SEA-URCHINS. 483 



of their evolution, to depart from the characters of a Cidarid, and 

 to approximate to those of the Irregular types. A recognition of 

 this direction of evolution in the group renders the interpretation 

 of the various sti'uctures more intelligible by including them all 

 in one coherent scheme. A complete reliance on this principle 

 would probably result in a misinterpretation of degenerate or 

 retarded development, so that the trend of evolution must be 

 considered in direct connection with stratigraphical evidence. 



To sum up, the chai-acters used here as indices of phylogenetic 

 development ai"e of two kinds. One series is available for tracing 

 the evolution of the group as a whole. Such features are (i.) the 

 gradual loss of masticatory structures and of pei-istomial Ijranchiie, 

 (ii.) the backward movement of the periproct, (iii.) the loss of 

 radial symmetry, and (iv.) the increase in density, and decrease 

 in coarseness, of the tuberculation. The other series is used to 

 indicate the intimate relations of the individual genera of the 

 group. These features are (i.) the plating structure of the am- 

 bulacral areas, (ii.) the composition of the apical system, and 

 (iii.) the stratigraphical sequence. 



2. The Origin of the Group. 



The oldest known member of the Holectypoida is Pygaster 

 reynesi, which occurs in the Middle Lias of Fi-ance. it will 

 therefore be necessary to look for the ancestor of this typically 

 Holectypoid form among the Regular Echinoids of the Liassic 

 or Triassic periods. It is unfoi'tunate that the origin of the 

 group should date from these periods, for, unlike the purer waters 

 of the Oolitic seas, the muddy shore-lines of the Liassic ocean, 

 and the saturated lagoons of the Triassic coiul-i'eefs, were un- 

 favoui'able to the free development, as well as to the ultimate 

 preservation, of Echinoids. Howevei", it is significant to find 

 that the earliest Irregular Echinoid appeared so soon after the 

 first stage of differentiation had begun among the Regular orders. 

 Its inception thus seems to have been an effect of that luirest in 

 structure and habit that usually accompanies profound changes 

 in the course of the evolution of a Class. 



The Liassic Regular Echinoidea seem to belong to two orders 

 only, the Cidaroida and the Diademoida. The former group had 

 become more or less stereotj'ped in character during the Permian 

 and Triassic periods, having been, as Bather (59) indicates, the 

 only surviving member of the varied Pala-ozoic types. The 

 Diademoida, as the same author has shown {t. c), were beginning 

 to assume the typical features of the order in Ti-iassic times, but 

 still retained features, such as a primary character of the am- 

 bulacral plates in the greater part of the area, and. a shallowness 

 of the branchial clefts, which are reminiscent of their Cidaroid 

 ancestry. 



From the Lias a considerable number of primitive Diademoida 

 are known, and they have been recently studied by Lambert (52), 



31* 



