1899 ] FLOWERS AND INSECTS 29 
another flower which she does visit for that purpose, and which, 
therefore, holds a more important relation to her species. Accord- 
ingly, I propose to consider those examples in which the female 
collects pollen of one species, or several species of the same 
genus or natural family, the relationship of the plants being such 
as to give significance to the cases. On the other hand, if a bee 
uses the pollen of only two plants of different families, I assume 
that it is essentially polytropic, and that the few visits are merely 
connected with the fact that it is rare or has a short flight. Of 
course there still remains a strong contrast between the visits of 
a bee which flies only a month or two and one which flies through- 
out the season. As a rule, if a bee has a long flight it must be 
regarded as polytropic, unless the flowers on which it depends 
have a long blooming time. Of the thirty-nine species of Halic- 
tus and the allied Augochlora and Agapostemon, I regard only 
one as oligotropic, Halictus nelumbonis. {t has a comparatively 
short flight, while the blooming seasons of the Nymphzacez are 
long. When a genus of plants has more than one closely allied 
species, the difference between a monotropic and an oligotropic 
bee may depend merely upon the accident that only one species 
occurs in the neighborhood. My observations show that an 
oligotropic American bee will gather the pollen of a closely 
related introduced European plant of the same genus. 
The relations of the host-bees to the flowers from which they 
get pollen are quite analogous to the relations of parasites to their 
hosts, of phytophagous insects to their food plants, or of pre- 
daceous insects to the insects upon which they feed or with which 
they provision their nests. How the bees maintain these rela- 
tions is much easier to understand, since the flowers are modified 
in such a way as to facilitate their visits. 
Any ecological position is of advantage only to a limited 
number of individuals. As soon as this optimum number is 
passed, anything which will enable a set of individuals to get 
along without coming into competition with the dominant form 
will be of advantage to them, and their preservation will depend 
"pon their adopting this course. A characteristic which would 
