June 28, 1918] 



SCIENCE 



627 



has done me the honor to endorse in two 

 reviews^ in tlie same journal. But I may 

 perhaps be permitted to ix)int out that, though 

 they object to the hypotliesis, they put nothing 

 in its place, and make no effort to explain the 

 figures upon which it is based, and which are 

 so clear and so consistent that if age and area 

 be abandoned, or a rival hypothesis be set up, 

 they can not be left without an explanation, 

 The hypothesis was originally based upon the 

 results obtained from tlie flora of Ceylon, where 

 distribution was only estimated, and though 

 there were many irregularities, the figures 

 came out clearly enough for it to be fairly 

 safe to publish the hypothesis. Since then, 

 actual measurements of area occupied have 

 been used — for New Zealand, the islands round 

 New Zealand, Jamaica, Hawaii and Australia 

 — and have given results which are far more 

 clear and decisive, following in the most extra- 

 ordinary manner what is required imder the 

 hypothesis. Its applicability has also been ex- 

 tended from the angiosperms to the Coniferse 

 and ferns. 



Professors Sinnott and Berry, it seems to 

 me, have to some extent misapprehended my 

 views. They certainly give them too wide an 

 application, and do not take sufficient notice 

 of the provisos with which I have hedged 

 round the use of age and area. For example, 

 as I have several times pointed out, age and 

 area must not be applied to single species (as 

 Professor Sinnott applies it), but only to 

 groups of 20 allied forms. In other words, it 

 is primarily a proposition in taxonomic plant 

 geography; and though I think that its ap- 

 plication may probably be extended, I have as 

 yet only applied it to groups of allied species 

 within a single country, and to genera (the 

 closest of groups) beyond. 



1 believe that in general, when there is a 

 genus covering the range of its family, that 

 genus is the parent of the rest of the family. 

 But of course not infrequently there are two 

 or more genera covering the family range, or 

 two or more dividing it between them (e. g., 

 one Old World, one New). In the solution 



2 Science, N. S., Vol. 43, 1916, p. 785; 45, 1917, 

 p. 641. 



of this problem, with which goes the parallel 

 fact that, as Professor Sinnott points out, 

 genera are often represented in a given coun- 

 try only by endemic species, and the similar 

 fact that species may be only represented by 

 endemic varieties, there lies, I believe, a very 

 large step on the way to modern theory of 

 evolution. 



There is no doubt that many species have 

 in past times died out, or been killed out, but 

 tliat is no proof that the process is going on 

 now, though man is no doubt responsible for a 

 great deal. Age and area, however, refers to 

 action imder practically imchanged conditions, 

 and the advent of man may completely alter 

 them. 



Professor Sinnott so far extends the ap- 

 plicability of age and area as to make it re- 

 sponsible for the deduction that herbs are 

 older than trees. In the first place these are 

 ecological, not taxonomic groups, and the 

 fact that they are undoubtedly very polyphy- 

 letic complicates the matter ; and in the second 

 place I have not said that the rate at which 

 a herb and a tree spread is the same, though 

 both are governed by age and area, and one 

 covering 1,000 square miles bears the same age 

 relationship to one that covers 500, in each 

 case. 



The attacks by Professors Sinnott and Berry 

 are upon my first three papers on age and 

 area.' Since then I have published* three 

 more. The first is controversial, dealing with 

 the inapplicability of natural selection, and 

 with the question of relative age (in one coun- 

 try) of endemic and widely distributed species. 

 I have shown by two crucial cases that the 

 former are the younger. In New Zealand the 

 widely distributed species take no notice of 

 Cook's Strait (between the chief islands), 

 while many endemics are held up there, having 

 evidently arrived too late to get across be- 

 fore the formation of the strait. Again, in 

 the Tristichacea; and Podostemacese, the prim- 

 itive genera, which resemble ordinary water 

 plants, are spread throughout the tropics, while 

 the extraordinarily dorsiventral forms, which 



3 Phil. Trans., 1915; Ann. Bat., 1916. 

 *Ann. Bat., 1917. 



