ECHINODERMA. 13 



In the original diagnosis of the genus I spoke of " a large keel-like plate on the 

 disc, which completely overshadows the radial shields," although I was aware of and 

 figured (Plate XXVL, fig. 3) an angle of the disc in which there appeared to be two large 

 radial shields ; in this case, however, three of the angles of the disc conform exactly to 

 the diagnosis. 



I have now before me a specimen in which all five angles have a pair of plates. 

 Were it not for the ' Southern Cross ' specimens it would not be possible to put this 

 specimen with the genus OjMosteira at all. 



When we are asked whether " specific characters are useful," we may retort that 

 generic characters even are not always constant. 



Winter Quarters, 13 fms. Coulman Island, 100 fms. 



Ophionotus victoriae. 



Opliionotiia vidoriae. Bell, Rep. ' Southern Cross ' (1902) p. 219. 



This species was not so abundant ; it was taken off the Barrier at 300 fms., near 

 Franklin Island, and from 254 fms. at an unrecorded locality. 



Ophiacantha imago. 



Ophiacantha immjo^ Lyman, ' CbalL' Rep. Ophiur. (1882), p. 180 iUgue dtatum. 



I hope I am right in referring to this species specimens from Winter Quarters, 

 30 fms. ; Hut Point, 77° 12' 12" S., 167° 27' W., 77° 50' 30" S., 165° 40' K, 100 fms. \ 

 but, in sooth, some of the ' Challenger' types of Ophiuroids are hardly mature. 



Ophiacantha vivipara. 



Ophiacantha vivipara, Ljungman, Ofv. Vet. Akad. Forh. 1870, p. 471 ; E. A. Smith, Phil. Trans. 108 

 (1879), p. 278, pi. xvii. fig. 3. 



A number of authors have mentioned this species, but, since Dr. Liitken was cited 

 by Mr. Edgar Smith, I.S.O. (Phil. Trans., vol. 168, p. 278), as dou])tful of Ljungman 's 

 original locality (" Altatam urbem mexicanam "), none seems to have remarked on 

 the apparent peculiarity of the distribution of this species. 



Prof. Thcel has been so good as to let me see Ljungman's Altaian specimen, and 

 at the same time to inform me that the types of Ljungman's OpJiiacantha rivip/ira were 

 Ijrought home from two quite different localities, Altata and Falkland Islands. 

 Ljungman does not give (op. cit.) the latter locality, but Liitken (Zool. Record 1872, 

 p. 448), threw grave doulits on the i\Icxican origin of the specimens, and suggested that 

 Patagonia was the "true habitat of the typical specimen." What is really more 

 interesting is the question whether the brood-pouch habit of a given species is found 

 in the confines of the tropics as well as in the colder waters of the globe.* 



* For a list of the viviparous Echinodcnus of warmer waters, and for tlicir preponderance in Arctics and 

 Antarctic Seas, cf. Ludwig, Zopl. JB. Suppl. Bd. vii. (1904), [i. 684. 



VOL. IV, R 



