136 AGE AND AREA, SIZE AND SPACE [pt. ii, ch. xiii 



Eupatorieae. The series here (29-7, 9-6, 7-8) also corresponds 

 completely with the series for area. 



Vernonieae. The series here (16-6, 4-7) also corresponds com- 

 pletely with the series for area. 



Helenieae. Grouping 1' with 1", and 3' with 3" as being of the 

 same age we get the series 7-3, 3-0, 2-9, which agrees completely 

 with the order of origin as given previously (p. 125). 



Arctotideae. Grouping 2' with 2" as being of the same age we 

 get the series 14-3, 13-4, which also corresponds completely with 

 the series for area. 



The prediction that the series for the average number of 

 species per genus will follow those for the average generic area 

 may, therefore, be said to be verified for the tribes on the whole, 

 ten out of twelve showing a similarity; and also for the sub- 

 tribes on the whole, with the exception of three tribes out of 

 fourteen. Further, the divergences amount to two out of four 

 sub-tribes in the Inuleae (limited); while the subdivision of the 

 other two tribes (Anthemideae and Cichorieae) is admitted by 

 Bentham to be artificial. For the Anthemideae he records (7, 

 p. 451): "In the Genera Plantarum we have, for convenience' 

 sake, classed the genera somewhat artificially," and (op. cit., 

 p. 450) "It is not easy, either, to group them into well-marked 

 sub-tribes." On the artificial subdivision of the Cichorieae he 

 has already been quoted (p. 130). 



The conclusion is, therefore, quite justified that in the Com- 

 positae on the whole both the average generic area and the 

 average number of species per genus are closely related to 

 absolute age. 



