REPORT ON THE CRINOIDEA. 41 



V.— CLASSIFICATION. 



Until the time of Johannes Miiller the number of recognised species of Comatulae was 

 extremely small, not more than a dozen, in fact. Retzius had described one, Linnaeus 

 two, Lamarck seven, and two more bore the names of Diiben and Koren ; but only 

 three of them had more than ten arms, viz., Comatula rotalaria with about twenty to 

 twenty-two, Comatula Jimbriata with twelve to thirty, and Comatula midtiradiata, with 

 forty to fifty. 



Under these circumstances the classification of the Comatulae presented no difficulties. 

 But Midler's descriptive work 1 raised the total number of species to nearly forty, 

 about half of them having more than ten arms. This very obvious character afforded 

 him the means of separating his species into two groups, which he further subdivided 

 according to the arrangement of the syzygies in the arms. Thus, for example, there 

 are two sets of ten-armed Comatulae, those like Actinometra pectinata (PI. LIII. fig. 15) 

 in which the two joints above the radial axillary are each traversed by a syzygy, and 

 those like Antedon eschrichti (PI. XXIV. figs. 10, 11) in which the first syzygy is in the 

 third brachial. In like manner the multibrachiate forms were separated by Midler into 

 two sets, those in which the brachial axillaries are syzygial joints (PI. LXVIII. fig. 2), and 

 those in which the axillaries are simple and not traversed by syzygies (PI. XLV. fig. 2). 



All the Comatulae known to Midler could be placed in one or other of these four 

 sets, no matter to which of the two subgenera they belonged, Alecto or Actinometra. 

 He never made any definite attempt to separate the species of Alecto from those of 

 Actinometra, no apparent system being determinable, either in the order of his 

 specific descriptions or in his tabular arrangement of most of the species in the 

 form of a key, a species of Actinometra not unfrequently intervening between two 

 of Alecto. For more than a dozen years after the publication of Midler's memoir the 

 classification of the Comatulae remained practically as he left it. No one took up 

 the subject, and no new species were described. In the year 1862, however, a step in 

 advance was made by Messrs. Dujardin and Hupe. 2 They divided the recent Feather- 

 stars into three genera, Comatula, Lamarck, Actinometra, Midler, and Comaster, Agassiz, 

 the last named being a type which Midler had been unable to recognise as generically 

 distinct from Comatula. Of his own subgenera, Alecto and Actinometra, the 

 latter was raised into a genus by Dujardin and Hupe, who referred to it three species, 

 while they limited Lamarck's name Comatula to the forms previously referred by Midler 

 to Alecto, and regarded them as constituting thirty-one species. These were divided into 

 groups having respectively ten, ten to twenty, twenty, twenty-six to forty, and more than 



1 Abhandl. d. k. Akad. d. Wiss. Berlin, 1849, pp. 237-265. 2 Op. cit., pp. 192-213. 



(zool. chall. exp. — rART lx. — 1888.) Ooo 6 



