SHORTER ARTICLES AND DISCUSSION 



SIMPLICITY VERSUS ADEQUACY IN MENDELIAN 

 FORMULAE 



In this journal for March, 1913, Professor William E. Castle 

 discusses and criticizes in a friendly spirit certain suggestions 

 concerning Mendelian nomenclature that I brought forward in 

 the January number of the same journal. There are so many 

 essential points on which we agree and so few on which we dis- 

 agree that I should like to make clear the necessity of having for 

 our work on Drosophila a dual set of symbols. Castle finds, on 

 the other hand, that for mice and for guinea-pigs a single set 

 of letters, dbc, suffices to make clear his results and to cover his 

 theoretical ideas. 



There are three reasons why in certain cases it seems necessary 

 to use more than a single system of lettering for factors. 



1. Castle's scheme gives us no way of adequately representing 

 heterozygous forms. In dealing with such combinations it is an 

 essential both to the author and to the reader to have the hetero- 

 zygote represented with its constituent allelomorphs. Instead of 



helpful. 



2. We are dealing in Drosophila with about one hundred 

 mutations, of which forty-five have been sufficiently studied to 

 show that they fall into three groups. Within these groups the 

 factors concerned show linkage to each other, but no factor of 

 one group shows linkage with any factor of any other group. 

 Linkage means some sort of relation which we interpret in terms 

 of a linear series. We further interpret this series in terms of 

 chromosomes, but even if the series is taken merely as an abstract 

 principle the need of a dual system of letters to express the 

 order of the factors in a paired linear series is imperative, 

 so that we may represent interchanges between the pairs. 

 To take the sex-linked group of factors, for example. In a 

 heterozygous female there are two linear series present, corre- 

 sponding to her duplex condition, or, as we think, to the two 

 homologous sex chromosomes. Any factor in the one series has 

 a correlative factor in the other series (in the other chromosome) 

 in a corresponding position, and in order to treat the linkage of 

 the factors we must have some method of representing and of 

 distinguishing them. If from the mother the factors aBcdE 

 enter the combination and from the father AbCDe, the hetero- 

 zygous female is represented by the two groups: 



