XXIV.] FOEMATION OF COMPLEX ORGANS. 327 



greatly modified, no doubt many parts would have to be simul- 

 taneously altered, in order that the organ should remain service- 

 able. But is this the case with smaller changes 1 " He then 

 speaks of the changes that would be needed in order to adapt 

 the eye of a diurnal species for a nocturnal life — a case which 

 presents no great difficulty ; and afterwards says : — " Amphibious 

 animals, which are enabled to see both in the water and in the 

 air, require and possess, as M. Plateau has shown, eyes con- Eyes of 

 structed on the following plan : ' The cornea is always fiat, or at ^™yg^' 

 least much flattened in front of the crystalline and over a space animals, 

 equal to the diameter of that lens,^ while the lateral portions 

 may be much curved.' The crystalline is very nearly a sphere, 

 and the humours have nearly the same density as water. Now, 

 as a terrestrial animal became more and more aquatic in its 

 habits, very slight changes, first in the curvature of the cornea 

 or crystalline, and then in the density of the humours, or con- 

 versely, might successively occur, and would be advantageous to 

 the animal whilst under water, without serious detriment to its 

 power of vision in the air." 



I confess I do not see much force in this argument. The 

 subject is most intricate, but I know of no reason for thinking 

 it possible that any apparatus consisting of lenses can be im- 

 proved by any method whatever, unless the alterations in the 

 density and the curvature are perfectly simultaneous ; and the 

 probability against this taking place by mere spontaneous 

 variation is practically infinite. 



1 The purpose of the flat cornea is evidently to produce as little differ- 

 ence as possible in the way in which the rays of light enter the eye from 

 air and from water. 



