A. D. Darbishire. 
274 
heredity the solution of which would be beyond our reach and no 
conceivable mating the exact result of which could not be predicted. 
There are very few, 1 imagine, who are so sanguine as to make 
this high claim for Mendelian principles. At the other extreme, 
however, there are those who assert that the Mendelian pheno¬ 
mena are of an anomalous nature and that the Mendelian principles 
are of very limited application. It shall be our endeavour to 
determine approximately where between these two extremes the 
truth lies. And perhaps the most convenient way in which we may 
tackle this question will be to trace the changes which the con¬ 
ception of the allelomorphic pair has undergone since Mendel’s 
papers were discovered. The earliest idea was that two 
characters were associated in a pair, of which the one happened to 
be the dominant and the other recessive; that the two characters 
were of equal value or weight—were homodynamous, if we may 
so express it; and that, beyond the fact that one happened to be 
dominant and the other recessive, and one—for instance—yellow 
and the other green, there was no intrinsic difference between 
them of such a kind as to give any a priori indication which could 
tell us, before they were crossed, which of them would be dominant 
and which recessive. The matter of dominance could only be 
determined a posteriori by the result of mating. 
This conception has now given place to a theory of the allelo¬ 
morphic pair, which seems to be in much closer accord with the 
true nature of the difference between the two characters which 
compose such a pair, but does not afford so sure an indication of 
the result of the meeting of the two characters in conjugation, as 
its enunciation would at first lead one to expect. It is however 
undoubtedly an advance. According to it the allelomorphic pair is 
not composed of a pair of homodynamous characters ; but consists 
of the presence of a single character and the absence of it. The 
presence is dominant and the absence is recessive. In this con¬ 
ception of the allelomorphic pair there is an intrinsic difference 
between the two component units of a profoundly important kind ; 
and with this difference the question of dominance is intimately 
associated. This theory may and perhaps even ought to, appear, to 
the reader not already acquainted with it, to have far exceeded the 
limits of legitimate hypothesis spinning. We shall therefore 
proceed to the consideration of an experiment of the kind to which 
the origin of the new conception was due and which, to a large 
extent, proves its correctness. And for this purpose one of the 
