66 
R. Ruggles Gates . 
One plant in a large culture of CE. rubrinervis 1 appeared with 
an extreme increase in anthocyanin pigment, this red being 
especially conspicuous on the sepals, flower stalk, stem, and the 
under surface of the leaves, although histological examination shows 
it to be present in increased quantity in all parts of the plant. The 
original individual was heterozygous, originating therefore through 
the union of a mutated with a non-mutated germ cell. The 
premutation must then be considered to consist in a change, 
probably chemical, in a locus of one chromosome, and this change 
probably occurred during some stage of meiosis. Farther than 
this the matter cannot be analysed at present. The altered 
chromosome, present in the nucleus of every cell through mitosis, 
increases the capacity of every cell for producing anthocyanin, as is 
shown for instance by comparing cells of the chalaza in the two forms. 
Through an unfortunate accident, only a few offspring of the 
original novum survived. They were 9 rubricalyx, 1 rubrinervis , 
and 2 rosettes classed as “ doubtful.” The next generation gave 
three families, with ratios 10 : 5, 14:6 and 33 : 11 respectively, all 
clearly representing 3 : 1 families. This leaves no reasonable 
doubt that the original rubricalyx mutant was heterozygous for 
one factor for red. For by no possibility can any of the three 
ratios 10 : 5, 14 : 6 or 33 : 11 be considered to represent a 15 : 1 
ratio, while their total, 57 : 22, is clearly close to the 3 : 1 
expectation. Also the ratio in the previous generation (which may 
be taken as probably 11 : 1), from which these three families were 
derived by selfing certain individuals, w T hile it might by itself 
conceivably represent a 15 : 1 family, cannot do so really. Because 
the chances are remote that, if the parent mutant were 
heterozygous for two factors for red, three of its offspring 
taken at random as mother-plants of the next generation should all 
happen to be heterozygous for a single factor (therefore giving 3: 1 
ratios) and none of them homozygous or heterozygous for two 
factors.* The evidence taken as a whole is therefore clear that 
rubricalyx originated in one individual which was heterozygous for 
a single factor for red. 
1 It now appears (Lutz 1917b) that the form which other investigators, at 
least MacDougal, Miss Lutz and myself, regarded as rubrinervis, differs from 
the Amsterdam rubrinervis, which is a less robust form. Further study by 
de Vries has revealed several types closely related to rubrinervis , which will be 
referred to later. 
2 If the original mutant were heterozygous for two factors for red, it 
would give in its offspring 1 plant which bred true to red, to 8 yielding 
rubricalyx and rubrinervis in the ratio 15R : lr, to 6 yielding ratios 3R : lr, to 1 
pure rubrinervis . It gave in fgct three families all of which were 3:1. 
