flowers. They fairly monopolized the staminate flowers, W 
40 BOTANICAL GAZETTE [JULY 
habit of sticking the pollen with honey, and so can use that of 
anemophilous plants. 
Those flowers, however, which, through their nectar and cor 
related modifications, were the best fitted to use the services of 
ordinary insects for cross-pollination, were the least fitted to | 
utilize the insects which were the highest product of anthophilow’ 
development. Strange as it may seem, the characters which hit- 
dered them from availing themselves of these services were the 
very characters which are considered the highest adaptations for 
cross-pollination, viz., diclinism, dichogamy, and large size. Ob | 
the other hand, the forms which have enabled flowers most readily 
to avail themselves of the services of bees are the very charac 
ters which have been interpreted as adaptations for self-polline 
tion and geitonogamy, viz., small size, homogamy, and the — 
aggregation of dichogamous and other flowers in close clustets. | 
If an insect in search of nectar visits a dicecious or othe 
diclinous plant, it is not hard to understand how it is likely t® 
the pistillate flowers were visited by an entirely different set 
insects. In the table there are six species of bees which 
their pollen exclusively from dicecious species, Salix and Spi 
Aruncus. Of the plants furnishing pollen to oligotropic — 
these are the least able to utilize these bees on account of thet 
dicecism. 
Dichogamous flowers are at somewhat of a disadvantage 
utilizing pollen-collecting bees from the fact that the bees 4 
more apt to pay attention to the flowers which are discharging. 
pollen and neglect those in the other stage. In Jmpatiens ful : 
and J. pallida | have observed that Megachile brevis collects * 
pollen from flowers in the first stage and avoids those * 
