extent occurs in maturing, is most probably deposited to fill up tubes, and may thus not greatly strengthen the mass; a hollow cylinder being stronger than a solid cylinder when extending horizontally over a considerable stretch, like a joist or beam; the mass may also become a little more fragile by maturing: besides a filling up is the result of some chemical change the {209} wood probably becoming slightly carbonized or approaching to that change which takes place when vegetables become peat. It is rather difficult to speak of the strength of timber, as different kinds of timber, and different parts and qualities of the same kind of timber, have different kinds of strength. Some kinds are stronger as beams or joists, other kinds as boarding; while, again, some kinds are better for enduring a regular pressure, others for supporting a sudden jerk or blow, either as beams or boards. Some kinds are also comparatively stronger, moist; others when dry—and some kinds retain their qualities of strength or toughness longer than others when moist, and others longer when dry, although no rot appear. No. 7. Purposely for experiment[45], we selected three ash trees, all growing in Carse clay, but differing the most in fastness of growth of any we could discover. We cut these down on the same day; two of them proved about 36 years planted, and the third 15; this, the youngest was of fast growth, and had layers of more than double the size of one of the {210} former, and about six times that of the other. We cut a number of pieces of exactly equal length and thickness (17 inches long, and nearly an inch on the side), from each of these, choosing them of clean straight fibre, at equal distance from the ground, and from the outside of the tree, and having their growths nearly parallel to one side, of course free of heart. We proved one of each immediately on being cut out while full of sap, with their growths on edge in horizontal position, supported at each end with a weight suspended from the middle. The smallest growthed, and the largest, weighed at the time of trial nearly equal; the medium growthed one-thirtieth more. The smallest growthed supported the weight about six minutes; the medium and the largest about half that time; the smallest growthed yielded the least before breaking, and the largest yielded the most. When completely dried, the weight of the medium growthed still continued greatest, surpassing the largest one-fourteenth, and the smallest about one-thirtieth. The smallest and medium supported nearly equal weight, during equal time, and outbore the largest about one-seventh[46]; when placed {211} with the growths on edge, they were stronger than when placed with the growths flat.