Sex ratio evolution is one of earliest topics in mathematical biology. Classic approach relies 
on assumption that fitness measure is number of grandsons. In literature we can find thesis 
that an individual sex ratio equal to 0.5 is an Evolutionarily Stable Strategy. 
A new model shows that 1:1 population sex ratio is not the product of selection of 0.5 female strategy, 
but  effect of self-regulation on the level of whole population (a kind of Evolutionarily Stable Set). 
A non-fisherian sex ratios can be produced by differences in mortality rates between sexes. 
It will also be shown where is the mistake in classic models.