142 TROPICAL NATURE, AND OTHER ESSAYS. 



so in the female, a further emigration westward took 

 place to the small island Mas-afuera, where they also 

 established themselves. Here, however, the change 

 begun in the larger island appears to have been checked, 

 for the female remains to this day intermediate between 

 the Juan Fernandez and the Chilian forms. More re- 

 cently, the parent form has again migrated from Chili to 

 Juan Fernandez, where it still lives side by side with its 

 greatly changed descendant. 1 Let us now see how far 

 these facts are in accordance with the general laws of 

 variation, and with those other laws which I have en- 

 deavoured to show regulate the development of colour. 2 



The amount of variation which is likely to occur in a 

 species will be greatly influenced by two factors the 

 occurrence of a change in the physical conditions, and 

 the average abundance or scarcity of the individuals 

 composing the species. When from these or other 

 causes variation occurs, it may become fixed as a variety 

 or a race, or may go on increasing to a certain extent, 

 either from a tendency to vary along certain special lines 

 induced by local or physiological causes, or by the con- 

 tinued survival and propagation of all such varieties as 

 are beneficial to the race. After a certain time a balance 

 will be arrived at, either by the limits of useful variation 

 in this one direction having been reached, or by the 

 species becoming harmoniously adapted to all the sur- 

 rounding conditions ; and without some change in these 



1 In the preceding account of the probable course of events in peopling 

 these islands with humming-birds, I follow Mr. Sclater's paper on the Land 

 Birds of Juan Fernandez, Ibis, 1871, p. 183. In what follows, I give my 

 own explanation of the probable causes of the change. 



2 See Macmillan's Magazine, Sept. 1867, " On the Colours of Animals 

 and Plants," and Chapters V. and VI. of the present volume. 



