OF DENTSON UNIVERSITY. 39 



from the most ventral part. As Stieda says, the cephalic portion of 

 the two striata are connected by the septum lucidum ; but in these 

 brains it is only the most ventral parts which are thus connected, since 

 the anterior cornua of the lateral ventricles separate them dorsally and 

 ventrad to these cornua the position of each is occupied by a loose 

 reticulum of fibres lying chiefly in the direction of the ventricle. 



The gray, or non fibrous portions of the striatum contain numer- 

 ous small cells, about half the size of the pyramidal cells of the cortex 

 adjacent. Their forms vary, being chiefly pyramidal or fusiform, with 

 the cell body almost perfectly transparent, but the nucleus and nu- 

 cleolus stain very deeply. The nucleus, however, is not elongated, as 

 in the pyramidal cells of the cortex, but circular, resembling both in 

 size and shape the Dieter's cells which are very numerous in the stria- 

 tum. 



The Fornix is very large in these types, as in all rodents, and ap- 

 pears to take the usual course. Fibres of the descending fornix tracts 

 are gathered into two firm bundles dorsad of the anterior commissure 

 and pass out caudo-ventrad from the ventro-lateral angles of the fornix 

 body. They pass rapidly caudad to the end of the cinereum, then dip 

 abruptly downward into the mammilliary. The latter body is unpaired 

 externally. Internally its two cell-clusters, or niduli, are rather ill- 

 defined and fuse in the meson ventrally. The descending fornix 

 tracts passing ventrad envelop the niduli of the mammillare on their 

 ectal and ventral aspects, at the same time sending fibres into their 

 substance. The fibres of the ascending fornix tracts gather on the 

 mesal and cephalic aspects of these niduli into close bundles which 

 pass cephalo-dorsad and at the same time laterad in such a way as to 

 form the usual figure-of-8 curve. This arrangement prevails in the op- 

 possum and conforms in the main with the description given by Meynert 

 for the human brain. * By reason, however of the great size of the ciner- 



: Stricker's Histology, American edition, 1872, p. 689. "The corpus candi- 

 cans (s. mammillare) is a ganglion, which lies in a loop made by the anterior 

 pillar of the fornix in twisting back upon itself (to enter the thalamus,) and by 

 its means a certain number of the fibres of the fornix are made to pass directly 

 into the tegmentum cruris cerebri. It is a mistake to suppose, with Jung, that 

 fibres from the fornix simply traverse the substance of the corpus candicans, and 

 that the superficial nerve-fibres which enclose the latter are to be referred to a 

 different source. On the contrary, the descending branch of the crus fornicis 

 first invests the outer and posterior surface of the ganglion, and then twisting 

 on itself, invests, under the form of the ascending branch of the same crus, the 

 inner and anterior surfaces."' 



