4 
462 JOURNAL OF SCIENCE. 
the ants are not only securely housed by the plant, but are provided 
with a bountiful supply of food.” This consists of a crater-like gland 
at the base of each pair of leaflets, which, when the leaves are young, 
secretes a honey-like liquid. While at the end of each of the small 
divisions of the compound leaflet a small yellow fruit-like body is 
developed, of which the ants are very fond, and which they carry off 
to their nests as soon as ripe. Here is a case of a plant paying 
black-mail to a colony of ants, and thus securing immunity from all 
other robbers. 
But besides keeping off creeping insects, flowers have often to 
keep off unsuitable kinds of flying insects, and though this is some- 
what forestalling my subject, I will refer to it here. Thus some 
flowers have developed large quantities of hair inside their petals, and 
this must render them rather obnoxious to many insects. One set 
have got their nectar in long tubes, and s0 are only visited by insects 
with long probosces; others have their nectar on flat discs, and a 
butterfly or moth would have about as much chance of making a meal 
off such a flower, as the stork had when the fox invited him to dine off 
the plate. Some, like our common frogsmouth, keep their lips so 
tightly shut that none but strong bees can get at the honey; while 
others produce nectar of such peculiar quality that if the wrong kind 
of insect visits them, it is intoxicated or even poisoned. 
Now I want to go back to our primary flowers at the stage at 
which we had left them, only you will notice that in this subject we 
are constantly liable to go off at a tangent and follow up side lines of 
evolution. The earliest differentiation of flowers then, appears to 
have been the production of either coloured petals from the outer 
stamens or of coloured sepals or bracts, of nectar and of scent, but in 
most cases only one or other of these developments was followed. 
But confining my remarks now to dicotyledons, | may state that very 
early two main lines of development were distinguishable, and our 
modern systems of classifying flowering plants are natural just so far 
as they recognize this fact. In one group the sepals became coloured 
or remained green and no petals were developed, and these constitute 
the group now termed Apetale ; in the other petals were developed. 
This group, showing as it did, the greatest power of differentiation, 
has more recently divided into two, one division termed the 
Polypetale retaining their petals all separate, while in the other,— 
which we may consider as the most highly developed group of 
flowers, the petals became united into a tubular corolla, forming a 
large division termed from this circumstance Corolliflore. The 
Polypetale again have divided into three groups; in the first of these 
all the parts are separately attached to the axis, that is to say the 
sepals stand by themselves, the petals by themselves, then the 
stamens, and lastly the pistil or carpels; this group is called the 
Thalamiflore. Ina different set called the Disciflore, a more or less 
well-detined dise or cushion has been developed on the bottom of the 
flower under the ovary, and on this the petals and stamens are placed. 
While in a third sct, the petals and stamens have become more or 
less closely attached to the calyx, and they constitute the group 
* Calyciflore. The accompanying table will help to show the lth 
genealogy of these groups— 
