most nitrogenous food, and near the flowers the nitrogen-catchers are situated.
Another peculiarity of P. arguta lies in its flowers, which are clus- tered in large and consj^icuous masses, and have petals that vary from pale yellow to primrose or almost white. This is a very interesting fact, because the native color of the potentillas is yellow; but the mountain species, and many other kinds, have varied to snow-white blossoms; and here we get a plant, as it were, in the intermediate or undecided stage between the two colors. Notice, too, that P. arguta is an herb of the rocky hill-sides, and therefore half-way toward be- coming a mountain species.
Now, on the summit of Mount Willard, just above the Notch of the White Mountains, I found another very beautiful member of this pretty group, the three-toothed cinque-foil (P. tridentata). This is one of your most northerly and mountain-loving potentillas, unknown in Europe, inhabiting the coast of New England from Cape Cod north- ward, and the mountain-tops of the great chains, from the Alleghanies to the Maine ranges, as well as in Canada, Labrador, and the extreme north of the continent. The three-toothed cinque-foil carries a step farther the same characteristic, for its flowers are pure white, as so often happens with mountain blossoms. Just in the same way, while almost all lowland buttercups are golden yellow, some of the Alpine buttercups are white as milk, and among these very potentillas there are a few lovely snow-white mountain species in Europe and Asia. One beautiful kind that I gathered on the Maritime Alps at Mentone (P. saxifraga) has a blossom as delicately mountainous in type as the saxifrages themselves, from which it takes its scientific name.
Of course, I don't for a moment mean it to be understood that I think P. tridentata is directly derived from P. arguta, or that the latter species is now on its way to merge into the former. My Mount Willard plant has palmate leaves of only three leaflets, while the com- mon P. arguta of the northern hill-sides has pinnate leaves of from three to nine cut-edged divisions; and in many other technical points they differ widely from one another. All I mean to suggest is merely that the yellowish-white P. arguta is now just passing through a stage which the ancestors of P. tridentata must have passed through long ago. On the whole, to put it briefly, the potentillas are a yellow lot; but a few advanced members of the race are white; and still fewer, like the ornamental P. nepdlensis and P. atvopurprirea of our gardens, are crimson, scarlet, or bright red. So far as I know, no potentilla is ever blue, which is the highest level of floral coloration.
The three-toothed cinque-foil has an almost shrubby and woody root-stock, and displays a tendency to assume the character of a true shrub. Biit its northern habitat and mountain manners keep it low and tufted, after the common fashion of upland vegetation. There is another of its kind, however (P. fridicosa), which really grows into a