Ernest Warren 
323 
axis, which matures before the ordinary zygomorphic flowers. This variety is 
known as monstrosa. 
When a peloric plant is crossed with a non-peloric plant the offspring are 
generally either completely peloric or completely non-peloric. There is no marked 
tendency for the ofJspring to be intermediate in character. Notwithstanding 
this fact the character is capable of dilution, since in the case of one family arising 
from such a cross nearly all of the offspring were more or less intermediate in nature. 
Certain of the internodes were much reduced, but not entirely suppressed, and the 
flowers involved arose in the form of petaloid blades which matured before the 
lower normal flowers, as in the case of a typical crowned plant. 
The gametic nature of the parents, as to whether they were piire or impure in 
the Mendelian sense with reference to the peloric character, could be judged either 
by raising families obtained by self-fertilization or by their action in the various 
crosses that were made. Families derived from self-fertilization from four parents 
were successfully raised, and in the case of the remaining parents the marked 
consistency in their behaviour in the dift'erent crosses indicated clearly as to whether 
they were homozygous or heterozygous. 
The non-peloric condition was dominant. 
On dividing the offspring into peloric plants (including those which showed 
any tendency towards a peloric condition) and non-peloric plants, the numbers 
were 83 peloric and 104 non-peloric. Calculating the Mendelian expectation 
for each family and adding the results the figures obtained were identically the 
same. 
(2) The results of reciprocal crosses appeared to be the same; that is the 
families obtained from ? ^ x c? -B did not differ in any constant manner from those 
derived from ^ B x A. 
(3) The results obtained by using pollen from the crown flower, the ordinary 
zygomorphic flower and of flowers with abnormally split open corolla did not appear 
to differ from one another in the least. 
(4) The colour of the seeds is very definitely related to the colour of the flowers 
of the seed-plant. Plants with white flowers had very pale buff-coloured seeds, 
those with light purple corollas had light brown seeds, Avhile those with dark purple 
corollas had very dark brown seeds. As might be expected, a plant with white 
flowers crossed with pollen from a plant with dark purple flowers produced pale 
coloured seeds, since the coloured testa of the seed is formed by the seed-plant and is 
not a result of fertilization. 
This observation emphasizes the fact that great caution is necessary in 
regarding characters as being independent of one another. It is often stated that 
an organism may be pure bred in one character and not in another ; but probably 
this is true only in a very limited sense. De Vries holds that a change in any one 
character involves some change in the constitution of the organism as a whole. 
