A. W. Wilmarth,[250] of Philadelphia, Pa., after a careful study of idiots and imbeciles ranging in intellectual power from the idiot, properly so-called, to the juvenile criminal, paranoiac, and "ne'er-do-well," finds that the brains of these vary greatly along the line pointed out by Spitzka and Mills. One type of brain in this class of children is very simple in its outward configuration. The convolutions are usually coarse, but little convoluted and comparatively free from secondary folds. The fissures tend to assume a confluent type. Another variety, found chiefly among the lowest grades, might well be termed "atypic." A brain without a corpus callosum is a marked example. In the frontal lobe of the right hemisphere the first frontal convolution is quite regular. Below this from the centre of the lobe seven fissures passed in different directions, cutting the tube into a number of radiating convolutions, entirely different from its usual appearance. The short fissure of Sylvius (about three inches in length) passed upward, turned sharply, and passed almost directly behind. Two parallel gyri curved around its posterior extremity. The arrangement of the convolutions of the temporal and parietal lobes were so exceedingly irregular and complex that it was impossible to classify them. In the occipital lobe, on the contrary, the gyri were complete in number and of regular arrangements. In the left hemisphere the arrangement of the frontal convolutions was more regular, but the temporal and parietal lobes presented the same complicated area of surface folding, bearing but little resemblance to the normal brain. The tendency of the convolutions to arrange themselves in parallel curves around the posterior extremity of the fissure of Sylvius was well shown in the brain of a boy of exceedingly low intellect. The frontal lobes in this brain are proportionately large; the convolutions straight, especially the third frontal, the fissures shallow. In the left temporal lobe they are nearly obliterated from pressure of fluid in the ventricles. The ascending frontal convolution on each side appears to be wanting. On the left side a large bridging convolution crosses the middle of the fissure of Rolando. Confluence of fissure is a decided feature of idiot brains. Even where confluence is not complete, the tendency of the principal fissure to cut through separating convolutions is very evident. Were the cases where confluence is nearly complete included, the number would be considerably augmented. The fissure of Sylvius passed into the fissure of Rolando in one case on both sides, in another on one side only. In two other cases they were connected by deep secondary fissures. The interparietalis has its origin in the fissure of Sylvius in four cases on both sides, in five cases on one side only. The calcarine fissure passed completely across the gyrus frontitatus on both sides in two cases, on one side in four cases.