74 OBSERVATIONS ON THE FERTILISATION OF SAXIFRAGA. 
parts one to another before we proceed further. Fig. 1 is a diagram 
that will answer our purpose. The sepals are not shown, as they do 
not affect the point we wish to elucidate. The outside ring of leaves 
represents the corolla, consisting of five petals. Next, there are the 
stamens, an outer whorl or circle of five long and an inner whorl 
of five short ones. Each of these consists of a filament or stalk, and 
a head or anther, which is two-celled, and contains the pollen. Then 
in the centre we have the ovary, or seed-vessel, which is two- 
chamhered. Connected with each chamber is a projection called a 
style, and at the tips of these two styles are the stigmas. 
We are now in a position to understand the description of the 
phenomena of the fertilisation of the Saxifrages. 
If we take a flower of the London Pride {Haxifraga umhrosa), 
which has not been too recently nor too long expanded, and tear away 
two of the petals with the stamens near them, we shall see something 
like what is shown in Fig. 2. Here we notice that the two styles 
touch at the tips. One of the longer stamens has lost its anther, 
another has assumed an almost erect position, with the anther cross¬ 
wise on the top of the filament, and the three remaining stamens of 
the inner whorl have grown out to about the length of the outer 
stamens, but still lie back upon the petals. If we had examined this 
flower a little earlier we should have found the stamens entire, and 
the styles a little apart all the way up. Later on, when all the pollen 
is shed, the petals and stamens will wither and fall, and the ovary 
assume the appearance of Fig. B. Here the styles have parted from 
each other and sprayed out, exposing the rough sticky stigmas at their 
tips, so as to brush off from any insect that may alight upon the 
flower some of the pollen with which it may have been dusted from a 
flower in the state shown at Fig. 2. Figures 4 and 5 represent the 
ovary of the mossy saxifrage ((S', muscoides) at an early and at a later 
stage of development ; and Figures 6 and 7 exhibit the same 
features in the White Meadow saxifrage (.S', granulata). The shape 
and size of the ovaries in these species differ from one another and 
from the London Pride, but essentially the same motions take place 
in each of the three species. At first, we have stamens entire, of 
two lengths, in two whorls, and ovary with immature parallel styles. 
Later on, the styles beiid over to each other, and press their stigmas 
together, while each of the stamens, first of the outer whorl then of 
the inner, in succession a^isumes an erect position, with the bursting 
anther lying across the top of the filament, just in the right position 
to besmear any hairy insect that may alight upon the flower, and at 
the same time incapable of dusting the stigmas by reason of their 
close contact throughout the whole of their surfaces. When each 
stamen has shed its pollen the filament either goes back to its former 
position or shrivels up ; and when all the stamens have performed 
their office the styles of the ovary part from each other and, widely 
diverging, fully expose the now ripe stigmatic faces. Finally, a 
