W. F. R. Weldon 
241 
cases lead to widely different results, and one great reason why this is so becomes 
evident on a little consideration. 
Mendel treats such characters as yellowness of cotyledons and the like as if the 
condition of the character in two given parents determined its condition in all 
their subsequent offspring. Now it is well known to all breeders, and is clearly 
shown in a number of cases by Galton and Pearson, that the condition of an 
animal does not as a rule depend upon the condition of any one pair of ancestors 
alone, but in varying degrees upon the condition of all its ancestors in every past 
generation, the condition in each of the half-dozen nearest generations having a 
quite sensible effect. Mendel does not take the effect of difference of ancestry into 
account, but considers that any yellow-seeded Pea, crossed with any green-seeded 
Pea, will behave in a certain definite way, whatever the ancestry of the green and 
yellow peas may have been. [He does not say this in words, but his attempt to 
treat his results as generally true of the characters observed is unintelligible 
unless this hypothesis be assumed.] The experiments afford no evidence which 
can be held to justify this hypothesis. His observations on cotyledon colour, for 
example, are based upon 58 cross-fertilised flowers, all of which were borne upon 
ten plants ; and we are not even told whether these ten plants included individuals 
from more than two races. The many thousands of individuals raised from these 
ten plants afford an admirable illustration of the effect produced by crossing a few 
pairs of plants of known ancestry ; but while the}' show this perhaps better than 
any similar experiment, they do not afford the data necessary for a statement as 
to the behaviour of yellow-seeded peas in general, whatever their ancestry, when 
crossed with green-seeded peas of any ancestry. 
When this is remembered, the importance of the exceptions to dominance of 
yellow cotyledon-colour, or of smooth and rounded shape of seeds, observed by 
Tschermak, is much increased ; because although they form a small percentage of 
his whole result, they form a very large percentage of the I'esults obtained with 
peas of certain races. The fact that Telephone beliaved in crossing on the whole 
like a green-seeded race of exceptional dominance shows that something other than 
the mere character of the parental generation operated in this case. Thus in 
eight out of 27 seeds from the yellow Pois d'Auvergne % x Telephone ^ the 
cotyledons were yellow with green patches ; the reciprocal cross gave two green 
and one yellow-and-green seed out of the whole ten obtained ; and the cross 
Telephone % x (yellow-seeded) Buchsbaum ^ gave on one occasion two green and 
four yellow seeds. 
So the cross Couturier (orange-yellow) $ x the green-seeded Express ^ gave 
a number of seeds intermediate in colour. [It is not quite clear from Tschermak's 
paper whether all the seeds were of this colour, but certainly some of them were.] 
The green Plein le Panier $ x Couturier ^ in three crosses always gave either 
seeds of colour intermediate between green and yellow, or some yellow and some 
green seeds in the same pod. The cross reciprocal to this was not made ; but 
Express % X Couturier ^ gave 22 seeds, of which four were yellowish-green. 
Biometrika i 23 
