muralis, which were exposed for eighteen months in the drier, after a few wettings resumed growth in all their parts. Other species of Barbula behaved similarly. A curious experiment was performed with Grimmia pulvinata, in which a stock which had been cultivated for some time in a moist atmosphere under a bell-glass was suddenly exposed to a warm and perfectly dry current of air. It became so dry in a short time that it could be pulverized. Then it lay in a drier for ninety-five weeks. But the quickening moisture was still competent to awaken it to renewed life. The most rapid drying which could be performed in the laboratory could not destroy the plant. It even showed greater power of resistance than would correspond with its real necessities, for so speedy and complete a drying out as was effected in the experiments never occurs in Nature. The fact that a property acquired by adaptation is so plainly manifested in excess is sometimes otherwise demonstrable, and is a hard problem for the theory of selection.
Those mosses which are not capable of drawing water in considerable quantities from the soil, are yet able to make the best use of the smaller quantities with which they are moistened. For this object their stems are furnished with provisions for the capillary distribution of the local water with which they are in contact. This capillary "outer water-conduit" was perceived several years ago in various mosses by C. Schimper; and it has recently been more closely studied with the aid of colored solutions by Fr. Oltmann.
The capillary spaces in which the water rises or, more generally speaking, diffuses itself, exist in different forms. In the simplest cases the leaves supply them; and of this kind there are, according to Oltmann's comparisons, several types. Thus in Hylocomium loreum, Hypnum purum, and similar forms, the leaves are so shaped and arranged with their opposite sides in close contact as to form a hollow cylinder around the stem, which is composed in its interior of a system of connected chambers. When enough water is present, the capillary space between the stem and the leaf is quite full; in other cases the water ascends only between the overlapping leaf-edges. In Plageothecium undulatum, Neckera crispa, etc., the leaves lap like shingles; in other cases they are small and thickly packed, so that a whole system of narrow capillary spaces is generated between them. The frequently observed phenomenon of the drying leaves erecting themselves and lying close to the stem, with wrinklings and curlings, involves, as Oltmann has remarked, an increase of capillary space. By these means the water, when a wetting takes place, is diffused more readily and completely over the surface of the plant.
In another series of cases, the capillary apparatus is formed by
