distinct reference to the habits and tastes of the fertilizing bees. It is a mountain plant by origin, belonging to a tribe which took its rise among the great central chains of Europe and Asia, and these Alpine races are usually highly developed in adaptation to insect fertilization, because they depend more absolutely upon a few upland species than do the eclectic flowers of the plains, which may be impregnated haphazard by a dozen different flies, or moths, or beetles. We can still dimly trace many of the links which connect it with very simple and primitive buttercups, if not directly, at least by the analogy of other plants. For all the buttercup tribe show us regular gradations in the same direction. The simplest kinds are round, yellow, and many-carpeled, like the buttercups. Then those species which display their sepals largely have dwarfed petals, like hellebore and globe-flower, or have lost them altogether, like marsh-marigold, which trusts entirely for color display to its big golden calyx. The still higher anemones have the sepals white, red, or blue; and the very advanced columbine has all the petals spurred, and developed into nectaries, like those of monk's-hood. But columbine still keeps to single terminal flowers, so that here the five petals remain regular and circularly symmetrical, though the carpels are reduced to five. Fancy a number of such columbine-flowers crowded together on a spike, however, and you can readily picture to yourself by rough analogy the origin of monk's-hood. The sepals would now become the most conspicuous part; the two upper petals would alone be useful in insuring fertilization, and the lower ones would soon shrivel away from pure disuse. The development of the hood and the lengthening of the upper petals would easily follow by insect selection. It is a significant fact that our only other spiked buttercup, the larkspur, has equally irregular and bilateral flowers, though its honey is concealed in a long spur formed by the petals, and accessible to but one English insect, the humble-bee.—Knowledge.
By M. W. SPRING,
OF THE UNIVERSITY OF LIÉGE.
VIEWED in relatively shallow masses, clear water appears wholly colorless. In our daily dealings with the liquid we seldom have occasion to observe it in great depths; hence it has been generally believed that water is quite destitute of color. The ancients were accustomed to explain the transparency of some bodies by assuming that they partook of the nature of water; and we now speak of a diamond as of the first water, to emphasize its perfect transparency and colorlessness. If, however, we regard the larger masses of water in