OPHIACANTHIDAE 249 



possible transitions between the forms with the disk covered by a uniform granulation 

 and those with numerous spines on the disk. Although these 5-armed specimens gener- 

 ally have more numerous spines on the disk besides the granules, there is thus no reason 

 for distinguishing them from the typical 6-7-armed O. vivipara ; and also in the other 

 characters they agree, on the whole, with the typical vivipara. Still the matter is not 

 quite so simple. 



It is undeniable that these 5-armed specimens in general have a much more robust 

 appearance than vivipara (cf. Plate VII, figs. 2-4); also the arms are longer and more 

 robust. This can hardly be due simply to the fact that they have fewer arms than the 

 typical form. Then it is a very noticeable fact that this large 5 -rayed form is not met with 

 among the specimens from the neighbourhood of the Falkland Islands, but only very 

 far south, from the Palmer Archipelago to the Graham Land region. (Koehler's 5-rayed 

 specimens are also all from this southern region.) If it were simply an individual varia- 

 tion of the 6-7-rayed vivipara, it would be hard to understand why such forms 

 should not occur equally commonly also in the Falkland region (the single 5-rayed 

 specimen I have seen from there is only a very young one). The typical vivipara also 

 occurs in the more southern region, together with the 5-rayed form; but whereas the 

 5-rayed form is of common occurrence in the south, it apparently does not occur farther 

 north. Further, there are among these large 5-rayed specimens several males, whereas 

 no male specimens were observed among the typical 6-7-rayed specimens. None of the 

 female specimens of the 5-rayed form have young ones in the bursae, so it is quite 

 possible that this form is not viviparous — at least there is no proof that it is viviparous. 



Thus, in my opinion, it is not justifiable simply to identify these specimens with the 

 typical 6-7-rayed O. vivipara. If it is really non-viviparous, it must represent a separate 

 species, but so long as we do not know this for certain, and in view of its resemblance 

 with vivipara in general structure, I think it the safest course for the present to designate 

 it as a variety of O. vivipara. 



From O. rosea, with which there is much general resemblance, it is distinguished 

 particularly by the outer mouth papillae being simple, not forming a cluster at the outer 

 mouth tube foot. From O. detisispina it diflfers in having both granules and spines on 

 the disk, and in the arm spines being more smooth and not joining in the dorsal median 

 line. Also the shape of the mouth shields is somewhat different (Figs. 5-6). 



Ophiacantha densispina, n.sp. 



(Plate VII, fig. i) 



St. WS 99. 19. iv. 27. 49° 42' S, 59° 14' W, Falkland Islands, 251-255 m. i specimen. 

 St. WS 248. 20. vii. 28. 52° 40' S, 58^30' W, Falkland Islands, 210-242 m. i specimen. 

 St. WS 825. 29. i. 32. 50° 50' S, 57° 15' W, Falkland Islands, 135-144 m. 2 specimens. 

 St. WS 840. 6. ii. 32. 53° 52' S, 61° 49' W, Falkland Islands, 368-463 m. 3 specimens. 



The type specimen (from St. WS 248) is a large, coarse specimen, 16 mm. in diameter 

 of disk. The arms are all broken, but judging from their size they must have been at 

 least about five to six times the diameter of the disk; they are somewhat flattened, 



7-2 



