BOTANY. 479 
the slightest touch. The sensitiveness will be seen to play a useful 
part in this fecundation. 
I will take the commonest species, M. moschatus, as a type. 
The flowers vary from erect in the bud to horizontal in the full 
blown flower, but never hang downwards. Of the four stamens 
the anterior, lower, and larger pair ripen after the posterior, 
upper, and shorter pair. Both pairs of anthers are held together . 
by hairs, and the longitudinal slits of the anther open towards the 
lower lip, and away from the base of the flower. The style is 
closely pressed against the upper lip of the corolla, and its stigma 
has two large flat fan-shaped lobes. In a very young bud these 
lobes are closed. In a hardly opened bud the lobes are beginning 
to open, the lower one bending back against the style; at this 
time it is that the shorter stamens burst, but as they are much 
Shorter than the style the pollen cannot reach the stigma, and its 
course down the tube is facilitated by the, at that time, slanting 
position of the flower. In a just opened flower the stigmas are 
fully open, parallel, and opposite to the lower lip of the corolla, 
its viscous surfaces being therefore both downwards; the shorter 
anthers are nearly empty, and the longer only just beginning to 
Split; the pistil is therefore synacmic with the shorter, and almost 
protogynous with respect to the longer stamens. 
a flower almost beginning to fade the longer stamens are 
still shedding their pollen, the shorter ones are withered, and thè 
stigma be-pollened and in many cases closed. This closing may, 
moreover, be experimentally produced by touching the stigmatic 
Surface with a pencil, in which case the stigmas will close in ahout 
thirty seconds. In faded flowers, whether from contact or other- 
Wise, the stigmatic surfaces have closed. 
From these facts it will appear that self-fertilization by the 
shorter stamens is impossible, and that it is rendered improbable 
by the longer stamens (1) by their bursting late; (2) by the di- 
rection in which the anthers open; (3) by their not reaching as 
far as the stigmas, and, as being anterior, by being some slight 
distance from the upper lip; (4) from the probability that the 
stigmatic surfaces may have been touched and closed before they 
burst at all. 
On the other hand, an insect attracted to the flower for the 
honey could hardly leave the flowers without some pollen on the 
Upper side of his body or on his proboscis. The hairs which hold 
