C «4P. VJII. 
SEXUAL SELECTION. 
317 
Section in the manner hereafter to be described. The 
Sa tQe train of reasoning is applicable in the above, 
as well as in the following case, if we assume that 
females instead of males are produced in excess, for 
8 <ich females from not uniting with males would be 
superfluous and useless. So it would be with poly- 
8 a mous species, if we assume the excess of females 
*° be inordinately great. 
An excess of either sex, we will again say of the 
tri ales, could, however, apparently be eliminated through 
Natural selection in another and indirect manner, namely 
an actual diminution of the males, without any in- 
crease of the females, and consequently without any 
^crease in the productiveness of the species. From 
variability of all characters, we may feel assured 
^at some pairs, inhabiting any locality, would produce 
H rather smaller excess of superfluous males, but an 
ec |Ual number of productive females. When the off- 
ering from the more and the less male-productive 
Intents were all mingled together, none would have any 
direct advantage over the others ; but those that pro- 
^Uc'od few superfluous males would have one great 
Mirect advantage, namely that their ova or embryos 
' v ould probably be larger and flner, or their young 
better nurtured in the womb and afterwards. We see 
jfes principle illustrated with plants ; as those which 
e ar a vast number of seed produce small ones ; whilst 
j^ose which bear comparatively few seeds, often produce 
ar ge ones well-stocked with nutriment for the use of the 
^edlings. 73 Hence the offspring of the parents which 
la< l wasted least force in producing superfluous males 
p " I have often been struck with the fact, that in several species of 
*' ia mla the seeds in the capsules which contained only a few were 
er y much larger than the numerous seeds iu the more productive 
Ca Psul es . 
