CH.xxii] GEOGRAPHICAL DISTRIBUTION 239 



a reasonable area, only individuals, and not the species, will 

 usually be affected by the struggle for existence. Only very 

 rarely will a new form overtake its parent over the whole or the 

 greater part of its range, and destroy it. We are no longer 

 obliged to regard a new species as coming into existence at the 

 expense of its ancestors. 



Another important general result of the work upon Age and 

 Area outlined above is to show that in any given countrv, and 

 therefore in the world in general, the "wides," which occupy 

 the largest areas (on the average), are the oldest forms, i.e. that 

 they were the first to appear. The facts set forth showing the 

 distribution of the various classes are indisputable at the stage 

 that the work has now reached, and they are wonderfully con- 

 cordant from one country to another. No one has attempted to 

 contradict them, but there has been much a priori reasoning to 

 the effect that this or that has not been allowed for, that it is 

 obvious that so-and-so must produce great effects, etc. None of 

 this reasoning, however, has attempted to explain the facts, 

 which are so striking and so consistent that they must have an 

 explanation, and that a mechanical one, on account of their 

 mechanical regularity. The only reasonable one is, as frequently 

 ]Dointed out, that the factors acting upon dispersal produce in 

 the long run a very uniform effect, so that age forms a measure 

 of dispersal. 



But if this be so, tliere is no possible and reasonable explana- 

 tion of the endemics, which in general are younger than the 

 wides, and occur beside or near them, except that they are 

 descended from the wides, directly or through other endemics. 

 But when a new endemic arises in this way, unless it is much 

 better suited to a variety of conditions than its parent, it will 

 never overtake the latter, and we have seen that there is little 

 reason to supi^ose a combat a Voutrance between them. The 

 parent will most often, probably, surA'ive beside the child. At 

 times it is possible that it may survive only beyond it; but the 

 distribution, for example, of the Ranunculi of New Zealand, 

 where the parental wides are just as common in the region where 

 the crowd of endemics occurs, as in the far north where there are 

 none (cf. map on p. 156), gives little evidence in favour of this 

 latter supposition. In my various papers I have assumed that 

 the wides give rise to the endemics, and have made nearly a 

 hundred predictions upon this basis. As these predictions have 



