254- PHYSIOLOGIC GENETICS 



The guinea pig is the species to turn to for studies of liposarcomas, 1198 and the 

 rabbit for studies of uterine tumors. 496 Even animals not yet introduced into the 

 laboratory may offer opportunities for the study of certain neoplasms. The rodent 

 Mastomys, recently brought into the laboratory in South Africa, develops carcinoma of 

 the glandular stomach in the laboratories there. 959 This neoplasm is rarely found in 

 any other species except man. 



ESTIMATION OF NUMBER OF GENES 



Incidence data can be used for estimating the number of genes involved in neo- 

 plasia provided one goes beyond the classical F l5 F 2 , and backcross ratios and obtains 

 breeding tests of the backcross or the F 2 segregants. In Wright's 1451 analysis of 

 Polydactyly, results of crosses between strains 2 and D simulated one-factor Mendelian 

 heredity with dominance of 3-toe over 4-toe in the F^ Segregation in the F 2 closely 

 approached a 3 : 1 ratio, and in the backcross to the 4-toed strain D, a 1 : 1 ratio. This 

 interpretation broke down, however, in the tests of the supposed segregants, and at 

 least 3 factors and probably 4 by which the parental strains differed was indicated. 

 He found a close approach to blending inheritance in a character which approached 

 alternative expression because of physiologic thresholds. 



In their classical studies of the genetics of leukemia in mice, MacDowell and co- 

 workers 835 were not content to classify their backcross segregants between the high- 

 leukemic strain C58 and the low-leukemic strain StoLi merely on the basis of whether 

 or not each developed leukemia. Instead, they subjected a series of backcross males 

 to progeny tests by mating them to StoLi females and observing the incidence of leu- 

 kemia in the family of each. The incidence in these families varied from to 42.8 per 

 cent with a fairly symmetrical distribution, a result explainable only on the basis of 

 multiple factors. 



Some of the early F 2 and backcross segregation ratios obtained by Bittner 107 for 

 spontaneous pulmonary tumors and by Andervont 29 for induced pulmonary tumors 

 suggested single-factor inheritance with susceptibility dominant over resistance. Such 

 an interpretation was not reconcilable, however, with the data on incidence for the 

 inbred strains. Not only were there highly susceptible strains such as strain A and 

 highly resistant strains such as C57BL, but other strains such as BALB/c and BL had 

 intermediate incidences. 



Estimation of the number of genes involved in pulmonary tumors was made 

 possible by transforming the phenotypic expression of the character from alternative to 

 quantitative expression. This was done by introducing a chemical carcinogen, in this 

 case dibenz[a, A]anthracene, and measuring response by counting the resulting 

 multiple tumors. 557 In crosses between the high tumor strain A and the low tumor 

 strain C57L, the mean of the F x was found to be intermediate between the means of 

 the parental strains when age was kept constant. The mean of the F 2 was likewise 

 intermediate, but the variance of the F 2 was greater than that of the F x . The A back- 



