war, when beech and other home-grown species replaced many tropical woods. Such replacement in normal times will be slow, as very many advantages are contained in the very remarkable assemblages of cells which go into the make-up of the different timbers.

The cell, or rather its skeleton, in wood has particular significance in this discussion. In woods produced by coniferous trees, known to us as “softwoods,” only two important types of cells are present—a rather thick-walled type (forming the bulk of the wood and combining in the tree the functions of conduction of food solutions and of mechanical strength to support the massive column which is the tree trunk) and thin-walled storage cells. Softwoods include the timbers best known to our western civilisation. The other vast group, “hardwoods,” have thin-walled cells forming tubes for conduction, thick-walled mechanical-strength cells, and thin-walled storage cells. Leaving aside the inherent differences in wood structure which distinguish the various timbers and determine to a considerable extent their properties and uses, I wish to stress the important fact that the bulk of the cells in woods are long and slender, with the long axis standing vertically in the trunk of the standing tree. The long axis of the cells is, in other words, along the grain of the wood. Wood has high tensile and compressive strength along the grain and low tensile and compressive strength across the grain. An individual cell may be likened to a rolled tube of paper in-so-far as these strength properties are concerned. If we increase the number of layers on the inside of the tube the enclosed space will be decreased, the wall thickness increased, and the weight and strength of the tube increased. It is mainly the ratio of air spaces to cell-wall volumes which determines the weight of a given volume of any timber, as the material constituting the walls is basically the same in all timbers. Deposits or secretions of gums, resins, oils, etc., do contribute in a small measure to the dry weight of the heavier timber—e.g., lignum vitae and black maire. The extremes in weight for a cubic foot of dry timbers of commercial importance are 6 lb. for balsa and over 80lb. for lignum vitae. In general it may also be said that strength varies directly with specific gravity.

I have said the cell-wall material, which I will call “wood substance,” is basically the same in all timbers. It consists principally of cellulose (50%) and lignin (26%), and has a specific gravity of 1.54 (i.e., it weighs about 96lb. per cub. ft.). It will be appreciated, therefore, that timbers weighing over 80lb. per cubic foot have not much room left in the cells for air space.

It will assist our understanding of a number of phenomena connected with timber if we have a brief glance at modern ideas concerning the makeup of the cell wall. Cell walls are made up of concentric layers which can be dissected into long, slender fibrils arranged spirally and forming a meshwork. A smaller unit of the cellulose is the crystallite, which in turn is composed of groups of cellulose molecules. This molecule has a length many hundreds of times its diameter—it is exceedingly strong in the longitudinal direction—but adjacent molecules are not strongly united crosswise. Lignin is generally believed to be amorphous and to exist interspersed between cellulose crystallites.

Another physical property of wood requiring mention is the relationship of wood and water. Some timbers when freshly sawn contain as much as 200% moisture content based upon the oven dry weight of the wood. Most of this water is in the cell cavities, but about 25% is in the cell walls actually between the crystallites. In drying out a great many problems arise both during loss of water from the cell cavities and the subsequent loss from the cell walls. The principal point to be made, however, is that when water is removed from between the crystallites the wood shrinks. Shrinkage is very slight along the grain—i.e, along the length of the cellulose molecules, but considerable across the grain, as the molecules are allowed to come closer to one another through the loss of the water films. In air-drying timber a point is reached at which the moisture content is in equilibrium with the average relative humidity. As an example this E.M.C. for timber out of doors in Wellington but not subject directly to rain wetting is usually assumed to be about 17%. Timber in use indoors has a lower E.M.C., which is easily attained only by kiln-drying. Winter conditions in a steam-heated building may require timber to be dried to less than 10% M.C. Shrinkage commences about 25% M.C. and continues until the wood is free from moisture. The most important point to make is that wood remains hygroscopic. With an increase in relative humidity of the atmosphere water is reabsorbed between the crystallites and swelling takes place. In passing it should be noted that both cellulose and water are polar, a fact which is well illustrated