The following is excerpted from “The Woodworker: The Charles H. Hayward Years: 1939-1967” Volume 4, The Shop & Furniture.
Although a back may not call for the high finish that is necessary for, say, a cabinet door, it needs to be strongly made and of a type to suit the particular job. “Craftsman” discusses here some of the points to be considered when deciding just what kind of back a job is to have. —Ed.
I AM afraid that many of us are inclined to let the backs of our cabinets take pot luck, as the saying goes. We make a job, say, in oak, possibly putting in oak drawer sides, and backs, but hesitate before going to the expense of oak for the back. The reason (or excuse, however you happen to look at it) is that it is seldom seen, has little or no wear to withstand, and that, since the cheap back answers the purpose just as well, it is clearly a waste to spend money on an expensive one.
Well, it is logical enough up to a point, and, providing that it is merely the material that is cheapened and not the method that is worsened, no great harm is done. In fact, there are many pieces of quite light woodwork in which a heavily built back seems almost out of place. Still, it is nice to have a piece of work in which nothing has been skimped, and the argument that a cheap back answers the purpose as well as a better one may not necessarily hold good, as we shall see later. The safe plan is to consider each piece on its merits, and give it the best back that it is worth.
BACKS OF OLD FURNITURE
If one goes back into the past one comes across some curious anomalies. Many of the antiques of the Queen Anne and mahogany periods of which we think so highly had wretched backs. I myself spent a good many years in a repair shop, and I can speak feelingly of the hours I devoted to gluing strips of canvas across gaping splits in panels and across open knot holes. I have seen a mahogany chest of drawers of the Chippendale period with magnificent show work—serpentine shaped drawers, fine carving, and so on—with a back consisting of pieces of 1/4 in. pine nailed across. An extraordinary inconsistency. Apart from its having no strength, the whole thing was bound to shrink and split.
Yet when we come to that much abused period of Victoria, we find exactly the reverse. Probably no finer cabinet backs have ever been fitted into furniture. Open the door of one of those huge Victorian wardrobes (there are plenty of them knocking about in seaside boarding houses). You will find the mirror back more strongly made than many a modern wardrobe door, and the carcase back a finely panelled framework often with moulded stiles or flush panels.
Perhaps one reason why there has been a tendency to fit lighter backs since Victorian times (apart from the all-round cheapening of materials and construction) is the introduction of plywood. It seems such an obvious use for ply, a material which is free from shrinkage and obtainable in such large sizes. Undoubtedly it is perfectly suitable for the purpose, providing the carcase is strong in itself, and does not rely upon the back to make it rigid.
TYPES OF BACKS
There are various considerations that affect the choice of a cabinet back. There is, for instance, the question of size. A single sheet of 3/16 in. plywood might make an excellent back for a little cupboard, say, 15 ins. high, but would obviously be absurd for a wardrobe. Apart from this, however, the first consideration should be: does the job rely upon the back for strength, or will the back serve merely to enclose a space? Fig. 1 shows the idea. At A the back is needed to prevent racketing and to stiffen the carcase generally. At B, however, the carcase is already strong, and only a light back is needed.
In the latter connection, of course, it is sometimes an advantage to build in the back with the carcase. Items such as sideboards are often made in this way. As a general rule, however, it is better to make the back separately, because it simplifies the subsequent fitting-up.
THE PANELLED BACK
For a thoroughly strong back the panelled type is undoubtedly the most satisfactory. It is perfectly rigid and is free from all shrinkage complications. It should always be used for pieces such as cupboards with large, heavy doors, which are particularly liable to distortion unless provided with a stiff back.
Fig 2 shows the usual form. The whole thing is put together with mortise and tenon joints, and the panels are grooved in. One point to note is that if there is a shelf in the cupboard, the middle cross rail should be arranged opposite to it if possible. It may not always be practicable, of course, but the advantage is that it gives a level surface against which the back of the shelf can face (see B, Fig 3). If this is not done there will be gaps opposite the panels as shown at A.
The same difficulty sometimes occurs in a bookcase or similar item, but owing
to the large number of shelves it is not practicable to arrange for many horizontal rails. The better plan is that in Fig 4, in which the panels are flush with the framework at the inside. It necessitates fairly thick panels, of course, but it gives a far neater result than cutting out the back edge of the shelf to fit.
A somewhat distant relative of the panelled back is the muntin type. It is nowhere near as strong, and is rather a doubtful member of the family. Like some relations, you can’t deny them (and they are useful sometimes), but you are a little shy about mentioning them in the best circles. It consists of a series of uprights, say 3/4 in. or 7/8 in. thick grooved at the edges to take thinner panels, as shown in Fig. 5. The ends of the muntins are cut away as shown inset, so that the panels can be fixed directly to the back of the carcase.
Now, as the panels are generally about 9—10 ins. wide, and of deal, it is inevitable that a certain amount of shrinkage will take place. Consequently it is a mistake to drive in nails right across the width because the wood would split in the event of shrinkage. The better plan is that in Fig. 6 in which nails are driven in near the centre only. The edges extending into the muntin grooves are free so that they can draw out. Note that the heart side is outwards so that the free ends are pressed tightly against the carcase by the natural twisting tendency of the wood.
If, owing to the presence of a number of shelves, it is desirable for the back to be entirely flush on the inside, the muntins can be rebated instead of grooved as shown in Fig. 7. The beads along the rebates are not entirely decorative, but they serve to render the gaps less noticeable in the event of the panels shrinking. All these details about shrinkage apply only when solid wood is used, of course. In the case of plywood it does not matter.
Speaking of plywood brings us to another variation of the muntin back. In its simplest form the plywood back is nothing more than a sheet of plywood nailed or screwed in a rebate. For quite light jobs this is satisfactory enough, but to give a neat finish the back in Fig. 8 is better. A series of grooved and rounded horizontals is screwed on. They can be arranged level with the shelves as shown. The plywood panels fit between them in the grooves. For a flush effect the rails can be rebated instead of being grooved (see D).
These are really in a class by themselves, for although they could be applied to pieces such as wardrobes, they are not so strong as a panelled back. One of two methods can be followed. That shown in Fig. 9 has the advantage of simplicity. The back is really a series of matched boards, tongued one into the other, with either a bead or a V worked at the joints. The boards are screwed or nailed directly to the top and shelves, and at the bottom to a rail specially fitted for the purpose. In the second method, Fig. 10, wide grooved rails are screwed at top and bottom and the matching fitted in the grooves. The wide rails give rigidity, the matching merely filling the space, so to speak. It can be either very thin as at A, or it can be stouter, the ends being tongued as at B.
Incidentally, a detail applying to all backs of any thickness is that the rebates in the ends should slope as shown at A, Fig. 11. If this is not done the projecting portion is liable to curl as shown at B.