Chromosomes and Genes 171 



(1950) and Pontecorvo (I952a,b; 1953) for biochemical mutants of 

 Aspergillus, namely, three adjacent loci afiFecting biotin synthesis ( bi ) . 

 Again the "position allelism" (to use Lewis' term in a purely descrip- 

 tive way) was ascertained. A comparable case in Neurospora, afiFect- 

 ing synthesis of nicotinic acid, has been reported by Bonner (1950). 



It would be of great interest to know whether cases of pseudo- 

 allelism involving more than three units exist. With the rigid methods 

 available in Drosophila, no such cases have yet been found. Only one 

 group of facts is known the explanation of which I expect to find 

 through a larger series, meaning, in my way of looking at it, a 

 larger chromosomal segment of allelic action. These are the facts 

 described for blood groups in cattle (see Stormont et ah, 1951; 

 Irwin, 1951; Stormont, 1950). There are more than thirty distinct 

 antigenic factors, each behaving as a Mendelian factor toward its 

 absence. They are arranged in subgroups, some of which are known 

 in diflFerent combinations while others exist only in definite combina- 

 tions. For example, the antigenic property called K appears only 

 with B and G as the complex BGK, while B, G, or BG can exist 

 separately. The details may be described in different ways: as a 

 group of closely linked genes, as a set of multiple alleles, or as a 

 set of "subgenes" of a single locus. Finally, a non-genetic explanation 

 by multiple reactions of the same antigen is possible. There is a 

 strong suspicion I think — Irwin has not overlooked this — that pseudo- 

 alleHsm in Lewis' terminology, with a considerable number of separa- 

 ble effects, is involved. A proof of the type of the Drosophila cases 

 is hardly possible; the main obstacles are the numbers needed to 

 demonstrate crossing over, and the peculiarity of loci producing 

 antigens, which exhibit not compound effects but separate effects for 

 each allele. 



We turn now in more detail to the interpretation of the group 

 of facts under discussion. The hypothesis that pseudoallelism — if we 

 use this now generally accepted term as a brief characterization of 

 the entire group of facts without accepting the implied meaning — 

 is the consequence of duplication of a gene without loss of allelic 

 action, provided the partners are located in different chromosomes, 

 is considered to be strengthened by cytological facts. Lewis has 

 pointed out that in three cases of two "position alleles" the salivary 

 gland chromosomes show doublets, capsules of two bands of typical 

 configuration. This is true for white-apricot (see also McKendrick 

 and Pontecorvo, 1952), Stubble-stubbloid, and Star-asteroid, which 

 seems to agree with Bridges' notion of repeats. Another case is 



