100 
THE ORCHID REVIEW. [APRIL, 1906. 
curiosities of hybridisation, and remarked that he had found it impossible 
to cross large and small flowers together, unless the larger was made the 
pollen parent. As examples, he mentioned certain Cattleyas and Lalias 
PRASash 
with Sop 
gl with Cochlioda. He had 
come to the conclusion that the smaller grains of pollen from the smaller 
flowers were not capable of producing tubes sufficiently long to reach the 
ovule of the larger flower, and were therefore ineffective. 
Mr. R. A. Rolfe, of Kew, remarked that there was something in the 
suggestion, for Mr. Charlesworth, of Bradford, had told him that when 
Cattleyas were fertilised with Sophronitis grandiflora the fertile seeds were 
always near the apex of the capsule, and when visiting the nursery he 
demonstrated the fact. 
Taking a mature capsule that had been so crossed, 
he divided it longitudinally, showing the contents at the base to be white, 
and those near the apex brown, the dividing line being well marked. At 
this point the capsule was cut transversely, and all the good seeds were 
found to be in the upper half. The pollen tubes had not reached the ovules 
in the basal half, in fact the different colour marked their limit. In the 
case of Brassavola Digbyana—and, by the way, its long beak, proved it to 
be a true Brassavola, 
not a Lelia—it was said that if this species were 
used as the seed parent the good seeds were usually found near the apex of 
the capsule, and thus the limiting influence of the long beak was apparent. 
But a notable exception to the theory occurred to him, for in the 
hybrids between Epidendrum radicans and larger - flowered _ species 
the Epidendrum was 
him that it had been found impossible to get capsules on Epidendrum ~ 
used as the pollen parent, and Mr. Seden had told 
radicans unless the pollen of another Epidendrum was used; the 
pollen of the larger species invariably proved impotent. It recalled 
the case of Mirabilis jalapa which was readily crossed with the 
pollen of M. longiflora, but the reverse cross would not take, as Kélreuter 
had proved by making the cross, if the speaker remembered rightly, 
some hundreds of times. M. longiflora had the longer style, which might 
be the reason for the peculiarity. But other difficulties might regulate the 
limits of crossing. The pollen tubes might sometimes be too short, oF 
they might be too large to penetrate the micropyle of the ovule. And they 
might develop too quickly or too slowly. An Orchid differed from most 
other flowers in the fact that the ovules were not developed at the moment 
of expansion, and could not then be fertilised. And without the stimulus 
afforded by pollination they did not develop. Mr. Veitch had shown that 
in the case of Cattleya Mossizx, fertilisation did not take place till about 
three months after pollination, and all this time the pollen tubes and the 
ovules were steadily developing. If a pollen parent were used which 
developed in a much 
shorter period failure might result from this cause- 
