170 
TOPULAR SCIENCE REVIEW. 
is always minutely described, but no mention is made of the 
all-important fact of the adherence of the contiguous cells of 
neighbouring anthers in their dehiscing portions. This, how- 
ever, is easily seen if one selects a young flower which has not 
yet been visited, and removes the corolla with due care. Ac- 
tually to see the bee strike the processes is of course impossible ; 
but I have often seen one come to a flower with a clean head, 
and leave it with a head dusted with pollen grains. 
In the fine-leaved Heath ( Erica cinerea) there is a me- 
chanism for fertilisation precisely similar to that of Tetralix. 
Probably the same is the case with all those Heaths in which 
the anthers are provided with processes. 
In the closely allied Vaccinium there is a somewhat different 
though very similar arrangement. Here also the anthers are 
furnished with processes, which act as in the Heaths. But the 
pores at the farther end of the cells are not placed on the 
sides, as in the Heath, but are actually terminal (fig. 10). They 
cannot therefore be closed as before by lateral adherence of 
contiguous cells. The closure, however, is effected in another 
way. The apex of the cell where the pore is placed presses 
against the style, so that the style itself serves the office of a 
cork, and prevents the escape of the pollen. The filament is so 
bent as to act like a spring, and keep the anther firmly pressed 
against the style (figs. 10 and 11). It is only when pressure is 
made upon some of the processes that the corresponding anther 
is tilted up, its terminal pores exposed, and the pollen allowed 
to escape. This pressure is applied, as in the Heath, by the 
proboscis of the bee. The result, as before, is a shower of pollen 
upon the insect’s head, which it carries off and leaves upon the 
slightl} 7 projecting stigma of the next flower it may visit. 
In the common Arbutus the mechanism of the anthers is 
similar to that in Vaccinium — that is to say, the pores are 
blocked up, while the flower is undisturbed, by the style itself. 
The arrangement is shown in fig. 12, which represents the 
position of the parts in a mature flower. In fig. 13 the same 
parts are shown in a yet unopened flower-bud. It will be 
noticed that the position of the anthers is exactly reversed in 
the two figures. In the bud the processes of the anthers and 
the parts in which the pores will afterwards appear are turned 
away from the mouth of the flower. In the mature blossom 
they are turned towards it. The anther, therefore, has under- 
gone a revolution, has been turned upside down during the 
ripening of the blossom ; and it is plain that if it had not done 
so, but had retained the position it held in the bud, the pollen 
could not have fallen out of the cells, seeing that the mature 
flower hangs with its mouth downwards. 
The way in which this revolution of the anther is brought 
