MONOGRAPH OF THE EXISTING GKINOIDS. 119 



Proccsx of arm redupUcotion. 



The original 10 arms are composed of numerous brachials and terminate in 

 a growing point, which continually adds new brachials, while in the adults there 

 are never more than four, and usually only two, segments in the second and 

 following division series. Obviously, therefore, the second and following division 

 series can not be formed by the branching of the growing point. 



This leaves two possibilities the proximal pinnules may transform into 

 arms, or the original 10 arms may break off at the base, and from the stumps 

 division series bearing new arms similar to those lost may arise. 



Pinnules have never been observed to transform into arms. It is true that 

 they may sometimes be replaced by arms, but in that case the supernumerary 

 arm is a true arm from its very first inception. 



If the first pinnule did transform into an arm, thereby forming a division 

 series of two ossicles, the original first two brachials, it would necessitate the for- 

 mation upon it of an ambulacral groove with the lappets, tentacles, nerves, and 

 correlated structures, the conversion of the first brachial syzygy into a synarthry, 

 the transformation of the muscular articulation between the fifth and sixth 

 brachials into a syzygy, a radial change in the structure of the first inner pinnule, 

 now become the first outer pinnule of the inner arm, including the loss of the 

 ambulacral organs, and a radical change in the disposition of the syzygies through- 

 out the arm, as the syzygies are always farther apart and more irregularly dis- 

 posed in multibrachiate than in 10-armed species, or in 10-armed young. Similar 

 and even more extensive changes would be necessitated by the development of the 

 first inner pinnule into an arm to form a division series of four elements. Such 

 changes are morphologically impossible : and moreover they could not occur with- 

 out being readily traceable. 



We are therefore compelled to accept the alternative hypothesis that the 

 original arms are cast off and from the stumps entirely new division series and 

 arms grow out to replace them. 



It has been frequently observed in several different species normally or ordi- 

 narily 10-armed that if one arm be lost at the base, a pair of arms will regenerate 

 from an axillary growing out from the stump, and that these arms will eventually 

 become exactly like the others except for the somewhat greater intervals between 

 the syzvgies and the slightly shorter brachials. 



Very exceptional in Antedon bifida, and, so far as we know, never occurring 

 in A. mediterranea or in A. adrintica, and reported but once or twice among all the 

 other macrophreate species, the tendency to develop additional arms after basal 

 fracture may be traced through easy stages in the Thalassometridae, Charito- 

 metridjfi, Calometrida?, Himerometridre, or Colobometridae, the percentage of fre- 

 quency of cases rising, species by species, until we find forms in which one or more 

 IIBr series are invariably found ; and similarly there is no break in the chain be- 

 tween species having ordinarily from 10 to 15 arms and those having from 15 to 20 

 or 30. Indeed, certain individual species may cover almost this entire range, though 

 142140 21 Bull. 82 10 



