CH. VI] AGE AND AREA 61 



Ceylon-Peninsular-India next, Ceylon endemics least, showed 

 not only for the grand total, but also for every family of 14 or 

 more species. It was clear that any one group of allied species 

 behaved like any other group, and it was therefore obvious that 

 nothing but a mechanical explanation Avould serve. Natural 

 selection could not act on all plants alike with even pressure. 

 The only possible mechanical explanation seemed to me to be 

 age, which would almost necessarily act alike upon all. If one 

 supposed the "wides" to be (on the average) the oldest, and to 

 have been the first arrivals in Ceylon, they were thus allowed 

 sufficient time to spread to the largest extent. On the way, they 

 would give rise, perhaps somewhere south of the middle of the 

 peninsula, to the species now found in Ceylon and Peninsular 

 India; these would be next oldest, and would spread in Ceylon 

 to the second degree of distribution. The Ceylon endemics would 

 arise in Ceylon, and on the whole probably later still, from one, 

 or more likely both, of these groups, and being the youngest, 

 would have spread the least. It seemed to me that I was at 

 last provided with a simple and feasible explanation of the dis- 

 tribution of species, though it involved a great break with the 

 older ideas, inasmuch as it indicated that the Ceylon species 

 were confined to Ceylon simply because they had been too young 

 to have had time to spread abroad. 



It is clear, of course, that age in itself can effect nothing, but 

 it allows time for the various factors that are active in distribu- 

 tion to produce their effects. The mechanical regularity of the 

 figures giA'cn above demands a mechanical exjilanation. and the 

 only possible one seems to be that age is mainly responsible for 

 the distribution, or in other words, that the various factors that 

 are operative produce an average or resultant effect — so much 

 dispersal in so much time. Dispersal therefore becomes a measure 

 of age, except in so far as barriers, physical or ecological, inter- 

 fere. Distribution is very slow, and probably the \'ast majority 

 of species have not yet reached the limits that they might reach, 

 if sufficient time were allowed. 



The greatest change from the older view of matters, however, 

 consists in the fact that since one can no longer acccjit either 

 the view of local adaptation or that of relic nature, for the great 

 majority of local species, and as these show definite numerical 

 relationships to those of wider distribution that occur beside 

 them, one must regard the two classes as related. But as area 

 goes vnth age, the endemics must be the younger, and must 



