IMPORTANCE OF HYBRIDISATION IN THE STUDY OF DESCENT. 279 
The question may here be raised in passing, whether the Mendelian 
“valency ’”’ of a character—I mean its condition as dominant or recessive 
—justifies a conclusion as to its phylogenetic age. With regard to the 
common inclination to regard the dominant character as the older, I have 
frequently insisted on the fact * that, though such a deduction is generally 
permissible, it is by no means without exception. Not only can the 
valency change with the constitution of the families, but phylogenetically 
new characters can also be dominant; for example, the beardlessness f of 
barley obtained by Rimpau as a novelty on crossing may dominate over the 
possession of beards or hoods; or petalody of the sepals may dominate 
over the normal flower-structure in Primula and Campanula. 
Already hybridisation derives a much greater importance as a means 
of forming new varieties from the possibility of a production of new 
Mendelian combinations of characters and component characters. First 
of all, from two parents which differ from one another in many ways a 
number of intermediate types may be sometimes produced: .a fact which 
is most- important for the practical breeder as well as for the student of 
evolution. Hybrid middle forms have apparently originated in great 
numbers from unintentional crossings, not to speak of those that have 
been artificially produced. But also of those species which in a wild 
state and under similar external conditions exhibit a number of constant 
varieties, very many may owe their variety of form, in part at least, to 
an original cross between distantly related varieties and to Mendelian 
segregation. 
In strictly Mendelian cases a series of intermediate forms resulting 
from a cross can be shown to constitute a discontinuous series (as 
opposed to a series of individuals resulting from spontaneous and con- 
tinuous variation) so soon as the attention is fixed upon single characters 
and not upon the collective impressions made by the whole. At all 
events in certain crosses which do not exactly follow the Mendelian 
scheme, but approach the Zea type of Correns, segregation results in 
the production of a whole series for continuous intermediate forms 
which carry the characters of both parents combined in yarious pro- 
portions. 
I have, for example, obtained such transitional series by the crossing 
of bearded and hooded strains with unbearded barley, and by crosssng 
barley with various numbers of rows, especially in regard to the degree of 
fertility of the lateral spikelets ; also with rye and wheat in regard to the 
type of the spikelet and shape of the seed. 
Anyhow the intermediate forms all appear to split further, but not all 
in the same way. It seems as if among the products of an impure or 
graduated segregation, fresh groups with various modes of transmission 
may arise in certain cases; and in every group it appears that greatly 
developed characters gain in their power of transmission. The hybrids 
with but a slight development of hoods or beards only produce few 
descendants with fully developed hoods or beards respectively. 
I have hitherto only spoken of continuous transition series, as they 
* Zeitschrift fiir das landwirthschaftliche Versuchswesen in Oesterreich, 1901, 
p- 1037; Beihefte zwm Bot. Centralblatt, Band xvi., Heft 1, 1903, pp. 16, 17. 
+ United with this is the cryptomeric possession of beard (in barley). 
