THE SUDDEN ORIGIN OF NEW TYPES 417 



situated excentrically and possesses the Cystidean character- 

 istic of being closed by a valvular pyramid of plates. In 

 Palaeodiscus ^ (also from the Lower Ludlow Shales) a close 

 approximation, on the other hand, is made to the Starfishes, 

 for a series of ambulacral plates like those of the Asteroidea 

 occurs inside the plates of the corona, and the tube-feet in 

 the oral portion of the radii apparently passed out between 

 the outer ambulacral plates, but the chief radial water-vascular 

 vessels seem to have run along inside the test ; the radioles 

 or movable spines are small and do not differ very greatly 

 from those of some Edrioasteroidea and Asteroidea. 



In the Devonian Lepidocentrus — the earliest and most primi- 

 tive of the Cidaroidea — the test is flexible and this flexibility 

 is still retained to a less extent in the Lower Carboniferous 

 Archaeocidaris in which there are 4-7 rows of large overlapping 

 plates in each interambulacral area and two rows of ambulacral 

 plates. Oligoporus has four ambulacral and 4-9 interambulacral 

 rows, whilst in the Carboniferous Melonechinus (Melonites) 

 there may be as many as 4-1 1 columns of interambulacrals and 

 5-14 ambulacrals; in Lepidesthes there are 8-18 ambulacrals 

 and 3-6 interambulacral rows ; in Palaeechinus (also Carboni- 

 ferous) there are respectively 5-7 rows of interambulacrals, 

 decreasing in number towards the poles and only two rows 

 of ambulacrals. It is evident that (with the exception of the 

 aberrant Bothriocidaris) the reduction to the optimum of two 

 rows took place with greater rapidity in the ambulacral areas 

 than in the intervening zones. The reduction in the number 

 of plates went hand in hand with an increased rigidity and 

 consolidation of the test as a whole, and an increase in the size 

 of the spines. 



If, as now seems likely, the Echinoids took their origin from 

 an early offshoot of the Asteroids rather than directly from the 

 Cystids, this would all the more point to the significance of a 

 multiple repetition of similar parts in inducing the evolution of 

 numerous divergent variations; Macbride- has suggestively 

 remarked that " when we recollect that some of the oldest 

 Asteroidea known to us had very narrow arms and interradial 

 areas edged by large square marginals, it does not require a 



' W. J. Sollas, "On Silurian Echinoidea and Ophiuroidea," Quart. Journ. 

 Geol. Soc. Iv. 1899, P- 7°i- 



* Camb7-idge Nat. Hist. \. p. 558. 



