330 AN INTRODUCTION TO MODERN GENETICS 



moner type of homozygous normal and have no affected children. On 

 the other hand, a quarter of their sibs are likely to be affected so that 

 the character appears as a "famiUal" one. 



A character determined by a recessive gene cannot be identified by 

 following it through several generations, since, as we have just seen, it 

 usually occurs suddenly in a lineage and then disappears again. If the 

 presence of a recessive gene is suspected, we should expect the ratio of 

 affecteds to normals, in families showing any affecteds, to be i : 3, 

 since these families will be produced by Aa x Aa matings. The 

 difficulty in testing this is that human families are so small that some 

 Aa X Aa matings will give only one or even no affeaeds. In casually 

 collected statistics as to the incidence of familial diseases, families with 



r^T 



6 



ff^ 



Fig. 136. A Pedigree of Albinism. — (Males squares, females circles, affected 

 individuals black.) The trait is inherited as a recessive. Note the inbreeding, 

 between an uncle and his albino niece at A and between cousins at 6. It is only 

 because of this unusual amount of inbreeding that the character appears in two 

 successive generations. 



(After Tertsch.) 



only I affected are often omitted, since they do not seem to show any 

 "familial" incidence. Even if the statistics have been carefully collected, 

 Aa X Aa families which happen to contain only normals and no 

 affecteds will not be recognizable. The proportion of affecteds among 

 the families which are recognized as involving the gene therefore tends 

 to be too high. Compensation can, however, be made for this. Agree- 

 ment with the theoretical i : 3 ratio is then often very good. 



The most conclusive test for a recessive gene is derived from a 

 consideration of inbreeding. Consider a heterozygous Aa individual. 

 He will only have affected offspring if he marries anotlier heterozygote 

 (or a homozygote, which is unlikely). Now since a is rare, Aa mates will 

 be fairly rare; but since the Aa individual has received this gene from 

 his parents and grandparents, it will also have been handed down to 

 some of his collaterals, such as cousins. Thus if the individual marries 

 his cousin, there is a considerable chance that the mating will be 

 Aa X Aa and that affected offspring will appear. Applying this argu- 



