210 Observation of the Origin of Varieties. 
results obtained in the years 1886-1893. They com- 
prise four generations, each of which produced only 
one or two peloric flowers amongst thousands of normal 
ones. The anomaly, therefore, seems to recur every 
year and is obviously due to the existence of some her- 
itable semi-latent potentiality which only very seldom 
becomes active. 
This result of the experiment supports the conclu- 
sion based on the repeated occurrence of isolated peloric 
flowers in nature. Linaria vulgaris hemipeloria is thereby 
shown to be a heritable form. The question whether 
it is identical with Linaria vulgaris itself, or constitutes 
a variety or a race of this, cannot be answered for the 
present. From it my L. vulgaris peloria arose, as I shall 
now show. 
In order to make this part of my experiment more 
easily intelligible I shall first describe it in the form of a 
pedigree. This contains the four generations already 
dealt with, and two further ones of the Hemipeloria 
(1-6), together with the first, second, and third genera- 
tions of wholly peloric plants (I-III). The meanings 
of the abbreviations are : 
h and H: Linaria vulgaris hemipeloria. 
p: peloria, 1st generation. 
P: " " 2d and 3d generations. 
Wherever necessary the number of plants is prefixed 
to these letters either in absolute numbers or in percent- 
ages. For the fifth and sixth generation I have, as will 
be seen, made repeated sowings in various years. The 
sign ( 2 ) means that the examples in question were the 
same as in the previous year, and bore seed a second 
time. Finally I have denoted by H the two plants of 
1893 which in their second year produced the seed from 
