191"] Ritter-Forsyth: Ascidians of Southern California 459 



Digestive system (pi. 42, figs. 41 and 42). — Situated on left pos- 

 terior half of branchial sac. Esophagus [e., pi. 42, fig. 41) emerges 

 from dorsal posterior part of branchial sac. Stomach wider than 

 long, its wall thrown into twelve to seventeen looped folds of varying 

 lengths depending on their position; eight of these folds heading 

 around the esophagus, the others, shorter, on opposite sidas of organ, 

 point toward a seam constituting a part of the cardiac end of stomach. 

 A sac-like caecum (c, pi. 42, fig. 41) on stomach near end of seam, 

 joined by a fine duct to pyloric gland which ramifies over intestine 

 (p. gl., pi. 42, fig. 41 ) . Intestine emerging funnel-wise from stomach, 

 regularly S-shaped. the dorsal half of the S twisted nearly to a right 

 angle with ventral half (int.. pi. 42, figs. 41 and 42) ; a narrow, thin 

 strip of epithelium running along entire convex side of intestine; 

 anus with a thickened bilobed margin, situated in atrial chamber near 

 atrial orifice. 



Reproductive system {o.,t., pi. 42, fig. 41). — Gonads in form of 

 "polycarps" attached to mantle on both sides of body. About three 

 ovaries anteriorly situated on each side of endostyle {o.. pi. 42, fig. 

 41). About five testis masses {t., pi. 42, fig. 41) on right side in a 

 row along the endostyle ; on the left side usually four or five masses 

 arranged around end of intestinal loop. From the summits of male 

 gonads extend short vasa deferentia into the peribranchial cavitv 

 (u. d., pi. 42, fig. 45). 



Breeding habits. — Embryos and advanced tadpoles retained in 

 peribranchial chamber; breeding during midwinter months. 



This species, the name of which we are glad to make stand as a 

 testimonial to Dr. Michaelsen's good work on this group of ascidians, 

 is rather sharply distinct from M. dura Ritter. its neighbor in habitat, 

 as well as in a number of structural features. "While dura never, so 

 far as our observations go, departs much from the completely aggi'e- 

 gated and fused type of colony (Ritter, 1896, especially figures 1 and 

 2, pi. 12), the colony of niichaclseni seems never to assume this form. 

 Not infrequently the zooids of the middle portion of the colony of 

 michaelseni are so close together as to be nearly or quite in contact 

 with one another. This seems to arise entirely from the intercalation 

 through budding of new zooids and never from the approximation of 

 the buds in the new parts of the colony. The trailing off of zooids in 

 these parts of the colony as shown in figure 14 of plate 39, is entirely 

 characteristic of this species. This difference may depend upon the 

 difference in habit of the two species, dura growing typically on sea- 

 weeds where, as a consequence, the room for expansion of a given 

 colony is limited, while michaelseni is almost if not quite restricted to 

 the surfaces of rocks and molluscau shells. As to the zooids them- 

 selves, the most positive difference is in the number of series of 

 stigmata, dura having twelve and michaelseni nine. 



