EEPORT ON THE CRISTOIDEA. 209 



converse is also equally true. All the three Challenger specimens of Antedon occulta 

 have post-palmar axillaries (PI. XLVIII. fig. 1 ; PI. XLIX. fig. 3), but it is quite 

 possible that other examples of the type may be eventually discovered in which these 

 are absent. It would be premature, however, to describe them as new, simply because 

 they did not agree with any of the species which had no axillary beyond the palmars. 

 In all such cases as these, and we shall meet with them frequently, the general characters 

 of the type must be carefully taken into consideration, apart from the frequency of its 

 ray-divisions, which is often much greater in some individuals of a species than in others ; 

 thus, for instance, examples of Actinometra parvidrra have been described with thirteen 

 and with thirty-nine arms respectively, palmar axillaries being altogether absent in the 

 former and abundantly developed in the latter. 



It is therefore desirable, so far as may be possible, to employ some other means of 

 classifying any particular series of multibrachiate Comatulaa than one which is based 

 solely on the number of post-radial axillaries ; and in the present case there is no difficulty 

 in effecting this object. 



We have seen that the bidistichate species of Antedon fall into three sets according 

 as there are one, two or three axillaries above the radials. Each of these sets contains 

 species which belong to two very different types of structure, the one with and the other 

 without an ambulacral skeleton. Most of the forms which have a distichal but no palmar 

 axillary (A. 2) resemble the members of the Basicurva-gvou]) in the ten-armed series in 

 the presence of an ambulacral skeleton and in the straight-edged and wall-sided nature 

 of the radial axillaries and of the next following joints. In fact, two members of this 

 group (Antedon flexilis and Antedon lusitanica) have been already noticed in connec- 

 tion with the Basiciirva-growp, owing to the occasional absence of any distichal axillary 

 (PL XXXIX. figs. 1-3 ; PI. XLIL). In like manner Antedon quinquecostata and 

 Antedon pourtalesi always have a number of distichal series, but these are not always 

 followed by palmars ; whde I have seen some examples of Antedon spinifera with one or 

 more post-palmar series, and others in which there are no axillaries beyond the palmars. 



All these species, together with others which will be described immediately, are the 

 bidistichate representatives of the Basicurva-group, and they form a very natural 

 assemblage which may be conveniently designated as the Spinifera -group. It thus includes 

 a variety of specific types which may have one, two, or three axillaries beyond the 

 radials, Antedon spinifera having sometimes two and sometimes three, and being the first 

 species in which an ambulacral skeleton was described. The remainder of the bidisti- 

 chate species of Antedon, which have neither an ambulacral skeleton on the pinnules nor 

 distinctly wall-sided rays, represent the 3£ilberti-gvou\) among the ten-armed species of 

 the genus in the characters and relations of their lower pinnules, and they may be 

 conveniently designated as the Pa hna to-group. Like the Sjnnifera-grou^ it includes 

 species with one, two, or three post-radial axillaries, but one or more of the pinnules 



(ZOOL. CHALL. EXP. — PART LX. — 1887.) OoO 27 



