344 BIOLOGY OF THE LABORATORY MOUSE 



the character, then offspring of individuals at one end of the variation 

 should differ from offspring of individuals at the other end. If there is no 

 significant difference in offspring from different types of matings (or if the 

 parent-offspring correlation is not significantly different from zero) , then the 

 strain may be assumed to be genetically uniform for the character in question. 

 Lack of variation in tissue specificity, indicated by ioo% "takes" in 

 transplants, is sometimes used as a rough measure of the likelihood of 

 uniformity in genes affecting other characters. 



Sublines 



When a strain is maintained with a large number of animals it should be 

 recognized that, unless matings are made up with reference to a pedigree 

 chart, the strain may break up into many separate lines. These lines may 

 have quite different characteristics if their last common ancestors were not 

 genetically uniform or if different mutations have become fixed. 



Among mice supplied to research workers there is, as yet, no universally 

 accepted way of designating the extent of dissimilar ancestry of a given 

 group of animals. One strain may be kept so that all individuals at any one 

 time trace back to a common pair of ancestors as soon as possible. Another 

 strain, perhaps listed under a single name, may have lines of descent which 

 have been separate for many generations. These are often listed as "sub- 

 lines," but this term may be used by one breeder to indicate five generations 

 of separate descent, by another to mean twenty generations, or by yet 

 another only when he has actually observed phenotypic differences between 

 the branches. 



The research worker who wants maximum genetic uniformity in his 

 material should, therefore, keep a check on the branching of his own strains 

 and should specify that animals supplied from other sources have a common 

 ancestry within a certain number of generations or exhibit no genetic varia- 

 tion affecting the character under investigation. 



Risk of Contaminatiox 



Since a high degree of homozygosis is obtained only after many genera- 

 tions of inbreeding, a single unfortunate outcross may undo years of work. 

 In a mouse colony in which different lines are maintained an accidental 

 outcross may occur as a result of faulty pens, into which stray animals can 

 enter, or to the returning of animals to the wrong pen after removal for any 

 purpose. Risk of the latter can be reduced to a minimum by handling 

 ' different lines and sublines at different times and by keeping them in 



