jbd IX STARRY REALMS. 



instance, a bird and a fish can have originated by natural 

 selection from some common ancestor. The whole question 

 is, no doubt, complicated ; but it is easy to show how 

 minute differences between one generation and the next, 

 all tending in one direction, speedily reach to an appre- 

 ciable aggregate. 



Let me give an illustration. I know some tender 

 mothers who like to have their darlings photographed 

 every year in order to preserve a permanent record 

 of their development. No doubt the mother would have 

 no difficulty in distinguishing between the photographs 

 of her child at two years old or at three, or even 

 between those of her boy at thirteen and at fourteen. 

 But suppose that, instead of having the child photo- 

 graphed only once a year, he were to be photographed 

 every week from birth until he was full grown. This is 

 not at all an impracticable suggestion ; there would be 

 little more than a thousand photographs altogether. An 

 album could easily be made which would hold them all. 

 Of course the prudent mother would mark the dates on the 

 back ; but suppose this was not done, and the whole thou- 

 sand photographs got into confusion, would it be possible 

 to arrange them all in order again ? Certainly no out- 

 sider could do it ; he could sort them in a general way, so 

 as to have the babies at one end, and the young men at the 

 other, and the boys in the middle. But could he put the 

 whole thousand in regular order from one end to the 

 other? He could not. I doubt very much whether 

 even the mother herself could do it without numerous 

 mistakes. 



Now if this be granted, the difficulty sometimes felt in 



