118 ORCHIDS OF THE CAPE PENINSULA. 
(in S. rhynchanthum) the column undivided, with a terminal stigma, 
and a very short rostellum. Anther placed in front, hanging under 
the rostellum, or horizontal, the cells nearly parallel; caudicles of the 
pollinia attached to two glands which point forward, and are placed in 
recesses of the margin of the rostellum, or more rarely are turned 
backwards; or rarely (in S. rhynchanthum) the caudicles united to a 
single gland. Ovary and pedicel straight, never twisted.—Terrestrial 
herbs, with the leaves mostly radical, or on the lower part of the stem, 
reduced to sheaths on the upper portion. Spikes usually densely 
flowered; bracts membranaceous or leaf-like, frequently erect and 
imbricate in the early stage of flowering, at length reflexed. (Name 
from the Satyri, the rural demigods of the Greek mythology.) 
The genus is very natural and distinct, and has experienced but 
few changes since its reformation by Thunberg and Swartz in the 
year 1800. The perianth is not, indeed, absolutely uniform, but the 
same general type is prevalent throughout. The lip is almost always 
galeate, or nearly so; in one species only (S. striatum, Thunberg), is 
it simply arched or vaulted. The other parts vary little in shape, but 
there are considerable differences in their degree of cohesion with each 
other, and in their setting. The column shows great differences in 
its length and curvature, in the shape of the stigmatiferous lobe, and 
of the rostellum. 
In 8. rhynchanthum, which was regarded by Lindley as the type of 
a distinct genus, Satyridium, the occurrence of a single gland (the 
only instance in the genus), combined with the anomalous structure 
of the column, seems to warrant its being regarded as a subgenus 
under that name. 
In seeking for characters upon which to found divisions of the 
remaining species, and which should also embrace those species of the 
genus which lie beyond our limits, it soon became apparent that the 
differences in the column, though apparently constant in the species, 
and yielding excellent marks for specific distinctions, do not appear 
to be correlated with differences in the perianth or in vegetative 
characters, and therefore do not indicate natural groups. Lindley’s 
divisions into Calcarate and Saccate are apparently the best and most 
useful for artificial sections, so long as no attempt is made to divide 
the former into subsections with long and short spurs. Then they 
break down. The length of the spur is variable even in the same 
species, and there is no broad division possible into long and short. 
There is a clear distinction, however, between the spurred and saccate 
species, because the shortest spurs are much longer than the longest 
sacs, and the former always taper to the extremities, while the latter 
do not, and they do not in any way shade off one into the other. The 
