ECHINOIDEA. I. J, 



matter of judgment, to which genus one species or another is to be referred, and most of the species 

 more frequentl\- mentioned Iiave also b>- and b\- been referred to ahnost all the different genera. So 

 far it is very consistently done by Duncan and Hell (73) quite to strike out all these undistingnish- 

 able genera, and only retain the old genus Cidan's\ but then on the otlier hand this wav of proceed- 

 ing means quite to abandon the pursuit. 



The reason why the result of the earlier attempts at classification has been so meagre, has to 

 be sought in the characters used. The most important ones have been, whether the two pores of 

 each ambulacral plate are connected by a groove or not, and whether the tubercles are crenulated or 

 not. Further the spines, the number of plates, the breadtli of the ambulacral area, and upon the 

 wliole the structure of the test have been considered of great importance. All these characters, 

 however, are insufficient or even unreliable. As lias been pointed out by both Doderlein and Dun- 

 can, it is often impossible to decide, whether the pores are or are not connected by a groove. The 

 crenulation is a very variable character; crenulated tubercles may be found in some indi\-iduals belong- 

 ing to species normally without crenulation. The structure of the test, the tubercles, the number 

 of plates etc. are vers- much dependent on the age of the animal. All these characters, says 

 Duncan, are of no physiological importance whatever ; any classification in which these 

 characters are used is artificial . On the other hand he thinks that the number of interradial plates 

 (is) of physiological importance; and there is a great temptation to consider t\pical Cidarids as having 

 but a few, say not more than seven, in a vertical row (132 p. 30). This character seems to be at 

 least as gratuitous, as the others criticised by Duncan are relative ones; neither .seems the result of 

 his systematic researches in any way to show that he has found here a systematic character of 

 any great importance. 



Among the characters hitherto used in the classification, the spines seem to be one of the most 

 reliable. They show a great richness of forms, but are at the same time of a rather constant form in 

 the separate species. Also their microscopic structure differs to a high degree, and here, perhaps, we 

 might find good generic cliaracters. There are in the literature not a few examinations of the struc- 

 ture of the spines in the Cidarids. Stewart'), Bell (57), and Agassiz (Revision of Echini and 

 Chall. Ech.) have figured transverse sections of the spines of different species; but especially H. W. 

 Mackintosh has rendered great services to the question by his excellent researches on the struc- 

 ture of the Echinoid-spines (264 — 65). The spines of the Cidarids differ from those of the other Echi- 

 noids by having a compact outer layer { osfracuin- Bell); (such a layer is also found, however, in 

 Salcjiia and Arbacia (on the point of the spines)); — <;acanthostracous; this kind of spines is called 

 by Mackintosh. Unfortunately it cannot with certainty be inferred from the existing examinations 

 whether the structure of the spines yields good generic characters. Mackintosh is decidedly of opinion 

 that the spines really yield characters of that importance; he finds instances < in which the acantho- 

 logical characters would seem to call for a change in the position of a genus (265 I p. 478), and he 

 lays stress on the importance of always mentioning the structure of the spines in the description of 

 Echinoids. Otherwise he has examined too few Cidarids to have got a sure impression of the 



') On the minute structure of certain hard parts of the genus Cidaris. Quarter!. Journ. Micr. Science. X. S. XI. 

 1871. p. 51 -55. pi. IV. 



