354 THE ORCHID REVIEW. [DECEMBER, 1909. 
Although the tribe is as a whole very natural and sharply defined, its 
division into genera and their arrangement in natural sequence is a matter 
of considerable difficulty, and in the delimitation of genera much diversity 
of practice exists. Bentham recognises four subtribes, namely Serapiadez, 
Habenariez, Disez and Coryciee. 
The subtribe Serapiadez is characterised by having the glands of the 
pollinia enclosed in distinct membranous pouches situated at the back 
of the rostellum, which protect a round ball of viscid matter serving to 
attach the pollen masses to the visiting insect. It contains the four genera, 
Orchis, Aceras, Ophrys and Serapias, the two former including some of our 
most familiar British species. The group is widely diffused in Europe, 
temperate Asia, and North America, and there are two somewhat anomalous 
representatives in North America. 
The structure of Orchis mascula was very fully described by Darwin, 
and is fairly typical of the whole genus. The sepals and petals form a hood 
over the column, and the lip forms a kind of landing stage in front, while 
its base is prolonged intoa stout spur. The flowers are fertilised by several 
species of humble bee, and Hermann Miiller caught ninety seven specimens 
which were visiting the flowers, and of these thirty-two had pollinia 
attached to their heads. The bee alights on the lip and inserts its proboscis 
into the spur, in doing which it touches the rostellum which projects into 
the mouth of the spur. The rostellum is protected by an exterior membrane 
which, when touched, ruptures in two lines, allowing the pouch of the 
rostellum to be depressed as the insect pushes forward, when it inevitably 
comes in contact with one or both of the viscid balls to which the caudicles 
of the pollinia are attached. So viscid are these balls that whatever they 
touch they firmly stick to. The viscid matter has, moreover, the peculiar 
chemical property of setting hard, like cement, in a few minutes time. As 
the anther cells are open in front the pollinia will be withdrawn as the insect 
retreats, firmly cemented to its head like a pair of projecting horns. And 
now another beautiful contrivance comes into play. The caudicles are 
hygrometric, and on exposure to the air begin to contract in a definite 
direction, causing the pollinia to curve forward through an angle of about 
ninety degrees, the movement being completed in about thirty seconds, and 
in the new position they are placedso that they come in contact with the 
stigma of the next flower visited. And here another very pretty adaptation 
is seen. Both the stigma and the gland of the rostellum are viscid, and 
when the insect retreats something must give way, and this is the elastic 
threads by which the packets of pollen are attached to the caudicle. All 
these points can be demonstrated artificially. Ifa pointed stick or lead 
pencil be pushed into the spur and then gently withdrawn the pollinia will 
be found attached to it, and the remarkable movement of the pollinia can 
