26 CERIANTHARIA. 



walled nematocysts (spirocysts) and a sprinkling of thick-walled ones (length 26 /<) and gland cells 

 which are principally large mncus cells. The "median streak" of Mc.Murrich, which is more strongly 

 ciliated than the lateral portions, is not deeply snnk, as the ectoderm is rather high, bnt broad; between 

 the supporting cells, mucus cells are interspersed now and then. The mesogloea and the musculature 

 stand as in Cerianthus lloydii. 



In the craspedoneme region the median streak is simple. In the coarser and shorter craspe- 

 donemes the endoderm between the ascending and the descending limbs of the filament is pretty well 

 developed: in the long narrow craspedonemes the endoderm is very slight. 



The cnido-glandular tract contains thick-walled nematocysts (length 24 — 26 m) in great numbers. 

 In addition nematocysts are found which are very small and without visible spiral-thread base. The 

 like occur also in the threads, craspedonemes, of the craspedia. In protomesenteries 2 and the meta- 

 mesenteries of the ist and 2nd cycle the filament forms — as in C. lloydii — aborally from the cnido- 

 glandular tract a more feebly developed part, which I have named craspedion region. In M this part 

 is especially long. The craspedion (fig. 3, pi. 5), which is plainly distinct from the endoderm of the 

 mesentery, shews much the same structure as the median streak. The nematocysts are principally 

 spirocysts and the gland cells principally mucus cells. In the most aboral part there is found on the 

 largest mesenteries of the ist cycle a more or less threadlike appendage (fig. 6, pi. 5 w/), in other words 

 the craspedion ends as a simple craspedoneme, which is formed by an ascending and a descending 

 limb of the craspedion separated by an endoderm portion (fig. i, 2, pi. 5). Just as the most aboral part 

 of the craspedia in C. lloydii shews a mixed character, in as much as thick-walled nematocysts there 

 l)resent themselves in great numbers, so here the ectoderm of the craspedia contains thick-walled 

 nematocysts (length (22)— 26/^) in very large numbers, whilst the thin-walled ones are very rare in 

 the greater part of the thread and become less so only at the base. The granular gland cells also 

 are more numerous here than higher up on the craspedion, at the same time that the craspedion gets 

 more sharply differentiated from the endoderm, in other words the craspedion thread acquires a 

 structure approaching that of the cnido-glandular tract. Compare further the section IV. 



As for the reproductive organ.s, they appear on protomesenteries 2 and on the metamesenteries 

 of the ist and 2nd cycles within the craspedia regions. Mc. Murrich states 1891 that the species 

 would be found bi-sexual. This is certainly not the case. Both in Aller's specimens from Beau- 

 fort and in a specimen from the same locality belonging to the Museum of Copenhagen both male 

 and female organs were found in the same individual. Certainly the ova in several specimens were 

 comparatively small, but there is no possible doubt that the species is hermaphrodite. Mc. Mur- 

 rich's statement is consequently based either on insufficient examination or else on the fact, that he 

 chanced upon a specimen whose male organs had not been developed. For the Cerianthidae are 

 probably proterandrous hermaphrodites. Still it seems to me that even in this case at least traces of 

 the male organs ought to have been observed. 



