THE PERISTOME. 



83 



by the buccal system of Cidaris. One range of interambulacral plates extending from the 

 apex to the actinostome intercalated between rows of ambulacra with two pairs of pores each." 



There are several objections to comparing the test of Palaeozoic Echini with the peristome 

 of Cidaris. Representative genera of Palaeozoic types have a peristome and ventral border 

 of the corona directly comparable to the same parts in living Echini. The family of the Palae- 

 echinidae (Palaeechinus, Melonechinus, etc.) which have hexagonal plates, do not have flexible 

 tests. Of course specimens may be crushed and plates displaced in fossilization, but they 

 cannot be described as flexible. Those Palaeozoic genera which do have flexible tests have 

 imbricate plates, but the imbrication in the interambulacra of the corona is always dorsal 

 (text-fig. 32, p. 75), as in the Echinothuriidae, not ventral as is the imbrication of all the plates 

 on the peristome of Cidaris (text-fig. 39, p. 75). While cidarids usually have a single column 

 of plates in each interradial area on the actinostome, this is not a constant number, and on the 

 aboral portion there may be two plates in a row, or as many as four or five, as in the Phylla- 

 canthus noted. 



The above three types of peristomes cover all the known cases in Palaeozoic Echini, but 

 there are yet three other types in post-Palaeozoic Echini 

 which have interesting relations to the Palaeozoic forms. 

 A fourth type is seen in the Centrechinoida (excluding the 

 Echinothuriidae), in which we have typically (Strongylo- 

 centrotus, text-fig. 50) only ten ambulacral plates around 

 the mouth. These plates have each two pores which are 

 superposed dorso-ventrally, as in Bothriocidaris. In the 

 young (text-fig. 49), as shown by Loven, these plates lie 

 close under the corona, as do the similar plates in young 

 cidarids and in adult Bothriocidaris (only in this last type 

 there are two rows of plates). In the adult (text-fig. 50) 

 the primordial ambulacral plates are usually far removed 

 from the base of the corona, and the surrounding mem- 

 brane is usually more or less covered by solid, scaly, 

 isolated, or granular plates which apparently cannot be 

 distinguished as ambulacral or interambulacral, but seem 

 to have developed independently in place. On the mar- 

 gin of the peristome in the Centrechinoida are found ten 

 peristomal gills. These may be very small, as in Phormo- 

 soma (text-fig. 44), or may be large and frondescent, as 

 in Strongylocentrotus (text-figs. 55, 50). 



While, as a rule, there are ten primordial ambulacral plates typically in all Echini, as far 



Text-figs. 55, 56. 



55. tSlrongytocentroliis drobachiensis 

 (O. F. Muller). York Harbor, Maine. 

 X 5. Peristomal gills in place, expanded. 



56. The same. Profile view of a gill. 



