XXIV.] NATURAL SELECTION. 315 



see any insuperable difficulty in further believing it pos- 

 sible that the membrane-connected fingers and fore-arm of 

 the Galeopithecus might be greatly lengthened by natural 

 selection : and this, as far as the organs of flight are con- 

 cerned, would convert it into a bat. In bats which have 

 the wing-membrane extended from the top of the shoulder 

 to the tail, including the hind-legs, we perhaps see traces 

 of an apparatus originally constmcted for gliding through 

 the air rather than for flight." 



It may be necessary to state distinctly, that no one Bat and 

 supposes the bats or the flying lemur to be descended from ^y^^S ^ 



^^ •" ^ lemiu- not 



a squirrel. The unlikeness of the other parts of the descended 

 organism excludes this. But Darwin believes, and I agree ™,]^i.i^el 

 with him, that their wings have been produced by the 

 improvement of an organ which originally resembled the 

 parachute of the flying squirrels. 



I mention this instance as the best I can find, of organs 

 that may have been produced by natural selection, and 

 cannot have been produced by any process of self-adapta- 

 tion. The membranes of the flying squirrel, the flying 

 lemur, and the bat, are simple organs ; and those of the 

 flying squirrel, at least, are not closely connected with the 

 other parts of the organism : that is to say, their presence 

 does not cause any very great deviation from the usual 

 structure of squirrels. These two facts, that the mem- 

 branes in question do not consist of a number of corre- 

 lated parts, but are simple organs, and that they have 

 no close correlation with the other parts, make this a 

 peculiarly good case to isolate in thought, so as to make it 

 the subject of reasoning. It is certain that the flying 

 squirrels can take much longer leaps than they could if 

 they had no membranes, and it needs no proof that this 

 power must be useful to them. But it is, I think, equally Mem- 

 obvious, that the habit of leaping cannot have any ten- pJlotjueed°* 

 dency to develop such membranes, or to increase their by self- 

 size after they begin to be developed. Self-adaptation tiou! but 

 thus fails to account for this very simple change. Natural ^^ natural 

 selection, however, here comes into play. I agree with 

 Darwin in thinking that the preservation, for generation 



