XIII 
between the pollen and stigma, by enclosing the former in an anther, 
from which it can only be released by the action of some external 
stimulus, such as might be received from a bird or an insect. 
On the other hand Dr. Brown, M. Adolphe Brongniart, and others, 
suppose that impregnation is effected here as elsewhere by contact 
between the pollen and stigma. It is alleged that pollen-tubes are 
emitted bv the pollen after it has adhered to the stigma, that these tubes 
descend the stigmatic canal, and mix themselves among the ovules over 
the surface of the placentae, and that in some cases the pollen-tubes 
have been traced into the foramina of the ovules. The latter is a remark 
of Dr. Brown’s, whose skill and faithfulness as an observer are above all 
question. 
For a long time I adopted the views of Mr. Bauer, but additional 
experience, a careful observation of Orchidaceae in a state of cultivation, 
and a verification of some of the facts last described, have caused me to 
alter my opinion, to acquiesce in the conclusions of Dr. Brown, and to 
regard the phenomena represented by Mr. Bauer as connected with cir¬ 
cumstances not belonging to impregnation. It is now well known that 
cultivated Orchidaceae, which rarely fruit if unassisted, may be made to 
bear fruit with certainty, if they are artificially impregnated. The effect 
of such impregnation is firstly to hasten by many days the decay of the 
flower, and secondly, to fill the stigmatic canal with pollen tubes, which 
may be traced without difficulty even up to the very maturation of the 
fruit. 
The determination of this point does not however diminish the value 
of Mr. Bauer’s observations; it only leaves the facts he describes still to 
be accounted for ; and it is a most curious subject of enquiry to inves¬ 
tigate the exact uses of so singular and peculiar a sexual apparatus as 
that with which Orchidaceae are furnished. 
The seeds are generally enveloped by a loose cellular testa, which 
is frequently open at the lower end, at the period of maturity. They con¬ 
tain a nucleus, whose apex is directed towards the base of the seed, and 
which is either covered by two extremely thin membranes, frequently 
