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



and 412), into modern terms (46), one would say that the Astereae 

 was predominantly a case in which a primitively world-ranging 

 type has been differentiated in situ with practically no spreading 

 of markedly new types from definite centres of origin. Some 

 such explanation is almost necessary for the frequency of inter- 

 mediate species and the grading of the Aster, Erigeron and 

 Conyza types into each other, 



Guppy's theory of "Rank and Range," which, although 

 similar to "Age and Area," is slightly different, is therefore 

 exemplified in this tribe of the Compositae; whereas the other 

 tribes are examples rather of "Age and Area." A detailed 

 examination of many of the sub-tribes in other tribes shows 

 that Avitliin the sub-tribe there are seldom groups of genera 

 which show markedly different average generic areas. The Eu- 

 gnaphalieae and Hehchryseae are exceptions. This leads to the 

 conclusion that, in spite of the large numbers of genera and species 

 in the Astereae, this group is really of the same "rank" (with 

 regard to differentiation and Age and Area statistics) as the 

 normal sub-tribes of most of the other tribes. 



Cichorieae. The ten sub-tribes into which this tribe is divided 

 are admittedly artificial. Bcntham {op. cit., p. 475) writes: "It 

 is very difficult to arrange these genera into sub-tribes; and 

 those we have adopted are in a great degree artificial, and have 

 little or no connection with geographical distribution; we must, 

 therefore, now consider the principal genera separately." 



The Lactucinae have been indicated (103, pp. 271 and 282. and 

 p. 125) as the primitive grou]3, while the Seorzonerinae have 

 been indicated as a fairly definite and advanced group (103, 

 p. 282). Scolymus is quite a distinct genus and the only one in 

 the Scolyminae. Grouping the other seven sub-tribes together 

 as one, we have the series— 6-2, 5-9, 3-7, 4-0. The last figure is 

 of little importance since it represents only one genus, while the 

 other figures are in the usual sequence. Of the seven sub-tribes 

 which are grouped only one. Hyoseridinae (42/10 = 4-2), has as 

 many as ten genera; while the other two sub-tribes which are 

 taken singly show the figures 69/11 = 6-2 (Lactucinae) and 

 37/10 = 3-7 (Seorzonerinae). The inclusion of the American 

 genera in Seorzonerinae is distinctly artificial, and if only the 

 Old-World genera are taken the average is 5-8. Then the series 

 for the Cichorieae reads 6-2, 5-9, 5-8, 4-0, and it is in complete 

 sequence and in perfect accord with the origins given on p. 125. 

 With appropriate statistical treatment, therefore, this tribe 



