02 l)i-. Til. Murtcnsen oti some Kchinolhurids 



tliat be a somewhat unjust reproacli ? The diagnoses of genera 

 must, of course, be made from a comparative study of tlie 

 species known, and then it has to be left to the authors 

 describing new sjiccies to emend tlie diagno-'^es if the new 

 species ]n'0ve the old diagnoses to be too narrow or too wide. 

 I*^uch has always been and will remain the case, as it 

 cannot be demanded that one should guess beforehand the 

 characters of species as yet unknown. It may well be said 

 to be a drawback to our classifications, but it is common to 

 all of them, and not ]ieculiar to my classification of Echinids. 

 AVhen, further, de ]\Ieijcre says {he. cit.), " Das System 

 liiln t also zu grosser Zersplittering, was wolil den Vorzug hat, 

 dass nur das sehr iilinliche zusanimen bleibt, aber auch den 

 Kaclilheil, dass die doch nahe Yerwandtseliaft mit der abge- 

 trcnnten Form aus dem Genus-Namen nicht melir zu erkennen 

 ist," I might remark that it is allowed and desirable to 

 make the now generic names such that they indicate to which 

 old genus the new one is allied — for instance, let the names of 

 (yidarids end in -cidaris, those of Diadematids in -diademi, &c. 

 The principal object, however, is to get the genera, and above 

 all the species, distinct and clear. 'J he sj)ccics are the units 

 with which we mu.'-t work, and when these units are com- 

 posed of different things — as was the case with several of 

 the Echinid species — much labour will be Lost {''. g., on the 

 geograj hical distribution). The arrangement of the species in 

 genera and the genera in families &c. is of secondary import- 

 ance, and differences of opinion on this subject are of far less 

 wide-reaching consequences. I may cite, in conclusion, the 

 Moidsof L. Agassiz*: — "Loin d'etre nuisibleauxvraisprogrfes 

 de la science, celte multiplication des genres, lorsqu'ils sont 

 dtablis sur des caracteres prdcis, ne saurait avoir d^iutre effet 

 que de rapprccher de ])lus au plus les es] ^ces que leurs 

 caraci^res naturels lient le plus etroitement. O^est-1^ le 

 giand a\antage des jietits genres, et cet avantage est surtout 

 sensible dans les families, dont toutes les csp^ces se ressemblent 

 juu leur aspect exterieur et par Fensemble de leurs caracteres.'" 



EXPLANATION OF THE PLATES. 



[The uiicrcscope used was a Zeiss instrument.] 

 Plate IL 



Fig. 1. Ar<Bo$(»na Owsteni, from the actinal side. 

 J !>/. 2. Same, fiom the abactinal side. 



♦ Introduction to Valentin's ' Anatomic du gf^nre Eihlmig,' p. x. 



