The Reactions of tlie Vertebrate Embryo to Stimulation etc. 569 



connecting links. There can be no reasonable ground for doubting 

 then that the network [a] in Fig. 14, Piate 23 is an integral part of 

 that uniting the two cells [h] and (e). If the idea is not abandoned, 

 in spite of the absence of any direct evidence showing that this network 

 [a] is the product of spongioblasts, it then becomes necessary to 

 attribute to neuroblasts the double capacity of producing both kinds 

 of fibrils. The manuer in which neurofibrillation takes place first 

 in the outer layers of the cord and then works in is well shown 

 in Fig. 10, Piate 23. Capping the round nucleus in the outer layer 

 of cells is a process, containing a number of fibrils which are con- 

 tiuued inwards almost to the edge of the canal, where the terminal 

 arrangements become too iudistinct to be carefully studied. Fibrils 

 Coming from the perijìhery sometimes seem to form a net within 

 this inner rim of protoplasm which bounds the centrai canal. I 

 bave no further knowledge of the relation of the neuro-reticulum 

 at this point to other structures than the absence of ali evidence 

 indicating that it is the product of spongioblasts. The cells of the 

 sympathetic nervous system are also united by the neuro-reticulum. 

 In Fig. 27 I, Piate 25 a bündle of fibrils may be seen Splitting 

 up [a] into several Strands which connect the bodies of the two 

 cells and then unite to form a common bündle (b) which runs in the 

 same piane uutil it finally bifurcates (e). The curious manner in which 

 fibrils sometimes encircle the nucleus of cells in the sympathetic 

 is represented in Fig. 27 II. An arrangement of the fibrils similar 

 to that shown in 27 I is reproduced in 27 III. 



The figures 7 and 11 of Piate 23 and 22 of Piate 25 aflord 

 iuterestiug comparisons. In the first, the deeply stained coarse fibrils 

 contrast sharply with those in the other two figures which represent 

 later stages of developmeut. The fact that the first drawing was 

 niade from a section of the cord of ScylUum canicula, and the other 

 two from Torpedo ocellata and Lacerta muralis respectively may 

 practically be disregarded, as the Splitting up of the fibrils is common 

 to ali species examined by me. In Fig. 22 the exceedingly fine 

 character of the reticulum fre(iuently exteuding down betweeu the 

 nuclei is apparent. If the whole section of the cord from which 

 this area is taken is studied carefully it at once becomes mani- 

 fest that fibrillation is definitely limited to the outer celi layers 

 in the vicinity of the ventral roots or to the immediate proximity 

 of nuclei lying in the marginai veil. The circumscribed character 

 of the areas of neurofibrillation is very striking, in connection with 



