GOO s. GOTO : 



tinctly convex on the ventral, and much more so on the dorsal 

 side, and is very different in general appearance from the speci- 

 mens described above, so mucli so that one is tempted to make 

 anotlier species of it. But from what we know about the 

 variability of tliis species, so ably set f jrth by Döderlein, it ap- 

 pears to me that it is merely a varietal form of tliis polymorphic 

 species. What strikes one at the first sight of this specimen is 

 the smoothness of the dorsal side, which is entirely destitute of 

 any coarse granules, except around tlie madreporite and the anus, 

 and a few other places at intervals. The poriferous areas are less 

 well developed than in the specimens previously described, the 

 intervening non-poriferous areas are relatively larger, and the 

 rosette-like arrangement of the pore areas is more or less evident. 

 The dorsal pedicellaria) appear to bo exceedingly few, if any.. 

 Spiniform granules of different sizes begin to appear towards the 

 margin of the dorsal side, mostly in the poriferous, but also more 

 or less in the non-poriferous, areas, and these gradually pass on 

 to the rounded coarse granules of the ventral side, which are 

 less well developed than in the specimens described above. The 

 ventral valvate pedicellaria) are also far less numerous in this 

 specimen. The inner adambulacral spines are only apparently four 

 in number, and a small fifth spine is always found covered over 

 by the abcentral spines of the next adcentral plate. The outer 

 spines are two and subequal and are more spiniform than in the 

 specimens 1-3. The oral spines arc eight or nine for each plate. 

 Another point which may be noted is the presence of an irregular 

 row of coarse granules directly outside the outer adambulacral 

 spines, separated from the general granulation of the ventral side by 

 a smooth space. The dimensions of this specimen are as follows : 



