The range for the outbreeder was about 6 to 37, with most values from 10 to 20, and for the inbreeder the range was 0 to 12, with most values from 1 to 6.
For instance, overdominance in fitness at the loci underlying quantitative variation (a mechanism proposed by Robertson ) would appear unlikely if inbreeders and outbreeders maintain similar levels of genetic variation.
As will be seen, it appears that heritability values are not much lower in inbred populations than in outbred ones, so that such data may be used to give a maximum estimate of the difference in genetic variability between outbreeders and inbreeders.
In the absence of data on similar characters in the different populations studied, there is a danger that different characters have been studied in populations with different breeding systems and that such differences produce a bias in favor of high variability in inbreeders. No such bias is evident from inspection of the characters studied, but it would certainly be preferable to make comparisons for individual characters studied in populations over the range of breeding systems.
There are unfortunately too few studies involving paternal family components of variance to provide useful comparisons of [V.sub.G] or of narrow-sense heritabilities between inbreeders and outbreeders.