126 AGE AND AREA, AND SIZE AND SPACE, [pt. ii 



of course, only one of rough approximation, but the inequalities 

 more or less cancel out when the genera are taken in groups of 

 ten or more as specified for the Age and Area hypothesis. 



The average generic area has been determined by adding up 

 the marks (= square miles in millions) for all the genera in each 

 group and dividing by the number of genera. This "average 

 generic area" has been determined for each tribe and sub-tribe 

 in the Compositae, firstly, for the whole world (using my own 

 notes); then for the Old World, all America and, in the sub- 

 tribes, for each of the twelve great regions into which Bentham 

 divided the world in relation to the Compositae (7), using the 

 data given by that authority. The use of two sources for the 

 data has introduced some slight discrepancies in the figures, but 

 the value of the check also introduced by this method makes 

 these slight differences of no real consequence. All these data 

 are presented in Table I, and on the whole, taken in conjunction 

 with the relative ages and sources of these groups as previously 

 determined (fig. on p. 125), they form a striking corroboration, 

 both for the Age and Area hypothesis and for the pre^-ious 

 phyletic conclusions. 



in accordance with the indications of a diphylctic origin of 

 the Inuleae (103, p. 301), that tribe has been given in Table I 

 as two, the Gnaphalieae (limited) which includes the first five 

 sub-tribes (with the Gnaphaliinae divided into Eu-gnaphaheae 

 and Helichryseac) together with half the Relhaniinae; and the 

 Inuleae (limited) which includes the last three sub-tribes to- 

 gether with the other half of the Relhaniinae (cf. fig. on p. 120). 

 The Inuleae as a complete tribe may, therefore, be omitted. 



Taking the tribes in order of origin, as given on p. 125, we find 

 (Table I, col. 16) that the average generic areas range from 7-9 

 (Senecioneae) through 6-5, 6--4, 6-2, 5-6, oo, 3-9, 5-3, 4-9, 4-5, 

 4-8, 3-6 and 3-8 to 3-6 (Calenduleae). In this Hst there are three 

 figures not in series — 3-8 for the Arctotidcae follows 3-6 for the 

 Helenieae, and 4-8 for the Vernonieae follows 4-5 for the Eupa- 

 torieae, but these two pairs of tribes are approximately of the 

 same age (p. 125), and the relative positions could be reversed 

 without any argument. The third figure not in series is 3-9 for 

 the Mutisieae, a tribe in which much geographical splitting of 

 the genera largely increases the number of genera in proportion 

 to the area occupied, thus decreasing the average generic area 

 for the tribe. 



The gradual increase of average generic area with geological 



