184 H. B. GUrPY, F.R.S.E.; ON I'LANT-DISTEIBUTION 



we only touch the fringe of a great problem. The view that 

 such investigations go a very little way towards explaining 

 the facts of general distribution is in accord with the view 

 elaborated in my book that nature lias niade no especial pro- 

 vision for seed-dispersal. The instability in the past as well 

 as in the present of the fruit as compared with the Hower, and 

 its relative unfitness for purposes of classification, are facts- 

 which point in the same direction. If we accept the principle 

 of the differentiation in siht of universally distributed types in 

 response to the secular differentiation of their life-conditions, 

 we see at once the accidental character of the working of the- 

 dispersing agencies. 



The Theory of Differentiation briefly stated. 



The hypothesis is one that connects the differentiation of 

 plant-types over the earth with the secular differentiation of 

 the life-conditions. "With the creation of these types we are 

 not concerned, since we are only witnesses of the processes 

 connected with their diversification. Althougli this view i& 

 advanced in the preface and final chapter of my book on Plant- 

 Dispersal, a work dealing mainly with insular floras, I had not 

 then sufficiently grasped the idea that whilst the study of 

 means of dispersal explains much in the case of floras of oceanic 

 islands, it goes a very little way towards solving the great 

 problems connected with continental floras. Only the later 

 phases in the history of plant-distribution are illustrated in the 

 islands. Principles of great weight in the stocking of oceanie 

 islands shrink considerably in their importance when we apply 

 them to the plant-distribution of the globe. 



Now what, we may ask, is the significance of a differentiating 

 world ? This process has been at work on our globe from the 

 beginning, and its operations are to be observed alike in the 

 infinitely great and in the infinitely small. In those first ages 

 when dense envelopes of mist and cloud screened off the direct 

 rays of the sun from the earth's surface, when the air was over 

 saturated with aqueous vapour, and when the life-conditions 

 were uniform over the globe, the same generalised plant- types 

 were distributed over the earth. Then ensued a process of 

 desiccation which is still in operation, and it is with a world 

 that has for ages been drying up that the significance of 

 differentiation lies. 



The origin of the tribes is to be connected with the earlier 

 stages of tlie process, those, for instance, concerned with the 



