MONOGRAPH OF THE EXISTING CKINOIDS. 103 



ThaZassometra (fig. 223, p. 175, and part 1, fig. 95, p. 157), in C'rlnometra (figs. 678, 

 683, p. 338), and in AteUcrinua (pai-t 1, figs. 14, p. 65, 123, p. 192). 



In very many species it is noticeable that the lateral angles of the axillaries 

 project beyond the sides of the ossicles preceding ; or that, if the axillaries are long, 

 their sides make with the sides of the preceding segments a more or less deep obtuse 

 reentrant angle. The first condition is well seen in Antedon bifida and in A. 

 petasus (part 1, fig. 103, p. 165) ; the second in Fontiometra andersoni (fig. 194, 

 p. 120), and in many of the Comasteridse, Himerometridse (fig. 123, p. 79), Maria- 

 metridse, and Colobometridae. This appears to be, like the intersegmental pores, a 

 provision for insuring an adequate supply of fresh water to the sides of the disk. 



POST-RADIAL HOMOLOGIES. 



Fimt T)rachial pair. 



The first two ossicles of the free undivided arms of the comatulids always 

 differ in size, form, ornamentation (if present), and mode of union from all 

 the succeeding brachials; the outer (second) is always situated just on the border 

 line between the perisome of the ventral disk and the ventral perisome of the 

 arm and bears a highly specialized pinnule; and the two following brachials are 

 always united by syzygy, forming the first syzygial pair. 



The constant structural peculiarities and associations of this pair of brachials, 

 the larger size, greater width, straighter proximal and distal borders, which are 

 more nearly at right angles to the axis of the arm, and consequently more regu- 

 larly wedge-like shape, and the possession of special ornamentation or modifi- 

 cation, ventrolateral, or dorsolateral flanges, more deeply and sharply flattened 

 sides, or synarthrial tubercles, combined with the presence of the only synarthry 

 occurring in the undivided arm between them, their invariably fixed position on 

 the border of the disk, the occurrence upon the outer of a highly specialized 

 pinnule differing widely in size, structure, and function from the succeeding 

 pinnules, and the invariable presence of a syzygial pair beyond them, leave no 

 room for doubting that this pair of ossicles is strictly homologous throughout the 

 whole group, and furthermore render it very readily identifiable. 



With the definite identification of a pair of ossicles wherever it may be 

 foimd, the elucidation of the brachial homologies of the comatulids becomes an 

 easy matter. 



But the study of the post-radial homologies is sometimes rendered more or 

 less difficult by the projection of certain of the characters of the first brachial 

 pair, though always in a modified form, upon several of the succeeding ossicles, 

 which then form a transition zone of greater or lesser length between the first two 

 and the distal brachials. Especially the pinnule upon the outer element of the 

 syzygial pair immediately following the outer element of the first brachial pair 

 usually resembles more or less closely the pinnule borne by the latter; but the 

 succceeding pinnules are ordinarily markedly different, although in large species 

 (as in the species of Heliometrinre) they may have some of the characteristics of 

 142140— 21— Bull. 82 9 



