24 ON A NEW SPECIES OF ENTEROPNEUSTA, 



from the " cord hollows." I have never been al)le to trace these 

 hollows in the roots as continuous canals throughout the whole 

 length of even the first "root," and am convinced that in none of 

 the " roots " of this species does a continuous canal exist such as 

 Spengel describes for the first "roots" in Pt. niiiiuta, clavig'^ra, 

 aperta, and bahamensis. Similar though smaller hollows may 

 exist in the " roots " succeeding the first, but just as in the first 

 " root " the hollows do not stand in continuous connection, but 

 are interrupted by anastomosing strands or in some cases by 

 branching ceils with large nuclei. I have likewise never observed 

 the " collar hollows " to pass into the roots. 



The roots end dorsally just as SjDengel describes (fig. 16, rt.). 

 The outer limiting membrane and the fibrous layer of the root 

 unite with the similar layers of the epidermis; while the cellular 

 tissue of the root projects in a cone-like mass between the epidermal 

 cells, the apex of the cone reaching to within a short distance from 

 the outer surface. The apex of this cone-like mass of tissue is 

 composed of a network of fine strands enclosing clear spaces, and 

 with a few very small nuclei in the angles where the strands meet. 

 Below this clear tissue and continuous with it, there occur more 

 deeply staining branched cells which are directly continuous with 

 the central tissue of the "roots." 



Nerve Stems of Trtink : These are related exactly as in the 

 described species, and the circular commissure placing the fibrous 

 layer of the collar cord in connection with that of the ventral stem 

 can be readily seen in horizontal sections occupying the deep 

 groove between collar and trunk. The cells covering the circular 

 fibre ring are entirely destitute of gland cells, and consequently 

 are sharply marked off from the adjacent epidermal cells. 



The dorsal stem in the branchial region is situated somewhat 

 below the level of the adjacent epidermis (fig. 20, dnv.), but 

 posterior to this region it forms a distinct ridge. The ventral 

 stem, on the other hand, occupies over its whole extent the bottom 

 of a shallow longitudinal groove (figs. 20, 24, vn.). The fibrous 

 layer of the dorsal stem is, as in Pt. minuta, somewhat triangular 

 in transverse section; it is thickest in the middle, and laterally 



