110 ROBERT TRACY JACKSON ON ECHINI. 



family. In this respect it is therefore primitive; only two specimens were seen. Centro- 

 stephanus rodgersi (text-fig. 96) has typically oculars I, V, IV insert, but in twelve specimens, 

 three have I, V, IV, II, and one has all oculars insert as progressive variants. The range of 

 characters is therefore the same as the dominant ones of Centrechinus setosus. The specimen 

 figured shows a secondary splitting of the madreporite and also unplated tissue ventral to the 

 apical disc, both of which features are seen in Echinothuriidae. In Centrostephanus longispinus 

 (text-fig. 97) all the oculars are insert; only four specimens were seen, Init if four agree in a 

 character, it is strong evidence that the typical feature is ascertained. The plates of the peri- 

 proct are numerous and minute; a few large plates lie on the periphery mainly over the oculars. 



Astropyga is a very interesting type as an extreme form of the family. All the oculars are 

 broadly insert (text-fig. 99). In this type the plates are imbricate (p. 74) as seen in text-fig. 

 100, and leathery tissue exists between the oculo-genital plates as in Echinothuriidae, yet the 

 genus is clearly one of the Centrechinidae as shown by the character of the peristome. 



In Chaetodiadema pallidum the plates are excessively thin and imbricating. The ocular 

 plates are all broadly insert (text-fig. 98), almost exactly as in Melonechinus (Plate 56, fig. 6). 

 No difference was observed in 36 specimens seen. The periproctal plates on the outer border 

 are exceptionally large, but minute in the center, in this feature being like the young of Centr- 

 echinus (text-fig. 88). 



Of the Echinothuriidae, only a few specimens have been studied. In Asthehosoma ijiinai 

 (text-fig. 101) the oculars are all broadly insert. Ocular and genital plates meet in a continu- 

 ous ring, but are deeply covered on their lateral borders, as are also in part the coronal plates, 

 by living tissue. Genital plates are split up ventrally into many minute plates and the 

 madreporite is split so that the pores exist in several dissociated plates. This type makes a 

 rather close approach to the character of Astropyga. 



In Phormosoma placenta, when 3 mm. in diameter, the genitals are in contact and all 

 oculars exsert (Plate 3, fig. 8), as shown by Dr. Mortensen (1903, p. 174; 1907, text-fig. 1, 

 p. 24). This is much like the condition in Jurassic Phymechinus (text-fig. 84). Later the 

 oculars travel in and meet the periproct, the oculo-genital ring being continuous as in pro- 

 gressive variants of Centrechinus (text-fig. 95), and in the typical condition of Chaetodiadema 

 (text-fig. 98). Mr. Agassiz (1904) shows this early stage in Phormosoma and in a beautiful 

 series of figures its development to the adult condition (text-fig. 170, p. 149), in which the 

 genitals and oculars are separated so that the periproct comes in contact with the interambu- 

 lacra (p. 63). This character of separation of oculars and genitals occurs as a feature in regular 

 Echini only in the Echinothuriidae, but has been seen as a variant in Strongylocentrotus 

 (Plate 6, fig. 5, and Plate 5, fig. 15). In the coronaof Phormosoma (text-fig. 170, p. 149), in 

 the placogenous zone, the ambulacral plates are simple and the young interambulacral plates 

 originate against the oculars as usual in Echini. The suborder Aulodonta exhibits a wide 



