700 HEREDITY AND CELL DIVISION 



carry out the reaction or produce the character for which the 

 enzyme was needed, and the heterozygote might be expected 

 to do so completely if one dose of the gene were enough, when 

 there would be complete dominance, or only partially when the 

 enzyme supply was inadequate and dominance would be incom- 

 plete. This explanation clearly cannot apply to multiple allelo- 

 morphs, where there are several possible states at one gene locus, 

 and in fact there is evidence from Drosophila that very few 

 recessive genes really do nothing. The gene scute— i, for example, 

 reduces the number of bristles on the fly ; by the action of X-rays, 

 which cause irregularities in meiosis, it is possible to get extra 

 fragments of chromosomes in the sperms so that more than one 

 gene of the allelomorphic series is present in a single nucleus. 

 When tw^o scute— i genes are present the number of bristles is 

 only a little less than normal ; with three it is greater than normal. 

 The recessive gene must be able to carry out the reaction per- 

 formed b}' the dominant, but less effectively. The simplest 

 chemical explanation that can be given of the change is that there 

 has been an alteration in a side-chain of the gene, such as a 

 jump from an ethyl to a methyl group, or a stereoisomeric change. 

 Similar changes are known to affect the activity of hormones. 

 Multiple allelomorphs would consist of a series of substitutions 

 which left the basic nature of the gene unchanged. 



The genes which we have been considering so far have been 

 those which cause relatively large and obvious changes in the 

 phenotype ; they are generally now called major genes. There 

 are others, the polygenes, which individually are of very small 

 effect, and can only be studied statistically ; there are usually 

 a number having similar effects, so that they are replaceable 

 and a very complex series of genetic types is possible. ]\Iuch 

 continuous variation, as of size and weight, is due to environ- 

 mental factors, but that part of it which is genetic in origin is 

 probably always controlled by poh'genic systems and is said to 

 be multifactorial ; other potygenes are modifiers, producing small 

 effects on the expression of major genes. Polygenes show seg- 

 regation and linkage, although their effects are too small to be 

 studied by the Mendelian method. 



The effects of a gene are influenced not only by specific modifiers 

 and variations in the environment, but by its position on the 

 chromosome and by the sum total of all the other genes present. 

 The position effect shows when by artificial means a gene is 



