1882.] CLASSIFICATION OF THE COMATtfL^.. 739 



beyond the palmars. It should read 3A'DP(P') ^. There may be 

 yet another in Act. bennetti and the number of arms reach 70 or 80, 

 which is only very rarely the case among the Comatulce. Prof. 

 Bell's formula (SA'RDP 2), however, only provides for 40, which is 

 a very common condition. The formula should therefore be 

 3 A'DPP'(P") ^, The same may be said of Act, schlegeli, the only 

 known example of which has two axillaries beyond the palmars and 

 over 80 arms. Its proper formula would be 3 A'DPP'P" ^, while 

 Prof. Bell gives 3 A'RDP ^^, which provides for less than half 



this number of arms, so that one of the most evident specific char- 

 acters is not taken into account at all. 



The ray-divisions of Act. pulchella have been thus described : — 

 "10-20 arms, most of the rays usually dividing twice, the first 

 division (distichals) consisting of two joints which are not united by 

 syzygy. "When the arms spring directly from the radial axillary, 

 the two lowest brachials are united by ligament, as in most Coma- 

 tulce, and the third is a syzygial or double joint. But in all the 

 arms which spring from a distichal axillary the two lowest brachials 

 are united by syzygy to form a double joint ; and the true third 

 brachial, which is also a syzygial joint, as in all Comatulce, thus 

 becomes the second arm-joint, as in Act. Solaris"^. Two formulae 

 are necessary for this type — one for the ten-armed form, and another 



for that with divided primary arms. The first would be 3 Ag, and 



the second \/l.2A'^ Prof. Bell, however, writes (1.2)3 A'g, thus 



taking no account of the presence of distichal axillaries in some 

 (often all) of the primary arms ; and his formula also implies that 

 the first, second, and third brachials of the same arm may all be 

 syzygial joints, which is never the case. 



One very serious objection to Prof. Bell's system of shorthand is 

 that (except in one case) it is only applicable to those Comatulm in 

 which the ray-divisions are regular, i. e. with the second and sub- 

 sequent divisions all resembling the first. For regular forms like 

 Act.parvicirra, Act. bennetti^ and their allies, which have three disti- 

 chals and three palmars with a syzygy in each axillary. Prof. Bell's 

 notation is probably as short a one as could be devised ; though it 

 gives no information respecting the number of joints in each series, 

 and does not always indicate whether palmars are pres ent o r not. 



Thus, for example, his formula for Ant. articulata is s/'SK^. This 



means that distichals are j)resent but do not " present a sygygial 

 joint." It does not indicate, however, whether the axillary is the 

 fourth joint (an actual case in another species), the first (which I 

 have never met with), or the third, or the second (as is really the 

 case). But no information is given at all respecting the presence or 



' Biill. Mus. Oomp. Zool. vol. ix. no. 4, p. 10. October 1881. Were I writing 

 now, I should insert the word "nearly" before "all i!omatid(S'' in the last 

 sentence. 



