450 STKONGYLOCENTROTUS TUBERCULA.TUS. 



system elliptical, covered by small plates carrying a few minute tubercles 

 near the edge of the anal system. Genital ring of nearly uniform breadth, 

 slightly narrower where the ocular plates reach the anal system. Madreporic 

 body large, pentagonal. Two main rows of tubercles in the amhulacral 

 and interambulacral space, each Hanked in the interambulacral space by 

 a smaller one uniting in a single vertical row in the median space. The 

 coronal plates are loosely covered by minute tubercles and few miliaries. 

 In the amhulacral space there is one irregular central vertical row of 

 small tubercles, with an exterior vertical row of tubercles somewhat larger, 

 from the base of which still smaller tubercles running obliquely separate 

 the arcs of pores. These small tubercles between the pores form in 

 larger specimens irregular vertical rows. Poriferous zone formed of arcs 

 of pores of from five to eight pairs of pores, more or less closed above the 

 ambitus, and. according to Battening of the test on the actinal side, more 

 or less petaloid. The notches of the actinostome are not deep, but broad 

 ami well defined. The color of the spines varies from dark violel to 

 black. Test when dry and denuded is usually greenish, the lower surface 

 whitish. Actinal membrane thin, covered by a few very distant elliptical 

 plates. 



'fhe variations of this species show, perhaps better than any other species 

 of the genus, that the genus Toxocidaris. which both Liitken and myself 

 had separated from Strongylocentrotus on account of the remarkably peta- 

 loid structure of the actinal part of the poriferous zone of several of the 

 species, can only be considered as a convenient subdivision, as we find even 

 among the adults of this species, in which the petaloid structure of the actinal 

 part of the poriferous zone is more marked than in any other, specimens in 

 which the spreading of that pari of the poriferous zone does not exist and 

 in •which the poriferous zone is no broader below than above the ambitus. 

 The extent of the spreading of the actinal part of the poriferous zone 

 depends entirely upon the greater or less flattening of that portion of the 

 test; this flattening usually takes place only after the specimens have 

 reached a certain size (say about 40""" in diameter), and is not a structural 

 feature, many of the specimens always retaining their globular outline, 

 and the poriferous zone never showing any trace of flaring. A fine series 

 of this species, collected by Mr. Dall and now in the Smithsonian collection, 

 shows the impracticability of considering this character as one of generic 

 value. 



