ANATOMY AND CLASSIFICATION OP THE ARENIOOLIDJl. 485 



course of the cell process which enters the giant-fibre it is 

 not always possible to trace the connection in one section, yet 

 it can be seen on following the process through one or two 

 adjacent sections. We have verified this point again and 

 again in the case of the majority of the giant-cells in one 

 specimen of Arenicola Grubii, by the examination of a con- 

 tinuous series of sections through the entire length of the 

 worm. Additional evidence on the point is given by prepara- 

 tions of the nerve-cord of the common lugworm (A. marina) ; 

 and as all species of the genus except A. Claparedii possess 

 giant-cells (arranged just as regularly as, but not always at- 

 taining the size of those of A. Grubii) we may fairly pre- 

 sume that this mode of origin occurs throughout the entire 

 genus. 



It is necessary, however, to go into the details of these cells, 

 their processes, and the connections of the median and lateral 

 fibres with each other and with the cord. 



B. Giant-cells. 



The giant-cells of the nerve-cord of Arenicola Grubii 

 are conspicuous by reason of their large size and singularly 

 regular arrangement. Throughout the greater part of the 

 length of the cord, as is shown on PL 29, there are, as a 

 rule, two of these cells placed in line an annulus apart, close 

 to the hinder limit of each segment. The limits of the 

 segments are marked externally by the gi'oove separating the 

 first from the second annulus behind the chsetigerous one. 

 The giant-cells are therefore placed in a fully extended animal 

 just in front of this groove, and where there is only a single 

 giant-cell, as is the case in the anterior one or two of the 

 branchial segments of the body, this single one usually corre- 

 sponds in position to the more posterior of the two present in 

 the majority of the segments. In the singular regularity 

 displayed by the arrangement of these cells Arenicola 

 Grubii appears to be paralleled by Polynoe (Rohde, 1887). 



We have been at considerable pains to ascertain and plan 

 out these cells from an entire series of sections of a complete 



