148 CO-OPERATION AND THE ORIGIN OF FLOWERS. 
large maxima of flowers. Here is a subject well worth 
investigation by younger naturalists of this Society.* 
Within limits, every insect seems always to find its daily 
honey. ‘Those factors which induce a flower to bloom seem 
also to govern the appearance of the necessary insect. { 
In the great Composite order, there is marked co-opera- 
tion between the flowers of the same plant. All the florets 
are massed together; thus the insect can visit a maximum 
number in the shortest possible time and with the least trouble 
to itself. 
Insects are often industrious: Bombylius Discolor has 
been observed to suck 34 separate flowers of Pulmonaria in 
one minute.! Moreover, even volatile or stupid insects, 
if they have once alighted on a Composite head are pretty 
sure to visit several flowers and may be therefore of some 
use. So it is not surprising to find that the insect clients of 
Composites are not only numerous but varied. The Dandelion 
has 111 species of insect visitors,? and other common flowers 
of this order are sought by almost a hundred distinct and 
separate kinds of insects. 
There are many other advantages in the Composite 
arrangement; the buds are easily protected; less material is 
required for forming pedicels or protective sepals, and so on.* 
In fact the Composites are a very successful order. Bentham 
& Hooker says of them—‘‘ Ordo omnium vastissimus.’’ The 
number of known species of Composite is given by Uphof 
as 14,324; the order is found all over the world and io practi- 
cally every association.4 
What was the origin of the capitulum of Composite? 
This obviously leads to a very interesting question. What 
was the original flower?{ What was the primitive 
inflorescence? Where, when, and in what sort of climate 
did it grow? 
It might perhaps be thought that these are not suitable 
* More observations are required here (first series). 
+ Here again more observations are required (second series). 
¢ In this paper by ‘‘ flower ’’ is to be understood ‘‘ angiosper- 
mous. ”’ 
