Chap. VIL] THEORY OP NATURAL SELECTION. 299 



transitions, through the abortion of one of the two 

 states of an organ. 



With respect to the steps by which these curious 

 organs have been evolved, Mr. Agassiz infers from his 

 own researches and those of Miiller, that both in star- 

 fishes and sea-urchins the pedicellarise must undoubted- 

 ly be looked at as modified spines. This may be inferred 

 from their manner of development in the individual, 

 as well as from a long and perfect series of gradations 

 in different species and genera, from simple granules to 

 ordinary spines, to perfect tridactyle pedicellariffi. The 

 gradations extend even to the manner in which ordinary 

 spines and pedicellarise with their supporting calcare- 

 ous rods are articulated to the shell. In certain genera 

 of star-fishes, " the very combinations needed to show 

 that the pedicellarise are only modified branching 

 spines " may be found. Thus we have fixed spines, with 

 three equi-distant, serrated, moveable branches, articu- 

 lated to near their bases; and higher up, on the same 

 spine, three other moveable branches. N^ow when the lat- 

 ter arise from the summit of a spine they form in fact a 

 rude tridactyle pedicellaria, and such may be seen on the 

 same spine together with the three lower branches. In 

 this case the identity in nature between the arms of the 

 pedicellarise and the moveable branches of a spine, is un- 

 mistakable. It is generally admitted that the ordinary 

 spines serve as a protection; and if so, there can be no 

 reason to doubt that those furnished with serrated and 

 moveable branches likewise serve for the same purpose; 

 and they would thus serve still more effectively as soon as 

 by meeting together they acted as a prehensible or snap- 

 ping apparatus. Thus every gradation, from an ordinary 

 fixed spine to a fixed pedicellaria, would be of service. 



