283 



wider question of the extent of the variability of these forms. If we 

 break up the specimens of Asterias glacialis in the British Museum 

 into different categories, we find as many as six sets at least. 



a. The typical form : here there is never more than occasionally 

 a single isolated rather small spine on either side of the well-marked 

 median row of spines which extends all along the ray. (Azores). 



h. Three to six spines may be found on one or both sides of the 

 median row, set regularly in series ; these may be found on some though 

 not all the rays, and they are always confined to the proximal half. 



(Azores) . 



c. On either side of the median row spines extend in a regular 

 series over more than half of the ray. (Madeira). 



d. A well marked row of spines may extend all along the ray, 

 on either side of the median row ; the median row is now less conspi- 

 cuous. (Portugal). 



e. Two rows are indicated in the place of one (Madeira) . 



f. Two complete rows of spines are now developed on either side 

 of the middle row, and are almost, if not quite, as conspicuous as it is. 



(Portugal) . 



Variations to this extent seem to me to deprive the character of 

 the number of rows of spines of any value as an agent in specific dis- 

 crimination, and they are of the more significance when we compare 

 them with what we already know as to the variation of the similarly 

 placed spines in A, Muelleri •. 



Prof. Greeff finds that two other characters remain as distingui- 

 shing A. glacialis from A. africana\ but as to the characters of the 

 pedicellariae I would submit to him that, just as there must be some 

 relation between the solidity of the skeletal plates and the strength of 

 the defensive spines 2, so there must be a very close relation on the 

 one hand between the concentration of large spines and the develop- 

 ment round them of a large number of these modified spines, which we 

 call , pedicellariae', and on the other, the diffused distribution of 

 smaller spines, less highly differentiated, and the less marked aggre- 

 gation at certain points of these pedicellariae. In other words, the 

 arrangement of the pedicellariae is dependent on that of the spines. 



The two specimens of A. africana in the collection of the British 

 Museum are dried, and are not in a condition to be completely exa- 

 mined ; from what, however, I have been able to make out, I am in- 

 clined to think that there is more variation in the shape of the spines 



1 Proc. Zool. Soc. London 1881. pp. 510, .511. 



2 Ann. Nat. Hist. (5) Vol. 8. p. 441. 



