160 A. B. DAWSON 



a. Tripolar cells 



The tripolar cells are found in locations where branchings of the 

 larg^ nerve trunks occur; that is, at the bifurcation of the lat- 

 eral nerve trunks to give rise to the dorsal and ventral rami of 

 the nerve rings and along the nerve rings at the points of origin 

 of the many peripheral nerves; but they are not found uniformly 

 in these positions. This is especially true along the nerve ring, 

 since four or five tripolar cells is usually the maximum number 

 found on any one ring. 



The cells approximate a triangular outline. Their nuclei are 

 spherical and in all methylene-blue preparations a well-defined 

 nucleolus and a delicate chromatin network can be distinguished. 

 Silver preparations leave the nuclei unstained, but show a dense 

 network of neurofibrils occupying the bodies of the cells (fig. 1). 

 In most of the tripolar cells the nuclei occupy central positions. 

 When eccentrically placed, however, they are invariably found 

 nearest that pole from which a proximal process arises. 



The tripolar cells which are found at the bifurcation of the 

 lateral nerve trunk do not always bear the same relations to the 

 nerve system. In some cells, one process extends proximally in 

 the nerve trunk, while the other two processes extend distally 

 and ventrally, and distally and dorsally, respectively, in the 

 nerve ring. In other cells, one process extends proximally, one 

 distally and ventrally and one peripherally. Again in others 

 practically the same relations are maintained as in the second case 

 except that one process, instead of passing ventrally, passes dor- 

 sally in the nerve ring. From the tripolar cells, situated at the 

 points of origin of the peripheral nerves, whether in the dorsal or 

 ventral portion of the nerve ring, one process always passes 

 peripherally toward the epidermis while one of the two others 

 passes proximally toward the ventral nerve cord and the other 

 distally toward either the middorsal or midventral line, as the 

 case may be. 



The processes passing peripherally were traced readily in 

 many cases to the bases of the epidermal cells, but were lost in 

 the basiepithelial network (Dechant, '06), so that positive evi- 



