HETEROCENTROTUS TRIGOXARIUS. 379 



The "Albatross" material is from the following places: — 

 Laysan, Hawaiian Islands. 

 Honolulu Market, Oahu, H. I. 

 Puako Bay, Hawaii, H. I. 

 Twenty-six specimens. 



Heterocentroius trigonarius Br. 



Echinus trigonarius Lamarck, 1816. Anim. s. Vert., Ill, p. 51. 

 Heterocentrotus trigonarius Brandt, 1835. Prodrom. desc. Anim., p. 266 (or 66). 



Plates 118-120. 



This species shows much greater diversity in every feature than mammillatus 

 does, but the number of pore-pairs in at least one arc appears to be always more 

 than 12, even in specimens, whose test does not exceed 20 mm. in length. In 

 specimens only half grown there is almost always at least one arc with 15 pairs, 

 but occasionally 14 is the maximum. The primaries range in form from long, 

 tapering, acute, trigonal spines to short, club-shaped spines with very thick 

 blunt tips. The actinal spines are of course more or less flattened, but the 

 amount of flattening and the number of primaries involved shows great diversity. 

 The secondaries show nearly as great diversity, for while they are usually more 

 or less acutely pointed, we have a number of specimens in which they are as thick 

 and truncate at tip as in any mammillatus. The color is also variable, and in 

 this particular there is a suggestion of local differences; specimens from Mauritius 

 are brown with distinct shades of green and orange-red on the primaries, while 

 specimens from the Paumotus are deep purple, with very little variation except 

 in shade. Specimens from intermediate localities connect the two extremes, 

 which will probably at some future time be designated as subspecies. 



Apparently trigonarius is more widely distributed or at least, more generally 

 distributed than mammillatus. Specimens are at hand from Mauritius, the 

 Philippines, New Guinea, the Carolines, the Gilberts, the Marshalls, the Fijis, 

 Baker's Island, Tongatabu, the Society Islands, and the Paumotus. As the 

 two species are so easily confused with each other, the published records especially 

 if they date back half a century or more are unreliable so that the question of the 

 geographical distribution of the two species of Heterocentrotus is still an open 

 one. The evidence however indicates that mammillatus is a northern species 

 ranging from the Red Sea (eastern Mediterranean?) to Hawaii, while tri- 

 gonarius is southern, ranging from Zanzibar to the Paumotus. The two occur 



