Bd. VI: 4) THE ECHINOIDEA. 25 



of canaliculata is somewhat exaggerated, on account of the previous confusion with 

 other species, I have shown above. That the length of the radioles may vary con- 

 siderably in canaliculata I do not deny (though I should like to examine for myself 

 some specimens of the longspined variety); but — so far as I know — no specimens 

 have been described resembling A. spinulosa in the coarsely thorned upper radioles. 

 This feature alone gives this species a very characteristic appearance, so that it is 

 easily distinguished at the first glance. Also the smaller number of coronal plates 

 and the not sunken abactinal area appear to me quite good specific characters, 

 whereas the differences shown by the pedicellarise are rather too inconspicuous to 

 be of much value. — It may be possible, of course, that a larger series of speci- 

 mens will reveal such as are intermediate between spinulosa and canaliculata. For 

 the present, however, such intermediate forms are not known, and judging from the 

 material at hand it is necessary to maintain spinulosa as a distinct species. 



The fossil Goniocidaris jorgensis described by De Loriol (Notes pour servir 

 a l'etude des Echinodermes. II. Ser. Fasc. I. 1902. p. 8) evidently also belongs to 

 the genus Austrocidaris. The few isolated interambulacral plates — all that has been 

 found of the species — show the inner edge smooth and depressed, »indiquant l'ex- 

 istence d'un sillon le long de la suture mediane des aires interambulacraires, identique 

 a celui que Ton remarque dans le Goniocidaris canaliculata Al. AgaSSIZ, par ex- 

 empted. As the fossil species proceeds from the Lower Patagonian, probably be- 

 longing to the Eocene, we may herein see evidence that the genus Austrocidaris 

 has originated in the same region, where it still lives (and the only region, where 

 it occurs) — and it would seem not unreasonable to suppose that the Austrocidaris 

 Jorgensis is the ancestor of the recent species of the genus. 



It may not appear very improbable to suggest that the Cidaris julianensis de- 

 scribed by De LORIOL from the same formation (Op. cit. PI. I. Fig. 13) might be 

 nearly related to the recent Ctenocidaris spcciosa. The interambulacra are evidently 

 very similar in structure, and the ambulacral pores, according to the figure given by 

 De LORIOL, are likewise separated only by a narrow wall. — This case is, however, 

 not so evident as Austroc. jorgensis. 



Fam. Arbaciidae. * 



Arbacia Dufresnii (BLV.). 



PI. V Figs. 4—12. PI. XV Figs. 2—3, 6, S— 10, 13. 



Echinus Dufresnii Blainville. 1825. Dictionnaire d. sciences nat. Vol. 37. p. 75. 

 — — — 1834. Manuel d'Actinologie. p. 226 



* The mode of writing iArbaciadae>, used by Agassiz & Clark is grammatically incorrect, in the 

 same manner as »Echinometradae» for Echinometridae. It does not appear why they do not write also 

 jSaleniadae* etc. 



4 — 100133. Schwedische Siidpolar- Expedition iqoi — 1903. 



