214 ROI5KRT TRACY .JACKSON ON ECHINI. 



feature characteristic of this family only. The lantern is inclined, with long interpyramidal 

 muscles (Plate 2, figs. 19-21), a feature like the Palaeozoic, and also the young of types that 

 have erect lanterns in the adult. It may be considered a retention in the adult of a youthful 

 character, seen also in Astropyga. Stewart's organs, more or less developed, are a feature 

 of this family, as of the Cidaroida, although they differ markedly, being vermiform or sausage- 

 shaped rather than frondescent as in the Cidaroida. Radial peristomal and somatic muscles 

 (text-fig. 226, p. 193) are peculiar features of the Echinothuriidae and known in no other Echini. 



The suborder Stirodonta differs from the Aulodonta in that the teeth are keeled, but 

 agrees with it in that the epiphyses are narrow (text-fig. 212, p. 184). In some the primary 

 tubercles are perforate as in the Aulodonta, but in most genera tubercles are imperforate as in 

 all the Camarodonta. This group, as most others, occupies an intermediate position, possess- 

 ing in some genera characters like the preceding, and in others characters like the succeeding 

 group. Such is found to be true of any classification if the groups are at all nearly related. 



The Saleniidae is one of the most interesting groups of Echini. The permanence in the 

 adult of the suranal is one of the most striking features (text-figs. 102-105, p. 111). It may lie 

 against genital 3 as in Salenia or facing ocular III (Peltastes), or may be one of several large 

 plates as in some Acrosalenias. This is apparently the first appearance of the suranal plate in 

 Echini, as discussed (p. 174). The ambulacral plates may be compound throughout, or only 

 ventrally and simple above, indicating extreme reversion. The large primary spines and 

 prominent central tubercle like Cidaris may fairly be considered a retention in the adult of 

 characters common to the young of the whole order, indicating with the simple ambulacral 

 plates, when they occur, a reversionary rather than a progressive type. The retention of a 

 single madreporic opening on the proximal side of genital 2 (Plate 4, figs. 1, 2) is a fact in 

 accordance with the same line of reasoning. Ocular plates are exsert or reach the periproct 

 in the sequence I, V, IV, II, III, a contrast to the Arbaciidae where the sequence is V, I, IV. 



The Phymosomatidae Mortensen (1904, p. 55) classes with the Saleniidae and Arbaciidae, 

 expressing closer relations with the latter. Duncan (1889a) also places this family next to 

 the Arbaciidae. Ambulacral plates are compound, of five elements each in the Recent Glypto- 

 cidaris crenulare, or with more or fewer elements in some fossils. One or more to all oculars 

 may be insert (text-fig. 106, p. 113). Mr. Agassiz's photographic figure (1873, Plate 7a, fig. 8) 

 of Recent Glyptoddaris crenulare shows ocular I only insert. This character, with the somewhat 

 prominent suranal plate opposite genital 3 in the figure cited, inclines one to affiliate the family 

 rather more closely with the Saleniidae than the Arbaciidae. The teeth are keeled, epiphyses 

 narrow, and spurs from the dorsal tips of the pyramids support the teeth as in Stomopneustes 

 (Plate 4, figs. 8-10). This structure is peculiar to these two families only, as far as known. 



The Stomopneustidae contains only the Stomopneustes variolaris. This genus has been 

 classed with the Echinometridae, but Mortensen (1903, p. 133) separated it as a distinct family 



