32 CHELTENHAM COLLEGE NATURAL HISTORY SOCIETY, 
power of movement. In the unripe condition they hang down, and as 
they debisce they stand up in turn in the path of the bee’s head, and 
when they are withered sink completely down again. When they have 
all had their turn, the carpels follow suit, and cross-fertilisation cannot 
fail to take place if a bee goes from a young to an older flower. 
In the Ranunculacee, then, we see that separate divisions have de- 
veloped very distinct advantageous modifications. Conspicuousness 
is ensured in Ranuncu/us by the petals, in Anemone and Ca/tha by 
the sepals, by both in Agurlegia and Delphinium, and by the stamens 
in Thalictrrum. Honey is wanting in Anemone and Zhalictrum : it is 
secreted by the petals in Ranunculus and Delphinium, and by the car- 
pels in Ca/tha. In some species it les patent, and in others it is 
deeply hidden. Easily accessible honey has the advantage of securing 
abundant insect visits, but the objection that the insects creep about 
in various ways, and also waste honey and pollen without fertilising 
the flower. Honey deeply placed shuts off the multitude of short- 
lipped insects, which is a disadvantage unless it attracts enough bees, 
etc., with their long proboscises, which move in a particular way, and 
ensure cross-fertilisation. And here we may opportunely note the other 
advantages that the visits of bees bring to a flower. The chief one is 
certainly that extreme diligence which struck Dr. Watts. Most of the 
flies waste a lot of time, and I have noticed even the highly developed 
drone flies hovering for some two minutes over a flower before alight- 
ing, whereas the bees always work steadily. Another thing is that the 
bees will only visit a single species on a single journey. I have watched 
them on a plot of seedlings, and noticed a line of flax entirely visited 
by one species of Aoemdus, and the next line of C/arkia by another, 
while the hive bees visited all alike, but no individual crossed from 
one line to another. This of course means a great saving of pollen, 
and the plant is not liable to be injured by extraneous pollen. 
We must not leave this order without noticing one fact which 
militates against the absolute superiority of cross-fertilisation. The 
cross-fertilised species are not necessarily the most abundant. The 
buttercup and celandine, in spite of frequent self-fertilisation, are far 
commoner than the columbine or larkspur. And we shall find this to 
be the case in most orders. It is only where nearly allied species come 
into close conflict that the self-fertilised one goes to the wall. The very 
fact of a species existing in spite of self-fertilisation probably indicates 
a hardy stock, or it may have other means for securing a change of 
condition, such as dispersion of seeds by wind or birds. ‘lhe time of 
development of plants has been enormously long, and the complexity 
