NATURAL SELECTION: II 505 



but there must have been a tachyteHc phase in the evolution of bats prior 

 to that time. Such a spurt in evolution, involving change from one way of 

 life to another, has been termed by Simpson "quantum evolution." 



Can we imagine the conditions which would lead to quantum evolu- 

 tion among certain arboreal insectivores, giving rise to bats? One environ- 

 mental niche is that of flying insect-eaters. That it is not an easy niche to 

 enter is attested by the fact that only three vertebrates have entered it: 

 small pterosaurs (Fig. 3.1, p. 22), some birds, and bats. Pterosaurs had 

 either disappeared or were about to disappear at the time bats were evolv- 

 ing. At any rate, bats would fill the niche much more eflRciently than would 

 these flying reptiles. Birds are highly efficient occupants of the niche, but 

 most insectivorous birds do their insect feeding by day (as, probably, the 

 pterosaurs did also), while bats are active by night. The environmental 

 niche found open by these early arboreal insectivores was, then, that of 

 nocturnal, flying insect-eaters. We recall that they were already insect- 

 eaters, and probably nocturnal, so the change called for was primarily de- 

 velopment of the power of flight. 



Most of the arboreal insectivores of that day had no capability for enter- 

 ing the vacant niche. But somewhere there must have existed a small group 

 of them, a subpopulation, in terms of our previous discussion, that under- 

 went rapid evolution in developing wings. We recall that division of a pop- 

 ulation into subpopulations affords optimal conditions for evolutionary 

 change. In some such subpopulation genetic variability of types pre- 

 viously enumerated (pp. 494-495) must have combined to alter the struc- 

 ture of the forelimb toward that of a wing. This alteration may have been 

 spread through the subpopulation by genetic drift; it was certainly favored 

 by natural selection. Probably, though not necessarily, the modified fore- 

 limb was first used for gliding from tree to tree; various vertebrates have 

 achieved one type of structure or another for gliding through the air. But 

 gliding does not provide the means for entering the ffying insect-eater 

 niche. The ancestors which essayed to enter this niche could have done so 

 only by developing forelimbs capable of true flight. Anything less would 

 have been too little. Insect-eaters in transition between life in trees, includ- 

 ing, perhaps, gliding from tree to tree, and life involving true flight would 

 have been in a most unstable position, not well adapted for any environ- 

 mental niche. Consequently, for them natural selection must have oper- 

 ated with extreme severity, resulting in rapid perfection of the flight 

 mechanism. 



Now we can understand why our fossil collections contain no transitional 

 forms between insectivores and bats. If our interpretation is correct, this 



