402 MR. H. J. VEITCH ON THE 
are grouped in clusters of no very definite form and outline; 
each ovule has the appearance of a single cell of ovoid form, but 
so minute are they at this stage, that no differentiation of parts 
could be made out under the low power to which we were re- 
stricted, although a faint reticulation was observable in some of 
the most advanced ; but this may have been an optical effect. It 
was, however, evident that the actual impregnation of the ovules 
by the pollen-tubes, which we were anxious to detect, was yet 
remote, or at least it was apparent that that event would not 
immediately occur. In fact, we were up to this time simply 
groping our way towards trying to ascertain what we hoped would 
prove to be a most interesting scientific fact, but with the road 
to which we were very imperfectly acquainted indeed. 
Our next examination was made fifty-five days after pollina- 
tion; but, to our surprise, although the pollen-tubes had pene- 
trated the ovary in countless numbers, and had completely 
choked up the canal leading from the stigmatic chamber to it, 
so that they had there taken the form of a bundle of fibres, 
minute as are the individuals composing it, and could be pulled 
away bodily in a coherent string, no actual impregnation of the 
ovules could be detected; the tubes lay along the sides of the 
placentas and among the ovules, and had reached as far as the 
bottom of the ovary. Moreover, all this while, and for some 
weeks afterwards, the summer weather was, for our climate, 
very favourable for the development of the impregnated flowers, 
the number of bright clear days being rather above the average- 
We waited 17 days longer before making a further examina- 
tion; we then found that the ovules were not only enlarged, 
but were also undergoing a change in form—a fact that induced 
us to believe that the long sought-for event was at hand, and we 
accordingly made a further examination three days later, but no 
perceptible advance could be detected. We therefore desisted 
from further trial for another fortnight, that is till 90 days after 
pollination, when, at length we were able to understand with 
tolerable certainty the process by which the impregnation of the 
ovules is effected, and to get an idea of the space of time required 
for its accomplishment. 
This will be best understood by reference to the diagrams. In 
fig. 9 are represented, two thirds natural size, two transverse 
sections of the ovary—D 55 days and E 72 days after the pol- 
lination of the flower; in fig. 10 a transverse section 90 days 
