288 Shull. 



literature from that which is already published, can produce the desired 

 simplification; but there are certain principles of conduct in the choice 

 of symbols which writers might elect to follow in the future, which 

 would avoid unnecessary addition to the confusion. 



(1) The most important of these principles is that of conformity 

 to those elements of past and present usage which have become so general 

 as to be fairhj considered "standard" . There appear to me to be two 

 such standard elements: namely, (a) The use of capital letters for the 

 assumed presence of determiners, and the corresponding small letters 

 for the assumed absence of those determiners; and (b) such a choice of 

 symbols that the specific reactions for which the designated genes are 

 believed to be responsible, shall be definitely suggested by the symbols, 

 thus assisting the memory and rendering less necessary the constant 

 reference to keys. The most feasible method of accomplishing this ob- 

 ject, and that wliich is so generally used as to be most nearly "stan- 

 dard", is the adoption of the initial letter, or of some other prominent 

 letter in the name of the reaction which the given gene produces, — 

 the name of the character being in that one of the three scientifically 

 international languages, English, French, or German, which is being 

 used at the time by the writer who proposes the symbol. Lang's sug- 

 gestion that the initial of the Greek name of the given character should 

 be used, does not appeal to me except in certain special cases, because 

 the Greek (or Latin) names are not as familiar to most geneticists as 

 the corresponding words in the three languages, English, German and 

 French, in which nearly all of the genetic literature is being published. 

 (2) A second princi])le whose general adoption would avoid much 

 confusion is the principle of priority. This should not be applied in 

 the rigid manner of the taxonomist, but whenever a writer is dealing 

 with genes which have been investigated and which already have pub- 

 lished symbols, he should use those symbols in preference to new ones, 

 unless, as will sometimes happen, a situation arises wliich makes the 

 use of certain prior symbols impossible in which case these unavailable 

 symbols, and these only, should be replaced by new ones. If a number 

 of formulations already occur in the literature, strict priority need not 

 be insisted upon ; but preference should be given to those symbols which 

 were used in the paper which has made the most fundamental contribu- 

 tion to the knowledge of the particular genotypic complex under con- 

 sideration, always, when other things are equal, giving the preference to 

 symbols which conform most nearly to the spirit of the preceding section. 

 If this mil' were generally adopted, regardless of the manner in which the 



