1899] FLOWERS AND INSECTS 35 
A bee may be regarded as oligotropic: (1) When the female 
collects the pollen of the plants in question and is not known to 
collect pollen of any other plants. (2) When the bee does not 
occur except during the blooming season of the flowers. If the 
female is shown to occur after the flowers have quit blooming, 
the case is very doubtful. (3) When the bee is frequent upon 
the flowers, and more or less rare upon other flowers, at any rate 
except inthe neighborhood of the food flowers. The case is also 
doubtful if it is shown that the distribution of the bee extends 
greatly beyond the plants upon which it is supposed to depend. 
~ In the more satisfactory cases, if any one should say that he 
had observed the bee collecting pollen from a quite unrelated 
flower, I would not accept the determination, or, if that were 
beyond question, the opinion that the pollen came from the 
flower on which the bee was taken. 
Below I give a list of bees which I regard as oligotropic in 
the above sense. When I have observed the female collecting 
pollen from more than one species of a genus, I give the genus; 
when from more than one genus, I give the family. The details 
will be given elsewhere. 
In Prosopis the females are destitute of pollinigerous appa- 
ratus, their nests being provisioned with a paste of honey and 
pollen. I know of no way to distinguish the flowers which the 
females visit for this purpose from those which they visit in only 
an incidental way, so. I assume that a species of this genus is 
oligotropic only so long as it is found exclusively on flowers of 
One species or group. This may be assumed for either sex so 
long as the condition holds, as in case of P. zllinoensis, of which 
I do not know the females. P. nelumbonis has always seemed to 
me to be the best case of an oligotropic Prosopis. 
I have never believed that our species of. Epeolus were 
cuckoos of Colletes, because there are more common species of 
the former than of the latter genus, and their phenological posi- 
tions do not show the same correlations which exist between 
Andrena and Nomada, Megachile and Coelioxys. Besides, the 
Maximum of Epeolus does not approximate that of any other 
