INADEQUACY OP NATURAL SELECTION, ETC. 611 



slight variations may be equivalent. This one has unusual 

 agility, and leaps a chasm which others balk at. That one de- 

 velops longer hair in winter, and resists the cold better. Another 

 has a skin less irritated by flies, and can graze without so much 

 interruption. Here is one which has an unusual power of detect- 

 ing food under the snow ; and there is one which shows extra 

 sagacity in the choice of a shelter from wind and rain. That the 

 variation giving ability to eat a plant before unutilized, may 

 become a trait of the herd, and eventually of a variety, it is need- 

 ful that the individual in which it occurs shall have more de- 

 scendants, or better descendants, or both, than have the various 

 other individuals severally having their small superiorities. If 

 these other individuals severally profit by their small superiori- 

 ties, and transmit them to equally large numbers of offspring, no 

 increase of the variation in question can take place : it must soon 

 be cancelled. Whether in the Origin of Species Mr. Darwin has 

 recognized this fact, I do not remember, but he has certainly 

 done it by implication in his Animals and Plants under Domestica- 

 tion. Speaking of variations in domestic animals, he there says 

 that " any particular variation would generally be lost by crossing, 

 reversion, and the accidental destruction of the varying individu- 

 als, unless carefully preserved by man." (Vol. II, p. 292.) That 

 which survival of the fittest does in cases like the one I have 

 instanced, is to keep all faculties up to the mark, by destroying 

 such individuals as have faculties in some respect below the mark ; 

 and it can produce development of some one faculty only if that 

 faculty is predominantly important. It seems to me that many 

 naturalists have practically lost sight of this, and assume that 

 natural selection will increase any advantageous trait. Certainly 

 a view now held by some assumes as much. 



The consideration of this view, to which the foregoing para- 

 graph is introductory, may now be entered upon. This view con- 

 cerns, not direct selection, but what has been called, in question- 

 able logic, " reversed selection " the selection which effects, not 

 increase of an organ, but decrease of it. For as, under some con- 

 ditions, it is of advantage to an individual and its descendants to 

 have some structure of larger size, it may be, under other condi- 

 tions namely, when the organ becomes useless of advantage to 

 have it of smaller size ; since, even if it is not in the way, its 

 weight and the cost of its nutrition are injurious taxes on the 

 organism. But now comes the truth to be emphasized. Just as 

 direct selection can increase an organ only in certain cases, so can 

 reversed selection decrease it only in certain cases. Like the 

 increase produced by a variation, the decrease produced by one 

 must be such as will sensibly conduce to preservation and multi- 

 plication. It is, for instance, conceivable that were the long and 



