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



I have not inserted these in any of the above schemes, since my work on the " Blake " 

 Comatula? is not yet complete ; but they will be fully described and illustrated in the 

 Report on the Collection, which will be published as one of the Memoirs of the Museum 

 of Comparative Zoology at Harvard College. 



Antedon, Series III. 



Two articulated distichals. 



Remarks. — This series includes all those multibrachiate species in which there are but 

 two distichal joints united by a bifascial articulation as the two outer radials are (PL XL. 

 fig. 1 ; PL XLIIL). The palmar and post-palmar series, when present, are normally of 

 the same character, though, like the distichals, they may be occasionally replaced by a 

 three-jointed series with a syzygy in the axillary. This, however, is far less common 

 than in Actinometra. The individuals of Antedon similis and Antedon occulta 

 represented in PL XL VII. fig. 3, and PL XLVIII. fig. 1, have their distichal and palmar 

 series very regularly constructed ; while in those of Actinometra elongata and of 

 Actinometra valida (PL LVII. fig. 2 ; PL LIX. fig. 3) two and four of the bidistichate 

 series respectively are replaced by sets of three joints, the axillary with a syzygy. 



The bidistichate forms of Antedon make up the largest series, next to that of the 

 ten-armed type, of all those into which the species of this genus naturally arrange them- 

 selves. It may be divided for the purposes of formulation into three groups, according 

 as there are one, two, or three axillaries respectively above the radials, as is shown on 

 pp. 54, 55. Several species of course appear in two groups, owing to the occasional 

 total absence of palmar or post-palmar axillaries in some individuals ; while others may 

 sometimes have only ten arms. Thus, for example, there are now and then no distichal 

 series at all in certain individuals of Antedon brevipinna, Antedon duplex, Antedon 

 Jlexilis, and Antedon lusitanica (PL XXXIX. figs. 1, 3). In like manner Antedon 

 pourtalesi and Antedon quinquecostata always have one and sometimes two post-radial 

 axillaries ; while there are always two and sometimes three in Antedon palmata and 

 Antedon spinifera. 



While, therefore, it is useful to have a system of formulation which gives some 

 information as to the extent of the ray-divisions, too much stress must not be laid upon 

 the presence of a palmar or post-palmar axillary as an aid to classification. In the list 

 given on pp. 54, 55, I have arranged the bidistichate species of Antedon in three groups, 

 according to the frequency of the ray-divisions. There are certain occasions in which 

 such a mode of grouping them is of considerable use ; but one must always bear in mind 

 that a species with (say) three post-radial axillaries which appears new at first sight, may 

 in reality be identical with one which is already known, but has never yet been found 

 with any axillary beyond the palmars, e.g., Antedon tuberculata (PL XLV. fig. 2). The 



& 



