On the mode of Inheritance of certain Chaéacters etc. 309 
which it meets the facts. But not only is this suggestion of “crossing 
over” unnecessary, it is contrary to the evidence. For if a sulphur- 
white produces occasionally functional pollen grains carrying whiteness, 
then when this form is pure-bred we should expect to find that a 
certain number of the white singles would dreed true to whiteness. 
Further that when used as the male parent in matings with a pure- 
breeding cream that a small proportion of the F, crossbreds would 
be white and not cream, since white is dominant to cream. And 
lastly, that when used as the male parent in matings with a form 
having white plastids that a small proportion of the F, crossbreds 
would give ox/y plants with white plastids in Fy. No such exceptional 
cases have ever been observed. Yet, on this hypothesis, they should 
not be so very rare. The sulphur-white strains appear to produce 
about 3% of double whites. But since on GOLDSCHMIDT’S view a 
sulphur-white produces the two kinds of ovules (single white and 
double cream) in equal numbers, and as presumably equal numbers 
of each kind of ovule will be fertilised by these pollen grains carrying 
the factor W, we must suppose that the number of these W pollen 
grains will be about twice as great as the number of double-whites 
recorded, i. e. about 6%. Let us now consider the evidence on this 
point (?). 
Of 82 sulphur-whites which were self-fertilised ove were found 
to breed true. (62 of these families are recorded in my paper in 
the Journ. of Genetics, Table III, p. 372, the rest have been recorded 
since.) 
Of 117 F, plants reaised from a mating with a pure cream as 9 
all were cream. 
Of 7 F, plants derived from a mating with a non-double-throwing 
form with white plastids as 9 a// gave some plants with cream plastids 
in Fy. 
From a mating with a non-double-throwing form with white 
plastids as © 42 F, families were raised. Of these 8 which were very 
small and did not include more than one double may be left ont of 
account. For where double creams and double whites both oceur it 
will naturally happen that when only one is obtained it will sometimes 
(1) It had been hoped that during the present season corroberative evidence 
would have been obtained from the ovules affording direct proof of the production 
of a small number of female gametes carrying the combination whiteness and doubleness. 
Unfortunately some of the commercial samples employed turned out to be impure so 
that the experiment had to be repeated and the results are not yet to hand. 
