294 FERTILIZATION OF FLOWERS BY INSECTS. 
different species, this or that part of the body has undergone 
special adaptations so as to be able to collect pollen with greater 
ease and in greater abundance. A sure criterion by which to com- 
prehend these differences thoroughly as well as to estimate cor- 
rectly the different grades of affinity between forms so varied, can 
be given only by the Darwinian doctrine. But we have no space 
_ to particularize. 
We will conclude by discussing some objections which can be 
urged against the explanations of facts, and against the general 
principles advanced in this discourse. 
It may be asked, what advantage can flowers and insects derive 
from having elongated respectively the melliferous tube and the 
proboscis instead of having them remain of a constant length? I 
answer that in order to comprehend the advantage of this elonga- 
tion, it is necessary to consider in one view the benefits and the 
injuries which different tribes of insects bring to plants. The Lep- 
idoptera are the only insects which, while aiding the plants by 
transferring pollen from one flower to another, do not cause injury 
by devouring the pollen. Therefore a plant which has modified its 
flowers so as to exclude bees and flies while admitting Lepidop- 
tera, has obtained a signal advantage. 
Suppose a plant develops a floral tube longer than usual so that 
the honey remains at a lower level; this variation will be an ad- 
vantage for that tribe of insects which lives on honey alone and 
can therefore adapt itself more diligently to this variation. The 
advantage in this case is for the Lepidoptera and will last until the 
proboscis of certain bees and flies equals that of the Lepidoptera. 
When this equalization has been completed, a further elongation 
of the floral tube will be useful to the plant, which will immediately 
be followed by a corresponding elongation of the proboscis of the 
Lepidoptera, and so on. In this way, by means of the rivalry be- 
tween the Lepidoptera, bees, and flies, the fact that the Lepidoptera 
do not consume pollen and can sooner adapt their proboscis to the 
variations of the flowers than their rivals coéperating, a gradual 
augmentation in the length of the tubes and spurs of flowers would 
become established, followed by a proportional elongation of the 
proboscis, concomitant in the Lepidoptera, later in bees and last of 
all in flies. It may be well to give in this place the measure of the 
longest proboscides of some of the Lepidoptera, bees and flies of 
our country. 
