566 Stewart Patoa 



ejected from the cells. In Fig. 4, Piate 23 the union between 

 neuroblasts and spongioblasts is clearly shown in a drawing made 

 from the section of a spinal cord of ScijlUum canicula, leugth 6 mm. 

 At the edge of the central canal (c. c.) is a large deeply stained 

 nucleus, not contaiuiug any nucleolus, whose surface near the circum- 

 ference shows a faint coarse mesh-like structure. Two distinct 

 fibrils pass outward from the outer end of the nucleus to become 

 united into a common Strand at jj. Wheu this point is studied care- 

 fully with the aid of the Zeiss Apochrom obj. 2 mm and Compens. 

 Oc. 18 there caunot exist any doubt in the mind of the observer 

 that this nucleus is directly connected with the network of neuro- 

 fibrils belonging to the large cell of Beard represented in the 

 ligure and whose nucleus lies just dorsal to the central canal. 

 A secoiid bündle of fibrils {p^) runs in the direction of the nucleus 

 of the spougioblast but ends abruptly before reaching it. A bündle 

 of fibrils (p^) the continuation of the fibril ]) marks the inner edge 

 of the fine network connected with the cell of Beard. All the 

 interveniug inter-nuclear substance is faintly tinged with eosin but 

 does not show any evidence of fibrillation. The long coarse fila- 

 ments already referred to, which are Seen to pass from the ventral 

 roots on one side through the anterior commissure and then to bend 

 in towards the central canal, where they end, either in a network 

 that surrounds the nuclei in the inner cell layer or spread out over 

 the protoplasm bordering the central canal may possibly be the 

 processes of neuroblasts. On account of the difficulty in following 

 the course of these structures as well as in determining the limits 

 of cell bouudaries the question caunot be definitely decided. As 

 development proceeds the connections between neuroblasts and 

 spongioblasts, at least the forms represented in Fig. 4, disa])pear. 

 Generally when the Selachian embryus bave attained a leugth of 

 7 mm. or 8 mm. many of the fibrils in and near the ventral roots 

 have split up, forming attenuated filaments at the same time that 

 difl"erentiation is proceeding more rapidly in the inner layer of the 

 cord. The fibrils which at first were not united by lateral connec- 

 tions, now begin to form a definite network, generally most marked 

 over the apical processes of the cells in the regions of the ventral 

 horus. After comparing the figures 1 — 5 in Held's jiaper, as well 

 as this writer's description with my own results there seemed at 

 first to be considerable discrepancy between us in regard to the 

 arrangement of the primary fibrils within the cell body. Held's 



