On Mendel's Laws. 
20 
the contrary that it does not, we are supposing that it is completely 
indifferent whether the ancestry beyond the immediate parents were 
normal healthy individuals or raving lunatics. For the working out 
of an actual case the reader may be referred to Mr. Galton’s paper 
on the inheritance of colour in Basset-hounds. 
It may be as well to point out that the law of ancestral heredity 
by no means implies that if the characters of all the ancestry are 
known, the estimate of the character of the offspring becomes an 
exact determination, That can never be the case, for the offspring 
of one individual, or of one pair, exhibit a range of variation as a 
rule only fractionally less than that of the race, although they all 
have the same ancestry. In point of fact the accuracy of estimate 
will sensibly, though not rigidly speaking, reach a limit after rela¬ 
tively few of the ancestry (say to the sixth generation or so) are 
known. Further it should be noted that the law of ancestral 
heredity is quite distinct from the alleged law of the prepotency of 
the phylogenetically older character. Apart from the fact that the 
latter is usually applied to cases of the crossing of distinct races 
with which the former has nothing to do, and to cases where the 
relative ages of the characters are measurable by geological time, 
while the former gives sensible weights only to the recent ancestry, 
the law of ancestral heredity is not a law of prepotency or 
dominance at all. The whole of a man’s ancestry on the female 
side might be sane ; only the last six generations on the male side 
insane. The law does not state that the offspring will probably 
be sane, sanity being the phylogenetically older character. Such 
a statement would be absurd. All that it affirms is that the more 
of the ancestry are insane, the greater is the chance of insanity 
appearing in the offspring—a very different thing. I can see no 
justification whatever for the confusion of the two laws by Mr. 
Bateson. Professor Weldon, in the Biometrika article referred to 
(January, 1902), states the law of ancestral heredity in these terms, 
“The degree to which a parental character affects offspring depends 
not only upon its development in the individual parent, but on its 
degree of development in the ancestors of that parent.” Apart 
from Professor Weldon’s use of the word “affects” which to some 
extent implies a direct physical influence, and for which I would 
prefer to substitute some such phrase as “indicates” or “serves as 
a basis for estimating the character of,” this law is, beyond question, 
as we have seen, of very general application. Yet Mr. Bateson 
comments on it in the following terms : “ Having rehearsed this pro- 
•CO 
