168 CELL HEREDITY 



synthesis of glutamic dehydrogenase complement one another in hetero- 

 karyons. So do a number of mutants affecting the synthesis of adenylo- 

 succinase in the same organism. In the latter case the more peripheral 

 sites show the greatest complementation. When the enzyme contents of 

 these frarw-heterokaryons are measured, there is only a partial restora- 

 tion of activity; at the most only 25 per cent of the wild-type activity is 

 attained. Such intra-allelic complementation is probably achieved by 

 the interaction between gene products. Isolation of nuclei from hetero- 

 karvons in which sites in the same cistron had been interacting in trans- 

 configuration, shows that their genetic structure is unchanged. In higher 

 forms, similar complementation exists between mutons within what seems 

 to be a cistron. These can be fragmented only rarely, if ever, by recom- 

 bination. The series of t sites affecting the development of the tail 

 and hind parts of mice is a case in point. 



Thus when two mutant functional units that are nonallelic are con- 

 fronted in transposition, complete complementation results; but when 

 two allelic functional units are confronted in transposition, only partial 

 complementation results at best. As long as this distinction holds in 

 practice, the functional unit can be operationally defined in terms of the 

 cis-trans test. But since the amount of partial complementation may in- 

 crease up to 25 per cent in proportion to the distance between the 

 mutant sites within the functional unit, we may ask whether a fourth unit 

 exists, smaller than a functional unit but larger than a recon and a 

 muton. In this event, there should be a cluster of mutons which com- 

 plement to the same degrees other groups located elsewhere within the 

 functional unit. The evidence available at present does not allow a 

 certain distinction between such a discontinous arrangement along the 

 functional unit and a continuous gradient in the quantity of comple- 

 mentation with distance. Elucidation of the intimate mechanism of 

 intragenic complementation promises a better understanding not only of 

 gene action but also of gene structure; it is further discussed in 

 Chapter 10. 



We can meanwhile construct an imperfect image of the gene. More 

 than one may be possible and the concept we will present is only an 

 approximation, accounting for a vast assemblage of facts. Its greatest 

 virtue is its specific nature, wherein also lies its greatest tendency for 

 error. But being specific, it is amenable to test by experiments which 

 will allow its proper modification in time. 



The cistron is considered to be a coded sequence of nucleotide pairs in 

 a DNA molecule. It must be larger than a muton and a recon, other- 

 wise single nucleotides could function as genes. It may carry the in- 

 formation necessary for the specification of the order of amino acids in 



