The Pangenes. 645 
Numerical changes of the pangcncs are therefore the 
basis of fluctuating variability. Changes in the position 
of the pangenc in the nucleus lead to retrogressive and 
degressive mutations, whilst to account for progressive 
mutation we must assume the formation of new types 
of pangenes. 
The facts which have been described in this work, 
conform to this hypothesis so closely that they may be 
regarded as proofs of the truth of the principle. It 
seems desirable to deal further with this point without, 
however, elaborating subsidiary hypotheses. 
MENDEI/S discovery goes far to support the theory 
of independent bearers of hereditary characters. Their 
independence stands out more clearly in this case than in 
any other, with the exception of the process of mutation 
itself. In my first preliminary note on this subject I 
have pointed out the great importance of his laws in their 
bearing on the doctrine of pangenesis; and since that 
time CORRENS, BATESON, CUENOT and many other authors 
have more or less subscribed to this view. CUENOT, who, 
with BATESON, was the first to demonstrate the applica- 
bility of MENDEI/S laws to the animal kingdom, calls the 
units which are concerned in these crosses "particules 
representatives."^ Whether these factors themselves are 
the pangenes of the nuclear threads, or whether these 
factors are composed of groups of similar units, is a 
highly important question which, however, can only be 
decided by means of future experiments. 
For, as BATESON points nut, it is still possible that 
the constant Mendelian hybrid races are not absolutely 
pure with regard to their individual characters : that is 
1 L. CUENOT, La loi de Mendel ct Vhcrcditc dc la pigmentation 
cliez les sonris, Arch. zool. experim. et generale, 1902, No. 2. 
