My bench inevitably has three squares close to hand, even though I could manage with just the one. Two of them are six-inch Starretts, any Starrett takes some beating, and my one old one was brand new in the box when I bought it in 1965. One of the Starretts was bought and given to me by a student in a New York, in a hands-on class in 2013; otherwise I would have just stuck with the two. My twelve-inch Rabone Chesterman has remained square and accurate throughout my sixty plus years of daily use. I doubt that many squares have seen more use than this one, and considering a knifewall pulled against the beam every few minutes throughout any given day, it’s a remarkable thing to know how well it has served me in my woodworking.
This last week I made these four drawers for my latest project. Front to back they are all the same depth and quite short. The overall project itself is a larger unit, about 1.3M (52″) square, roughly, and 40cm (16″). It comprises a series of half a dozen frames housing plywood panels that lie nigh-on flush to the outside faces of some of the frames. During the construction, after the joinery to each frame is concluded and the clamps are applied to close the joint lines, I usually check for squareness. I know. They can’t be far out, if at all, if my shoulder lines are dead on square, which they always are. In times past, in my hands-on classes, measuring corner to corner diagonally, first one side and then the other, usually prompted the question, “What are you doing there?”
Checking for squareness by measuring from corner to corner is a common trade practice. Large-scale measuring, the foundation for a building, such like that, massive frames of timber buildings and so on too, can be accurately checked or pegged out with stakes and nails to establish the exact lines for the building walls to ultimately be aligned to, etc. At the bench, no matter the size of the frame or box or drawer, it’s the same. Internal measuring from corner to corner when the opposite sides are equidistant gives us precision we cannot get any other way, and especially is this so with wood, which in reality moves with every tide of atmospheric change.
In the reality of life, our wood is often moving without us knowing. I planed my wood dead flat, square edged and straight before I started the dovetailing. The following morning, the very top board was cupped. Why was that? Well, it’s all to do with the bottom half dozen boards being stacked beneath being protected by the top one. The outside top flat of the top board was allowed to take in moisture from the atmosphere in my workshop. And my shop is climate controlled, btw.
The wood on the outside surface (the top one in the picture) can either expand or shrink according to the relative humidity and atmospheric moisture content, whereas the inside face remains similar to what it was after my initial truing work. To stop this from happening, I should have placed another protective layer on top of the stack or slipped the boards into a plastic bag, covered with a blanket or such like that. I don’t do that because I don’t concern myself with a cupped board that I know will not resist being straightened as I work it. A cupped board with the dovetails cut, that cupped after the dovetails were made, will
almost always go together. You simply straighten them with a clamp. But be prepared for a little stress because if the board has cupped, then it’s width has also expanded, albeit in a small distance. Compression needs to happen too.
Wood, like steel and less like iron, flexes too. No matter the care, the wood, the construction type, etc, Wood moves and especially so in its earlier days where flex is a ‘built-in’ feature of its DNA. The stiles and rails of frames can and will distort, whereas the corner to corner distance rarely ever moves without mechanical pressure. That’s the reason we must very carefully place clamps as we tighten them to ensure that they remain parallel to the rails. If they are out of parallel, the force on the shoulder lines can compress the face grain adjacent to it and within the joint the glue solidifies and sets, and the frame can be clamped out of square.
The above measurements are two opposite corners and the measurements measure dead on 16 3/16″ so the drawer is dead square, however, look below, and you can see the square tells a different and inaccurate story. There is a 1/32″ gap over at the far end of the square beam. Two very slight distortions in the drawer sides are then exaggerated by the short stock of the combination square. Would a longer square stock make a difference? No, not at all.
Using the short stock of the square to register against is usually too short a distance, even though in the theory of it, it should still work. It’s when the wood moves even a fraction along its length into a hollow or a camber, as above, which all woods inevitably do and at any given point in time that movement will by any variable amount. That being the case, only the corner to corner measurements, or equal diagonal distances, can be trusted.
So many of the subassemblies in this project rely on frames for the carcass superstructure, and then too the two doors. All must be square and checked for square in the dry assembly and then the actual glued up frames before resting them to dry. The drawers too must be square, so corner to corner cross-referencing is ideal. In some cases, for various reasons, I switch to what we call a squaring rod that became known to me in times past as an apprentice. In those days, we generally used folding rules as tape measures were quite uncommon rather than the norm they are today. Working on massive frames for doors and window frames stretch far beyond the four-fold three-foot ruler I used so we took a length of wood around 1/2″ by 3/4″ and cut a wedge shaped point on the end so it could go right into the corners by marking one diagonal distance, corner to corner, and then trying across the opposite one, we could check to see if the distance was the same. If it was, the frame was square
We often fail to understand that applying pressure to joints through clamps will distort the long axis of the rail according to any existing bias. If the length is even slightly biased, applying pressure develops the bow even further. Offering a square to the corner will then show a major fault when, in reality, the corner to corner reading will likely be the same distance and give the accurate reading. That said, knowing the sides are curved is helpful in some cases (pun intended).
Large or small, marking the corner distances will show discrepancy, and moving the clamps out of parallel can correct almost all variables. Try it on your next frame, on a dry run, when you have time. These things are worth noting and storing.
Squares in squares in squares is the way of cabinet making. Boxes in boxes in boxes in boxes is what we are all about. How do I describe the feeling this evening when I close the workshop door and the light fades around my day’s work? A hundred per cent of my wood was resized on my 16″ bandsaw for a rough-sawn surface more than adequately close, and then I hand planed all four surfaces of every single piece and cross-cut them with my handsaws.
The accuracy of my hand work is reflected in the finalising work with the fitment. I can’t explain the result of my sharpening my planes a few minutes ago. The end grain of the stiles and then the reflected light from medullary rays only, only oak can give. There can be no frowns now. Smiles in oak reward all the days and weeks of work. My drawer fronts look beautiful in their simplicity and clean lines they will soon become as they move in and out to open and close. I call to Joseph, Natalie and Mark to come and see. I want them to enjoy my joy with me. Such things are shared experiences if possible.
Oh, one last thing. My last fifteen years of reaching out to woodworkers around the world has brought me equal measure of thankfulness and joy. Thank you for all of your support in joining me on the journey in real woodworking!
