1907] MACDOUGAL—HYBRIDIZATION OF WILD PLANTS 47 
as final proof that it may really have come by one of the rarer com- 
binations which he has missed. On the other hand, success may 
come with the first cross and in the first generation. 
In some instances the result of a hybridization is a single type 
which offers the qualities of the parents locked in a stable combina- 
tion in the first generation and reproducing without separation in 
successive generations. It is this type of hybridization that is implied 
in the general assertions as to the hybrid origin of any plant, and it is 
a type of which we have the fewest illustrations in breeding experi- 
ments. To recur again to the genus furnishing the example previously 
given, O. cruciata varia was suspected by the author to be a combina- 
tion of O. Lamarckiana and O. cruciata, and in the synthetization test 
the good fortune was encountered of selecting the one of the three 
known elementary species of O. cruciata which had originally entered 
into the union. The egg of O. Lamarckiana and the pollen elements 
of this form entered into a stable combination which has the distinct- 
ness and fixity of a species, and as a matter of fact this hybrid has 
been long mistaken for the true O. cruciata by a great number of 
European gardeners and botanists. 
In addition to the difficulties of hybridization and interpretation 
of the results described above, it is also to be taken into account that 
in some instances a long period ensues between the act of pollination 
and the perfection of the fruit, and then a long time is necessary for the 
germination of the seeds and development of the progeny. Ten, 
fifteen, or even twenty years might be necessary to make an applica- 
tion of this method to some of the species of trees, which would 
obviously make it unavailable except under sctracedioucy circum- 
stances 
The examination of the anatomical characters of a plant to deter- 
mine its ancestry is a method which has become of less esteem in the 
light of modern additions to information as to the character and 
behavior of hybrids. Of the various types of hybrids described it 
is of course the fixed hybrid which is most likely to come under exami- 
nation, and while it occasionally presents a fair average of the char- 
acters of the parent, more frequently it is goneoclinic to one or the 
other, and may be so near one parent that a gross or minute estimate of 
the tissue structure would offer nothing better than a guess as to the 
