instance, in the fly-flora of New Zealand and the bird-flora of Juan Fernandez. Indeed, M. Sevali claims that grasshoppers fertilize the Leguminosæ of New Zealand, and Delpino thinks that Rhodea, which is self-sterile, depends for the production of seeds upon snails!
In order that the selection of insects may cause change in the characteristics of flowers, two things are necessary on their part. Their visits must be methodic—we have seen that they are so for the most highly specialized groups. They must be frequent. Any one who has followed a "busy bee" for half a day will be ready to witness that they are. A bumble-bee in mid-Sweden was seen to suck honey from the monk's-hood at the rate of from 960 to 1,200 visits an hour; a butterfly visited 194 violets in six minutes and three quarters; 2,155 bees were actually counted on a single head of the "honey plant" between 5 a. m. and 7 p. m., the thirty heads of one plant furnishing supplies for over 64,000 bees in one day. Some one has calculated that 2,500,000 visits are made to the red clover for every pound of honey. The United States, by the census report of 1880, produced 25,743,208 pounds, representing, therefore, 64,358,020,000,000 visits of hive-bees alone in this country, and this includes only that used for economic purposes, not at all that kept by the bees themselves. Add to these visits those of the wild bees, bumble-bees, flies, butterflies, birds—which are by no means indiscriminate—and surely here is a force whose selective influence must be enormous, which might easily bring about a comparatively rapid evolution. May not this shed some light on the mystery of the rapid floral development of Phanerogams after their initiation in geologic ages? Gardeners and farmers have, in relatively short time, been able to introduce and establish new forms of flowers, fruits, and grains, but the results of the industry of this vast army of workers must have been inconceivably larger.
This brings us to the other side of the question. Something is necessary on the part of the flowers, which must themselves be capable of great variation in color and form in order that selection may have material to choose from. Every one knows that they are so. As the latest production of the vegetable world they are the most plastic, the most easily influenced by alterations of the environment. (Old traits are not easily changed.) Many of them in their individual development pass through a series of changes which sometimes, perhaps, represents the color phylogeny of the species—e. g., several species of honeysuckle are white the first day, become yellow the second, wither the third. A certain hibiscus is white in the morning, rose-color during mid-day, and red in the evening, repeating these transformations each day as long as the flower lasts. Such changes are supposed to be