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



forty arms. But the French authors altogether gave up Midler's method of grouping 

 both the ten-armed and the multibrachiate Comatulse according to the arrangement 

 of the syzygies in the arms and their subdivisions, placing for example Comatula 

 brachiolata and Comatula Solaris with syzygies in the first and second brachials between 

 Comatula adeonse and Comatula echinoptera, both of which have the third brachial a 

 syzygy. In like manner Comatula fiagellata with no syzygies in the brachial axillaries 

 is placed between Comatula japonica and Comatula timorensis, in both of which the 

 axillaries are syzygial joints. While therefore Dujardin and Hupe made a distinct 

 advance on Miiller's classification in recognising two generic types of Comatulse, 

 their rejection of the characters on which he relied, and rightly so, as being of much 

 systematic value was a decidedly backward step. For all subsequent work has shown 

 that the position of the first syzygy in the free arms and the presence or absence 

 of syzygies in the brachial axillaries are characters of very considerable systematic value, 

 without the aid of which the classification of the hundred or more species comprised in 

 each of the genera Antedon and Actinometra would be even more chaotic than it is. 



For some fifteen years after the appearance of Dujardin and Hupe's Histoire 

 Naturelle, systematic work on the " Comatulse progressed with extreme slowness, the 

 most important step being Norman's restoration of the generic name Antedon, owing to 

 its priority over both Comatula and Alecto. 1 New species were described by Bohlsche, 

 Grube, and Pourtales ; but they were never figured, and no attempt was made to assign 

 them places in the system either of Midler or of Dujardin and Hupe. Dr. Liitken had 

 examined from time to time a considerable number of Comatulse which had been collected 

 among the Pacific Islands by the agents of the Godeffroy Museum ; and he arrived at 

 the conclusion that the real distinction between Antedon (or Alecto) and Actinometra lies 

 in the central or excentric position of the mouth, the number of groove trunks reaching the 

 peristome being a character almost entirely devoid of the systematic importance attributed 

 to it by Miiller. Liitken's views were never published, and I only learnt of his holding 

 them after myself arriving at the same conclusions ; but he was good enough to inform 

 me at the same time of a character then unknown to me, which I have since found to be 

 of almost invariable occurrence in Actinometra, viz., the presence of a terminal comb on 

 the lower pinnules (PL LVI. figs. 2, 4). These facts were published in my memoir on 

 Actinometra, 2 where I also endeavoured to classify the species of the genus that I had 

 been able to identify, by an extension of the method employed by Miiller. 



While recognising the systematic importance of the presence or absence of syzygies 

 in the arms of Comatulse, Miiller made no attempt to classify the multibrachiate forms 

 according to the number of joints between the successive brachial axillaries, though he 

 furnished the means for doing this in his descriptions of many species, a process which 



1 Ann. and Mag. Nat. Hist, 1865, ser. 3, vol. xv. p. 98. 



2 Trans. Linn. Soc. Lond. (Zool.), ser. 2, 1877 [1879], vol. ii. pp. 18-29. 



