346 BIOLOGY OF THE LABORATORY MOUSE 



It is hoped that the earher part of this chapter has made it abundantly 

 clear that phenotypic variation is usually present in an inbred line as a 

 result of environmental causes and that, although it is usually less, it is 

 sometimes actually greater than that in random bred stocks. 



It might be thought that this variation could be reduced by giving more 

 attention to the uniformity of the laboratory environment. This would 

 have little effect, however, on the many characters whose variation is due 

 largely to intangible factors in the maternal environment. Wright's elabo- 

 rate search for the environmental causes of variation in white spotting in 

 the guinea pig (22) ended with 89% of the variability due to causes which he 

 could classify only as "developmental accidents." Wright has suggested 

 that the degree of irregular asymmetry in the expression of a character 

 serves as a rough estimate of the importance of these factors and, therefore, 

 of the extent of variation to be expected after inbreeding. 



Misinterpretation of Differences between Strains 



There is a common belief that a character occurring only in alternate 

 categories (as opposed to the other extreme: a continuous distribution) 

 must be due to alternate genes in the same way that agouti and black coat 

 color are. This has sometimes led to a hunt for a single major gene differ- 

 ence as the cause of a character difference that shows no, or little, overlap 

 in two inbred strains. Thus, attempts have been made to find a single 

 pair of alleles responsible for the difference between "high" and "low" 

 tumor strains. Such a hunt is all right if it is critical. It can only be 

 critical if cognizance is paid to the fact that, owing to the common occur- 

 rence of biological thresholds, of all-or-none processes in development, many 

 characters are necessarily alternate in expression. Many genes may be 

 involved, the effects of some combinations falling below the threshold, while 

 the effects of the others fall above. If this fact is realized it will be appre- 

 ciated that apparent dominance in the Fi of a cross between strains, a 

 3: 1 ratio in the F2, and a i : i ratio in the backcross, are not critical criteria 

 of the presence of a single major pair of genes. Many genes may be involved 

 and the above generations happen to be cut by a threshold of effect into 

 approximately the above proportions. In one of his crosses between three- 

 toed and four-toed strains of guinea pigs, Wright (20) actually obtained 

 the above ratios, but was able to show that at least four factors were 

 involved. The critical experiment is to test the genetic nature of the types 

 apparently segregating in the backcross or F2 by breeding them with the 

 "recessive" stock. 



