IN THIS TROPICS. 
339 
the ease of schizogonous characters upon the mode of origin 
of the particular allelomorph—whether arising in a homo¬ 
zygote or in a heterozygote. 
In certain cases there is greater or less correlation between 
different sets of allelomorphs. In such cases, as Bateson 
and Saunders point out, Correns’ terms “ halb-identisch” 
and “ conjugirt ” are already well enough expressed by the 
terms “ in perfect ” and “ in imperfect correlation.’* 
VI.—LATENCY OF CHARACTERS. 
I have already had occasion to account for the absence of 
a character—the presence of purple spots (on the testa in 
peas)—in certain individuals in which it should have 
appeared according to the simplest form of Mendel’s theorv 
for dihybrids. The explanation tentatively put forward was 
that this character was unable to manifest itself except in the 
presence of another character, or rather group of correlated 
characters including the greenish grey colour of the testa 
and the dimpled form of the seeds. This assumption having 
been made, it was shown that the example was in perfect 
accordance with Mendel’s Law. 
There is evidence, however, that this explanation may cover 
a considerable number of supposed exceptions* Thus Tscher- 
mak has described (63) a series of cases which show con¬ 
siderable resemblance to the above, but in which a new or 
atavistic form appears in Fj. Before describing Tschermak’s 
cases an example of this kind of reversion may be given. 
Example VIII .—I crossed Barr’s dwarf kidney bean “ Green 
Harieot” with pollen from “ Canadian Wonder.” The former 
plant has a pale green testa and pale greenish-white flowers, 
and the pods are green. 
Canadian Wonder has deep crimson seeds and pale lilac 
flowers, and the pods are also green in colour. 
The hybrid plants had perfectly black seed coats. The 
flowers were of a fairly deep crimson eolour, like that of the 
