TORSION OF THE CRUSTACEAN LIMB. 137 



question. But the height of the difficulty is not yet reached, 

 for upon the selection theory, the torsion of such limbs must 

 have arisen gradually, through successive fractions of a degree, 

 until the member had moved through a quadrant of arc ; further- 

 more we are required to believe that each successive position was 

 so much more favorable than the last, that forms showing such 

 as well as possibly other correlated variations would outstrip their 

 competitors, and alone leave descendents. As already suggested, 

 the problem is somewhat analogous to the migration of the eye 

 in the metamorphosis of flat-fishes, attempts to explain which in 

 terms of the selection theory have not been very successful. 



The difficulty of the problem is not lessened when we reflect 

 that the slaughter of the young in a form like the lobster, takes 

 place on a tremendous scale and is of the most indiscriminate 

 character, before the prehensile organs are fully developed. We 

 can only guess at the causes of such a variation. Its great antiq- 

 uity, dating from as early at least as the Jurassic period, and 

 other considerations already touched upon, render it highly prob- 

 able that it first arose as a discontinuous variation. That it was 

 not indispensable, is suggested by its erratic distribution at the 

 present day. 



BIOLOGICAL LABORATORY, WESTERN RESERVE UNIVERSITY, 

 February, 1905. 



