56 
in retracting thus within itself, draws with it a few grains 
of pollen, which thus appear to penetrate the tissue of the 
style, but which in fact are always on the outside of the hair. 
With care these hairs may be pulled out again by the point 
of a needle, and then the pollen-grains which appear to have 
penetrated the style are immediately expelled. Such pollen- 
grains undergo no change during their application to the 
collecting hairs, nor even when they are drawn inwards by 
the latter during their act of retraction. 
There is therefore no communication between them and 
the interior of the style. 
As to the immediate cause of this retraction of the hairs, 
without pretending to give a certain explanation of it, I think 
it may be ascribed to the absorption of the liquid contained 
both in the hair and in the cavity at its base, an absorption, 
the effect of which will be to pull back the hair into the cavity, 
at least I see no other part whose action can produce the phe- 
nomenon. 
An examination of the structure of the external stratum 
of the style and stigmatic arms, has already tended to show 
the baseless character of the opinion held by those physiolo- 
gists, who think that fertilization can take place by the action 
of the pollen upon this part; an opinion offered with doubt . 
by Cassini, and Alphonse DeCandolle, admitted on the con-: 
trary in the most positive manner by Treviranus, who, in his 
Physiology, vol ii. p. 843, considers the internal stigmatic 
surface to be formed of papille analogous to those which 
sometimes terminate the petals, while, according to him, the 
hairs covering the external surface of the style and stigma, 
perform the part of the stigmata. Link (Philosophia Botanica, 
2nd edition, vol. ii. p. 229), also admits that fertilization takes 
place by these hairs, whose points he says are destroyed while 
the base remains, and so present a large opening which leads 
into the style. 
We therefore see that the most distinguished Botanists 
entertain opinions either doubtful or contrary to the most 
probable analogies. Nevertheless in dissecting the true stig- 
ma of Campanulas, that is to say, the inner face of the stig- 
matic arms, after their divergence, we find that the grains of 
pollen scattered over the surface adhere to it, as to all true 
stigmas, first by aid of the fluid that lubricates them, and 
finally, by the production of pollen-tubes which penetrate it, 
X 
