of Dutchman's pipe, Bryophyllum, and Cirrhopetalum, and the spathes of the skunk-cabbage; the glittering, dew-like drops of butterwort and the unicorn plant occur again in the false flower-glands of Parnassia and the deceptive, sparkling ovary of Paris; and that disagreeable odor is a common characteristic in both classes strengthens the belief that in both carrion-loving flies (or beetles) are the objective points of the attraction, with this unessential difference that in the one class the plant feeds the flies, in the other the flies feed the plant.
VI. Attractive Color.—Leaving aside the negative evidence derivable from cleistogamy and the existence of only inconspicuous flowers in places where insects and flower-frequenting birds are absent, it remains to prove that attractive qualities actually have reference to the visits of animals, thereby establishing their usefulness—i. e., their eligibility as characters upon which natural selection may work.
Where there are bright flowers there are color-minded animals. All orders of insects are represented to an altitude of 2,300 metres, butterflies, flies, and certain bees even to 4,600 metres. The number and kind of insects are in close relation to the number and kind of flowers and their hours of waking and sleeping. Climate affects the color of all parts of a plant. Cereal grains are said to be brighter in the North. Fruits are invariably so. Many travelers have observed the intensity in the color of Alpine flowers up to certain limits of temperature. But if the brighter color is useless it will not be retained after a few generations, as the inconspicuous character of the flora of insect-poor Greenland shows; if serviceable, it will be not only preserved, but deepened as time goes on. The absolute number of flowering plants decreases with increase of latitude or altitude. As men flock to cities until the average compensation becomes equal to less than what they can obtain in the country, so all insects would stay on the plains or in the tropics until their number, becoming disproportioned to that of the flowers, better rewards can be obtained in less crowded regions. To this interdependence of insects and plants, and to the constancy of the numerical relations between the two, inherited intensity of color must be largely due. If the insects are greatly in excess of the attractive flowers, inconspicuous and conspicuous blossoms would be searched and fertilized alike in the resulting scarcity of food. If the number of flowers is much larger than that of the insects supported by them; the latter, becoming fastidious, frequent only the brightest or more fragrant, neglecting the others, which accordingly remain inconspicuous and self-fertile. When both insects and flowers are scarce, the former will remain only so long as the greater attractiveness of the latter makes it as easy to obtain the same amount