150 ENDEMISM AND DISTRIBUTION: SPECIES [pt. ii 



Iberian peninsula contains about 800, or roughly the same as 

 Ceylon, which, however, has only one-ninth of the area. The 

 really large numbers are south of the tropic of Cancer. The 

 Hawaiian Islands have 600, Ceylon 800, New Zealand over 1000, 

 Australia about 7500, Mexico and Central America about 8000, 

 and Brazil perhaps 12,000. They are especially common in 

 momitainous country, and it is worth noting that most islands 

 are also mountainous. 



No country or island has all its species endemic, though in 

 several or most places where there is a very large proportion of 

 endemic species, like Hawaii or New Zealand, it is very common 

 to find genera with all their species endemic (cf. reply to objec- 

 tion 28, p. 95). St Helena, with a very small flora, seems to 

 have perhaps the highest proportion of endemic species, but of 

 countries with any large number, West Australia, with 85 per 

 cent, of its species endemic, takes the first place. The Hawaiian 

 Islands, with 82 per cent., are close behind. New Zealand (37; 

 has 72 per cent., the Galapagos 46 per cent., the Bahamas 14 per 

 cent., thus illustrating the fact that on the whole the further out 

 and more isolated an island is, the greater is its proportion of 

 endemic species. Fiji and Tahiti have much smaller proportions 

 than the Hawaiian Islands, but Fiji, with 50 per cent., is much 

 nearer to the mainland than Tahiti with 35 per cent., so that this 

 alone is not sufficient explanation. Nearly half the ferns and 

 lycopods in the Hawaiian Islands are peculiar to the group, in 

 Fiji and Tahiti only about 8-9 per cent. 



A study of the areas occupied by endemic sjDeeies soon shows 

 that the}- may be of any size from a few square j' ards upwards, 

 and that there is no difference to be seen between them and 

 species tliat are not usually considered endemic, and which may 

 have areas of larger and larger size, up to one of a large portion 

 of the globe. It was these extraordinary differences in area 

 occupied, between species closely resembling one another, and 

 differing only in characters which could not, by any stretch of 

 imagination, be looked upon as fitting or unfitting them in any 

 way for the struggle for existence, that first caused me to begin 

 studj^ing areas, and searching for some more potent agent in 

 distribution than adaptation, a search which ultimately led me 

 to Age and Area. 



This new point of view, that the mere area occupied by a 

 species has some more definite immediate interest than simph^ as 

 an expression of some unknown character in the protoplasm, or 



