FROM AN OLD STANDPOINT. 



109 



example, inosL of tlie tribes established Ity the systematist for 

 living- forms are to be found in the Mesozoic deposits (see l^ilger's 

 *' Taxaceie," Das Pjiaiizenreicli, iv, 5). AVith a fainih' like the 

 AceracetT, which practically consists of a single genus (Acer), the 

 sections or subgenera based on the characters of existing species 

 include all the Tertiary forms ; and, stranger still, most of the 

 sections of the genus that were confined to one or other side of 

 the Atlantic in Tertiary times possess the same distribution now 

 (see "Aceracect" by F. Pax, JDas FJlanzcnreich, W, 163). We 

 seem indeed to be rarely able to get at the beginning of things 

 in the distribution of the flowering plants, whether it be a 

 family, a tribe, or a genus. 



The FiitST Postulate of the Theory of Differentiation. 



In those families where we get a glimpse of the differentiation 

 of the tribes we are apparently brought face to face with the 

 differentiation of a world-ranging primitive stock. This is a 

 point of the greatest sigiiiticance in connection with the stand- 

 point adopted in this paper. If behind the facts of distribution 

 lies the cardinal principle that the farther we ti'ace a type back 

 the more generalised are its characters and the wider is its 

 range, then we should be justified when working out the 

 history of a family in postulating a world-ranging primitive 

 parent type with the subsequent development of centres (>f 

 differentiation over its area. The means of dispersal would then 

 take a very secondary place as determining distribution except 

 in the case of insular floras.' This is the position which I 

 will first endeavour to establish in the elaboration of the theory 

 of differentiation. It will involve the possibility of the 

 development of tribes and even of genera in more than one 

 locality in the area of the family. 



The Views of Mr. Bextham and Professor Huxley. 



I will first refer to some of the indications supplied l)y the 

 great group of the Composita3. ^s'otwithstanding that it makes 

 a poor show in the fossiliferous deposits, Mr. Bentliam, the 

 monographer of the family (see Journ. Linn. Soc. Bot. xiii, 187->), 

 arrived at the conclusion not only that it is a very ancient 

 plant-group, but that its primitive stock was already widely 

 dispersed at an early period of its history. Both the Old and 

 the New World possessed the family at the earliest recognisable 

 stage, America, South Africa, the Mediterranean region, and 

 Australia serving subsequently as " centres of differentiation " 

 and becoming the homes of the tribes. The possibility of 



M 2 



