468 JOURNAL OF SCIENCE. 
produced. Of these the larger and more conspicuous form has six 
(or more) white sepals, and a central tuft of stamens. The smaller 
and greener form has also six sepals, a single row of stamens, and a 
central tuft of carpels. In structure it is apparently hermaphrodite, 
but as the anthers have no pollen, it is functionally a pistillate form, 
so that we always call Clematis indivisa a dicecious species. Lastly, 
the common bush-lawyer or bramble (Rubus australis) has two forms 
of flowers; in the larger and more conspicuous of these there is inside 
the petals a number of stamens, but no trace of a pistil beyond a 
small projection to mark its usual position; in the other and smaller 
form there is a tuft of separate carpels but no trace of stamens. 
Here therefore we have a flower, both structurally and functionally 
dicecious. Such examples as these are very numerous in our flora, 
but the three adduced are so common and so easily examined, that we 
need to go no further for types. It is just noteworthy that in all 
such cases the staminate flowers are larger, brighter coloured, and 
generally more attractive than the pistillate, and this probably because 
while a single grain of pollen falling on a stigma is presumably sufficient 
for each ovule, and therefore one or two visits of pollen-bearing 
insects will suffice to fully fertilize the pistillate flower, it is essential 
on the other hand that such insects be first well dusted with pollen, 
and this is attained by the flowers being rendered specially attractive 
to them. Therefore in all or nearly all cases of dicecious flowers 
which depend on insects for their fertilization, the staminate flowers 
are the most conspicuous, or most fragrant, or most richly provided 
with nectar. Finally, for the full comprehension of what has here 
been said on the subject of insect fertilization, it must be remembered 
that insects do not visit all flowers from which they can obtain honey 
indiscriminately. Many species of insects confine their visits to a few 
species of flowers only, and do not fly from one to other of these 
selected kinds indiscriminately. A bee generally begins work in the 
morning on some particular kind of flower such as dandelion, and it 
will continue if the flowers of that species are sufficiently common to 
visit one after another of them, without as arule going off to the 
daisies and gorse or broom which may happen to be growing about. 
But of course if the bee did carry pollen from the daisy and leave it 
on the stigmas of dandelion, it would exercise no more fertilizing 
influence there than so much sand would do. We therefore sce that 
a certain quantity of pollen is always wasted, and this is one of the 
reasons why flowers produce so much. But the more special the 
mode of fertilization of the flower, the smaller the quantity of pollen 
produced. In this respect again the extreme point has been reached 
by Orchids and Asclepiads, where as a rule the pollen adheres into 
two or four masses, which are sometimes all removed by a single visit 
of an insect. 
I cannot well leave this subject of the relationship of insects to 
flowers, without pointing out another mode in which they have largely 
influenced development. Very many kinds of insects appear to 
depend almost exclusively on sight in their visit to flowers, and it is 
in relation to this sense that we refer the development of petals from 
stamens. But many flowers are so very small that even with their 
petals fully developed, they are still inconspicuous objects. Now 
most of the earliest flowers seem to have been produced in tho axils 
