1877.] The Colors of Animals and Plants. - 723 
rivalry for mates (to which advantage the excess of color and 
plumage might sometimes conduce), there seems nothing to pre- 
vent a progressive development of these ornaments in all domi- 
nant races, that is, wherever there was such a surplus of vitality 
and such complete adaptation to conditions that the incon- 
venience or danger produced by them was so comparatively small 
as not to affect the superiority of the race over its nearest allies, 
If, then, those portions of the plumage which were originally 
erected and displayed became developed and colored, the actual 
display, under the influence of jealousy or sexual excitement, 
becomes intelligible. The males, in their rivalry with each other, 
would see what plumes were most effective, and each would en- 
deavor to excel his enemy as far as voluntary exertion could effect 
it, just as they endeavor to rival each other in song, even some- 
times to the point of causing their own destruction, 
There is also a general argument against Mr. Darwin’s views 
on this question, founded on the nature and potency of ‘ natural ” 
as opposed to “sexual” selection, which appears to me to be 
itself almost. conclusive of the whole matter at issue. Natural 
selection, or the survival of the fittest, acts perpetually and on an 
enormous scale. Taking the offspring of each pair of birds as, 
on the average, only six annually, one third of these at most 
will be preserved, while the two thirds which are least fitted will 
die. At intervals of a few years, whenever unfavorable conditions 
occur, five sixths, nine tenths, or even a greater proportion of the 
whole yearly production are weeded out, leaving only the most 
perfect and best adapted to survive. Now, unless these survivors 
are on the whole the most ornamental, this rigid selective power 
must neutralize and destroy any influence that may be exerted by 
female selection. For the utmost that can be claimed for this 
is that a small fraction of the least ornamented do not obtain 
- mates, while a few of the most ornamented may leave more than 
the average number of offspring. Unless, therefore, there is the 
strictest correlation between ornament and general perfection, 
_ the former can have no permanent advantage; and if there is 
= (as I maintain) such a correlation, then the sexual selection of or- 
nament, for which there is little or no evidence, becomes need- 
less, because natural selection, which is an admitted vera causa, 
will itself produce all the results. In the case of butterflies the 
_ argument becomes even stronger, because the fertility is so much 
greater, and the weeding out of the unfit takes place, to a great 
extent, in the egg and larva state. Unless the eggs and larve 
