414 bulletin: museum of comparative zoology. 



i. e.\ do not parallel those of A. tetragona (Bigelow, 1911b). In all 

 cases the stem is much contracted, or broken off; as a rule only very 

 young siphons and tracts being visible. But in one specimen from 

 Station 10,180 a group of appendages, or "eudoxid," so far advanced 

 as to be nearly ready to assume its independent sexual existence, is 

 still intact. In its main features (Plate 6, fig. 1), this eudoxid closely 

 resembles the corresponding phase of Abyla Icuckartii (Bigelow, 1911b, 

 pi. 13, fig. 8), the somatocysts of the two agreeing even to the re- 

 curved tip of its descending branch. And the bract is similarly 

 asymmetrical, consisting of three facets, dorsal, and two laterals, of 

 which the left hand is much the smallest (the apical region is still open, 

 for the passage of the stem). But in Ceratocymba the margins of the 

 bract, and its left lateral ridge, are very strongly toothed, whereas in 

 the slightly younger specimen of A. huckartii they were smooth, or 

 only minutely denticulate. 



The chief interest of this eudoxid is its corroboration of Moser's 

 (1911, 1912a, 1913a) statement that the free-swimming sexual phase 

 of " Diphyabyla," like that of Abyla Jeuckartii (Bigelow, 1911b), is a 

 "Ceratocymba." Two species of "Ceratocymba" are known, C. 

 sagiliata Quoy and Gaimard, well described by Chun (1888, 1897b) and 

 by Bedot (1904), which according to Moser is the eudoxid of Diphy- 

 abyla, and C. asyvictrka Lens and Van Riemsdijk, the latter the 

 eudoxid of Abyla leuckarfii (Bigelow, 1911b, p. 219).^ 



According to Moser (1913a), their bracts are exactly alike. But, 

 though both are essentially similar, with the same peculiar asymmetry, 

 my experience has been that they are readily separated by charac- 

 teristic, if minor, features. Most important of these is the fact that 

 whereas the left lateral (the asymmetrical) ridge of the latter invariably 

 joins the left apical ridge, and the apical facet is correspondingly 

 quadrilateral in outline (Bigelow, 1911b, pi. 15, fig. 4), in C. sagittata 

 the lateral ridge never reaches the apical ridge, and the apical facet 

 is triangular (Plate 5, fig. 5). Furthermore, while the apical facet of 

 the eudoxid of Abyla leuckartii is fiat, or even slightly convex, in 

 Ceralocymba sagittata it is deeply concave, with prominent lateral 

 angles or horns. This concavity is greatest in small specimens, but 

 it characterizes the largest also, though to a less and varying degree 

 (Chun, 1888). Finally, as a general rule at least, the somatocyst is 

 relatively smaller in the latter than in the former. The gonophores 



1 Following Lens and Van Riemsdijk (1908). I formerly (1911b) referred the C. sagillala Bedot 

 (1904) to C. asymelrica. But actual examination of the two forms shows that there is no 

 connection between them; C. sagiliata being identified at a glance by its characteristic gono- 

 phores, as described below (p. 415). 



