4104 “PRE. URGCHID REVIEW. [JULY-AUGUST, 1919+ 
hybrid seeds as well as pure’seeds are developed and may germinate side by 
side. If the hybrids are fertile the intercrossing may be carried further, 
and there is ample opportunity for plenty of mixed ancestry and reversion. 
The plants are brought home by collectors, and on flowering present the 
different characters with which weare.so familiar. There is no record of 
their origin, and they can only be classified by the characters they possess, 
pure or mixed as the case may be. Some of them have been unwittingly 
described as species, as the plants now known as Leliocattleya elegans 
and Lc. Schilleriana were, and those who recognised their mixed character 
have had to fight for their recognition as hybrids, until some experiment 
placed the matter beyond dispute. Others were recognised as hybrids from 
the outset, especially where they appeared in importations of known origin, 
while those of less definite character have sometimes been regarded as 
varieties of recognised species, occasionally with the cryptic remark that 
they showed a “touch of hybridity,’”’a recognition of the possibility of their 
being hybrid reversions. 
In the case of garden hybrids there need be none of this uncertainty. 
Everything is within rigid control. Nature knows no selection of parents, 
_and keeps no records. The hybridist can do both. He can cross species © 
which do not grow together, and he can recombine the offspring in any way 
he chooses, in the hope of getting the desired results. . There are limitations, 
of course, and the result cannot always be foreseen, certainly not controlled, 
but plenty of good material is obtained for further experiment in whatever 
way improvement seems desirable. Surprises there will be, and plenty of 
variation and reversion, and it is this which supplies the clue to what is 
going on in nature. The results would be strictly comparable if ex- 
periments were limited to the species which grow together in a wild state, 
and the crosses made without selection of parents, A few more experiments 
on these lines might be carried out with the object of settling doubtful 
questions, and it would eliminate some of the difficulties of nomenclature, 
The object of horticultural experiment is novelty and improvement, and the 
inferior forms are inevitably neglected, and ultimately disappear. Varietal 
names can be given to the best and most distinct, the others being chiefly 
interesting as variations. In the absence of record we can only suggest 
classifying them under the recognised hybrid that they most resemble, as 
has to be done in the case of natural hybrids. 
In the case of variations among seedlings from the same batch, or from 
batches of identical parentage, the adopted course is to bring them all under 
the same specific name, with the addition of distinguishing varietal names 
where desirable or necessary. 
