June 9, 1923 
849 
Rogue Types in Garden Peas 
nuclear divisions would have to take place too regularly and too often 
to make the explanation seem reasonable. 
As an alternative hypothesis, the idea of ‘‘mass'* somatic mutation 
is advanced to account for the anomalous behavior of the rogues in 
heredity. Using the factor symbols here adopted the x factor of Gradus 
is assumed to mutate occasionally to the X factor of Gradus rogue. 
The factor pair xx when in a homozygous condition is relatively stable 
and mutation to X rarely takes place. However, the heterozygous Xx 
combination is very unstable and x mutates with great frequency to X, 
creating a homozygous and more stable condition of the germplasm in 
regard to the X factor. Mass somatic mutation of x to X occurs in the 
soma of the hybrid at an early stage in its development and would 
account for the prevailing absence of the x factor among the F^ gametes. 
The same phenomenon would be true of Gradus plants in which the 
primary mutation of x to X took place. Such a mutation occurring early 
in the ontogeny of a Gradus type plant would produce an individual 
which would gradually become more roguelike as it matured. Or when 
the mutation takes place late in the development of the plant, the 
form of the individual would not necessarily be changed but the issue 
derived from the cell containing the X factor, would, if it entered into 
gamete formation, cause rogues to appear in the next generation. 
The fact that occasionally an F^ hybrid of Gradus X Gradus rogue is 
Graduslike at maturity rather than roguelike is interpreted as indicat¬ 
ing that rare Gradus gametes are produced, as a consequence of mutation 
or otherwise, in which the X factor is replaced by an allelomorph com¬ 
parable in stability to the x' of Mummy. 
It is probable that the rogues are an extreme manifestation of a series 
of mutations originating by changes in the x factor of Gradus. Thus the 
various sorts of intergrading intermediates described by Bateson and 
Pellew (i) may well represent various modifications of the factor x 
which may be designated as x", x'", etc. The xx" or x" x" 
combinations produce intergrading intermediates in which mutation of 
either of x or x" to X is more frequent than in Gradus type. The 
differences in the stability of the various modified x factors as exhibited 
in the rate of change of these factors to X, accounts for the existence of 
high and low rogue-producing strains of intermediates. 
The difference in the proportion of rogues produced from the lower 
as compared with the upper nodes of intergrading intermediates, besides 
being due to a multiplication of the somatic cells in which a primary 
mutation of x to X has occurred, may conceivably be caused by an increase 
in the number of somatic mutations brought about by physiological 
changes in the protoplasm as the plant matures. 
The mutation of x to X in Gradus, in the F^ hybrid and in the inter¬ 
grading intermediate is comparable to what Emerson (5) calls a recurring 
somatic mutation. The varieties in which rogues arise are of the sort 
earlier described by de Vries (jj) as “ever-sporting,” meaning that it is 
impossible to free the stock of the tendency to produce “sports” (muta¬ 
tions). 
The results in the F2 generation of crossing Gradus type with Mummy 
and Gradus rogue with Mummy have led to the assumption of a factor 
pair Yy in addition to the xx factors as partly responsible for stipule 
shape in Gradus and Gradus rogue. In Gradus there are at least two 
factor pairs xx and yy determining the ratio of width to length of stipule. 
