68 THE DESCENT OF MAN. 



which it consumes, or passively to the surrounding con- 

 ditions, cannot have been thus acquired. We must not, 

 however, be too confident in deciding what modifications 

 are of service to each being; we should remember how little 

 we know about the use of many parts, or what changes in 

 the blood or tissues may serve to fit an organism for a new 

 climate or new kinds of food. Nor must we forget the 

 principle of correlation, by which, as Isidore Geoffrey has 

 shown in the case of man, many strange deviations of 

 structure are tied together. Independently of correlation, 

 a change in one part often leads, through the increased or 

 decreased use of other parts, to other changes of a quite 

 unexpected nature. It is also well to reflect on such facts, 

 as the wonderful growth of galls on plants caused by the 

 poison of an insect, and on the remarkable changes of 

 color in the plumage of parrots when fed on certain fishes, 

 or inoculated with the poison of toads; * for we can thus 

 see that the fluids of the system, if altered for some 

 special purpose, might induce other changes. We should 

 especially bear in mind that modifications acquired and 

 continually used daring past ages for some useful purpose, 

 would probably become firmly fixed, and might be long 

 inherited. 



Thus a large yet undefined extension may safely be given 

 to the direct and indirect results of natural selection; but 

 I now admit, after reading the essay by Nageli on plants, 

 and the remarks by various authors with respect to ani- 

 mals, more especially those recently made by Prof. Broca, 

 that in the earlier editions of my "'Origin of Species" I 

 perhaps attributed too much to the action of natural selec- 

 tion or the survival of the fittest. I have altered the fifth 

 edition of the ''^Origin" so as to confine my remarks to 

 adaptive changes of structure; but I am convinced, from 

 the light gained during even the last few years, that very 

 many structures which now appear to us useless, will here- 

 after be proved to be useful, and will therefore come within 

 the range of natural selection. Nevertheless, I did not 

 formerly consider sufficiently the existence of structures, 

 which, as far as we can at present judge, are neither bene- 

 ficial nor injurious; and this I believe to be one of the 

 greatest oversights as yet detected in my work. I may be 



*"The Variation of Animals and Plants under Domestication," 

 vol. ii, pp. 380, 383, 



