124 PKOCEEDINGS OF THE ACADEMY OF [Feb., 



system of ramifying intracellular spaces or canals, forming a veritable 

 labyrinth. These canals are filled with a pale staining granular sub- 

 stance, probably the precipitate of a fluid. 



After sending ofT its branch to the lateral glandular portion of the 

 nephridium, the inner canal continues backward a short distance, its 

 diameter meanwhile diminishing to ca. 0.5 m, then bends almost directly 

 downward, widening as it does so, and passes into the transverse canal 

 (text fig. IV, figs. 25e and 26a, t.c.n.). The length of the latter is about 

 the same as that of the inner canal, but its diameter is much less, 

 probably in most cases not exceeding a micron. Owing to its minute- 

 ness it is difficult to determine whether it is furnished with cilia or not, 

 but its appearance in fig. 25e seems to indicate that it contains at least 

 a single cilium or flagellum. The transverse canal terminates in the 

 antero-lateral portion of the ventral glandular portion of the nephridium 

 (text fig. lY, figs. 25c?, 25e, and 26, v.g.n.). This body is very similar 

 to the lateral glandular portion, and will therefore require no further 

 description. It lies just above the lateral nerve cord of that side, and 

 posterior to the middle of the second trunk segment, whereas the lateral 

 glandular portion is situated at precisely the middle of the segment. 



The transverse canal serves to link together the two glandular por- 

 tions of the nephridia, but these are further connected by a large 

 ciliated canal, the diagonal canal (text fig. IX, figs. 256, 25c, and 26, 

 d.c.n.), which arises at the anterior end of the lateral glandular portion 

 and passes diagonally backward to the anterior end of the ventral 

 glandular portion. It enters the latter near the termination of the 

 transverse canal; the two canals in fact being connected with one 

 another at this point. 



The duct of the nephridium springs from the ventral glandular 

 portion at the point where the latter is connected with the transverse 

 and diagonal canals (text fig. IV, figs. 25a-c, d.n.). The duct is here 

 relatively large, but narrows rapidly as it passes cephalad. A short 

 distance from its point of origin it dips downward and becomes buried 

 in the substance of the body wall laterad of the ventral nerve cords, and 

 is recognizable in sections only as a brownish ovoid area in the hypo- 

 dermis (fig. 5, d.n.). It thus pursues a direction cephalad and slightly 

 mesiad until it reaches the middle of the first trunk segment. Here 

 it bends sharply inward (mesiad), widening considerably at the same 

 time, and opens to the exterior at the postero-lateral border of the 

 mouth (text fig. IV). The duct is similar in structure to that of 

 unicellular glands, such as the salivary gland of Dinophilus, and 

 appears to be merel^^ a process of one of the cells of the ventral glandu 



