66 CALIFORNIA ACADEMY OF SCIENCES. 



These again are separated from one another by a small 

 quantity of tine granular matter — intercellular substance, I 

 take it to be. The bodies c, c. are probably small cells, 

 since in man}^ of them what seem to be nuclei are 

 found. The large, distinct cells are undoubtedly digest- 

 ive cells, since the secreted matter can often be seen 

 forming' o-lobular or semi-o-lobular masses at their 

 inner extremities. They seem to be similar to the gland 

 cells found in the stomach of other Tunicates (Sheldon, 

 '88; Roule, '85; R. Hertwig, '72, and others). It is 

 possible that Roule ('85) may have seen cells similar to 

 the small, clear ones described above, in Polycarpa var- 

 iaiis. After speaking of the glandular and hepatic 

 cells in the stomach of this species, he sa3's : " Cer- 

 taines cellules meme sont entierement hyalines, et 

 possedent I'aspect typique des cellules calicinales a 

 mucus" (p. 65). This structure of the stomach of Pero- 

 phora as shown in iig. 28, pi. iii, resembles considerably 

 a tigure representing a surface view of the wall of one of 

 the coecal appendages of the stomach of Salpa, given by 

 Dolley ('87, fig. II, pi. xiii). In the text, however, he 

 describes the " corsely granular pyramidal cells " as be- 

 ing " separated from one another by lighter, finely gran- 

 ular spaces, which, when viewed from the surface of the 

 coecum (fig. 11), present a reticulated appearance" (p. 



303)- 



As seen by fig. 22, pi. iii, the lumen, which leads from 

 the stomach into the duodenum, is veiy narrow (.v' is at 

 entrance of this passage). The walls of this section are 

 composed of cells of quite a different character from 

 those which enter into the structure of the stomach. 

 They are scarcel}-, if at all, granular, and their nuclei are 

 generally invisible. What I suppose to be the cell bound- 

 aries are very distinct, particularly in about the outer half 



