1881.] PROF. F. J. BELL ON THE GENUS ASTERIAS. 499 



of greater value which neither specimens nor opportunities yet 

 allow'. 



1 now proceed to the consideration of the heads under which the 

 numerous species of this genus are to be grouped. The genus Aster- 

 acanthion of Miiller and Troschel was divided by a single coup into 

 a group with spines on the back, and a group with stalked knobs on 

 the back ; those of the former group alone are now members of tlie 

 emended genus. No other zoologist has, so far as I kuow, attempted 

 any convenient grouping of the species, though Dr. Stimpson has 

 rendered some assistance in this direction by pointing out the affi- 

 nities of the species he describes. 



It is, at the same time, obvious enough that there is a very large 

 number of species in the restricted and emended genus Asterias ; and 

 the number is unwieldy even now, when our knowledge of the varia- 

 tions that may be found in it is very possibly altogether elementary. 

 No good, however, is gained by hastily accepting or proposing ill- 

 defined generic coups; the more necessary work just now is to sort 

 out the species into different categories. These must be as natural 

 as they can be ; but where nature fails us, or becomes too obscure for 

 our vision, we must make use of what are not so satisfactorily 

 natural characters. 



It is necessary to make some such preface as this, because I have to 

 propose a primary subdivision of the genus to which, from purely 

 theoretical considerations, it would be possible to raise some not un- 

 important objections ; for I propose, first of all, to separate the species 

 into those in which there are developed more than five rays, and those 

 in which, so far as we know, the number five is constantly retained. 

 To these two groups I propose to apply the terms Heteractinida and 

 Pentactinidd^ , 



It is, so far as is yet known, only among the former, or Heterac- 

 tinida, that the presence of more than one madreporic plate has been 

 noted, though in forms which belong to the Heteractinida, but in 

 which five arms only are developed (and this is not only a possible, 

 but an actual case) there may be, and at times are, two madeporites. 

 The plurality of madreporic plates affords a good secondary point of 

 difference ; and I propose, therefore, to form subdivisions which may 

 respectively be known as polyplacid and monoplacid. 



The third distinctive character lies in the number of spines which 

 border the ambulacra — theadambulacral spines : as an ordinary rule, 

 one finds one or two rows of these ; and the forms which belong to 

 them may be distinguished respectively as Monacanthida and Dipla- 

 canthida. 



It is a difficult matter to say exactly what is the real significance 

 of the difference in the number of the rows of adambulacral spines. 

 In the first place, it is necessary that we should have for a number 



'■ This might almost be an echo of the words of Prof. Ales. Agassiz (see 

 'North-American Starfishes,' p. 122). 



2 Whether this division be natvu-al or artificial, it will, for the great majority 

 be found, I think, to work pretty satisfactorily : the problem of heteractiiiism 

 is still unsolved ; and it is a serious question whether the heteractinism of e. g. 

 A. temdsjnna is exactly comparable to the polyactinism of e. g. Heliaster. 



