On Mendel's Laws . 
-o j 
first generation of offspring will exhibit X as the dominant character; 
but if x l and x' 2 or x 2 and.r'j. are the dominant forms then the first 
generation may exhibit a character that is a “ blend ” or intermediate 
between the two parent races, or a character greater or less than 
either. In the second generation again the parental forms will not 
he the only ones to appear; the new forms x ± -f x \ and x\ + x 2 
will also present themselves. Two doubly-compound characters 
will therefore give rise after crossing to four somatic forms ; two 
triply-compound characters to eight ; two characters compounded of 
n such units to 2" forms. But how great is n likely to be in such a 
case as stature, assuming that it can he analysed into a set of Mendelian 
units ? As Mr. Galton has remarked (( 2 ) p. 83) “ . .. human stature 
is not a simple element, but a sum of the accumulated lengths or 
thicknesses of more than a hundred bodily parts, each so distinct 
from the rest as to have earned a name by which it can he specified. 
The list includes about fifty separate bones, situated in the skull, the 
spine, the pelvis, the two legs, and in the two ankles and feet.” 
Surely it would he a very moderate estimate that the number of 
units could not be less than 50 ? Yet this would suffice to give, 
on the simplest Mendelian assumption that each unit can only 
exhibit two types, not some mere ten thousand different values of 
stature, the run of which would be quite indistinguishable from 
strictly continuous variation, but over a thousand-million million 
different types ! Even then if the variations of “ units ” do take 
place by discrete steps only (which is unproven), discontinuous 
variation must merge insensibly into continuous variation simply 
owing to the compound nature of the majority of characters with 
which one deals. There does not seem any escape from this 
conclusion. Continuous variation, in the present state of our 
knowledge, we can only say may be due to continuous variation of 
the elements of the germ cell (determinants or what not), or may be 
due to the compounding in some way of the discontinuous variations 
of a number of such elements. 
Precisely similar considerations hold good for the case of 
blending. It is quite possible that characters behaving in other 
respects as Mendel’s Laws would lead one to expect i.c., “ unit 
characters,” may in some cases give a blended form for the individuals 
developed from the heterozygote. But in any case compound 
characters must blend—more or less. This is obvious if dominance 
be absent; two pure forms x L -f * 2 and x' x -f x' 2 would then 
produce with equal frequency offspring of the somatic types at x -f * 
