When a truly flat board of considerable width is required it would be unwise to employ one plank, even supposing we could find a piece of sufficiently large growth suitable for our purpose. We have already stated that the woods are very liable to warp and become otherwise distorted in the cross-way of the grain, consequently little dependence can be placed on the material in this direction. It should not, therefore, if unassisted, be employed of great width in works where permanence of form is important. It will frequently be found that the plank becomes convex on the side which was towards the centre of the tree, provided both sides are subjected to similar conditions.

If the wide plank be cut into several narrow ones and the alternate pieces be reversed, and then glued up edge to edge, the wide plank thus formed will retain its shape much better than it could have done in its natural state. A wide plank not artificially formed would become curved, like a large bow, one side being convex and the other concave. But in a "made-up plank" the component parts being comparatively narrow, the curvature is shorter, and changing the sides of the alternate pieces has the effect of imparting a slightly serpentine outline instead of a large curve to the surface of the work, if it become warped.

Since the direction in which the wood is employed is of so much importance, it may be observed that those parts upon which the most reliance is placed to retain the work in shape should always be constructed in the length-way of the grain. For instance, the length and width of a box are usually much greater than its depth, and as the true figure of the work depends on the permanence of the side and end pieces, their longest directions must be made with the grain. If all the dimensions of the box be equal, the length-way of the grain must still be at right angles to the depth of the box. The grain of the end pieces must always be in the same direction, or plain, as that of the sides; the combination of the four pieces will then mutually support and retain each other in shape.

The liability to warp is not removed by changing the alternate sides of the narrow pieces or strips into which the wide plank was divided, and although the evil may be mitigated to some extent by this treatment, yet we must resort to the more powerful restraints of clamps to retain the work in shape. The simplest forms of clamp are shown in fig. 35, which may be supposed to represent the ordinary drawing-board used by mechanical draughtsmen. The board is constructed of two pine planks, A B, free from knots, and thoroughly seasoned. The edges which are to be united must be very truly planed, and a shallow groove formed in each for the reception of a thin strip of wood, t, called a slip-feather, which materially strengthens the joint. The groove cannot be satisfactorily made without a grooving-plane, of which we shall speak presently. In the absence of this tool the joint may be sufficiently well made for ordinary purposes, by simply glueing the edges together.

Fig. 35.

When the glue is thoroughly hard, the board must be planed upon the sides and edges. The clamp, C C, which is made of a well-seasoned straight-grained piece of "stuff" is fastened by screws across the grain of the board a short distance from the end. It is always necessary to have two clamps - one near each extremity of the board, - and if the length of the latter be considerable, a third clamp may be required in the centre. As the board is liable to shrink or become narrower, it is subjected to a considerable strain by being rigidly fixed to the clamp, C C, which is incapable of shrinking in the direction of its length. The clamp being unable to yield in the length-way of the grain, it can afford no liberty to the cross-grained wood of the board, which is apt to split and tear away from the fastenings by which it is confined.

A better kind of clamp which effectually preserves the truth of the board, and at the same time allows it to contract without restraint, is seen at D D. Instead of securing the clamps by screws or other rigid connections, a dovetail, d, is formed on the side next the board, in which latter is made a corresponding dovetail groove. The dovetail on the clamp and the groove in the board are made with a slight taper in the length-way to enable the clamp to draw itself tight when driven forward. The board is quite free to shrink throughout the entire length of the dovetail, and the truth of the surface is also strictly maintained. If the shrinking of the board be so considerable as to loosen the clamp, it may be immediately tightened by a few light blows of a hammer on the wide end of the dovetail. This form of clamp is very superior to the one first named, but it is much more difficult to make; for if the dovetail be badly fitted, it will fail to afford the requisite support to the board. The clamp, C C, is a very ready and convenient device for strengthening the covers, and also the bottoms and sides of rough boxes or packing-cases.

Fig. 36 represents a wide board formed by narrow strips glued edge to edge. This board is strengthened by a clamp, E E, which is secured to the end instead of the side. The ends of the board (only one is shown) are rebated on each side to form the tongue, T, which should be one-third of the thickness of the material.

Fig. 36.

A groove is planed in the edge of the clamp, E E, to receive the tongue, which should fit it moderately tight. If there be any doubt respecting the complete seasoning of the wood of which the board is constructed, it would be imprudent to glue the clamp to the board, except in the middle of its length. The shrinking of the pieces on either side of the middle strip would occasion sufficient contraction in the width of the board to cause it to separate from the glue which bound it to the clamp. This description of clamp is much used in cabinet-work, and as the seasoning is then tolerably complete, the work may be glued throughout with little fear of subsequent failure.