THE MENDELIAN CLUE 233 
represented by differences in the ultra-microscopic 
architecture. It is quite likely that several factors 
may be concerned in one character, or that one factor 
may influence more than one character. 2. The 
second idea in Mendelism is that of dominance. 
When Mendel crossed a pure-bred tall pea with a 
pure-bred dwarf pea, the offspring were all tall; and 
he called the quality of tallness dominant to the 
recessive quality of dwarfness, which the hybrid 
offspring keep, as it were, up their sleeve. The 
dwarfness is not expressed, but it is certainly in the 
inheritance, for it reappears in a quarter of the 
progeny of the hybrid generation, if these are inbred 
or allowed to self-fertilize. If a Japanese waltzing 
mouse is crossed with a normal mouse, all the hybrid 
progeny are normal, the waltzing peculiarity being 
recessive to normality. If the hybrids be inbred, 
some of their progeny are waltzers in the average 
proportion of a quarter—and these waltzers might 
be sold as pure waltzers—although both their parents 
and one of their grandparents were normal. Simi- 
larly, about a third of the rest of the progeny are 
purely normal, while two-thirds are like the first 
generation of hybrids—to all appearance normal, but 
with the waltzing character up their sleeve. It often 
happens that the two parents differ, not in presenting 
a pair (or more) of contrasted or alternative char- 
acters, but in the one having certain unit characters 
which the other has not. This works out in the 
same way—the unit character that is present being, 
as it were, dominant to its own absence, In 
