Having finally reached the commercial stage, the lumber is shipped away from the mill either by water or by rail to the lumber yards of the country.

Here it should be seasoned. In the past this process consisted of piling the planks in the open air in such a way that air could circulate freely through the pile, allowing the sap to evaporate and the wood to dry evenly. This was a sure but slow process, and in the hurry of modern life quicker methods have been tried.

One of these is known as kiln drying, by which the time is reduced to a few weeks. It consists of piling the wood in a room like a kiln and drying it by artificial heat. The result is not so satisfactory as the natural method, because the sap near the surface hardens and prevents the inner moisture from escaping, so that kiln-dried lumber while dry at the surface is "green" inside. When planed till part of the surface is removed the green wood is brought near to the air again, and warping is liable to occur.

Other methods have been tried, such as steaming to vaporize the sap, and soaking in hot water for the same purpose. Of course these processes all add to the cost of lumber, yet so valuable is time that it is difficult to obtain good old-fashioned seasoned wood unless it has lain for some time in a local yard.

In order to understand the phenomena of warping, shrinkage, checking, shakes, etc., it is necessary to know something of how the tree grows. Like all living organisms, it is made up of minute cells. The new cells are formed on the outside of the tree under the bark, and here the sap is most active. The cause of the flow of sap is not very clearly understood, but it corresponds to blood in the human body, in that it carries the nourishment that forms the cells. As a new mass or layer of soft new cells forms each season, the layers may be distinctly seen and counted, but the line of separation is not a sharply drawn one, as we find by examining a cross section of wood with the microscope. However, the layers or annual rings are distinct enough to be counted, so that the age of the tree at the time it was cut down may be readily discovered.The new or sap wood, then, is further from the centre each year, and while the old cells may not be dead, they contain less and less sap, are therefore drier, and after a few years change colour, becoming darker.

There is often a very great contrast between the colour of the heart wood and that of the sap wood, although the latter may be represented by several years of growth.

These annual rings are not actually circular, but very irregular, and often wider in some parts than in others. The study of these rings is very interesting, and it shows that the tree usually increases in diameter more rapidly during the first few years than later. Very often, after growing slowly for several years, the tree will apparently grow rapidly again. The cause of this cannot be determined without a knowledge of the tree's history.

It has been proved by experiment that thinning the forest increases the growth of the remaining trees 18 per cent., and these peculiarities in the rings may have been due to some like cause. The bearing of this fact on the peculiarities of warping and shrinkage is that when cut down the log is drier at the heart and more sappy at the outside, so that evaporation occurs near the surface.

The effect of it is shown in Fig. 239. The outside drawing together has opened the wood, or "checked" it, most at the outside, diminishing toward the centre. The evaporation would have occurred just the same had the log been cut into planks, causing them to curl as shown at a. This is known as warping, and it is one of the troubles of the woodworker. In construction it must be constantly guarded against, and overcome as far as possible. It cannot be entirely prevented, but if the wood has been well seasoned before it is used a large part of the warp will be taken out in the planing mill, or in the squaring up.

Twisting, winding, and warping are also caused by the two sides of a board having been subjected to different degrees of heat, moisture, etc. If a plank is laid on the floor, the upper part is more exposed to the air and to changes of temperature and humidity; therefore it curls.

If a board is stood on end or placed in a rack where there is a free circulation of air, the curling will be much less. Even in a rack, if several boards are piled one on another, the top one will have different conditions from the others and be apt to curl or wind.

Shrinkage is a term applied to the decrease in diameter of the tree, due to sap evaporation.

In the case of the board it means a decrease in width, and it varies greatly in different trees and their woods. As shrinkage is always across the grain, its effect on a common joint may be illustrated in Fig. 240. At a is shown a middle lap joint just put together. If the wood is not well seasoned, shrinkage will in time change it to the form shown at b, which is exaggerated to make the meaning clear.

A square piece of timber, one corner of which is the centre of the tree, will change from c to d. Shrinkage as well as warping must be taken into consideration in construction.

The development of the panelled door is a good illustration. Suppose we wish to close a space with a door, knowing little about shrinkage. Let us construct it by the simplest method, say four vertical boards. If the width of these boards equals the opening when the door is built (Fig. 241) there will soon be an opening wide enough for the fingers to enter and lift up a latch on the inside. The door is very much of a failure. We notice, however, that there is no opening at top or bottom. An idea! We construct a door with planks placed horizontally. Now although we find after a while no opening at the sides we do find openings at top and bottom. The panelled door is not constructed solely for beauty but to overcome shrinkage as far as possible. Fig. 241 shows the various parts. The rails maintain the width, the only shrinkage being in the cross grain of the stiles, and they preserve the height except for the small amount in the rails. The remaining spaces are panelled, the construction being shown at a. Both stiles and rails have a groove plowed out to receive the edges of the panel. This should be free to shrink in the grooves, where it is invisible, but if the mistake is made of fastening the panel edges rigidly in these grooves the panel will shrink anyway and frequently split from top to bottom.

Many other forms of construction which we have seen daily as long as we can remember have equally sound reasons for their form. No piece of woodwork should be designed without considering how it will be affected by shrinkage and warping.

In selecting lumber always look out for "shakes." This is a defect caused by the separation of the annual rings. A tree may be considered as a series of irregular cylinders of diminishing diameters. The forest-grown tree is much more spindling, tall, and straight than the low-crowned, heavy-branched specimen grown in the open, where there is no crowding.

The swaying of the forest tree in the wind, especially when its neighbours have been cut down, is sometimes sufficient to make the rings separate and slide one within the other. This is more noticeable in some species than others and it gives the wood a serious fault. (Fig. 239).

"Winding" is the result when the ends and sides are no longer parallel. Like all peculiar characteristics of wood, this varies greatly in lumber of various kinds, and may be largely avoided by exposing both sides to the same conditions, or keeping equally distributed weight on it until used. When winding becomes excessive, the board is useless for any kind of work.