^Ol- 55'] SILURIAN ECHINOIDEA AND OPHIUROIDEA. 705 



definitely indicated by the nature of the buccal apparatus, the in- 

 vestigation of the ambulacra acquires additional interest, since it 

 affords an opportunity of discovering the precise stage at which these 

 important organs had arrived in the course of evolution during early 

 Palaeozoic times. The structure of the ambulacra is more compli- 

 cated than has been supposed, or than is suggested by a study of 

 the fossil in its usual state of preservation. All that can be usually 

 seen in ordinary casts of the oral surface is a furrow, which 

 represents the ambulacral groove, whence the plates have wholly 

 disappeared, leaving, however, impressions, which prove clearly that 

 they did not lie flush with the plates of the interradii. In better- 

 preserved examples the furrow is seen to be bordered, even on the 

 oral side of the test, by a row of pores, 0'19 mm. in diameter, 

 usually single, more rarely double, situated at the point of flexure, 

 where the ambulacral plates bend inward to form the ambulacral 

 groove. The form of the groove in transverse section, as shown in 

 fig. 9, can readily be made out by taking a cast in wax, and slicing 

 this across. The ambulacra do not appear to have been limited to 

 the oral surface, since, iu addition to the impressions which regularly 

 radiate over this aspect, fragments of others may be detected : these 

 can only have been derived from the aboral surface which has been 

 squeezed down into contact with the oral. In the examples just 

 described the ambulacra present a remarkable resemblance to those 

 of an Asteroid, although I think the pores do not, as Dr. Gregory 

 supposed, lie between adjacent ambulacral plates, but are confined 

 each to a single plate, as in modern urchins. 



In unusually well-preserved specimens additional features may be 

 observed. A shelf of the matrix, bearing on its outer surface the 

 impressions of additional plates, is found to roof over the ambulacral 

 furrow, and partly conceals the ambulacral plates or their casts, 

 which dive beneath it and appear as a row of slanting tubes 

 (figs. 9 & 10). 



A single specimen exists in the Oxford University Museum 

 wherein some of the original plates of the test, displaying the 

 characteristic mesh-structure of recent Echinoids, are preserved. An 

 examination of these fully confirms the inference that might be 

 drawn from the best-preserved casts, namely, that each ambulacrum 

 was provided with two series of paired plates, one forming its roof, 

 the other its floor. The last set correspond to the ambulacral 

 plates of an Asteroid, the former to those of an Echinoid ; and in this 

 observation we find the clue to that long-standing riddle, the corre- 

 lation of the ambulacra of Asteroids and Echinoids. 



In some specimens the ambulacral plates are almost entirely 

 concealed on the outer surface by the ' velvety pile ' of thin spines, as 

 Wyville Thomson happily phrased it ; on the interambulacral plates 

 these spines are far more sparsely scattered. 



As I was led to investigate the ambulacra of Palceodiscus by a train 

 of a i)riori reasoning, it may not be out of place to indicate at this 

 point the general direction of my argument. If we consider the 

 relation between function and structurein the Ophiuroidea, 



