MY LITERARY WORK 99 



very much higher than ours. It contained also a chapter on 

 humming birds, as illustrating the luxuriance of tropical 

 nature; and others on the colours of animals and of plants, 

 and on various biological problems. 



As soon as we were settled at Croydon, I began to work 

 at a volume which had been suggested to me by the neces- 

 sary limitations of my " Geographical Distribution of Ani- 

 mals." In that work I had, in the first place, dealt with the 

 larger groups, coming down to families and genera, but taking 

 no account of the various problems raised by the distribution 

 of particular species. In the next place, I had taken little 

 account of the various islands of the globe, except as forming 

 subregions or parts of subregions. But I had long seen the 

 great interest and importance of these, and especially of Dar- 

 win's great discovery of the two classes into which they 

 are naturally divided — oceanic and continental islands. I had 

 already given lectures on this subject, and had become aware 

 of the great interest attaching to them, and the great light 

 they threw upon the means of dispersal of animals and plants, 

 as well as upon the past changes, both physical and biological, 

 of the earth's surface. In the third place, the means of dis- 

 persal and colonization of animals is so connected with, and 

 often dependent on, that of plants, that a consideration of the 

 latter is essential to any broad views as to the distribution of 

 life upon the earth, while they throw unexpected light upon 

 those exceptional means of dispersal which, because they are 

 exceptional, are often of paramount importance in leading to 

 the production of new species and in thus determining the 

 nature of insular floras and faunas. 



Having no knowledge of scientific botany, it needed some 

 courage, or, as some may think, presumption, to deal with 

 this aspect of the problem; but, on the other hand, I had 

 long been excessively fond of plants, and was always in- 

 terested in their distribution. The object, too, was easier 

 to deal with on account of the much more complete knowl- 

 edge of the detailed distribution of plants than of animals, 

 and also because their classification was in a more advanced 



599316 A 



