46 
POPULAR SCIENCE REVIEW. 
than the others. Their anthers are lodged in the closed part of 
the hood, filling completely all the vacant space. They are 
slightly attached by their external surface to the walls of the 
hood, while their dehiscence is on their internal surface ; so that, 
if you split open a flower longitudinally, as in fig. 1, you 
expose the interior of the anther cells on one side to view. The 
dehiscing surface, therefore, of the anthers of the right side is 
face to face with the dehiscing surface of the anthers of the left. 
Now the edges of the open anther cells on one side exactly cor- 
respond with the edges of the open anther cells on the other, 
just as the edges of one valve of an oyster exactly correspond 
with the edges of the other. It will thus be seen that the 
pollen grains of the opposite anther cells lie in a common 
cavity, and so long as the opposing edges of the cells are held 
in close contact, the pollen is kept securely imprisoned. How 
is this holding together brought about ? As to the upper 
anthers there is no difficulty, for the part of the hood in which 
they lie is so excessively flattened as necessarily to press them 
together. The two lower anthers are in rather a wider space, 
and might perhaps retreat from each other, and let the pollen 
escape, were it not for a number of stout hairs underneath them, 
which by their elastic pressure prevent the separation. These 
hairs spring from the filament of the upper anther on either 
side, which passes underneath the lower anther ; and it is to be 
noted that no such hairs grow on the filaments of the lower 
anthers, where their presence would serve no useful end. The 
four anthers with the style so completely fill up the closed part 
of the hood that all is held tightly packed, and it is next to im- 
possible for the parts to get disadjusted. 
We have now to consider what is the use of this exquisite 
workmanship. It is very simple. The flower is visited by 
large humble bees, which are attracted by the nectar of the 
tube. As the bee approaches the mouth of the tube, it strikes 
the projecting stigma with its head, and then settles on the 
lower lip. It cannot now reach the nectary without inserting 
its head into the fissure of the hood ; but this, owing to the 
flattening already described, is so narrow that the broad head 
dilates the hood, and forces the two sides more widely asunder. 
The necessary result of this can be foreseen. The anthers are 
attached, as I have said, externally to the inner surface of the 
hood; consequently, the widening of the hood draws the oppo- 
site anther cells apart. A little fissure is formed between the 
edges, which were before in close approximation, and through 
this the dusty pollen falls in a little shower on the back of the 
bee’s bead. With this adhering to it the bee flies off, and, 
striking the stigma of some other flower with its pollen-daubed 
head, fertilises it. I am unable to assert positively that the bee 
