XV 
a state of confluence, I also supposed the position of the stigmata to be 
opposite the petals ; being led to that conclusion by the constant position 
of the stigmatic arms of Ophrydz. That opinion I now retract, in con- 
sequence of the position of the stigmata in Cypripedium, which C. specta- 
bile shews most clearly to be opposite the sepals; and therefore the 
stigmatic arms of Ophryde are to be understood as side lobes of that stigma 
which is opposite the dorsal sepal. This circumstance, however, only con- 
firms the accuracy of my view of the true nature of the stamens, which 
are certainly all opposite the lobes of the stigma in Cypripedium. 
While, however the untenableness of my first opinion, concerning the 
relation borne by the stigmata to the other parts of the flower, is thus 
admitted, there remains a difficulty that opposes itself to the view I 
now take in common with Brown, and which must not be overlooked. It 
is that the placentiferous pieces of the ovary are not opposite the stig- 
mata, but alternate with them, ‘while the seedless pieces of the ovary 
are ina line with the stigmata! This seems to show that the ovary 
is composed of 6 carpellary leaves, of which three bear stigmata without 
ovules, and three bear ovules without stigmata. However paradoxical 
this may appear, it is by no means incompatible with the due perform- 
ance of the functions of impregnation; for the carpellary leaves do 
not adhere into a solid mass, either in the ovary or in the style. On the 
contrary they form a cavity open from the stigmatic apex down to the 
ovules, and the whole of that cavity is lined with a lax conducting tissue, 
which may nevertheless be exclusively furnished by 3 stigmas only, and 
may become so confluent with the placente as to form a perfect channel of 
communication for the pollen tubes in their descent into the ovules. 
The order owes its chief peculiarities to the following circumstances : 
Jirstly, to the consolidation of all the sexual organs into one common mass, 
called the column ; secondly, to the suppression of all the anthers, except 
one in the mass of the order, or two in Cypripedee ; thirdly, to the pecu- 
liar condition of its pollen, and the anther which contains it ; and fourthly, 
to the very general developement of one of the inner leaves of the perianth 
or petals in an excessive degree, or in an unusual form. These peculiarities 
are in most cases so striking, and all so strongly nranifested in the same 
flower, that the inexperienced botanist may be unable to discover their real 
character. We find, however, that the true nature of each part is indicated 
by special cases of structure occurring in different parts of the order. 
Thus in Cypripedium not only are two lateral stamens furnished with 
anthers, while the central stamen is antherless, but the stigma and style 
Separate from the filaments nearly to the base, and the triple nature of the 
former is distinctly shown, together with the relation of its lobes to the 
other parts of the flower. The pollen, which has so anomalous an appear- 
ance in its waxy or sectile state, presents the usual appearance of that 
substance in Goodyera, and many Neotteze. And the irregularity of the 
labellum disappears in such genera as Thelymitra, Paxtonia, Macdonaldia, 
Hexisea, and some others, whose flowers are almost as regular as those of 
a Sisyrinchium. 
se is indeed to the latter genus, more nearly than to any other, that 
rchidaceze seem to approach in structure, except to Zingiberaces ; so 
- * 
