Bd. VI: 4) THE ECHINOIDEA. 31 



cleaned. On cleaning them I find the colour to be green as in Dufresiiii. (Also 

 LoVEN states allernaiis to be green.) In the elaborate description of the two spe- 

 cies given by LOVÉN I am likewise unable to find any other more important dif- 

 ference indicated than that of the tuberculation. 



It is my decided opinion therefore that alternans cannot be maintained as 

 distinct from Diifresnii, not even as a variety. 



With Avbacia africana (Troschel) the case is different. The type spe- 

 cimens appear to have become lost. They are no more in the collections of the 

 University in Bonn, and not in Stockholm either, where LovÉN once had them for 

 examination. On the other hand I have received from the Berlin Museum the spe- 

 cimens mentioned by LovËN (Op. cit. p. 105), as also another specimen in alcohol 

 (2310; Schinehoxe; Africanische Gesellschaft). As these specimens (in any case the 

 dried ones) have been directly compared with the type specimens by LovÉN, we 

 may safely assume that they really belong to A. africana. 



A mere glance at these specimens (one of them represented in PI. V. Figs. 

 13 — 15) shows that they have nothing at all to do with A. Diifrcsnii. The ab- 

 actinal interambulacral tubercles are strongly developed and cover most of the inter- 

 ambulacral area, leaving only a small space naked, just as in A. lixiila. The colour 

 of the test is reddish or brownish, with a faint olive tint, very different from the 

 green colour of Dufresiiii. The spicules are like those of A. lixula. The pedi- 

 cellariae show only unimportant peculiarities; tridentate pedicellarias were not found. 



From what has been pointed out here, as well as from the figures given 

 (PI. V Figs. 13 — 15) it is evident that A. africana has nothing to do with A. Dii- 

 fresnii. It is decidedly nearest related to A. lixula — if not a synonym only of 

 that species. To decide this question more and better material would be desirable; 

 on this occasion I must content myself with having sho-wn that the alleged syno- 

 nymy of A. africaiia with A. Diifresnii., so highly improbable from a zoogeo- 

 graphical point of view, was really erroneous. 



The geographical distribution of Arbacia Diifresnii is thus again restricted to 

 the South American Coasts, where it is known to range from the southern extremity 

 to the La Plata River (37° 42' S. »Hassler» Echini) on the East Coast, and to 42" S. 

 on the West Coast (Puerto Montt; PLATE (MEISSNER)). Further it was taken by 

 the French Antarctic Expedition (Exp. Charcot) at the Island Booth Wandel near 

 the Antarctic Continent. — The bathymetrical range of the species, as far as known, 

 is from the shore down to 175 fathoms (»Challenger»). 



In the »Challenger5 Echinoidea (loc. cit.) Arb. Diifresnii is further mentioned 

 from Nightingale Island (Tristan da Cunha), 100 — 150 fathoms. This occurrence 

 would be easily explained, if the species had pelagic larvae; but this seems not to 

 be the case. Studer (»Gazelle» Echinoidea p. 868) gives the following statement: 



