FERTILISATION OF VARIOUS FLOWERS BY INSECTS. 
171 
about appears to me to be as follows. In the bud the sharp- 
pointed end of the anther (fig. 13) adheres to the style, with 
which it is in contact, by means of a sticky, gum-like fluid. 
Sometimes the anther-point does not seem to be in actual 
contact with the style, but to be united to it by a slender white 
thread (fig. 13). This thread, however, I take to be some of 
the same gum-like fluid, which has been accidentally pulled 
out and then has dried. Be this as it may, in some way or 
other the point of the anther is at this period attached to 
the style, and therefore fixed. The other end of the anther 
is jointed movably to the filament. Consequently, as the 
growing filament increases in length, a bend necessarily occurs 
at this joint, the lower end of the anther being tilted up, and 
the anther made to rotate round its fixed end. This motion 
continues till the anther reaches the position it holds in the 
mature flower (fig. 12), when the opposing style hinders any 
further rotation. Such lengthening of the filament as may 
occur after this can only take place by the filament itself 
curving outwards ; and the more it is thus curved, the more 
firmly will it press the anther against the style, and the more 
securely, therefore, will the pollen be retained. No pollen 
escapes during the rotation, for the pore remains closed until 
it comes into contact with the style. 
It may seem strange that Nature should have adopted so 
apparently roundabout a plan for bringing the anthers into 
their necessary position. Why should they not have been 
developed from the first inverted, and with their pores in 
contact with the style, in which case no rotation would have 
been requisite? In answer to this, it may be said that Nature 
by no means invariably selects the most direct path to reach 
her end, but, as the student of development well knows, fre- 
quently proceeds in a zigzag in the production of an organism ; 
and not rarely, after following some definite path for a certain 
distance, may even retrace her steps, obliterating as she does so 
her footprints, and start afresh in a new direction. 
The flowers with which I have now dealt in this and two 
former articles are numerous, and belong to many different 
orders. In all we have seen that there are special arrangements 
to ensure the transfer of pollen to the body of an insect, which 
is the same thing as saying that there are special arrangements 
to ensure at least occasional intercrossing”.* If to these be 
added the Orchises, Tynums , Ty thrums, Primulas , &c., &c., 
which Mr. Darwin has long since shown to be dependent upon 
insects for their due fertilisation, and to these again such other 
instances as have been described by other observers, we have 
* See “ Popular Science Review,” July 1869. 
