466 University of California Publications in Zoology [Vol. 16 



As pointed out by Bancroft (1899), D. occidentalis is easily dis- 

 tin^iishable from all the other species of the genus, with the possible 

 exception of D. rosea. There seems, however, no doubt about the 

 distinctness of these two. Although the color of occidentalis is so 

 varied that not much reliance can be placed upon it for taxonomic 

 purposes, yet since rosaceous is one shade that seems not to occur in 

 occidentalis, the color distinction is worth something. The most 

 positive differences between them so far as recognizeed are the absence 

 in rosea of internal vessels crossing the stigmata ; the smaller number, 

 usually twelve, of stigmata in a half-series in occidentalis; rosea 

 having, according to Herdman, from twenty to thirty; and the pres- 

 ence on the inner surface of the stomach wall of occidentalis of the 

 network of low but distinct ridges. 



Color. — Examination of an abundance of colonies by the senior 

 author at Point Conception in January. 1908, some of which were 

 nearly white while others were brick-red, led to the supposition that 

 the red color was an old-age mark. There was ample evidence of 

 degeneration in the red colonies, but not in the light ones. But besides 

 the degenerated zooids in the red colonies, there were also many partly 

 gro\Mi ones, indicating that the colonies were undergoing degeneration 

 and regeneration simultaneously. On the whole the zooids of the light- 

 colored colonies were considerably larger than those of the red ones, 

 though partly grown zooids were also present in the light colonies. 

 The pigment of the red colonies was confined to the test. No larvae 

 were found in any of the animals at this time. 



Embryogenesis. — The escape of the larvae from the parents and 

 their behavior during their brief existence before becoming attached 

 were observed by the senior author on July 24, 1896. The young are 

 retained in the parent until the tadpoles are fully formed. They 

 emerge from the common cloacal orifice posterior end foremost. Their 

 exit is entirely passive, the egg membrane being still intact and the 

 larva's tail closely folded aroimd the body. Frequently the indi- 

 viduals are shot out with considerable force. So far as could be 

 ascertained by watching the escaping larvae, they are forced into the 

 common atrial chamber by muscular contraction of the individual 

 parent zooids, and are then expelled to the outside by a contraction 

 of the common chamber. The exact nature of this second phase of the 

 expulsion is not clear and merits study, particularly as to the 

 mechanism by which it is accomplished. 



Immediately upon reaching the outside world most of the larvae 



