Theory of Plant Breeding 
sults were as follows: 1141A, 2267Aa, 1140a. A anda 
being the dominant and recessive characters employed. 
Perhaps the most interesting of the recent papers on 
this subject is another by the same author.¥ Of 44 
pairs of Mendelian characters, he shows that 41 (no 
less than 93 per cent.) may be explained on the theory 
that the dominant character is presence and the recessive 
is absence of some particular unit character which is in 
most cases a colouring substance. 
Suppose one represents the first generation of hybrids 
by two buckets with a number of yellow and green 
counters (half yellow and half green in each). We 
may then make up the second generation by taking 
one counter from each bucket at random. By the 
ordinary laws of probability, the result of a very large 
number of trials would be such a percentage as 25YY, 
25YG, 25GY, and 25GG, which is 25 pure yellow, 
50 yellow-green, and 25 pure green, giving the three 
to one proportion (for yellow dominates green). 
So that in any case of this sort where two races are 
crossed which only differ in that one possesses a colour- 
ing substance, this enzyme (or whatever it may be) would 
be transferred in three cases out of four. 
Of course it is rash to suppose that the inheritance 
can be quite so simple as this. The colour is probably 
produced as a result of enzymes formed by a series of 
chemical changes. The originator of the series will be 
passed on to the germ of those that are to possess it. 
When one uses two pairs of characters, the Mendelian 
results are a little more complicated, Such proportions 
occur as 9: 3: 3: I, or 9 to 7 when one dominant 
is alone considered, 
One of these complex cases has been very beautifully 
worked out by Miss E. R. Saunders in the case of certain 
stocks. The characters used were the glabrous or 
393 
