The measurements in the foregoing table were taken at consecutive distances of 3 inches from the pith of trees having well-formed concentric rings, and by doubling this, the diameters of 6, 12, 18, 24, 30, 36 inches, &c., were obtained. There are only two or three of these results that can be compared with the list furnished by Decandolle, but severally they will be found useful for reference; and, later on, I shall have occasion to revert to some of them, when treating of the characteristics of the individual kinds to be noticed in these pages.
The proportion which the width or thickness of each layer at the upper bears to that at the lower extremity of the tree, varies considerably in the several species, the difference being the least marked where there is the greatest length of clear stem. Indeed, as the sap which goes to form a branch must be regarded as so much matter diverted from the trunk itself, it follows that in a tree in which the branches occur low down, the stem will taper more than in one which has them only near the top. The diminution in the thickness of the layers will be most apparent in the trees which produce the largest branches, and will be regular or irregular according as the branches are thrown out at regular intervals or otherwise.
It might be supposed that as every layer from the pith, or medulla, to the bark is in a different stage of perfection, the innermost or earliest, being the most matured, would be the strongest; but experience teaches us that this is only true up to a certain period of growth, and that in the majority of cases the maximum of strength and toughness lies nearer the more recently-formed heart-wood, or duramen. For this reason it becomes a matter of great importance, in selecting timber for use—especially if it be intended for works of any
