144 CALIFORNIA ACADEMY OF SCIENCES. 



cells. The septa posterior to somite xiv again are not miieli eiipped, hut each one is 

 lined on each side hy a single hxyer of very thick tall perittuioal cells (fig. 1 "»o). 



Alimentary canal. Phnnjnn is furnished with a large upper chamber and is 

 apparently only developed superiorly, allliDiij^^h it possesses a small suhpharyngeal 

 gland close to the ventral nerve cord. 



CEaophtgus rises diagonally upwards and joins the single gizzard situated in 

 V. The tiibulnr-intpMlne extends to xiv. The saccfilalcd-intestine commences in xv. 

 There is a ty[)hlosole in the donsal wall of tlic iiile-tine in .vvi to xix. The typhlo- 

 sole presents a network of fibres resembling in a general way the structure of a 

 sponge. This typhlosole does, however, not descend into the canal, but partakes 

 more of the nature of a wide continuous blood-sinus. 



Salivary glands. The suprapharyngeal glands form a mas.s with five distinct 

 lobes, of which the posterior one, as usual, is the largest and the anterior one the 

 smallest (fig. 149). There is also a suhpharyngeal gland very low but rather long 

 (fig. 14!)). A thin but wide septal gland is found in vi, posterior to the gizzard, 

 while smaller septal glands, which are principally developed ventrally are found in 

 vii and viii. 



Spennalhecie consist of two pairs of long and narrow organs in viii and ix 

 opening in the intersegmental grooves between vi and viii and viii and ix. In my 

 specimens they were probably rather undeveloped and did not show any trace of 

 diverticula either externally or in the wall. 



Testes are in x and xi and ovaries in xiii. The nndnct opens between the 

 setns and septum. SpermiducH I funnels or rosettes are ver}' small, thick and compact, 

 and situated in x and xi. The spermducts run separately backwards between the 

 longitudinal layer and peritoneum and open jointly on the center of xviii as usual. 



Prostates are in foiir pairs as has been already stated, two and two opening in 

 each pore in line with setie 1 and 2, these setae, however, not being present in these 

 somites (fig. 1 oO). The prostates showed no glandular part, the whole being muscu- 

 lar (fig. 1.j4). They were very thin, tubular, the two prostates in each twin couple 

 running entirely parallel and close together along the septa, as far as the line of setiC 

 3 and 4. A large pait of this distance the muscular part is surrounded by regular 

 peritoneal cells, not by the large glandular ones. Each prostate remains separate 

 from the other and even their external pores, though situated close to each other, are 

 not strictly joined, though they are surrounded by a common thicker lip (figs. 1 oO 

 and ]/>lj. I have already referred to the duplication of prostates in one species of 

 Kerria, otherwise it has not been found with certainty in Acanthodrilidse. Some of the 

 earliest described Acanthodrilides were, however, supposed to have a prostate and 

 spermducts open jointly, and it does not seem unreasonal)le to suppose that in some 

 case at least a <luplicate prostate existed, and one was mistakeuM'or a spermdiict. This 

 would be (|uite easily done in specimens poorly preserved, especially if the prostates 

 should be entirely muscular as in Aca nthodrilus Vasliti, in which species they are 

 also very narrow and thin and not really wider than the spermducts of many forms. 

 Ude, in descril>ing Ucodntus singularis from Danville, Illinois, mentions as one of the 



