That the accurate localization of tumors of the brain is often a matter of extreme difficulty is well recognized. False localizing signs, particularly in relation to cranial nerve palsies, have been the subject of a number of communications, but it has rarely been emphasized with what frequency a mass lesion in one area may give definite symptoms of a lesion elsewhere in the brain. Apparently, statistics covering a large number of cases, which have betrayed how often a mistaken localization between a supratentorial and a subtentorial lesion has led to a cerebellar exploration when the neoplasm actually lay in the cerebral hemispheres. have not been compiled from any clinic. An opportunity to assemble a group of cases in which this error in localization was made has been afforded by a recent study by Lehmann1of the results of cerebellar operations in Dr. Cushing's clinic. In the course of this