330 BIOLOGY OF THE LABORATORY MOUSE 



linkage, be considerably higher than would be expected purel\' from the 

 original number of heterozygous gene pairs. On the other hand, the remain- 

 ing organisms will carry more heterozygosis than expected. In a similar 

 treatment Bartlett and Haldane (i) have discussed the effects of forced 

 heterozygosis. For brother-sister mating of yellow mice, for example, they 

 have estimated the probability of finding heterozygosis due to linkage with 

 the yellow locus. 



A factor that may affect the rate of increase of homozygosis is described 

 by Haldane (5) as follows: "A breeder will probably select the most vigorous 

 individuals as parents. He will eliminate a number of weak or infertile 

 recessives, which will be homozygous for particular genes, and probably so 

 for genes closely linked with them. He may also select for vigor due to 

 heterozygosis as such. Hence at least during the first five to ten genera- 

 tions, when the population is still appreciably heterogeneous, progress 

 towards homozygosis will be slightly slower than the above calculations 

 would suggest." 



Haldane gives formulae by which, under various systems of inbreeding, 

 the frequency of heterozygosis, at any locus, due to mutation after inbreed- 

 ing has begun can be estimated from the mutation rate. Unfortunately, 

 little is known about mutation rate in mammals. Haldane (4) estimates 

 that the gene for haemophilia arises by mutation in the population of 

 London about once in 50,000 life cycles. He concludes (5): "If this is 

 generally true for mammals, and the number of genes is not less than in 

 Drosophila, we may expect that as the result of mutation most members of 

 a mammalian pure line will be heterozygous for at least one gene as the result 

 of mutation. Since after 30-40 generations the majority of animals in such 

 a line have lost all their original heterozygosis, the line is then as pure as it 

 is every likely to be." 



The last sentence sums up the practical conclusions to be drawn from 

 this section. A later section will show that the genetic uniformity of a 

 given strain in regard to a given character can usually be tested statistically. 



Phenotypic Effects of Inbreeding 



So far we have considered only the genetic effects of inbreeding. The 

 experimentaHst does not work on genotypes, however, he is concerned with 

 characters or phenotypes. It is important, therefore, to consider how the 

 genetic consequences of inbreeding may, in turn, influence phenotypic or 

 character variation. As an introduction to this, it seems desirable to 

 digress briefly on the general causes of phenotypic variation. 



