the Mediterranean to the shores of the Atlantic, and, crossing the ocean with seed-corn and fodder crops, have clogged the steps of the intrusive white man through all his colonies and settlements elsewhere. These cosmopolitan weeds succeeded in America to the soil once covered by forest-trees, whose indigenous undergrowth could not stand the garish sunlight of the open clearings. But nowadays, the weedier types of the Western prairie-belt are moving eastward, as farms move west; and being accustomed by nature to open plains, they will probably, in many cases, succeed in establishing themselves side by side with the older plagues of the long-suffering farmer. Potentilla paradoxa is one of the first crop of these weedy immigrants, and its appearance already on the shores of Lake Ontario is the signal for its future advance in a formed phalanx against the tilled fields of New York and New England.
This Western immigrant departs widely in one respect from the type of all the potentillas we have yet considered, and that is in the arrangement of its five, seven, or nine leaflets. In the true cinquefoils, and all their like, the leaflets are arranged, as we say, palmately—that is to say, all start together, like the lobes of a horse-chestnut leaf, from one point. In the P. paradoxa they are arranged pinnately—that is to say, they start in opposite pairs or singly, from a common midrib, like the barbs of a feather or the leaflets of a locust-leaf. The same arrangement, a more convenient one for long leaves, reappears in P. Pennsylvanica, which (in spite of the name incorrectly bestowed upon it by Linnæus) is a Northwestern species. But as I have not seen this last-named plant in the living state, and as I do not like to write about what I have only examined in a dried-up herbarium (a bad habit of the old-fashioned, purely structural botanists), I will say no more at present about it.
On the rocky hills of the North and West there occurs in July a rather pretty, half-shrub-like potentilla (P. arguta), which presents several other interesting peculiarities. This plant has brownish, hairy stems, covered with a viscid, clammy exudation, something like that which covers the young branches and buds of the clammy rose acacia (Robinia viscosa). As I observed that insects are often caught in this clammy secretion, exactly as in the case of the common catchflies (Silene noctiflora Virginica, regia, etc.), I have not the least doubt that the potentilla eats and digests the creatures it entraps, in order to supply it with nitrogenous material for its own pollen, ovules, and seeds. This is the more probable, as the clamminess increases near the flower-buds and blossoms, and is scarcely at all noticeable near the base of the stem. How the potentilla digests its food I do not know, but long observation has fully convinced me that whenever a plant has viscid, glandular hairs or secretions upon its penduncles, pedicels, calyx, and flower-buds, it is invariably an insect-catcher, and an insect-eater too. The flowers are the part that require the
