THE TBOPICb. 
341 
On crossing A(B) (B being latent) with ab* we get?-^- 
^1 ABab. (B latent haviDg become B active). 
And the further account is identical in the two cases. 
Tschermak also described a number of more complicated 
cases of a similar nature, but in his preliminary paper largely 
confined himself to demonstrating the existence of crypto- 
meric characters in plants. He pointed out that extracted 
Mendelian dominants and recessives require testing for traces 
of the opposite parental character. Thus we may hope, on 
crossing with a properly chosen race, to demonstrate the 
latency of the character purple spots in some of the white 
forms in the case already described. 
Among examples of “ Kryptomerie ” in existing literature, 
Tschermak points out that a number of the spontaneous 
Oenothera mutants studied by de Vrie B were constant when 
inbred ; but on crossing showed a partial atavism. De Vries 
further upholds a very far-reaching Cryptomerism, namely, 
the inclusion of all the varieties of a species one within the 
other. 
The special importance of Tschermak’s contribution is, as 
he himself points out, the establishment of the fact that 
certain races and characters exist, which show atavism in a 
definite and regular way, following Mendel’s Law in the 
majority of cases examined. 
Cuenot (24) has already carried an investigation similar 
to this one a step further in the case of animals. This 
writer has worked out a series of cases of latency of colour 
characters in races of mice in a very ingenious fashion. 
Cu6not’s hypothesis is nearly identical with the one suggested 
above. He supposes that the colours yellow (J), grey (G), or 
black (N) can only make their appearance when the same 
zygote includes a perfectly independent colour bearing 
character (C), which is allelomorphic to and dominant over 
the albino character (A). According to Cuenot’s scheme all 
