3 2 ° 
A. D. Darbishire. 
consistent with the results both in F, and P 2 ; but we did not 
commit ourselves to the conclusion that this was a demonstration 
of the correctness of this theory. It will, however, be seen that the 
theory readily admits of being tested; what I mean is this:—the 
existence of 4 whites in every 16 in F 2 is consistent with the theory 
put forward to account for this case, but we should not rest 
satisfied with it until we have obtained more evidence that these 
two types of whites really do exist within the four. All that has 
to be done is to cross the whites which appear in this generation 
with pure greys. Those carrying P, that is to say 3 in every 4, 
should give purple spots in F x , whilst those of the formula gp 
should give simple dominance of grey. 1 
This experiment is not a difficult one to carry out, because 
there is a correspondence between the colour of the flower of the 
plant and the characters of the seed-coat which will be manifested 
when the fruit is ripe. Plants which will produce GP and Gp 
seed-coats always have purple flowers, whilst plants which will 
produce white seed-coats, whether of constitution gP or gp always 
have white flowers; so that it is possible to know whether the plant 
will have white seed-coats or not, by the colour of its flowers ; 
although this does not enable one to determine whether, if 
coloured, they will be GP of Gp. 
I have myself made, this year (1909) a number of crosses 
between these extracted F 2 whites and certain pure and extracted 
grey types. The only fact which affords some kind of clue as to 
what the result will be, suggests that it will be in a direction 
contrary to that expected by the theory we have considered. I 
refer to the fact that hitherto, so far as I have been able to 
discover, all crosses between greys and whites have given F t with 
purple spots, which means that all of the whites used must have 
been of the formula gP. That is to say, supposing the whites 
employed in such crosses may be considered as fairly represen¬ 
tative of white-coated Peas in general, none of the white-coated 
varieties on the market are pure white, i.e., of formula gp ; and the 
difficulty of imagining any a priori reason why they should not be 
so, suggests that the Mendelian hypothesis put forward to explain 
this case may not closely represent what actually occurs. This, 
however, is only a very faint indication, and will, of course, count 
for nothing as against the result of the experiment which I said I 
am conducting. I have dealt somewhat fully with this case for a 
number of reasons. In the first place it is an instance of the 
