STADLER 



thetical gene is unitary by definition. 

 But the genes identified in our experi- 

 ments cannot be made unitary by defi- 

 nition. The five genie elements repre- 

 sented in the diagram are not actually 

 parts of one gene; they are five genes. 

 But if certain multiple allelic series 

 have a basis of this type, it would be 

 possible to establish the fact experi- 

 mentally only in the cases most favor- 

 able for analysis. Accordingly, there 

 might be many cases in which the seg- 

 ment of the gene-string identified ex- 

 perimentally as a single gene might 

 actually be a cluster of genes of iden- 

 tical or similar effect. 



The notion of the compound gene, 

 or some equivalent unit, may prove to 

 have significance, since there may be 

 special relationships among the clus- 

 tered elements that mark them off as 

 a group from adjoining unrelated ele- 

 ments. One of these may be interrela- 

 tionships in gene action between the 

 clustered elements, which could lead 

 to the occurrence of position effects 

 when members of the cluster are sepa- 

 rated by crossing over or transloca- 

 tion. This may be a basic factor in the 

 explanation of position effect in gen- 

 eral. Another relationship to be ex- 

 pected is synaptic equivalence, leading 

 to the opportunity of unequal crossing 

 over. It is the latter that concerns us 

 here. 



A striking example of minute defi- 

 ciencies simulating gene mutations is 

 provided by the "crossover-mutants" 

 of R'^ (24). Certain i?'' alleles consist 

 of at least two independently mutating 

 genie elements: (P), determining an- 

 thocyanin pigmentation of certain 

 plant tissues and of the pericarp, and 

 (S), determining anthocyanin pigmen- 

 tation of the endosperm and embryo. 

 The crossover-mutants 7?" and r^ re- 

 sult from unequal crossing over and 

 must, therefore, involve the loss of 

 (P) in the one case and of (S) in the 



255 



other. They give no cytological or 

 genetic indication of deficiency, and 

 they are wholly normal in develop- 

 ment in the haploid gametophyte, as 

 is shown even by the very sensitive 

 test of competitive pollen-tube growth 

 in the transmission of the mutant 

 through male germ cells. The cross- 

 over-mutants are wholly indistinguish- 

 able in appearance and genetic be- 

 havior from the noncrossover mutants 

 occurring in the same cultures. 



The occurrence of unequal crossing 

 over within the R complex yields some 

 interesting indications of the genetic 

 nature of multiple allelic series and of 

 the possible role of gene losses in re- 

 lation to seemingly qualitative muta- 

 tions. In addition to (P) and (S), there 

 are other phenotypically recognizable 

 genie elements of the R complex. In 

 certain R' alleles of dilute pigmenta- 

 tion, both plant and seed color are de- 

 pendent upon a single genie element 

 (D). In various i?*" alleles of unusually 

 strong pigmentation, there appear to 

 be additional elements determining 

 certain aspects of plant-color expres- 

 sion. In addition, there are various dis- 

 tinguishable aleurone-eolor types such 

 as "Stippled," "Marbled," "Navajo- 

 spot," and so forth, some occurring 

 with plant color and some without. 

 Each of the distinguishable complexes 

 may be regarded as one of a long series 

 of multiple alleles of the gene R. 



Let us pause a moment to clear the 

 terminology. To avoid confusion I 

 shall refer to the recognized alleles of 

 R under their customary italicized 

 designations (R\ R", r, 'R^\ and so 

 forth), although the analysis shows 

 that several of these so-called "alleles" 

 are actually complexes of two or more 

 genes. 



The term gejiic eleme?it will be used 

 for any gene-like constituent identi- 

 fied as a component of one of the R 

 alleles. The use of this term does not, 



