FERTILIZATION OF FLOWERS BY INSECTS. 293 
there is a close relationship between them and Lepidoptera, a rela- 
tionship which is further attested by the lepidopterous appearance 
of the genus Psychoda, the tipulaceous habits of Pterophorus, the 
similar venation of the wings in many Tipulariz (Limnobia, Cten- 
ophora,) and the Phryganeide, the aquatic habitat of the larvae 
of the Tipul, and, finally the circumstance that it is far easier to 
deduce morphologically the proboscis of the Tipule from the buc- 
cal ae of the Phyganeide than from those of any other order 
of inse 
dione according to my opinion, the stock or kindred com- 
mon to the Diptera, Lepidoptera, and Phryganeide, in its manner 
of life, and the structure of its body would be very closely allied to 
the Phryganeidee of to-day, living in water in the form of sheath- 
bearing larvee, and in the perfect state remaining in the vicinity of 
the water. Its posterity divided at first into two branches, to wit, 
the conservative one par excellence, of the Phryganeide, which con- 
tinuing in the same mode of life as its ancestors, has undergone 
very few variations ; and the branch of those insects which suck 
the honey of flowers, which have gradually removed from their 
aquatic abode, have developed by natural selection the sense of 
colors, and acquired through sexual selection a squamose cover- 
ing. This second branch again divided into two, one of which 
accustomed itself to feed exclusively upon the honey of flowers 
and produced Lepidoptera; while the other, less exclusive in its 
tastes adapted itself to imbibe all sorts of fluids as well as to 
pierce the more tender tissues, and produced the Tipulariz. One 
part of these besides sucking different juices, grew accustomed to 
eating pollen and thus little by little the proboscis of the Tipulæ 
was transmuted by natural selection into that of flies equally well 
adapted to suck honey or eat pollen. 
The Hymenoptera which visit flowers, the bees, being given 
exclusively to floral food not only in the perfect state but also 
while larvæ, present the greatest possible variety of adaptation. 
Starting from the mouth of the fossorial Hymenoptera adapted 
only to bite and provided with a very short tongue, we arrive, 
through numerous transitions, to the highly developed proboscis of 
the Anthophore and Bombi which can extrude their tongue to a 
length equal to that of their body, and then coiling it up, draw it 
back again into its cavity so as to give free play to the action of 
the mandibles. Furthermore, in different ways, according to the 
