— 241 
the average number of these which did not germinate was 
8. This comparative result may be briefly summarised : 
cross-pollination leads almost constantly to fertility ; self- 
pollination almost constantly to sterility. 
In the course of experimentation it was found that the 
pollen from the semi-aborted (outer) stamens was some¬ 
times potent. 
The above experiments were carried out under protection. 
In the open air the flowers are visited by humble bees in 
numbers during sunshine, and so thoroughly do they 
perform their work that on one inflorescence there were 
17 and on another 12 fine capsules due to their visits. 
Sometimes the bee when it first arrives tries to get to the 
nectar through the slit between the inner segments, but 
finding this impracticable it immediately proceeds to the 
top of the flower and opens it by pushing its head 
•against the deflexed apices. Müller (1) in explanation 
of this interesting phenomenon (in connection with 
Bryonia dioica) says : — « To push the head between 
closely approximated parts of a flower is a charac¬ 
teristic action among wasps and bees, which acquire the 
habit in constructing the chambers for their young. >■> The 
bees enter the flowers sometimes with the dorsal, some- 
times with the ventral, surface directed to the pistil. They 
push the body well into the flower until the thorax is 
under the anther and the proboscis within reach of the 
nectar at the base of the ovary. As far as was observed 
nectar was sucked from only one of the glands during a 
visit, the two opposite the undisturbed segments being left. 
When the bee forcibly pushes back the segment, the stamen 
(1) The Fertilisation of Flowers, Eng. edition, 1883, p. 269. 
16 
