232 G. Udny Yule. 
general still. In any case then where it is only possible to deal 
with attributes , and not measurably variable characters, it seems 
impossible to disprove the existence of segregation; it may 
occur (as Mr. Bateson seems to consider probable) or may not. 
Nor have we exhausted the ways in which segregation may be 
masked, apart from any question of its partial failure such as 
might be invoked to account for some of the divergent results 
obtained by Mr. Bateson and others. In supposing dominance to 
fail, we have still assumed the inheritance to be exclusive ; this is a 
logical necessity if predetermination hold, but, if predetermination 
fail, as well as dominance, the inheritance may become blended, i.e. 
the heterozygote may produce, on the average, a per-centage of 
A's characteristic neither of the ZTcells nor the 6-cells (in our 
previous notation), but intermediate between the two ; such blending 
would give rise to yet another series of forms of ancestral heredity. 
So far, however, we have dealt solely with cases of inheritance 
of attributes, without considering the individual variations to 
which the attribute may be subject within the race; but the relation 
of these individual variations and their laws of inheritance to the 
phenomena considered by Mendel is obviously a question of first- 
class importance. They present two features not directly treated 
by Mendel at all, but noticed at some length by Mr. Bateson—a 
sensible continuity of variation in the first place, and, so far as our 
experience goes, blended inheritance in the second, the offspring of 
two widely different parental forms shewing no tendency to revert 
to such forms, but resembling the offspring- of an intermediate type. 
There can be no doubt, of course, as to the existence of such 
individual variations in many of the characters dealt with by 
Mendel and his followers—length of stem in peas may be cited as 
an instance where the variations are not merely conspicuous, and, 
one would imagine, susceptible of easy quantitative measurement— 
but they are necessarily neglected by hybridisers, who unfortunately 
rely on their unaided judgments (no sarcasm is intended). The 
attitude of Mr. Bateson towards these individual variations, and 
his views on the bearings of Mendelian phenomena on the con¬ 
ceptions of variation in general are rather difficult to follow. He 
believes, as I gather, that the origin of races or of species is due 
solely to large and marked variations, and that small variations 
are of no importance, and speaks of Mendel's discovery as “that 
discovery which, once and for all, ratifies and consolidates the con¬ 
ception of discontinous variation” (Mendel’s Principles, p. 116). It 
