148 CALIFOKNIA ACADEMY OF SCIENCKS. 



are distinctly paired. On the median line below the intestine the ends of each 

 glandular mass meet and join into one duct, the one I have jnst referred to. The free 

 ends of the olanihilar masses are attached by mesenteric tissue to the ventral or sub- 

 intestiiial blood trunk. 



A much smaller tubular gland of the same nature runs between the dorsal 

 vessel and the intestine; discharge duct unobserved. A yet smaller gland is seen 

 above the subventral vessel in the same somite as the former; its discharge duct could 

 not be followed. 



These glands stain exactly as the common supra-pharyngeal glands and septal 

 glands, but they show no similarity as regards reagents with the chloragogic cells of 

 the intestine and blood vessels, in corresponding places in the somites posterior to xi. 

 The septal glands appear to be of about equal size, a close examination being im- 

 possible, from want of sufficient material. 



Other glands are found in the epithelial walls of the intestine, arranged in 

 clusters, like the cloves in a garlic. They are scattered about at short intervals among 

 the epithelial cells, and appear of the same nature as those I have described in Ar- 

 gilophilus, but they are not as numerous as in that species. They do not stain freely, 

 but stand out bright and pellucid among the darker staining cells. 



lyphlosole not present. 



Gizzards are connected by a very thin wall of the same nature as the oesophagus. 

 As far as I can make out the gizzard must be in v and vi, at least the posterior gizzard 

 is bound by the septum separating vi and vii. The circular muscular layer is about 

 30 strands wide, and is at the widest place about four times thicker than the epithelial 

 layer and cuticle together. The whole width of the gizzard wall is little more than 

 twice that of the body-wall in that somite. The longitudinal muscular layer of the 

 gizzard is only one single strand thick, and the thickness of this strand is less than 

 any one strand of the circular layer of the gizzard. The epithelial layer is com- 

 paratively thick, about one-fifth of the whole gizzard. It contains the same peculiar 

 glands as I figured in Benhamia. 



Sjtermatliecce. The absence of diverticula is interesting in as much as most 

 species of related genera possess them. There are, however, some warty elevations. 

 The muscular or basal duct is very long, slender and tubular, several times longer 

 than the upper ovoid sac (fig. 74). The muscular part oflfei's no peculiarities of 

 structure. The spermatlieene occupy each only one somite. They stand upright fol- 

 lowing close to the anterior surface of the septum. 



Testes are greatly lobed and are situated high up on the septum, just as are the 

 ovaries. 



Spermdncts and ciliated rosettes. The spermducts are separated, but enclosed 

 in a common muscular sheath until somite xxi is reached. Between xx and xvi the 

 two ducts fuse into one lumen, which opens out into the center of somite xxi. The 

 double lumen runs along the circular muscular layer forward until somite xiii/xii, 

 when the respective ducts rise upwards following the septum to the ovary and testes. 

 From here on forward each duct is thicker and muscular, and instead of following the 



