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



exigua (PL XXX. fig. 1 ; PL XXXII. fig. 4) ; while iu the group of species allied to 

 Antedon eschrichti they are more or less flagellate, consisting of a large number of 

 relatively short joints (PL XXIV. figs. 1-3, 7-9 ; PL XXV. figs. 1-3 ; PL XXVII. 

 figs. 8, 9). There is much less variation in this respect in the oral pinnules of Actino- 

 metra, which are always provided with a terminal comb (PL LIII. figs. 3-6), a character 

 which never occurs in Antedon. 



On the other hand, the sacculi which are almost invariably present in this genus, 

 never occur in Actinometra, even when species of the two genera are living side by side 

 in the same locality ; and this fact is a very strong argument against the theory of Vogt 

 and Yung l that the sacculi are symbiotic Algae, as I have explained elsewhere. 2 There 

 are a few species of Antedon, e.g., Antedon quinquecostata, in which they are small and 

 poorly developed, though they are abundant in others obtained at the same localities ; 

 and in some other instances the condition of the specimen has been such that I have not 

 been able to assure myself satisfactorily of the presence of sacculi. This is the case for 

 example with the two specimens of Antedon abyssicola from 2900 fathoms, the greatest 

 depth at which Comatulas have been obtained ; but they are fairly abundant in another 

 individual of the same species from 2600 fathoms in the Southern Sea. There are few 

 species of Antedon in which they are not present ; though they are more variable in their 

 occurrence among the species of Eudiocrinus, as has been already explained. 



The very definite relation of the sacculi to the side plates of the ambulacra in those 

 species of Antedon which have a highly differentiated ainbulacral skeleton is a very 

 strong argument against the views of Vogt and Yung that they are symbiotic Algae. It 

 was pointed out on p. 127 of Part I. how the distal edges of the side plates are notched 

 for the reception of the sacculi, and figures were given on pi. liv. illustrating this 

 character in four species of Antedon. But it does not occur at all in species of Peata- 

 crinus and Metacrinus which live at the same localities as these Coniatuhe (Stations 170a, 

 175, 192, 214), and I cannot think, therefore, that the problem of the nature of the sacculi 

 has been solved by Vogt and Yung. These are not the only difficulties which suggest 

 themselves in connection with the details of their theory as I have explained elsewhere. 3 



Classification. — It has been shown on a previous page how the numerous recent 

 species of Antedon may be associated together into groups of variable size, according to 

 the characters of the rays and of their subdivisions. The first group to be considered 

 includes those species in which the two outer radials are united by a syzygy and not, as 

 is most frequently the case, by a bifascial articulation. Five of the eight recent species 

 of Pentacrinus are distinguished by this character, and it occurs in several species of 



1 Op. cit., p. 570. 



2 On the Supposed Presence of Symbiotic Algae in Antedon rosacea, Quart. Journ. Micr. Sci., 1S87, new sen, 

 vol. xxvii. p. 386. 



3 Ibid., pp. 380-384. 



