Variation and Heredity a7 
but at least one new whole vertebra was added at a time; 
and we know several cases in which the number of vertebrze 
in the neck has suddenly been increased by the addition 
of one more than normal, and the new vertebra is perfectly 
formed from the first. 
In cases of this sort we can easily understand that the 
inheritance must be either of one kind or the other, since 
intermediate conditions are impossible, when it comes to 
the question of one or not one; but if one individual had 
one and another six vertebrze, then it would be theoretically 
possible for the hybrid to have three. 
This brings us to a question that should have been spoken 
of before in regard to the inheritance of discontinuous varia- 
tion. It sometimes occurs that a variation, which appears 
in other respects to be discontinuous, is inherited in a blended 
form. Thus the two kinds of variation may not always be 
so sharply separated as one might be led to believe. There 
may be two different kinds of discontinuous variation in re- 
spect to inheritance, or there may be variations that are only 
to a greater or a less extent inherited discontinuously ; and 
it seems not improbable that both kinds occur. 
This diversion may not appear to have brought us any 
nearer to the solution of the difficulty that Darwin’s state- 
ment has emphasized, except in so far as it may show that 
the lines are not so sharply drawn as may have seemed to be 
the case. The solution of the difficulty is, I believe, as 
follows : — 
The discontinuity referred to by Darwin relates to cases in 
which only a single step (or mutation) has been taken, and tt 
is a question of inheritance of one or not one. If, however, six 
successive steps should be taken in the same direction, then 
when such a form ts crossed with the original form, the hybrid 
may inherit only three of the steps and stand exactly midway 
between the parent forms; or it may inherit four, or five, or 
three, or two steps and stand correspondingly nearer to the one 
