44 BULLETIN : MUSEUM OF COMPAEATIVE ZOOLOGY. 



the blood vessel, pressing in its walls at the point of contact, so that 

 the short processes of the nephridial canal project fi-eely into the 

 Inmen of the hlood vessel, thus forming the so-called nephridial 

 gland (PI. 14, iigs. 84, 85). In this way the lumen of the blood 

 space is greatly decreased, and at the same time the very numerous 

 finger-like projections bring a vast surface of the nephridial canal 

 into contact with the blood contained in the lateral vessels (PI. 15, 

 % 92). 



The i^osterior two thirds of the nephridial canal is large and 

 smooth and entirely Avithout branches. In Carinoma mutahilis 

 the tube passes backward for some distance in the parenchyma 

 near the lateral vessels, then l)ends forward for a short distance, 

 and ends in a narrow efferent duct which passes through the mus- 

 cidar layers of the body wall to open by a small excretory pore situ- 

 ated on the dorso-lateral surface of the body (PI. 14, fig. 84). 



In Carinella sexlineata^ as described in this jjaper, and in other 

 species of the genus, the relations are similar, the main canal on each 

 side passing forward from the efferent duct quite to the region of 

 the " nephridial gland " without branches. Here it sends off several 

 branches wliich subdivide into the finger-like processes in the wall 

 of the adjacent lateral blood vessel, as described above. In C. rubra 

 the efferent ducts connect with the main canal by a broad, sieve- 

 like funnel (Text-fig. 14) which projects into the ^ddened lumen of 

 the canal. In C. alhocincta there are 5 to 8 longitudinal canals on 

 each side, and these do not join until near the efferent duct, where 

 they unite to form a rather large lacuna fi"om which the efferent 

 duct passes to the exterior. 



In Carinella and Cakinomella the pair of excretory pores lie 

 in the vicinity of the lateral sense organs (PI. 15, fig. 91). 



B. Hoplonemertea. — In this order the nei)hridial canals branch 

 j^rofusely in the body parenchyma, and these branches lie in close 

 proximity to the lateral blood vessels. Burger ('91) descril^es the 

 branches of the nephridia in Drepaxophorus as twining about the 

 lateral blood vessels very much as a Aane twines about a tree. Each 

 of the small branches terminates in a small chamber in which swings 

 a tuft of long cilia. 



In many hoplonemerteans the branches extend all through the 

 parenchyma on each side of the body, a portion of them lying at the 

 side of the proboscis sheath far remoA^ed from the lateral vessels, as 

 in A. gelatinosus (PI. 19, fig. 120). 



