HUMAN GENETICS 



329 



affected individual, and there will be half this proportion among his 

 nephews, nieces, uncles and aunts, and a quarter among his first 

 cousins, and so on. On the other hand, if the gene was a recessive, most 

 of the affected x normal matings will be aa x AA and there will be 

 no affecteds among the offspring, or among the parents who would 

 usually be Aa x Aa. If the character depended on two complementary 

 factors, the affected : normal ratio would fall off very rapidly in passing 

 to more distant relatives of an affected individual. Thus this test may 

 suffice to identify a dominant even of low penetrance.^ 

 A true understanding of the heredity of a disease is of course more 



-O-n-O 



k k ^-r-O ^44 



MANY 



NORMAL 



DESCENDANTS 



o-T-[5 S f 



-D 



~i~^ 2>ra 



cr561)i 6 i i 6 AiT~i 



Fig. 135. A Pedigree of Glaucoma. — The character behaves as a dominant; it Is 

 usually handed on only by individuals who themselves show it. But at A and 6 

 it is transmitted by non-affected individuals, so that its penetrance is not complete. 

 Males squares, females circles, affected individuals black. 



(After Holmes, data of Courtney and Hill.) 



difficult to obtain when penetrance is low, but it possesses peculiar 

 importance in such cases. Some non-affected relatives of an affected 

 individual will in reaUty carry the gene, and if they can be detected it 

 may be possible to help them to adjust their environment so as to 

 minimize the chance of their gene being effective. 



3. Autosomal Recessives 



The equilibrium proportions of a random mating population can be 

 written i AA :2uAa: u^ aa, and if a is rare, u is small, and u^ very 

 small. Thus most recessives will be produced by matings of the type 

 Aa X Aa, which will occur in a relative frequency 4u^, whereas 

 matings Aa x aa will only occur with frequency 2u^. Recessives 

 therefore rarely have aflfected parents, and usually marry the com- 



^ Levit 1936. 



