24 GEOGRAPHICAL DISTRIBUTION [Chap. VII 



Letter 396 sufficient for this ? and what became of the wonderfully rich 

 Cape flora, which, if the temperature of tropical Africa had 

 been so recently lowered, would certainly have spread north- 

 wards, and on the return of the heat could hardly have been 

 driven back into the sharply defined and very restricted area 

 in which it now exists. 



As to the migration of plants from mountain to mountain 

 not being so probable as to remote islands, I think that is 

 fully counterbalanced by two considerations : — 



a. The area and abundance of the mountain stations along 

 such a range as the Andes are immensely greater than those 

 of the islands in the N. Atlantic, for example. 



b. The temporary occupation of mountain stations by 

 migrating plants (which 1 think I have shown to be probable) 

 renders time a much more important element in increasing 

 the number and variety of the plants so dispersed than in the 

 case of islands, where the flora soon acquires a fixed and 

 endemic character, and where the number of species is 

 necessarily limited. 



No doubt direct evidence of seeds being carried great 

 distances through the air is wanted, but 1 am afraid can 

 hardly be obtained. Yet I feel the greatest confidence that 

 they are so carried. Take, for instance, the two peculiar 

 orchids of the Azores {Habenaria sp.) What other mode of 

 transit is conceivable? The whole subject is one of great 

 difficulty, but I hope my chapter may call attention to a 

 hitherto neglected factor in the distribution of plants. 



Your references to the Mauritius literature are very 

 interesting, and will be useful to me ; and I again thank 

 you for your valuable remarks. 



Letter 397 To J- D - Hooker. 



The following letters were written to Sir J. D. Hooker when he was 

 preparing his Address as President of the Geographical Section of the 

 British Association at its fiftieth meeting, at York. The second letter 

 (August 1 2th) refers to an earlier letter of August 6th, published in Life 

 and Letters, III., p. 246. 



4, Bryanston Street, W., Saturday, 26th [Feb., 188 1]. 



I should think that you might make a very interesting 

 address on Geographical Distribution. Could you give a little 

 history of the subject. I, for one, should like to read such 



