The Reactions of the Vertebrate Embryo to Stimulation etc. 567 



preparations imdoubtedly convey the impression that from the 

 veiy beginniug the fibrils are arranged in the shape of nets, 

 while in my sections (compare Figs. 3, 5, 8, 9, 12, Piate 23) the 

 first are nearly always parallel without any cross counections. (The 

 celi represented in Fig. 6 is a much later stage than those where 

 the fibrils are parallel.) One explanalion for these apparent dis- 

 crepancies is that the method employed by me seems to be capable 

 of staining the fibrils at a period antecedent to that represented by 

 Held. The longitudinal arrangement of the primitive fibrils is well 

 shown for example in Figure 9 at the points indicated by the letter f. 

 In i>reparations where the diflferentiation is not well marked and 

 the sections are more than 5 /< in thickness it may sometimes be 

 difficult to recognize this primary longitudinal arrangement, as short 

 sections of fibrils lying in qnite another piane may be superimposed 

 on each other so as to give the appearance of a network. The 

 relative vaine of the method described in Seetion 3 as compared 

 with that of Ramon y Ca.jal in difterentiating the neurofibrils may be 

 seen when applied to the study of the processes of the ventral horn 

 cells. It is not difficult to show that the eoarse bundles are made 

 up of fine fibrils, while the arrangement in the celi processes as re- 

 presented in Figure 8 of Held is probably the result of an im- 

 perfect differentiation, by which individuai fibrils are glued together 

 by the excessive Silver deposit. 



Prior to the period when nets appear in the ventral horns the cells 

 of Beard in the posterior columns bave passed rapidly in their develop- 

 ment from the stage represented in Fig. 12, Piate 23, to that shown 

 in Figs. 4 und 14 of the same Piate or Fig.15, Piate 24. In the one 

 instance there are only a few fine fibrils Ijing parallel to each other 

 within the apical processes, while in the others there is a diffuse 

 network extending over the processes, nucleus and body of the cells 

 and in two of the figures they are shown to form an intimate union 

 between two cells lying on opposite sides of the dorsal commissure. 



The importance of the sections represented respectively in 

 Fig. 14, Piate 23 and Fig. 15, Piate 24 is very great, because these 

 two preparations seem to me, as well as to the others who bave 

 studied them, to demonstrate conclusively that definite and very 

 distiuct connections exist between the neurofibrils in one neuroblast 

 with those in another. It is comparatively easy in sections stained 

 by the method I bave employed to find many cells where the 

 general arrangement and distribution of the fibrils seems to indicate 



