14 S. LOVEN, ON POURTAI.ESIA, A GENUS OF ECHINOIDEA. 



two first plates of 4 and 1 are followed in the anterior series on either side by ten 

 or nine large oblong hexagonal, finally irregularly pentagonal or s<juarish plates. 



The hinder series, beginning outside the seconds of the ambulacra I and V, con- 

 tain each seven separate hexagonal plates, '2—8, among which the \ a :i and 4 b 3 

 are lengthened, the following slightly diminishing, and rather squarish. 



But, after all, strangely displaced and brought out of their normal position in 

 the peristome as are the lateral interradia 1 and 4 in Pourtalesia Jeffrevsi, they still 

 retain a peculiarity eminently characteristic of the Spatangidte. In my former memoir 

 on Echinoidea ') I called attention to the singular deviation invariably met with in the 

 structure of their interradium 1, the one on the right side of the animal. In all the 

 known genera of Prymnodesmic Spatangidas, from the extinct Micraster of the Chalk to 

 Lovenia, ") this heteronomy is so strikingly displayed as not to be overlooked. It is 

 brought about by the constant combination of the plates 1 a 2 and \ a 3 into one 

 single compound plate 1 a 2 + 3. It is no less apparent, and effected in a similar manner 

 in most of known Prymnadete genera, in Hemiaster, Abatus, Agassizia, Schizaster, Moira, ^) 

 and, among the Adete forms, in Echinospatagus. ^) In Paheostoma '"'), exceptional in other 

 respects also, the plates 1 a 2, \ b 2, and 1 b 3 are all united into one, while Faorina 

 and Desoria '') have the plates \ a 2 and \b 2 compounded into one single binary plate. 

 This same disposition holds good also in the almost extinct group of Adetes, such as 

 Anancites, '), Holaster, **j Cardiaster, Offaster, characterised by the plates of the inter- 

 radium 5 following the labrum not being united into a shield-like double sternum, but 

 separate, transversel}^ pentagonal and wedge-like. In times farther back yet, among 

 the Collyrites, there seems to have been no trace of the heteronomy. 



This striking feature presents itself under a new form in Pourtalesia Jeffreysi, 

 PI. 1, Jig. 2; II, 9; 111, 10. If in this species the plates of 4 a and 1 h are counted, 

 there are ten in the former, and only nine in the latter. This may depend on one 

 plate being wanting in the 1 b series, or on two plates having been united into one. 

 That this latter is really the case is seen plainly enough. The two hinder series, 1 a 

 and 4 b, consist each of eight plates, the first, marked 1, 1 and 4, /, belonging in 

 common to the two series 1 a and 1 6, 4 a and 4 b. When the plates are compared one 

 by one from side to side, as their numbers correspond, they are found symmetrically 

 conformable and evident counterparts, from 4 b S and 1 a s to A b 3 and 1 a 3. But 

 4 6 :^ is not conformable to 1 n 2, the former being distinctly hexagonal, the latter as 

 evidently pentagonal. This is so because the former, 4 h 2, has to face adorally the 

 three plates 2, 3, 4 of the series 4 a, while the \ a 2 has to front two plates only. 

 But the plate 4 a 4 evidently answers to the third in order in 1 b, so this latter must 

 be marked i h 4 and the plate between it and the plate 1, / set down as 1 b 2 + 3. 

 Pourtalesia laguncula Al. Ag., PL VI, Jig. 37, presents exactly the same heteronomous 

 disposition of the identical plates, 1 b: 1, 2 + 3, 4=4 a: 1, 2, 3, 4, and presumably it 

 will be found to pervade the whole group. It seems sti'ange that the plate 1 b 2 + 3, 



>) Etudes, p. 49—52. -) Ih. pi. XXXII— XLIII. ■') Ih. pi. XXYI, XXIX, XXX, XXXI. ") Ih. p. 58. 

 ■') lb. p. 50, pi. XXXII, fio- 197. 6) [1, pi XXVII. XXVIII. ") 11). pi. V, fisr. 51: pi. XXIV. ii<r. 181. 

 ») lb. pi. XXV, fig. 182—184. 



