498 P. W. GAMBLE AND J. H. ASHWORTH. 



have shown that from several, if not from all of the meta- 

 merically arranged large unipolar ganglion-cells, there arises 

 in each case a single process which gives off fibrillations to 

 the cord, both to the right and to the left sides. The pi'ocess 

 then enters a giant-fibre, which may divide, some of its 

 branches entering the spinal nerves, and others, proceeding 

 for a longer or shorter distance, anastomose with the other 

 giant-fibres, and again give off branches to nerves. The 

 structure of the cell-process and giant-fibre is undoubtedly 

 that of an axon, and corresponds with what Apathy calls 

 " sensorische Schlauche," but there is no proof that an inti- 

 mate fusion of the process of one cell with that of a suc- 

 ceeding cell takes place. A more successful topographical 

 method must first be employed to decide this point, and also 

 in order to determine in what manner the fibrill^of the cord 

 enter the giant-cell, how they are connected through the 

 collaterals with the axon, and in what way these fibrillaerun a 

 recurrent course along the cell process to the giant-fibre and 

 spinal nerves, and so reach their termination. 



D. The Occurrence of Giant-cells in Arenicola. 



With reference to the presence and development of giant- 

 fibres and giant-cells in the different members of the genus 

 Arenicola, we have ascertained the following particulars. 

 In the common lugworm (A. marina), a specimen 17"5 mm.^ 

 long showed an arrangement of the giant-cells precisely 

 similar to that described above in A. Grubii. In the great 

 majority of the segments there were two of these cells placed 

 in line and near the hinder border of the segment ; in the 

 remainder only one was present. The caudal region was 

 carefully examined, and giant-cells were found apparently 

 regularly (in the middle of the tail) two annuli apart. The 

 cells measured 'O^ mm. in length and about '02 in breadth. 

 They were pyriform, unipolar, and their process extended 

 towards the dorsal surface of the cord, but was not ascer- 

 tained to end in the giant-fibre. The nucleus is large, and 

 » PI. 29, fisr. 80. 



