1 86 THE FOUNDATIONS OF ZOOLOGY 



If we suppose that while the guns in each generation differ among 

 themselves like the original guns, each is more likely than a gun 

 taken at random to resemble its parent, is it not clear that if the 

 guns in each generation and the screens in the series are both 

 innumerable, the ultimate outcome of this survival of the fittest 

 will be the production of a race of guns adapted for sending their 

 balls through the holes in this particular series of screens, and 

 that if another series of screens arranged in a different line or of 

 a different size or shape were set up, the guns in later generations 

 would become adapted for sending their bullets through them ? 

 It is not necessary for us to know anything about the mechanism 

 of guns, or the reason why they differ among themselves, or any 

 data which might enable us to predict the paths of the bullets, 

 in order to see that this result may be expected to follow, in 

 course of nature, if only the trials be innumerable ; if all the balls 

 which fail to go through a hole are- counted out, and if each gun 

 is, on the average, more like its parent than a gun selected at 

 random. 



Each discharge of a bullet is an event ; the race of guns adapted 

 for driving the bullets through the holes is an adaptive type, and 

 the series of screens is the equivalent of those conditions of life 

 which, in course of nature, form a zoological type, or species, out of 

 the events which the infinite diversity among living things affords. 

 I have used the illustration as the simplest way to show the error 

 of the opinion that natural selection does not account for the origin 

 of species unless the differences between individuals are adaptive 

 prior to selection ; for it is plain that, in our illustration, the result 

 is independent of the nature of the projectiles, and equally inde- 

 pendent of the mechanism by which they are propelled, since our 

 reason for expecting the result would be the same even if they were 

 unknown projectiles propelled by unknown means. It is also clear 

 that one who witnessed the process from the far end, through the 

 holes in the targets, might suppose that the course of adaptive 

 modification had been directed, from behind, to a definite end, since 

 none of the balls that failed to go through the holes would be 

 visible from this point of view; nor would the discovery of fossil 

 machine guns do much to correct this error; for the difference 

 between the exterminated guns and the survivors in the same gen- 



