164 Dr. F. A. Bather— A New OpJdimd 



The ai^pearance of a transverse division in the vertebra, between 

 the proximal and distal processes, suggests that each segment really 

 consists of two ambulacral pairs, of which only the distal one is 

 associated with adambulacrals, the proximal pair of adambulacrals 

 being suppressed. Each of these supposed ambulaci'al pairs 

 possesses a single pair of lateral processes, but those of the distal 

 pair are partly overlaid by the adambulacral elements. In the 

 distal region of .the ray the proximal and distal halves of th& 

 vertebra are distinguishable; but nearer the disc and within it they 

 have come closer together, the transverse division has disappeared, 

 and the side-processes are reduced to structures which appear as 

 bars, but are probably flattened laminse. Thus the vertebrae seen 

 in ray v are, on this explanation, of compound origin. 



The preceding interpretation may be compared with Dr. Gregory's 

 interpretation of the structure in the Lower Devonian Ophiurid, 

 BiindenbacMa ' (see our Fig. 5). Here " the body of the [ambulacralj 

 ossicle is apparently divided into two pieces by a transverse 

 depression. The adambulacral plates are small and narrow, and 

 support a triangular spine-bearing plate." If, however, as seems 

 legitimate, this last plate be regarded as itself an adambulacral 

 element, then each segment will consist of two ambulacral pairs, 

 and two adambulacral pairs of which one only is spiniferous. 

 The fact that in Bundenbacliia the ambulacrals still alternate does 

 not affect the composition of the segment. Biindenbachia is not to 

 be regarded as a direct ancestor of the contemporary Epphaven 

 fossil, but it suggests the possibility that, during the race-history of 

 the Ophiurids, there may have arisen forms with a tendency to 

 suppression of alternate adambulacrals and to fusion of adjoining 

 ambulacrals. 



In Biindenbachia there appear to be preserved both adambulacral 

 and ambulacral elements of the subsequently suppressed or coalesced 

 segment. The Silurian Protaster SedgwicM appears to retain the 

 ambulacral elements less reduced than in Biindenbachia, while it 

 has lost the adambulacral elements. In the proximal arm-segments, 

 however, as figured by Dr. Gregory (op. cit., fig. Ic), the existing 

 adambulacrals are forked, a peculiarity suggestive of their compound 

 origin. 



Now comparison of the vertebrae of recent Ophiurids with the 

 arm-segments, especially the proximal ones, of our fossil leaves 

 no room for doubt that the depression marked p is for the reception 

 of a podium, while that marked q is for the insertion of the ventral 

 inter- vertebral muscles. A single arm-segment of a recent Ophiurid, 

 such as Ophiarachna or Amphiura, is therefore homologous with any 

 one of the whole segments here numbered 1-8. If one of these is 

 really composed of a distal and a proximal half (lettered a and b 

 respectively in Figs. 1 and 2) the same statement would apply to 

 the recent Ophiurid, and some trace of this origin might be seen in 

 the development. The observations of Professor Ludwig - and 



1 Proc. Zool. Soc. London, 1896, p. 1033 ; 1897. 

 - Zeitschr. wiss. Zool., vol. xxxvi, p. 181 ; 1881. 



