128 AGE AND AREA, AND SIZE AND SPACE, [pt. u 



artificial, while several contain genera which are exceptionally 

 widespread because of special dispersal mechanisms, we cannot 

 expect a complete correspondence. On the other hand, a family- 

 tree is available (p. 125) which was worked out in the first place 

 from the morphological characters of the styles and stamens 

 (103, fig. 7) and subsequently modified only slightly as a result 

 of the consideration of the most extensive data. It is, therefore, 

 interesting to examine the deviations from the numerical se- 

 quence in average generic area for the other seven tribes. 



Senecioneae. The Tussilagininae have been shown to be a 

 somewhat mixed group of genera, separated from the Senecio- 

 ninae in a rather artificial way (103, pp. 39 and 298), and these 

 two sub-tribes are, in fact, fused by Hoffman in the Pflanzen- 

 familien. The proper statistical procedure is, therefore, to take 

 them as one group for comparison with the other groups within 

 the tribe; then we obtain another unbroken sequence — 8-9, 6-8, 

 3-6 (Table I, col. 17). 



Gnaphalieae (limited). With the Eu-gnaphalieae and Heli- 

 chryseae as two distinct groups the series for the Gnaphalieae 

 reads 9-8, 8-8, 9-3, 4-3, 1-6, 1-7, 3-2. There are in this case two 

 marked exceptions to the sequence. The first is the Filagininae 

 (9-3) Avith only eleven genera including Micropus as a widely 

 spread weedy type, and Filago also of the weedy type and a 

 distribution suggesting either early dispersal by man or-a poly- 

 phyletic origin. If this genus were broken up into three, as was 

 done by many of the earlier syhantherologists, the average 

 generic area for the sub-tribe would be 7-8, and the sequence 

 would be imbroken except for the last sub-tribe. The second 

 exception is the Angianthinae (3-2), chiefly an Australian group 

 with only ten genera, the distribution of which in Australia may 

 be somewhat less on the whole than has been estimated. Such 

 a reduction in this sub-tribe would bring the average for the 

 Gnaphalieae to about 6-3, but a similarly careful revision in 

 detail of the other sub-tribes might result in raising the average 

 one or more decimals, so that such changes may be considered 

 negligible when the broad outlines of the history of the family 

 are being considered. 



Heliantheae. The series for this tribe appears rather irregular, 

 running thus: 6-9, 9-6, 3-7. 6-5, 2-7, 3-2, 12-8, 3-8, 6-0, 1-3; but 

 most of the sub-tribes contain less than ten genera. It is, there- 

 fore, advisable to group them; the first three groups are of early 

 origin (see p. 125), while those numbered 6', 6", 6'" in Table I 



