EcMnoclermataj hy Prof. F. J. Bell. 443 



secondaiy or smaller spines ; in 8. Raschi, on the other hand, 

 this difference is, typically, hardly noticeable, and in corre- 

 spondence with this the tuberculation is much more uniform. 

 In one of the specimens of S. Raschi now lying before me the 

 spines are as long and as prominent as in a specimen of S, 

 purpiireus of nearly the same size ; in another, somewhat 

 larger, the spines are much longer than we generally find 

 them in 8. Raschi] but they are much more uniform in size 

 than in either the first-named specimen or than in 8. piirpu- 

 rens, and, so far, the latter could not be confounded with the 

 more common species. Nor could the first-named, but for a 

 different reason ; it is much higher than a 8. imrpureus of 

 the same length, but the second specimen, though some 10 

 millira. longer, is about 2 millira. less high, and, of course, 

 looks much less high than its smaller companion. 



With tlie difference in the size of the spines there is, of 

 course, correlated a difference in the size of the tubercles which 

 bear them 5 an inspection of Prof. Loven's figure* shows that 

 the difference is not very marked in his type specimen. I 

 removed the spines from a specimen which, in its spinulation, 

 most closely resembles 8, purpiireusj and I find on cleaning 

 the test that some of the tubercles are more than ordinarily 

 larger than the rest ; the general facies of this test is, however, 

 distinctly that of 8. Raschi. 



So, again, it may be noted that while some tests are less 

 deep than others, others are more rounded; again, variations 

 may be seen in the depth of the peristome. On the whole 

 the most constant character of the deeper-water species appears 

 to be the form of the labrum ; this is always more pointed 

 and convex than in 8. purpureas. 



We may, then, observe with regard to a number of the so- 

 called specific characters of 8. Raschi that they vary within 

 very wide limits. Of the specimens collected not one would 

 be assigned to any other species, the general facies of 8. 

 Raschi being maintained throughout; but on analysis the 

 several " specific characters " are found for the most part to 

 vary considerably. 



These observations seem to me to have some bearing on 

 the question of the utility of specific characters, for they show 

 that we must exercise the greatest caution in the selection of 

 the points of structure which we use as such marks. It 

 would be preposterous to imagine any zoologist more capable 

 than Prof. Loven of discriminating between two species of 

 Echinoids, and yet among the characters by which his species 



* Ofv. Vet.-Akad. F5i-liaiidl. 18(>i.), pi. xiii. 



^2* 



