EXPERIMENTS IN HYBRIDISATION, &c. 125 
lums of the second generation, raised from a first hybrid, out-crossed again 
with a third species, are more vigorous even than the first hybrids, being 
quite abnormal in this respect. This seems to show that out-crossing 
continues to increase the vigour of hybrids abnormally, while inbreeding 
tends to reduce their vigour to a normal state. 
Tue Liuits oF CROSSING. 
In referring to the question of the fertility of hybrids, we saw upon 
what slight conditions sterility sometimes depends ; and yet, if the con- 
ditions be favourable, it is remarkable what extreme forms of plants can 
be united by hybridisation. For instance, in the Orchidex, during the 
past decade not only have many distinct species been successfully united, 
but, as we have already seen, numbers of distinct genera hybridise to- 
gether with ease. 
The four genera Lelia, Cattleya, Epidendrum, and Sophronitis are 
now all united by hybridisation, as are Zygopetalum, Colax,and Bate- 
mannia. These hybrids have all flowered in gardens, and show the 
characters of their parent genera. In addition to these are numerous 
records of generic crosses, yet unflowered, though plants or seeds have 
been produced.* 
The most extreme of the flowered generic hybrids in Orchidee are 
apparently those between Phaius and Calanthe, genera belonging to dis- 
tinct sub-tribes, so that the generic limit even has now been surpassed, 
and that of the sub-tribe reached. Is it possible to go farther than this ? 
I think so, the relationship between many of the generic hybrids, yet 
unflowered, being much more remote. 
I have made a number of experiments in crossing distinct genera in 
the Orchidez, and, though many failures have resulted, yet some have 
proved rather interesting. For instance, Sophronitis x Cattleya, 
Sophronitis x Lelia, Paphiopedilum x Phragmipedilum, Cypripedilum 
x Phragmipedilum, Paphiopedilum x Cypripedilum, Oncidium x 
Odontoglossum, all produced healthy capsules, containing some good 
seeds, many of which are now germinating; while Lycaste x Lelia, 
Angrecum x Lelia, Angrecum x Vanda, Paphiopedilum x Odonto- 
glossum, and Paphiopedilum x Dendrobium all produced fully developed 
capsules, which opened naturally in due time, but which contained no 
good seeds. Again, Ada x Lelia, Cattleya x Angraecum, Dendrobium 
x Odontoglossum, Epidendrum x Dendrobium, Epidendrum x Odonto- 
glossum, Cattleya x Dendrobium, all produced partly formed capsules, 
which ultimately faded ; while Lelia x Lycaste, Dendrobium x Cattleya, 
Paphiopedilum x Cattleya, Epidendrum x Dendrobium, and Epidendrum 
x Paphiopedilum all failed to set capsules. 
Incompatibility of structure, apart altogether from systematic affinity, 
may have something to do with the limitation of crossing, for I observe 
in the above experiments that in all the instances where Sophronitis was 
crossed with Cattleya, when Sophronitis was the seed parent, good seeds 
were obtained, which duly germinated ; while the reverse crosses, made at 
the same time and under the same conditions, in every case failed to set 
a capsule. 
* See Jour. Roy. Hort. Soc. xxi. [1898], p. 468. 
