Origin and Development of the Composite. 5 
The truth and value of this contribution to the subject is 
obvious when it is noted that the composition of and many other 
points concerning the flora of a country can be predicted with a 
reasonable degree of accuracy. It has been applied very success¬ 
fully to the evolution of the Podostemaceae by Willis (IV, 93) and 
has been confirmed for the Graminese of Australia by Breakwell (10). 
Taylor (70) dealing with only 22 endemic species, finds it true in as 
detailed a fashion as can be expected when dealing with single 
species, but he transgresses the Willisian dictum that comparisons 
must be made in groups of 20 or more species. Under these 
circumstances the exceptions he finds are .quite in accordance with 
expectations. Taylor effects a compromise between the views of 
Sinnott (65) and Willis. He finds that of 22 endemic species 16 
are young and 5 relicts, the remaining case being doubtful. Of the 
16 young species 14 are considered to be due to generic or specific 
instability and 2 to the direct action of the environment (cp. Chap. 
XI, B). 
The theory of “ the differentiation of primitive, world-ranging, 
generalised types in response to the differentiation of their conditions” 
(Guppy, IX, 22, p. 313) is quite in accordance with the facts of Age 
and Area, and the idea of the differentiation, especially of insular 
endemic species, in situ is in close agreement with what is known of 
epharmonic variations. This also solves one of the problems of Age 
and Area, i.e. the occurrence on islands of endemic species of wide 
spread genera, which are there the only representatives of these 
genera. If the individuals were confined on arrival to a particular 
area they would naturally all show the same epharmonic variation, 
with the result that the original species as represented on the island 
would be entirely transformed into the new species. 
Guppy, however, seems to use primitive in a peculiar sense 
(op. cit., p. 315) since he classes the Compositae as primitive! He also 
appeals very much to geological changes, but no authentic fossil 
Composites are known below the Oligocene, and^the differentiation 
of climate since has not been sufficient to account for all the 
differentiation in the family. 
Other accounts of phenomena bearing on Age and Area and 
Differentiation are given by Kroeber (46), Andrews (4), A. Jordan 
(44, pp. 18-19), D. S. Jordan (45), Lloyd (47) and Samuelsson (59) 
The most important contribution for present purposes is that by 
Bentham, who was never an enthusiastic Natural Selectionist. He 
says (I, 7, p.481) “The result of the best-founded opinions on this 
