IOO 
V. H. Blackman. 
The extreme case of this tendency is obviously that of a 
heterodynamic pair of characters where the one completely masks 
the other:— 
H. ( 3 ) 
A a 
Again, of the two homodynamic characters, sometimes one, 
sometimes the other, may be more strongly developed, the hybrid 
then shows many forms and the pair of characters may be termed 
poikilo dynamic :— 
.H’.H” 
A a 
Lastly, an extreme case of the fourth class is possible, in which 
sometimes the one character, sometimes the other, is in abeyance 
and all intermediate forms are wanting; pairs of characters behaving 
in this way are aptly termed by Correns dichodynamic , thus:— 
H.-...H ( 5 ) 
A A 
The last two cases are considered by Correns as special cases 
of homodynamic characters.* 
Under the first two classes are included many of the characters 
of hybrids generally, and many of the characters of Hieracium 
hybrids and of Zea Mays races. The third class is well known and 
is observed in many of the characters of Pea hybrids. Under the 
fourth class must be included many of the characters of Hieracium 
hybrids also ; as these characters are also homoiogonic there may 
arise from a single cross a number of new and constant (when self- 
fertilised) races of hybrids. The classical examples of the fifth 
( dichodynamic) class are the strawberry hybrids described by 
Millardet ( 1894 ); these are of two forms and exhibit either all the 
characters of the one parent or all those of the other, so that 
without a knowledge of their origin it would be impossible to 
tell that they were of hybrid nature. On account of this striking 
behaviour, Millardet was led to call these hybrids “faux hybrides.” 
The very interesting cases of “mosaic” formation in hybrids 
may be considered as special cases of either class 4 or 5 , in which 
*It is obviously possible to consider the third case, that of 
heterodynamic characters, as only a special case of homody¬ 
namic characters in which one is always in abeyance, the 
difference between the two main classes would then be only 
one of degree. Correns (1901 b), in fact, considers that the 
two classes are not fundamentally different, as is suggested by 
the change from one class to the other mentioned above. The 
object of the classification is mainly for convenience of 
reference. 
H”. 
H’”. 
H”’\ 
(9 
