290 THE VOYAGE OF H.M.S. CHALLENGER. 



spherical or pyriform portion, the thickened exoderm of which is full of cniclocysts. 

 From its base arises the single tentacle (figs. 35, 37, 42, t). The basigaster is separated 

 from the pedicle as well as from the stomach by an annular constriction (sphincter). The 

 basal sphincter is a very strong ring-muscle, and it is very probable that by its sudden 

 contraction the three distal segments are frequently detached from the proximal pedicle. 

 In my preliminary examinations of the large and well-preserved specimens of Rhodalia 

 miranda collected by the Challenger, I could find in them neither siphons nor tentacles. 

 I saw only the pedicles of the siphons attached to the corrnidia, and judged them 

 to be the highly contracted siphons, and their opening (the pylorus basalis) to be the 

 true mouth. I was thus led into the same error as Gegenbaur thirty years before in 

 Stephanospira. Some time afterwards I examined accurately the masses of horse-hair 

 covering the bottom of the vessel in which the Rhodalia had been packed by the 

 naturalists of the Challenger. There were entangled between the horse-hairs some 

 irregular whitish lumps composed of interwoven long filaments and nodes. Further 

 careful examination convinced me that the long coiled up filaments were the tentacles 

 of Rhodalia, and the nodes were the detached siphons connected with the former 

 (PI. IV. fig. 20). Along time afterwards I received from Dr. John Murray the complete 

 specimens of Stephalia corona taken in the " Triton" Expedition (1882), and in these the 

 majority of the siphons and tentacles were still connected with the corrnidia (PI. VII.). 

 Supported by this confirmation of my suggestions, I was able to restore the anatomy of 

 Rhodalia, and to draw the entire corm with that completeness which is figured in PI. III. 

 At the same time this experience teaches afresh the lesson that much care and critical 

 judgment must be employed in the anatomical examination of preserved specimens of 

 Siphonophorse, and of such specimens as come up in the tow-net or trawl from the deep- 

 sea. Many parts of the corms, especially the nectophores and tentacles, but also often 

 the siphons and palpons, are so easily detached, that they seem to be entirely wanting. 

 I have no doubt that the " deep-sea Siphonophorse, without tentacles," which have been 

 described by Studer (40), Fewkes (45), and by former authors, are corms which have lost 

 the tentacles during capture. 



The stomach (sm), as the third and largest portion of the siphons, is a long cylindrical 

 or spindle-shaped tube, often ovate in the inflated state, and separated by an annular 

 constriction from the two neighbouring segments, the proximal basigaster (sb) and the 

 distal proboscis (sr). It is easily distinguished from both by the dark longitudinal liver- 

 stripes, which extend parallel and equidistant in its whole length (PI. IV. fig. 20, sh ; PI. 

 VI. figs. 35, 38, sh ; PL VII. fig. 42, sh). The number of these hepatic ridges seems to be 

 variable, sometimes eight, at other times twelve or sixteen. After removal of the glandular 

 entoderm, the remaining exoderm of the stomach exhibits a large number of longitudinal 

 parallel muscle-bands (PI. IV. fig. 19, ml). No doubt the siphons are very expansible and 

 contractile, as usual. 



