170 BULLETIN OF THE 



ill size it is no longer possible for the veiitral line with the entire nervous 

 mass to be retained in its usual position. It is therefore forced up- 

 ward into the body cavity^ and fills a considerable portion of it (Fig. 9) . 

 In fact, near the end its diameter is half as large as that of the body 

 at that point. 



Seen in cross section (Fig. 96) the nerve cord itself presents little' 

 here that is different from its general character throughout the body. 

 It is the peripheral layer which demands particular attention. This 

 is crescent-shaped, and composed of a dense matrix in which are embed- 

 ded many nuclei. Along its periphery the matrix is finely striated per- 

 pendicularly to the surface, and is separated from the underlying nerve 

 cord by a narrow space filled with loose fibrous tissue (Fig. 97). The 

 horns of the crescent are turned ventrad around the nerve cord, and are 

 connected with the tissue lying below it. In these horns one sees a 

 fine longitudinal striation ; occasionally more plainly marked fibres, 

 coming from the cells above, may he traced into them. The matrix, 

 which usually takes a faint stain, contains numerous oval nuclei (Plate 

 V. Fig. 71) with a sharply defined nuclear membrane. The nucleus in 

 general does not take up the stain, so that the one or two large nucleoli 

 stand out in strong contrast to the rest. One can neither find any cell 

 boundaries in the matrix, nor determine how much, if any, of the sur- 

 rounding protoplasm belongs to each nucleus. In fact, the fibres which 

 one finds often appear to extend from the nucleus itself around into the 

 horn of the crescent. 



In addition to this thick peripheral layer, which I regard as nervous, 

 there are in the anal ganglion a few large cells. Some are wedged in 

 below the middle portion of the cord like those which have already been 

 described. Others may be found in the space between the peripheral 

 layer and the cord (Fig. 96), with the process directed ventrad. These 

 cells do not as a rule appear to be paired. 



At the extreme posterior end of the cord one finds somewhat different 

 conditions. Here there is a mass of large ganglionic cells of varied size, 

 closely crowded together, and with their processes (Plate VII. Fig. 90) 

 mostly directed forward into the ventral nerve cord. The space between 

 the cord and the peripheral layer does not exist, and, curiously, the 

 ganglion cells of the latter have nearly always two nucleoli instead of 

 one as is usual elsewhere. The relation of the cells and processes is very 

 complicated here, and the gradations of size are so fine that with the 

 material at my command I was unable to determine the exact number 

 of cells, or the plan on which they are arranged. The nuclei of these 



