view of the absence of premonitory symptoms and in the absence of intracranial pressure. As for alcoholic pseudoparesis it is true that the patient was excessively alcoholic. However, against these two diagnoses and in favor of the diagnosis of NEUROSYPHILIS, are the positive serum W. R. and the pupillary reactions (although these are short of the true Argyll-Robertson phenomenon). Dizziness with retention of consciousness and associated with the paralyses mentioned suggests rather a subcortical than a cortical lesion. We are inclined to regard this lesion as probably THROMBOTIC, and to place it possibly in the region of the internal capsule. We are inclined to regard the phenomenon as purely vascular and as not in this case associated with an encephalitis. We are, however, not entirely satisfied with the diagnosis. 2. What shall be said as to treatment? A full-blown left-sided hemiplegia may be produced even when the thrombotic lesion is itself exceedingly small. It is common to explain this on the basis that there is an area of collateral edema about the small necrotic, thrombotic, or hemorrhagic area responsible for the lesion. In short, numerous neurones are functionally rather than structurally affected, or at all events capable of early restitution of function. 3. What is the prognosis in such cases? It appears that now and again patients run for several years without further trouble, both with and without treatment. We are inclined, however, to advocate treatment rather than absence of treatment for a variety of reasons. In the first place, vascular lesions may at any time become associated with meningitic lesions, and treatment by salvarsan may perhaps be counted on to head off this process; secondly, the treatment with iodids may possibly aid in the resolution of a local thrombotic process.