126 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
Now Sophronitis has a very short column, while Cattleya has a 
decidedly long one, and one can quite understand that while it would be 
an easy matter for the pollen tubes of Cattleya to reach the ovules of 
Sophronitis ; yet, on the other hand, it would be much more difficult for 
the pollen tubes of Sophronitis to reach the ovules of Cattleya, owing to 
the length of the column duct in the latter. 
Incompatibility of colour seems to have something to do with the 
limitations of crossing, for Darwin gives us a large number of facts where 
colour varieties of the same species were more or less infertile when 
crossed. 
The limits of crossing do not seem to be determined so much by 
systematic affinity or natural relationship; for, as we have seen, some 
very distinct genera cross with ease, while some closely allied species 
refuse all attempts to unite them. I have found Oncidium flexuosum 
always infertile when pollinated with its own pollen, yet quite fertile 
when fertilised with pollen of O. Forbesii, a distinct species. Nor do the 
limits of crossing depend upon constitutional differences, for annuals 
can be crossed with perennials, deciduous trees with evergreens, plants 
from the tropics with plants which reach to the Arctic Circle. To sum up 
the whole question, we can only ascertain the true limits of crossing by 
actual experiments. 
It is encouraging to find that the more experiments made, the more 
barriers there are removed, the wider become the limits of crossing. 
The main thing is not to be discouraged by failures, but to try again and 
again, and above all to keep precise records, both of the successes and the 
failures, which records may prove to be of inestimable value to science. 
(End of the first day’s Conference.) 
