THE SIPHONOPHORA. 339 



extends from end to end of the nectoiihore. The relationship of nectosac to hydrcccium, 

 and the form of the latter, can hest be seen ia theaccomjianying photographs (fio's. 1, 2). 

 The typical somatocyst of other monophyids is here represented by a series of canals, 

 which form not the least interesting feature of the nectophore, and suggest the 

 condition in Stephanophyes superba, Chun (1891). The arrangement of the canals is as 

 follows : — from the point on the inner surface of the hydroecium where the stem is 

 attached, a canal runs on either side over the inner lateral surface of the hydroecium, to 

 terminate blindly before reaching its ventral margin. Each of these canals, at a point 

 nearly opposite the dorsal margin of the nectosac, gives off a branch which runs in a 

 direct line to the dorso-lateral angle of the nectophore. The ascending branch, 

 corresponding in location to the ascending branch of the somatocyst of Rosacea and 

 Pray a, is so much injured that it is impossible to tell how far it extended normally. In 

 the specimen it can be traced about halfway to the apex of the hydroeciutu. The 

 descending branch, after giving off the trunks to the lateral walls of the hydroecial 

 cavity, follows the dorsal margin of the hydroecium, as in the Prayinre ; but instead of 

 connecting with the nectosac by a single canal, as is usually the case, it sends no less 

 than four to that structure. These canals, clearly seen in the photographs, are a pair 

 of laterals, arising together, a dorsal and a ventral each arising separately, opposite the 

 dorsal or ventral margin of the nectosac, as the case may be. The canals do not join 

 the nectosac at its apex, but about at its mid-level. But in spite of their unusual 

 arrangement, they no doubt correspond to the radial subumbral canals of more typical 

 forms. In its further extension the descending branch of the somatocyst reaches nearly 

 to the ventro-basal angle of the nectophore. 



Stem and Appendages. — The stem is very much contracted, but when detached from 

 the nectophore the following features could be determined. There are no buds for 

 reserve bells, nor could I distinguish any structure suggesting that a second bell had 

 been attached, but had been lost. The evidence of a single example, so contracted as is 

 this one, is perhaps insufficient to make it certain that the species is a monophyid. But 

 so far as it goes it is positive, and unless other specimens are found with posterior 

 nectophores, or with the buds for sucli structures, no course is open except to refer 

 Nectopyramis to the Monophyida3. Besides a considerable number of very young 

 appendages and buds, there is one group in which siphon, tentacle, two gonoiihores, and 

 bract are well developed. 



The bract is very characteristic. In general form it is saucer-shaped, or scale-like, 

 bluntly pointed at the superior, transversely truncate at the inferior end. Though 

 licit, there are three dorsal ridges, a single descending and two ascending, and near the 

 centre, where the latter arise, there are two triangular prominences. The canal system 

 of the bi-act consists of a single descending, and two diverging lateral branches, each of 

 which gives off a short, nearly vertical branch running up into one of the dorsal 

 prominences. 



It thus resembles even to minor details the canal system of the bract of Praya 

 cymbiformis, so often described and figured, and, though less closely, that of llosacea 

 plicata. 



51* 



