species of Panurginus, Perdita and Andrena. These bees have a weak flight and are not fitted to travel long distances. It is known that in some instances they build their nests near the flowers they visit. Probably this generally true. The medium-sized oligotropic bees, belonging to the genera Colletes, Epeolus and Melissodes, fly in the fall and visit chiefly the Compositæ, a family which offers a wide choice of flowers. It is not always easy here to draw the line between an oligotropic and a polytropic bee.
There are still in existence many intermediate stages between monotropic, oligotropic and polytropic bees. While many bees visit a great variety of flowers, others visit only one family, as the Compositæ or Nymphæaceæ, others only a single genus, as Salix, and others only a single species, as the violet, strawberry or spring-beauty. Many exceptions no doubt occur and will be recorded when the habits of these bees have been more carefully observed. For instance, I have often seen the loosestrife bee on the umbels of the prickly sarsaparilla. It is evident that if a monotropic bee extends into a region, where the flower it visits elsewhere does not occur, it must of necessity visit other flowers. Dr. Graenicher writes me that the pickerel-weed bee (Halictoides novæ-anglicæ) is found in Wisconsin; but the pickerel-weed does not flourish in the same locality, and so this bee is compelled to visit the blossoms of other plants. Evidently this habit did not originally exist among bees, but has gradually been acquired.
We may sum up the matter as follows. All bees including the honey-bee show a strong tendency in collecting both nectar and pollen to be constant to one species of flower. This is manifestly for the advantage of both insects and flowers. In the case of a number of bees flying for only a small part of the season this habit has become so specialized that they visit only one or a few allied species of flowers, which offer an abundance of pollen and nectar. Primarily it seems to be the direct advantage gained rather than the avoidance of competition that has led to the rise of the oligotropic habit. As the honey-bee for a time restricts its visits to the white clover, so in like manner a monotropic bee visits but a single kind of flower. But in the former case the bee flies throughout the whole season, but in the latter when the flower fades the bee's period of flight is over.
The idiosyncrasies of bees in visiting flowers present many remarkable peculiarities, and undoubtedly offer an attractive field for observation. There are certain bees, which though they are not oligotropic obtain the larger part of their supplies from comparatively few flowers, as the plums, thornbushes, cornels and Viburnums. In this locality one of the leaf-cutting bees (Megachile melanophæa) shows a decided preference for the purple vetch (Vicia cracca), and if I desired a specimen I should look for it on the blossoms of this plant. As the parasitic bees do not provide stores for their brood and seek nectar for themselves alone, they show little preference for special flowers. For a simi-
