as "elaters." They move with great vigor, and probably serve the same purpose.
In much more numerous cases, seeds are carried by the wind. For this, of course, it is desirable that they should be light. Sometimes this object is attained by the character of the tissues themselves, sometimes by the presence of empty spaces. Thus, in Valerianella auricula, the fruit contains three cells, each of which would naturally be expected to contain a seed. One seed only, however, is developed, but, as may be seen from the figure given in Mr. Bentham's excellent "Handbook of the British Flora," the two cells which contain no seed actually become larger than the one which alone might, at first sight, appear to be normally developed. We may be sure from this that they must be of some use, and, from their lightness, they probably enable the wind to carry the seed to a greater distance than would otherwise be the case.
In other instances the plants themselves, or parts of them, are rolled along the ground by the wind. An example of this is afforded, for instance, by a kind of grass (Spinifex squarrosus), in which the mass of inflorescence, forming a large round head, is thus driven for miles over the dry sands of Australia until it comes to a damp place, when it expands and soon strikes root.
So, again, the Anastatica hierochuntica, or "rose of Jericho," a small annual with rounded pods, which frequents sandy places in Egypt, Syria, and Arabia, when dry, curls itself up into a ball or round cushion, and is thus driven about by the wind until it finds a damp place, when it uncurls, the pods open, and sow the seeds.
These cases, however, in which the seeds are rolled by the wind along the ground are comparatively rare. There are many more in which seeds are wafted through the air. If you examine the fruit of a sycamore you will find that it is provided with a wing-like expansion, in consequence of which, if there is any wind when it falls, it is, though rather heavy, blown to some distance from the parent tree. Several cases are shown in Fig. 12; for instance, the maple, a, sycamore, b, hornbeam, d, elm, e, birch, f, pine, g, fir, h, and ash, i, while in the lime, c, the whole bunch of fruits drops together, and the "bract," as it is called, or leaf of the flower-stalk, serves the same purpose.
In a great many other plants the same result is obtained by flattened and expanded edges. A beautiful example is afforded by the genus Thysanocarpus, a North American crucifer; Th. laciniatus has a distinctly winged pod; in T. curvipes the wings are considerably larger; lastly, in T. elegans and T. radians the pods are still further developed in the same direction, T. radians having the wing very broad, while in T. elegans it has become thinner and thinner in places, until at length it shows a series of perforations. Among our common wild plants we find winged fruits in the dock (Rumex) and in the common parsnip (Pastinaca). But though in these cases the object to be obtained—namely, the dispersion of the seed—is effected in a