CO-OPERATION AND THE ORIGIN OF FLOWERS. 149 
questions for a local Natural History Society. That is not 
my Opinion; reference will be found throughout this paper to 
important points on which local observations are very much 
required. Indeed, I hope that some of our members may be 
induced to carry out observations either in the field or in 
their gardens. 
For the most primitive type of flower, we will accept 
the evidence and description given by Messrs Arber & 
Parkins in a valuable paper with illustrations.° The most 
archaic orders agree in the following points. There is no 
clear distinction between sepals and petals, and there is an 
indefiniteness in their number. The parts of the flower (or 
some of them) are attached to the axis spirally, not in circles 
as in all the more usual modern types. Indeed in 
Calycanthus, there is a continuous spiral attachment of 
sepals, petals, stamens, and carpels. All these flower leaves _ 
were free and independent of each other; there was neither 
cohesion nor adhesion. The flower-axis was more or less 
elongated with internodes between the insertions of the 
various parts. The pollination was carried out by insects, 
and some of the perianth parts were, or rather became, more 
or less coloured to attract insects. 
In the paper above referred to there is continual reference 
to a possible Cycadlike ancestor allied to the fossil Bennet- 
tites. 
Others have tried to show that the ancestor was either 
a tree or, at any rate, a woody shrub or undershrub (Sinnott 
& Bailey).6 But, as a matter of fact, many genera to-day 
contain both herbaceous and woody species; plants usually 
herbaceous in our climate become woody perennials at the 
Cape of Good Hope. Then again, is a Birch or Oak seedling 
in its first year a herb or a shrub? 
We may perhaps assume that the original ancestor had 
a rosette habit (Cycadlike), but it may quite well have been 
something very insignificant; it had two cotyledons 
It may also have had a single terminal flower.” As to 
the time when, it is clear that in the Neocomian or Aptian 
strata at the base of the Cretaceous era, true flowering 
