[Illustration: PLATE I.] [Illustration: PLATE II.] [Illustration: PLATE III.] [Illustration: PLATE IV.] [Illustration: PLATE V.] [Illustration: PLATE VI.] There is more evidence in favour of attaching some importance to the shape of the brain, to its relative development, to the condition and relations of its convolutions. Broca, Topinard, and many other eminent anthropologists and anatomists have attributed great value to these relations. Gall was perhaps the first to suspect their significance. Benedikt, in 1879, published some interesting generalisations on the brains of criminals which he had examined. He found special frequency of confluent fissures; that is to say, according to his own description, if we imagine the fissures of the brain to be channels of water, a swimmer might with ease pass through all these channels. Benedikt also found in the brains of his criminals that the frontal lobe frequently presented four convolutions, a peculiarity which he considered as a reversion to the carnivorous type; the investigations of Hanot and Bouchard confirmed these results. But Benedikt neglected to make an adequate comparison with the normal brain, and Giacomini, Corre, Fallot, and Féré have shown that these peculiarities are not very rare in ordinary subjects. The question of confluent fissures had before this time attracted the attention of Broca, and his conclusions may probably still be accepted:--“One or more of these communications,” he said, “do not prevent a brain from being at once very intelligent and very well balanced, but when they are numerous, and when they affect important parts, they indicate defective development. They are often seen in the small brains of the weak-minded and idiots, very frequently also in the brains of murderers, with this difference, that in the first case they are related to the smallness of the convolutions and of the brain generally; while in the second case they coincide with convolutions for the most part ample in size, and bear witness to irregularity in cerebral development.” Flesch studied the brains of fifty criminals, and found that every one presented some anomaly, sometimes of a remarkable character, as incomplete covering of cerebellum by cerebrum. He found two kinds of deviations common, one characterised by less richness of convolution than is found usually in ordinary brains, the other characterised by much greater richness of convolution than he had ever observed in normal brains. On the whole we may agree with Hervé, that “what the brains of criminals present, not characteristically but in common with those of other individuals badly endowed though by no means criminals, is a frequent totality of defective conditions from the point of view of their regular functions, and which renders them inferior.”[20]