106 The American Naturalist. [February, 
Aculeate Hymenoptera (Fig. 25, Plate IX), the butterflies 
(Fig. 23, Plate IX), and the flies (Fig. 22, Plate [X)—all on the 
scale of five species to the millimetre. No curve is made out 
for the whole because these curves agree in showing a 
maximum for July, which, of course, would determine the 
position of the general maximum.’ The bees are by far the 
most important, since they depend upon flowers both for 
their own food and for that of their young. Asa rule, except 
in the case of the cuckoo bees, which lay their eggs upon 
food deposited by the host bees, the female bees are provided 
with brushes of hair upon which they carry pollen, the essen- 
tial part of the bee-bread, upon which the larve feed. 
In a previous examination of the curve for the Choripetalae 
(Fig. 2, Plate VIII) there was observed a maximum in Au- 
gust, a secondary elevation in May, and an intervening depres- 
sion in June. 
With the principal exception of the Leguminosae (Fig. 15, 
plate VIII), these plants have horizontally expanded regular 
flowers, with readily accessible nectar and stamens exposed so 
that the pollen is easily collected or eaten. The Leguminosae 
generally have lateral irregular flowers, with the nectar con- 
cealed and deep-seated, and intricately concealed pollen, for 
which reason they will be separated for special consideration.. 
Now, since the maximum for the Choripetalae coincides with 
that of the Leguminosae, the separation of this family will 
change the maximum of the group to the secondary point. 
There are two families of insects which are particularly fond of 
simple flowers with easily accessible nectar and pollen—the less 
specialized bees (Andrenidae, Fig. 26, Plate IX.—2 spp. per 
mm.) and the flower-flies (Syrphidae, Fig. 36, Plate X.—2 spp. 
per mm.)—and they both have more species flying in early 
spring. There is no question but that the strong pre- 
dominance of the more simple Choripetalae is, to a great ex- 
tent, correlated with the early predominance of the Andreni- 
dae and Syrphidae. The flowers of the buttercup family 
: (Ranunculaceae, Fig. 9, Plate VIII) and of the Rosaceae (Fig. 
5 My phaenological observations are most defective for August. I expect 
to find the maximum of the general anthophilous insect fauna a little later. 
