ii 4 SECRETS OF EARTH AND SEA 



larger region than our little island namely, at those of a 

 great continent like Africa or South America we find that 

 there are species which show gradations. Thus at a series 

 of points, A, B, C, D, separated by some hundreds of miles 

 from each other, we find a corresponding series of butterflies 

 which are apparently closely similar species of one genus, 

 differing by a few spots of colour, or darker and lighter 

 tint, much as our Large White, Garden White, and Green- 

 veined White differ. But when the butterflies are caught 

 which occur at points intermediate between A and B, B and 

 C, C and D, we find intermediate varieties, and, in fact, if 

 we get a very large number from intermediate regions, we 

 can, in some instances, arrange them in line so that they 

 constitute a graduated series of forms, each being scarcely 

 distinguishable from the one before or the one behind it, 

 yet differing clearly from one a dozen places away. In 

 such cases there is often evidence to show that the variety 

 found at A breeds with that found at B, that of B with 

 that of C, of C with D, so that they form an inter-breeding 

 group, though perhaps the varieties at D will not pair with 

 those at A, or even with those at B. Then sometimes we 

 find in such a series, otherwise complete, a gap. Let us 

 suppose it is between the butterflies of B and C. We 

 find the series of gradations nearly complete, but some 

 natural condition such as the encroachment of the sea, 

 or the slow elevation of a mountain range, or the climatic 

 destruction of the necessary food-plant has " wiped out " 

 a few forms somewhere between those of B and C. They 

 no longer exist. The series is no longer connected by 

 inter-breeding forms; those occurring from A to B and 

 some distance beyond are one " species " varying in the 

 direction of the series C to D, but abruptly broken off 

 from the latter. The series C to D is also a " species " 

 with graduated varieties, but distinct ; it is cut off from the 

 lot once in continuity with it by the destruction of the 



