78 THE VOYAGE OF H.M.S. CHALLENGER. 



taken the trouble to make himself sufficiently acquainted with the works of his pre- 

 decessors, and has therefore committed himself to various statements which will not bear 

 investigation. 



I do not know what grounds of fact he has for his assertion that the species of 

 Actinometra swim but little. If, as I believe, it is merely an inference from the supposed 

 small size of the muscular bundles between the arm-joints, his premises are wrong, as I 

 have explained above ; while I do not know that he has ever been able to observe living 

 species of the genus, and to notice their abstention from the performance of swimming 

 movements. On the other hand, Professor Semper has kept various forms of Actinometra 

 in an aquarium for weeks together, and his observation of the regular alternating 

 movements of their arms while swimming was mentioned by myself as long ago as 

 1877. 1 I pointed out in the same memoir, and again five years later'" that the cirri of 

 Actinometra are few in number, and almost entirely limited to the margin of the 

 discoidal centro-dorsal ; while those of Antedon are numerous and more or less extensively 

 distributed over the under surface of the centro-dorsal. But yet Perrier tells us that 

 ' ' surtout les Actinometra " as compared with Antedon are adapted to fixing themselves 

 by their cirri. The extreme inconsistency of this assertion with the real facts of the case 

 becomes still more apparent, when it is remembered that in many species of Actinometra 

 the cirri borne on the centro-dorsal duricg early life drop off, and their sockets 

 become gradually obliterated (PI. LTV. figs. 1-9 ; PL LXV. figs. 1-6). It was 

 mentioned in my preliminary Eeport 3 that I had found the centro-dorsal of many 

 Actinometra-s\>ecies to be in the form of a simple flat plate, more or less stellate in form, 

 but entirely devoid of cirrus-sockets ; while in other individuals only a few imperfect 

 sockets are present, owing to their not having been completely obliterated. The 

 occurrence of a fossil Actinometra presenting these characters was also noticed ; 4 and 

 other references were made to this peculiarity as it was found in a successively increasing 

 number of species of the genus. 5 Copies of the papers in which this character was 

 described were sent to Professor Perrier, who seems nevertheless to be altogether 

 unacquainted with its occurrence. For it is difficult to see how Actinometra paucicirra or 

 Actinometra divaricata (PL LIV. fig. 1 ; PL LXIII. fig. 8), with its perfectly flat centro- 

 dorsal entirely devoid of cirri, can be regarded as one of those Comatulae which are 

 especially "organises pour s'accrocher solidement aux corps sous-marins." 



After making these somewhat ill-considered remarks, Perrier goes on to describe the 

 disc of Eudiocrinus atlanticus, which is not more than 5 mm. in diameter and is thus 

 very small in proportion to the size of the arms, which attain 120 mm., while the cirri 

 are from 15 to 20 mm. long. 6 Perrier then adds " II resulte de ce que nous venons de 



1 Journ. Linn. Soc. Lond. (Zool.), 1877, vol. xiii. p. 446. * Bull. Mus. Gomp. Zool., 1882, vol. ix. No. 4, p. 13. 



3 Proc. Roy. Soc, 1879, pp. 389-391. 4 Quart. Journ. Zool. Soc, 1880, vol. xxxvi. p. 51. 



6 The Comatulae of the Leyden Museum. Notes from the Leyden Museum, 1881, vol. iii. pp. 196, 208. 

 QComptes rendus, 1883, t. xcvi. No. 11, p. 727. 



