218 ACTINIiE COLLECTED IN THE NORTH SEA 



in preserved specimens. B. longicornis is not easy to keep alive, and 

 soon after capture usually protrudes the stomach walls to an enormous 

 extent, collapses to an abject flatness and dies. 



RHODACTINIA CRASSICORNIS (0. F. Miiller). 



In 1902, 0. Carlgren, in his paper on the Actinice of the Olga expedi- 

 tion considered that the name Urticina (Tcalia) crassicornis really 

 covered three species belonging to two genera, i.e., Rhodactinia crassi- 

 cornis (0. F. Mull.), Tealia coriacca (Cuv.), and Tealia lofotensis 

 (Dann.). 



Unfortunately I did not see this paper until after the voyages, and 

 regarded all the forms obtained as varieties of Urticina crassicornis. 

 merely making notes on colour and external characters at the time of 

 capture. In these notes I however distinguished three forms : (a) U, 

 crassicornis, the normal form of the littoral area ; (h) the large deep-water 

 form, the Tealia tuhcrculata of Cocks and Cunningham , and (c) a form 

 with small warts, occurring on the Great Fisher Bank. A subsequent 

 anatomical examination of such specimens as were preserved shows this 

 last to be B. crassicornis and the other two to be forms of T. coriacea. It 

 is thus almost impossible to assign all the numerous records to their real 

 species, only those cases where my notes actually notice the size of 

 the warts and the preserved specimens can be safely noticed. 



B. crassicornis occurred at several stations during Voyage XC about 

 the region of the Great Fisher Bank, together with B. longicornis and 

 Chondraciinia digitata. 



Almost all the specimens were large, having an expanse of from 12 

 to 14 cm. They were attached to valves of Cyprina islandica and 

 Modiola, both dead and living, and Fusiis antiquus, either living, empty, 

 or tenanted by Eupagurus hernhardus, and once upon the shell of a 

 living F. turtonis. 



The small warts which beset the upper part of the column are 

 arranged in irregular vertical rows, and during partial contraction the 

 arrangement frequently appears annular. The colouration is very 

 variable ; the following were noted : — 



(«) Much resembles Gosse's description and plate of Bolocera eques. 

 The margin was frequently not retracted, even when the tentacles had 

 been withdrawn. The disk, however, was never " pellucid," nor was 

 the scarlet tentacular ring bounded by white, as described by Gosse. 



When I first saw this I took it to be B. eques. 



ih) Column dull orange, the summit white. Tentacles dull pink, 

 the scarlet ring indistinct. Disk pale orange. 



