XXU THE VOYAGE OF H.M.S. CHALLENGER. 



II. Reasons for considering previous Classifications invalid. 



As the foregoing is the latest scheme of classification, and is, in fact, the embodiment 

 of all that have gone before, I propose to examine briefly the fundamental characters upon 

 which it is based. 



Perrier's classification is based upon the character of the pedicellarise. He considers 

 that the pedicellarise furnish characters of the highest taxonomic value ; in other words, 

 he regards them as class characters, upon the modifications of which divisions of ordinal 

 rank may be made. He assigns to the pedicellarise this importance on the ground that they 

 are the degenerated rudiments of organs whose functions were more important in the 

 ancestral forms than those which pedicellarise now perform. Perrier states that the pedi- 

 cellarise appear earlier in the embryo of Echinids than the spines. He regards them as 

 more ancient organs. He rejects the view that they are modified forms of spines, as sug- 

 gested by A. Agassiz. He considers that they furnish positive ordinal characters in the 

 Echinoidea. He asserts that these statements are even more clearly applicable to the 

 Asteroidea than to the Echinoidea. He regards the more complex forms of pedicellarise 

 as older than the simpler forms, and believes that the forcipiform pedicellarise are older 

 and more typical than the more simple forficiform pedicellarise. 



Confining my remarks to the Asteroidea, I venture to think that facts do not support 

 any one of these statements, so far as that class is concerned. 



(1.) Eespecting the priority of appearance in the pedicellarise and spines, I may say 

 that in no starfish embryo which I have examined have I found anything to warrant the 

 assumption that pedicellarias appear before spines ; in fact, my observations indicate 

 unequivocally that the spines are formed before the pedicellarias. In Asterias, a form 

 which is crowded with pedicellarise when adult, and is one grouped by Perrier amongst 

 what he considers (erroneously in my opinion) the oldest forms of Asteroidea, this is 

 certainly the case. Neither has any other observer who has written upon the development 

 of starfishes recorded, so far as I am aware, the appearance of pedicellarise before spines. 



(2.) As to whether pedicellarise are modified forms of spinelets, and as to whether the 

 older forms of the organ are simpler or more complex than the more recent, I consider 

 that those Asterids, which I believe to represent the most archaic forms, distinctly support 

 the views, (i.) that the pedicellarias are modified spinelets, and (ii.) that the older forms 

 of the organ were simpler and less complex than the more recent. As to the mode in 

 which the more complex forms may have been evolved it is unnecessary to speculate here. 

 The further outcome of the argument that the most complex form of pedicellarias indicates 

 the most ancient organism would logically lead to the conclusion — although such an opinion 

 is not definitely expressed by Perrier — that the Echinoidea are phylogenetically older than 

 the Asteroidea, for I imagine that it will be generally admitted that the pedicellarias of 

 Echinoidea are more complex than those of Asteroidea. Such a conclusion, I venture to 



