BIGELOW: MEDUSAE AND SIPHONOPHORAE. 413 



Most striking of these is that the pointed apex of the nectophore above 

 the level of the somatoc^'st is longer in the adult than in the young; 

 the nectosac being correspondingly longer, the soraatocyst propor- 

 tionately smaller, the triangular hydroecium proportionately broader 

 at its mouth. But the same characteristic arrangement of ridges and 

 facets is to be seen at all stages in growth. And this is likewise true 

 of the basal teeth of both nectosac and hydroecium. In the young 

 specimen listed here, and in the still smaller Albatross example 

 the apex of the nectophore is bent ventrally, above the level of the 

 somatocyst. But since it bends dorsally, though at about the same 

 level in all the large specimens, both Siboga (Lens and Van Riemsdijk, 

 1908) and Bache, it is a question whether this bending is normal, or 

 the result of contraction or preservation. 



Inferior nectophore. The inferior nectophore of Ceratocymba 

 has not previously been described, though it is figured by Moser 

 (1912b, fig. 23). The inferior nectophores listed above (p. 412) oc- 

 curred side by side with the anterior ones, and, from this fact, from 

 their large size (20-45 mm. long) and resemblance to the gono- 

 phores of Diphyabyla, and from Moser's (1912b) figure, probably 

 belonged to it. The general outline with its long, narrow nectosac, 

 is better shown (Plate 7, fig. 1), than verbally. Its most diagnostic 

 feature is its curious asymmetry, not sufficiently illustrated by Moser. 

 The hydroecium is open, ventrally, from end to end, enclosed, however, 

 by two broad flaps. Of these, the right hand one terminates in a very 

 long basal tooth (Plate 7, fig. 1) whereas the left hand one has no 

 basal tooth but merely makes a right-angle with the short dorso- 

 basal wall of the hydroecium. The latter, as well as the basal parts 

 of the hydroecial flaps, bears a series of prominent, hook-like serrations. 

 The mouth of the nectosac is armed with two very prominent triangu- 

 lar teeth, right and left lateral; and a very much smaller dorsal tooth. 

 Besides the flaps, the nectophore has two ridges, so prominent as to 

 be better described, perhaps, as wings, at first sight apparently a 

 dorsal and a left lateral. But the former is in reality the right lateral, 

 for it terminates in the right lateral basal tooth. The dorsal ridge is 

 almost entirely suppressed, except close to its base, being traceable for 

 most of its course only as a faint line. The nectophore is thus triangu- 

 lar in cross-section, like that of Abyla trigona. But the resemblance 

 between the two is purely superficial, for in A. trigona it is the right 

 lateral ridge, not the dorsal, which is suppressed. Above the level 

 of the nectosac the apex of the bell is prolonged in a narrow triangle. 



The subumbral canals (Plate 7, fig. 1), are of the usual abylid type, 



