FERTILIZATION OF FLOWERS BY INSECTS. 291 
` the stomach, it appeared very large and was full of a yellow sub- 
stance which consisted of hundreds of thousands of pollen-grains. 
T have had since then many opportunities to observe this eating of 
pollen, not’ only in all the species of Eristalis, but also in the 
genera Rhingia, Syrphus, Volucella and Scatophaga. This chew- 
ing of pollen alternates with sucking honey if the flowers have 
any, and I am of the opinion that the singular structure of the 
proboscis of flies cannot be fully explained without taking into 
account its double function of sucking honey and eating pollen. 
In the Tipularize and also in those flies which do not eat pollen but 
live exclusively upon juices, for instance, Bombylius, the two 
valves of the proboscis serve no other purpose than to protect 
and guide the sucking tubes, but in the flies which devour pollen 
besides this function there is also that of grinding the pollen, for 
which they have special adaptations, for the margins of the two 
valves at the point of union are transversely dentate with fine and 
parallel bands of chitine. Probably the greater or less distance 
of these bands in different species is related to the different size of 
the pollen upon which they feed. 
Since the proboscis of the Tipulariz often possesses one simple 
function and has in accordance with that a very simple organiza- 
tion, we may consider these Tipulariz as the most ancient branch 
of the stock from which Diptera are derived. A fact casually dis- 
covered by me and of which I find no mention hitherto, seems to 
me of great importance in the systematic disposition of this order. 
In the spring of 1868, while engaged in examining the head of a 
gnat, with a view to ascertain whether or not the valves of its 
proboscis had the transverse bands of chitine, I was surprised to 
discover that the proboscis and palpi were clothed with scales en- 
tirely like those of butterflies. 
I find no mention of this important fact in the special works 
of Meigen and Schiner which are in my possession. Meigen 
simply points out that in Culex, Anopheles, and Corethra, scaly 
productions are observed on the venation of the wings, and he 
figures some of them which, however, being quite narrow and 
ae two sharp points, have no analogy with real lepidopterons 
The gnat-scales observed by me and accurately 
ae TD closely resemble the most characteristic lepidopterous 
scales. They suddenly dilate from a short and narrow peduncle 
to a large scutiform surface which is traversed longitudinally 
AMER. NATURALIST, VOL. V. 19 
