On Mendel's LazuS. 
is not easy to see in what way. As he states elsewhere ( Report , p. 
150) “ Mendel’s discovery_applies only to the manner of trans¬ 
mission of a character already existing. It makes no suggestion 
as to the manner in which such a character came into existence,” 
and the question of the origin of varieties stands, therefore, pre¬ 
cisely where it did. Mendel’s work has only ratified the conclusion 
of the biometric school, that, beyond question, large variations are 
rare, for as we have already pointed out, Mendel’s Laws cannot 
hold in any case where J's give rise ton’s during the course of the 
experiments, and Mendel himself states that no such large variations 
were remarked in thirty-three varieties of peas during ten years. I 
do not wish to imply that such variations do not occur at all—on the 
contrary I think that, e.g., the chemical properties of the soma 
necessarily imply discontinuity of variation in some respects, a dis¬ 
continuous origin of colour varieties such as those of the Iceland 
Poppy (Papaver nudicaule) being highly probable—but all the 
evidence we have goes to indicate that large and sudden variations 
are most exceptional, the highly divergent individuals forming a 
vanishingly small proportion of any race. 
Mr. Bateson reaches his conclusion as to the importance of 
Mendelian phenomena for the theory of discontinous variation only, 
I fancy, by tacitly assuming that the germinal determinant of a 
character is a structure—or whatever one may term it—incapable 
of small variations, and afi fixed and stable as the ideal molecule of 
a chemical compound; this I take to be the meaning of his reference 
“ to physical and chemical laws,” as rendering continuous variation 
in the germ cell improbable (p. 22), in the sentence already quoted 
from Mendel’s Principles But this assumption is in no way justified 
by the results of Mendel’s work; all that is predicated by his 
hypothesis is the existence of some sort of separable determinant 
for each character for which the law of segregation holds; con¬ 
cerning the variability of that determinant nothing is postulated 
but the fact that it shall not he so large as to render the boundary 
between the two races obscure. Surely the very fact that the germ- 
plasm gives rise to a long and complex ontogeny indicates that its 
molecules differ in some way from the simple molecules of water, 
salt, or sulphuric acid ? A dozen or two of atoms may be sus¬ 
ceptible of only a few stable groupings, but can the same assumption 
be made as regards a molecule built up of some thousands ? 
Richter’s Lcxikon dcr orgmiischcn Vcrbindungcn gives for instance as 
the formula for Haemocyanin C 807 H 1303 0., 58 N 223 S 4 Cu 
