310 ANGELOPSIS DILATA. 



hypocystic cavity is significant for classification, because in Fewkes's specimens, 

 which were almost certainly mature, it was not only much more extensive but 

 more nearly subdivided than it is in the "Albatross" specimen. 



The only account of Haeekel's species which has yet appeared, is his sum- 

 mary of its more important generic characters, and the statements that it is 

 allied to Fewkes's Angelopsis glohosa, that it resembles Stephalia corona Haeckel 

 externally, and that its tentacles resemble those of Rhodalia. This brief notice 

 is of course entirelj' insufficient for specific diagnosis. But as it applies per- 

 fectly well to A . glohosa, so far as it goes, and inasmuch as there is no geographic 

 barrier between the localities of capture of the two, it is probably better to unite 

 them. 



The specimen in the "Albatross" collection agrees in .general structure 

 with A. glohosa, so far as the latter is known, but differs from it, as already 

 mentioned, in the form and extent of the hypocystic cavity. Unfortunately, 

 however, it is somewhat fragmentary. We have here another of those cases, 

 so commonly encountered by the student of the pelagic Coelenterata, where 

 it is difficult to decide whether the cause of science is best served by creating a 

 new species, by referring the specimen to an old species on doubtful grounds, 

 or by leaving it without specific identification. The difference in the hypo- 

 cystic chamber is probably sufficiently important for recognition and may there- 

 fore justify a new species. But I must add the warning that research on better 

 material may well prove it unfounded; therefore it should not be used as an 

 instance of geographic distribution until tested further. 



Angelopsis dilata, sp. nov. 

 Plate 21, figs. 6-8; Plate 22. 



Station 4641 633 fathoms to surface. 1 specimen 12 mm. high by 6 mm. 



in diameter. Type preserved 

 in formalin. 

 In the single specimen all the nectophores, except two very young ones, 

 and most of the cormidia are lost. Fortunatelj^ however, it is in sufficiently 



m 



good histological condition to allow an investigation of the aurophore. After 

 being photographed (Plate 21, fig. 6), one of the cormidia was detached for 

 study, and the corm was divided longitudinally a little to one side of the mid- 

 plane; one of its parts was sectioned in the radial, the other in the transverse 

 plane. 



