APRIL, 1912.] THE ORCHID REVIEW. 103, 
This was shown by dividing the capsule longitudinally, the fertilised seeds 
being marked by their quite different colour. All the lower part contained 
colourless empty chaff, and was now cut off and thrown away. On the 
other hand, Brassavola Digbyana is not a good seed parent, because of its: 
very long beak above the ovary, and when so used one generally only finds 
good seeds at the apex of the capsule. Cochlioda Neetzliana is a good seed 
parent, but its pollen usually fails to fertilise Odontoglossum, where the 
pollen tubes have a much longer column to traverse. The cause of such 
cases is clear, but it is believed that other limitations are connected with 
the relative sizes of the pollen tubes and the micropyle of the ovule, the 
former sometimes being too large to enter. This could only be demon- 
strated by microscopical examination. A third set of limitations is con- 
nected with the time required for the development of the ovules and pollen 
tubes after pollination. 
The latter reveals a very remarkable condition of things. An Orchid 
flower is not mature on expansion, as in the case of a lily, as may be seen 
by cutting the ovary across, when the ovules are seen to be very minute 
and altogether undeveloped. Their development lags behind in a perfectly 
normal way, and only proceeds after and as the result of pollination, which 
introduces a totally new phase in the process of fertilisation. The result 
of pollination is immediately seen in the discoloration and withering of the 
floral segments and the thickening of the column. The pollen tubes now 
commence their growth down the tissue of the style, while the ovaries 
grow and mature. Both processes require a considerable time for their 
completion, varying from a week or ten days, in the case of the European 
terrestrial Orchids, to about three months in that of Cattleya Mossiz, which 
latter has been worked out by Mr. Harry J. Veitch in an elaborate paper. 
This development of the pollen tubes and ovules proceeds together, and 
when both are mature and fertilisation is effected—the capsule by this time 
appearing about half grown—the ovules then develop into seeds in the 
usual way. Now it is obvious that if plants were crossed whose pollen and 
ovules matured at relatively very different periods, fertilisation might fail 
because the pollen tubes matured before or after the ovules were ready, and 
thus fail at the critical period. It is often remarked that capsules go off 
when only half developed, or open somewhat later, and are found to 
contain nothing but chaff, and the cause of this is evidently due to the 
fact that fertilisation has not been effected. Pollination and fertilisation 
are totally different processes, and the stimulus to the development of the 
ovules may be given by many different kinds of pollen, which are after- 
wards utterly incapable of effecting their fertilisation. 
Hybridisation and selection have been among the most potent causes 
of improvement of garden plants. One can hardly turn anywhere without 
