CH. I] INTRODUCTORY 5 



theory that was growing up in my mind in respect to it. Almost 

 simviltaneously Copeland (18) presented evidence for the same 

 view, practically enunciating the hypothesis itself, though not 

 in definite arithmetical terms. 



In various subsequent papers I published further suggestions 

 in regard to Age and Area, and the other hypotheses that I had 

 associated with it, but it was not until in 1912 I actually worked 

 over the complete flora of Ceylon with respect to local distribu- 

 tion that I discovered that the effects of mere age upon dispersal 

 were so clear and unmistakable that they could be expressed in 

 figures. My paper embodying these results was published in 

 1915, and has been followed by many others upon the same 

 subject. 



While the distribution of any single species is due, as has been 

 said, to the complex interaction of man}^ factors and barriers, 

 it must be remembered that only in the case of a group of allied 

 species will these be likely to act with some uniformit^^ Age, 

 on the other hand, pulls all alike, so that if one deal with groups 

 of allied species, and call the various factors a, b, c, d, e, etc., 

 while some Avill probably pull different ways on different species, 

 and so cancel one another, others will pull the same way upon 

 all, so that the dispersal of one group of ten may be due to 

 {a + b + e + f) X age = 10, and of another (alhed to these) 

 {a + b + e + g) X age = 20. The latter will evidently be of 

 much greater age than the former, as it occupies twice the 

 area, and the factors other than age are much the same. But if 

 one take two groups of unallied types, e.g. one of Leguminosae 

 and one of Gramineae, or one of trees and one of herbs, one may 

 have in one case (a + c + <Z + e) x age = 10 and in the other 

 (b + d + f + g) X age = 20, and a comparison as regards age 

 alone will evidently be impossible. 



A very excellent illustration of the principle here involved is 

 given by the tables of expectation of life published by the in- 

 surance companies. In no single case does "age" alone deter- 

 mine the period to which a man will live, yet by taking averages 

 of men of the same race it is possible to say with perfect accuracy 

 how long an average man of 45 will have to live, or a man of 

 46, etc. 



If one be deahng with one species only (or one life only), then 

 the interaction of many factors, including age, will be so com- 

 plex that one cannot say to which the distribution (or length of 

 life) is actually due. It must always be remembered that the 



