176 University of California Publications in Botany [Vol. 5 



less to meeting some such list of technical requirements as is given 

 by East and Hayes in a recent paper (1911, p. 6). Perhaps all 

 of the tirst six of the desiderata there stated may have influenced 

 Mendel's choice but undoubtedly the first was most prominently 

 in his mind — i.e., "the genus or species under investigation 

 should be variable. There should be a goodly list of types which 

 are differentiated by definite characters easy of determination" 

 {loc. cit., p. 6). The genus Pisum was variable and supplied 

 a goodly list of types which were differentiated to Mendel's mind 

 by very definite characters. These definite characters later in- 

 vestigators have chosen to speak of, often rather loosely, as "unit- 

 characters" (cf. Darbishire, 1911, p. 131). 



Such definite characters which seem to be "unit-characters" 

 and to be transmitted as "independent units in inheritance" have 

 many times been dealt with in the past and will continue to be 

 dealt with in the future. For when "definite characters are suf- 

 ficiently constant to be expressed by a fixed standard . . . the 

 whole heredity short-hand is . . . simple" (East, 1912, p. 648). 

 On the other hand there are quantitative characters which are so 

 inherited that their mode of inheritance can only be explained in 

 accordance with the "]\Iendelian notation" by the assumption 

 that within the character a "multiplicity of factors exist, each 

 independently inherited and capable of adding to the character 

 in question." The "philosophical query as to whether the char- 

 acters of the organic complex of which living organisms are com- 

 posed can in any sense be dissected and analyzed into the units of 

 heredity which are the basis of Mendelian inheritance" seems 

 out of order at this point. 



The fact remains that some real and definite distinction has 

 been made in the past and will be made in the future, so far as 

 an actual hybridization experiment is concerned, between char- 

 acters which are sufficiently definite to be represented in the germ 

 cells by hypothetically fixed determiners or genes and charac- 

 ters, mostly quantitative, which are heritable potentialities only 

 and are represented in the germ cells by independent and inter- 

 changeable units functioning to produce addition, and thus seem- 

 ingly subtraction, phases of the original character. To the 

 writer's mind this distinction has been brought to light, and will 



