140 



ECHINOIDEA. I. 



Colobocentrotus Brandt (enieiid.). 



Pores multigeniiiiate; primary tubercle 011 all the ambulacral piates. The pore areas on the 

 actinal side petaloid. The buccal menibrane with nnmerous fenestrated piates, partly hidden in 

 the skin. Spines both on the buccal piates, and on sonie of the piates outside of these. The test 

 oblong, flat. The spines very short, thick, truncate, form a dense mosaic on the abactinal side. The 

 spines on the ambitus longer, flat; those on the actinal side of the common form. 



Species: 6'. atratiis (L.), Mcrtcnsii Brandt. 



Distribution: The Indo- Pacific Ocean. Littoi^al forms. 

 Iitccrtæ sed/s: 



EcliiiiHs »i/ilh'color Yoshiwara. 



Toxopiiciistcs inaculatiis (L,amk.). 



Strongyloccntrotus mcxicaiius (Ag.). 



— iiudiis (x^g.). 



— globidosus (Ag.). 



The system given liere is, I think, in all essentials an expression of tlie natural relation of 

 these forms. To be sure, we must a priori hesitate before building up a s}stem chiefh' on so minute 

 things as pedicellariæ and spicule.s. But the result is the best possible one: no undoubtedh- connected 

 forms are separated; on the other hånd, forms hitherto placed very far from each other in spite of their 

 great similarity as to habitus, are now put together {Parcchinns and Loxccliiiius). That the boundary 

 line in one place is somewhat arbitrary is no important objection to the system — this will be the 

 faet everywhere, where transitional fonns are found. The genus Pseudechinns is liere referred to the 

 Ec/iiiiometrid(r; but there eau scarceh- be any doubt that it is also closely allied to the EcJiiiiidcr^ it 

 seems especiallyto be a near relation oi Strrcchiinis iiiagcllaiiiais. Here it has been referred \.o\\\(t Ecliino- 

 metridæ especially for practical reasons, it being then possible to give a quite certain character of 

 these two families: in one teeth on either side of the blade of the globiferous pedicellariæ, in the 

 other only one unpaired lateral tooth. Pscndrclnims forms the coiinecting link between the two 

 families, and it is especially worthy of notice that in this genus may sometimes be found an iudication 

 of a lateral tooth also on the other side of the blade of the globiferous pedicellariæ. 



The family Toxopiintsfidæ is sharply limited from the other two families, without transitional 

 forms. Objections can scarcely be raised against the subfamily Scliizechininæ — all the genera 

 referred thither, are evideutly closely allied. Less sure are the subfamilies Parasa/fi/iiicc and Stroiigy- 

 loccntrotiiiæ. Pos.sibly the feature whether the globiferous pedicellariæ have a neck or not, is not of 

 so great importance, as has here been supposed; but I think it impossible to decide this faet with 

 certainty, as long as only so few forms belongiiig here are known. 



That no other outer characters are found in these forms, which may be used in the classifica- 

 tioii, I think to be certain; both the test and the spines have been studied rather tlioroughly, so that 

 anything new of importance is scarcely to be expected here. It is hardly probable that tlie inner 

 anatomicai structure will \-ield systematic characters of any greater importance, but tliis question, at 



