THE ORCHID REVIEW. - 72% 
here again is an absolute difference. The first subdivision of the Orchideze 
is into those with two or more stamens—the Diandre—and those with- 
only one—the Monandree—the former with simple pollen “grains, the 
latter having them united in fours, or still further aggregated in masses.’ 
The Diandre are further divisible into two marked tribes, Apostasiez 
and Cypripedieze, whose differences may be briefly defined. In Apostasiez 
the perianth is regular or nearly so, the column very short, the anthers- 
elongated, the pollen dry, and the style very slender; while in Cypripediez 
the perianth is highly irregular by the modification of one petal into a 
pouch, the column more developed, the anthers globose, and the pollen 
grains agglutinated together by a viscid exudation. There are other 
characters, but all are liable to exception. For example, Apostasiez have 
a three-celled ovary with axile placentation, but so has Selenipedium, 
while the dorsal stamen may be modified into a staminode in Apostasiez, 
though it is never shield-shaped, as in Cypripediez. 
In Apostasiez we find three distinct types of structure, which evidently 
represent the same number of genera. In the first the three stamens are 
all perfect, and this characterises Neuwiedia, a genus of six known species. 
In the second type the dorsal stamen is modified into a subulate staminode, 
and the anthers are linear-oblong with unequal bases, but in the third the 
dorsal stamen is entirely suppressed, and the anthers are equal and cordate. 
at the base. The second type is that of the genus Apostasia, as originally 
described, and the third was called Apostasia section Adactylus, in order 
to distinguish it from the original. But it is evidently a good genus, which 
may bear the latter name. Apostasia, as thus limited, comprises four 
species, three Indo Malayan and one tropical Australian, while Adactylus 
has three, A. nudus, A. Lobbii, and A. latifolius (which formerly bore these 
respective names under Apostasia), all of them Indo-Malayan. 
In the Cypripediez we find simply a higher development of the same 
structural type, and one of the two genera—Selenipedium—has retained 
the ovarian character of the more ancestral Apostasiee. To separate 
Apostasiee as a distinct Order, while at the same time including Cypri- 
pediez among Orchids, is simply to ignore the facts. The gap which 
separates Monandre from Diandre is far wider, but, taking all things into- 
consideration, I should consider the two as distinct suborders only. 
In the Monandre the aggregation and displacement of parts has been 
carried much further, and the dorsal stamen of the outer whorl—the 
staminode of .Cypripediee—is alone perfect, though others are potentially 
present as wings of the column and other staminodial appendages. In 
addition to the compound pollen-grains, it is in this suborder alone that: 
one of the three pistils becomes modified into the rostellum, while the two- 
remaining ones are represented by the stigma. This highly specialised 
group has grown and multiplied exceedingly, and comprises the four great 
