174 University of Calif oriiia Puhlications in Botany [Vol. 5 



actually taken place. The point brought up in this connection is, 

 to the writer's mind at lea.st, of somewhat broader significance 

 than the mere fact that certain experimental results in Fj may 

 or may not be susceptible of a particular interpretation. We are 

 involving, in jNIendelian interpretations of variation which are 

 apparently continuous, the old question of what is a "unit-char- 

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



In 1909 when this breeding experiment with flower size 

 varieties of N. acuminata was suggested to me by Professor 

 Setchell, the facts of size and form inheritance had not received 

 the amount of attention which has more recently been accorded 

 them. For this reason, perhaps, the crossing of a large flowered 

 form with a small flowered form was taken in good faith as likely 

 to involve a single Mendelian pair of characters and a simple 

 mono-hybrid ratio was anticipated for the results of measurements 

 of corolla diameter of flowers on F^ individuals. When one bears 

 in mind that the parental varieties are in general indistinguishable 

 one from the other except on the basis of this corolla diameter 

 character and that the greatest variation in corolla diameter of 

 their flowers has not exceeded 6 mm. during three consecutive 

 years (on the basis of over 3000 measurements) it is not strange 

 that orthodox ' ' unit-character ' ' behavior was looked upon as the 

 basis for the Mendelian ratios which were vaguely anticipated. 

 Nothing, to the writer's mind has been — and still is in some 

 quarters — so firmly fixed in the "genetical" mind as the "Law 

 of Dominance," the "Law of Segregation" and the "unit char- 

 acter" conception. The "Law of Dominance" never was a law 

 and is "no inseparable attribute of Mendelian inheritance" 

 (cf. Bateson, 1909, p. 13 and p. 53). This seems to be generally 

 recognized at present, so thoroughly recognized indeed that the 

 pos.sible deeper and broader significance of "dominance" in Fj in 

 general is in danger of being almost entirely overlooked. With 

 reference to the "law of segregation" a leading student of the 

 problems of heredity has recently said "there can be no reason- 

 able doubt that ^Mendel's law is of fundamental importance in 

 genetics" (Castle, 1912, p. 352). And the most important con- 

 tribution of ^Mendel's discovery to the study of genetics is the 

 assumption that "segregation of potential characters in the germ 



