268 TROPICAL NATURE, AND OTHER ESSAYS. 



has been found to be more universal and more complex 

 than could have been anticipated. Whole genera and 

 families of plants have been so modified as, first to attract 

 and then to be fertilized by, certain groups of insects ; 

 and this special adaptation seems in many cases to have 

 determined the more or less wide range of the plants in 

 question. It is also known that some species of plants 

 can be fertilized only by particular species of insects ; 

 and the absence of these from any locality would 

 necessarily prevent the continued existence of the plant 

 in that area. 



In this direction, I believe, will be found the clue to 

 much of the peculiarity of the floras of oceanic islands ; 

 since the methods by .which these have been stocked 

 with plants and with insects will be often quite different. 

 Many seeds are, no doubt, carried by oceanic currents, 

 others probably by aquatic birds. Mr. H. N. Moseley 

 informs me that the albatrosses, gulls, puffins, tropic 

 birds and many others, nest inland, often amidst dense 

 vegetation ; and he believes they often carry seeds, 

 attached to their feathers, from island to island for 

 great distances. In the tropics they often nest on 

 the mountains far inland, and may thus aid in the 

 distribution even of mountain-plants. Insects, on the other 

 hand, are mostly conveyed by aerial currents, especially 

 by violent gales ; and it may thus often happen that 

 totally unrelated plants and insects may be brought 

 together, in which case the former must often perish for 

 want of suitable insects to fertilize them. This will, I 

 think, account for the strangely fragmentary nature of 

 these insular floras, and the great differences that often 

 exist between those which are situated in the same 



