Progressive, Retrogressive, Degressive Mutations. 577 
ferent conditions in the two parents; as for instance in 
the union of latent characters with active, of semi-latent 
with semi-active, and so forth. The opposite visible qual- 
ities, determined by the two different conditions of the 
same internal factor, constitute a so-called pair of char- 
acters. Therefore, racial forms which differ from one 
another by such qualities only, constitute a group for 
themselves, in the theory both of hybridization and of the 
origin of species. 
Uni-sexual crosses are of an entirely different nature 
and lead to the direct production of constant hybrid 
races. They occur when at least one character is present 
in one parent, but is absent from the other. Hence MAC- 
FARLANE'S name uni-sexual. In more complicated cases 
one form may have a larger excess of factors; or again, 
each of the parents may possess factors which are absent 
from the other. The excess will be unilateral, if one of 
the forms has arisen directly from the other, but bi-lateral 
if both have arisen in diverging directions from the same 
ancestor. In these cases vicarious pairs of characters do 
not exist in the two forms crossed, although externally 
thev may seem to do so. 
r j 
Therefore uni-sexual crosses correspond to progres- 
sive mutations ; the Mendelian law, however, to the retro- 
gressive and degressive forms of differentiation. Con- 
versely we may infer that characters which, when crossed, 
behave in a Mendelian way, are in a latent or semi-latent 
condition, and that in the uni-sexual crosses an internal 
factor is entirely absent on one side. I willingly admit 
that the main purpose of this discussion is to make my 
principles clear, and to show how the two great branches 
of the theory of mutation may, in spite of the vast differ- 
ence in their points of departure, ultimately lead to the 
