FKOM AN 01-D STAND I'OINT. 



185 



emergence of the continents and tlie first development of 

 climates. AVitU the further differentiation of the life-condi- 

 tions within the tribal area is to be associated the birth of the 

 genus ; and in response to the further specialisation within 

 the generic area of the conditions of existence arose the 

 species. In the plant-world the process of change has ever 

 been from the general to the special, the family becomes 

 specialised in its tribes, the tribe in its genera and the genera in 

 its species. AVith each step in the (hfferentiation scale, the 

 geographical range would become more and more contracted,, 

 until whilst a family occupied a continent or ranged the world 

 a species would be usually restricted to a very limited area. 

 This is a theory somewhat idealised, but nature has ever been 

 best symbolised by broad ideas. The dispersing agencies, for 

 instance, would tend to blur the outlines, but the main features 

 of distribution would remain unchanged. What explanation, I 

 would ask, that assumes only a single centre of development 

 could explain the behaviour of families distributed in both the 

 Old and the New World ? Here we find, as I have befoi'e 

 remarked, that whilst most of the tribes occur in both hemi- 

 spheres, most of the genera, and almost all of the species, are 

 restricted either to one or the other. 



The theory here advocated is concerned only with the normal 

 difl'erentiation of primitive types in response to the secular 

 differentiation of the physical lif© conditions. It does not 

 concern itself directly either with the abnormal plant-forms 

 that have arisen under exceptional conditions of isolation as on 

 oceanic islands, or with the floral modifications and monstrosities 

 that have been developed in later ages through the establish- 

 ment of a close biological relation with insects, birds and other 

 creatures, but it holds that we ought in all cases to be able to 

 penetrate the disguise. Such forms have not been produced on 

 ihe lines of development which begin with the differentiation of 

 a prinutive family type and represent the response of tlie 

 plant-world through the ages to the differentiation of the 

 physical world. They are essentially distinct, and the generic 

 value cannot possibly be the same for genera of such different 

 histories. 



Gexekal Application of the Theory of Diffeuentiatiox. 



I will conclude the first part of this paper with a few 

 general reflections. In the first place I would say that if we 

 are not too curious about beginnings the theory of differentiatioa 



N 2 



