This type of square was produced originally for engineers, but woodworkers have come to realise that it can be very handy to them. Of course the high precision tool needed for engineering is extremely expensive, but a low cost square of sufficient accuracy for woodwork is available, and this article deals with some of its advantages.
It will be seen that the blade of the square is separate from the stock, and is free to slide to give any required amount of projection. It is fixed by means of the small thumb screw. The advantage of this is that it becomes two squares in one. Every practical man knows that, whereas the 12 in. square is necessary for larger work, the 6 in. size is handier for general bench work. Thus, in the ordinary way the square is used with about 6 in. projection as in Fig. 1. When wanted for larger jobs it is merely a matter of sliding it along.
It can be used equally well for mitres as shown in Fig. 2. For this it has not the same usefulness as the set mitre because it can be used for the acute angle only, not the obtuse one, but it is suitable for the majority of jobs.
Apart from its general bench handiness, however, it can be used for purposes for which the ordinary square would be useless. Suppose you have to work a large rebate. How would you test it to see that it is square? The ordinary square could not be used because the blade projects too far. By giving the blade of the engineer’s square projection slightly less than the rebate width it can be used to test as shown in Fig. 3. In the same way if lines have to be squared across a rebate the square can be used as shown in Fig. 4. The advantage is that the butt of the square rests on the edge of the wood, not the bottom of the rebate. It might be that the rebate is tapered, and in this case the only way of marking a square line would be to work from the edge, not the rebate. This is made clear in Fig. 4 (inset).
The square can be obtained with or without a spirit level. Fig. 5 shows it in use to test a horizontal surface. The blade is either pushed in level or withdrawn entirely. For testing the vertically of a post or whatever it might be the square is held as in Fig. 6, the blade resting against the post. As a still further use, the blade of the square is marked out as a rule.
Having cut out the seat, the first job is to clean up the sawn edge. Here is a wooden jack-plane in use. I very rarely buy new tools. They are expensive and mostly made of cheese. A plane like the one I am using costs about £10, and the blades are thicker and made of much superior steel, either cast or laminated. Put an edge on that and it will last several times longer than a new iron for a modem plane. There are many good dealers who sell old tools. I often find interesting items on market stalls.
Having cleaned the seat up, I have to decide which is the top and bottom. Only experience can tell me this. What I am trying to do is get a good ‘picture’ on the completed seat, a nice grain pattern. As it has to be scalloped out, this is not always easy to tell. I mark the top and bottom with a pencil line 2″ from the edge. The marker gauge is a simple oak tool that one of my sons made many years ago.
A sculptor’s adze in use to ‘bottom’ the seat. This is not a traditional way of doing this. I have and can use the full size gutter adze. However, if you look at the average dining-chair seat you will notice very gradual curves and probably a maximum depth of 1/2″, possibly only 3/8″. I like to cut in from my pencil line at a much sharper angle, and by using a small adze I can more easily control this. Also, I leave less to do with the finishing tools.
The small adze has a curve across the blade. It is used across the grain, and several passes must be made to achieve the right depth. All wood of the same species varies. An oak on a hill, in a valley, parkland oak, oak grown in windy conditions; they are all a bit different. None varies more then elm. Even a different part of the same tree can have totally different characteristics. I recall sailors talking about a ‘soldiers’ wind – blows both ways at once. Well, the grain in elm is similar. This is a nice piece and is behaving well.
Chamfering the underside of the seat. This is one of the ways of taking the chunkiness out of the seat. Sometimes I round the bottom, some chairs I leave the edge at its original thickness, just taking off the arras top and bottom. All this is purely for appearance, and I go through phases, sometimes favouring one way, sometimes another. I have to do this part after chopping as otherwise the seat is difficult to hold in the vice.
I screw a block of wood under the seat to hold it in the vice. The tool I am using here is a small ‘round both ways’ plane with handles. This again is a home-made tool. Chair-makers traditionally used travishers. Travishers are very hard to come by. I suspect this is because most of them are hanging neatly on the walls of the High Wycombe chair museum, or, because of their scarcity, are in collections. I abhor this situation; tools are for use. On the three visits I have made to the chair museum at Wycombe, I have been the only visitor, and I stand with mouth watering looking at literally dozens and dozens of travishers. Such is life. My tool works well, however.
The next stage in smoothing the seat. This is a scorp. It’s a drawknife bent to a semi-circle and is really a cooper’s tool. The scorp has a bevel on the outside, so it is reasonably easy to grind and sharpen. I find that for this work an engineer’s vice bolted on top of the bench is essential. All the work is at a reasonable height, and my back does not suffer. I line the jaws with wooden blocks, and there is hardly a job in chair-making in which I do not use this vice.
I decided to put a bead around this seat. I don’t know why. It was quite unnecessary. I am using a block of wood with a screw protruding about 3/8″. The edges of the screw head are filed sharp. Of 400 or so chairs I have built I have perhaps beaded a dozen. On this chair it looks alright, but it would have been just as alright without it. There’s a lesson here somewhere.
God forbid that I should ever have a fire in the workshop, but if I did, and had to get out in a hurry, I’d make sure my dumbscrape was in my pocket. This is a magical tool. Called a cabinet scraper in the tool catalogues, it is sharpened to have a wire edge with a burnisher of hard steel. It cuts like a plane – see the curly shavings on this seat. When they come from the shop they are oblong, square-sided. For this kind of work the edges need grinding to a gentle curve. It is a most pleasing business using a scraper.
Ah, man’s vanity! This is the signature I put on all my work. I did it. Well, I did part of it. I have a partner – the Great Chair-maker – so I put a cross, as we Celtic people see it. There is no point in smoothing the underside of the chair. It does not show unless the chair is turned upside down, and working with hand-tools only, one tends not to spend time on unnecessary things.
Craftsmanship is a tough concept to get your head around. Even the dictionary gives it short shrift. “Skill in a particular craft.” Pretty lame. This is a bit better: “The quality of design and work shown in something made by hand; artistry.” Much closer. I guess it’s one of those abstract impressions that’s hard to define, but you know when you see it. It has to do with skill, accuracy, artistry, expertise, technique, workmanship and sometimes even design. That’s thanks to a hunt through the thesaurus. What all those words have in common is a connection to the human hand and heart. That, I think, reiterates the notion of practice stated in Chapter 2: Do anything long enough and you become good at it. You develop and become proficient at craftsmanship.
I think that brings up the very important notion of standards. What are your standards when building a desk, cutting dovetails or finishing? Chances are that your standards are not mine. We all strive for a different benchmark. I once worked for a fellow whose idea of a finished piece of furniture was anything sanded to #100 grit. He’d become frustrated and pass it off to someone else to complete the sanding process. Let me tell you what a pain it is to sand a piece that is already assembled. In my shop, I generally sand to #400 or #600 grit, depending on the wood and the surface. Yet other woodworkers will insist on sanding to #1,200, while some insist that #200 is smooth enough and you can’t feel the difference. It gets waxed, anyway. The reason I go to #400 or more is that figured cherry sanded to #200 or less appears as “blotchy,” while sanding to #400, #500 or #600 shines the wood and really pops out the grain, so it looks like tiger’s eye when viewed from different directions. When sanding, I work my way up through the grits from #120, #150, #180, #220, #320 and #400, then polish with #0000 steel wool. I start with each grit on an orbital sander, then hand-sand with that same grit to remove swirls. When completed, there will be no swirls, and the surface will be smooth and polished to reflect light. If it looks good and feels good, it is good. Would sanding through to #1,200 make a difference? I’m not sure that a discerning individual could tell the difference. Is that extra little bit worth the added time it would take? That depends on your definition of perfection, your standards.
Cherry sanded only to #180 or #220 grit looks blotchy.
Some folks prefer handplaning to sanding. I have no argument with that. I, too, prefer it when possible. Handplaning may be a bit more difficult, but if done right yields a smooth reflective surface. If you experience tear-out, either your plane is dull or your technique needs upgrading.
But in some instances, sanding is preferable – for example, dovetailed drawer sides. Block planes with low angles are perfect for the end grain of pins and tails. If, however, the flat grain is figured or goes downhill, the block plane will tear it to shreds. So you go for a high-angle smoother, right? It will work wonders on the figured flat grain, but tear your end-grain pins and tails to shreds. Ever try to plane between the pins and stop exactly at the half-blind front? I’m not saying it’s impossible, just frustrating and time-consuming. Sometimes the sander is faster and more efficient, no matter which way the grain runs. To me, it’s a job to be done to perfection, not therapy.
A higher sanding grit reflects light better and shows the chatoyance of undulating grain. When viewed from side to side, dark areas become light, and light areas become dark. A little extra effort really pays off visually.
I don’t sand every surface to a high polish. The bottoms of cabinets and the undersides of my tables are smooth to the touch, but those get sanded only to #220 grit. That’s not to say I try to cut corners; rather, I try to work efficiently. My prices are pretty dear, and adding 10 or 20 percent more time to an already time-consuming project begs the question, “When do you quit?” To put it more precisely, I try to put my efforts where they will show and have the most impact.
I guess that everyone’s standards are different. That’s why I don’t have employees. I know of one woodworker who in fact does sand to #1,200 grit. Honestly, I can’t see or feel the difference between #600 and #1,200. How much time do you want to spend to achieve perfection? If your piece is 99.9 percent perfect, is it worth another three or four hours to get to 99.99 percent perfection? Who decides? Of course if time is not an issue, then it’s a moot point. I’ll go into this a bit more in Chapter 7.
On the other end of the spectrum are the historical purists, who insist on building the way some craftsmen in the 18th and 19th century worked. One woodworker stated that historical mistakes were fine to be reproduced. “I prefer to go with what was original. The panel may crack a little, but as they say here in Maine, ‘it won’t bother.’ It’s normal to see modest splits in the sides of period case work…” Really? I know for a fact that doesn’t jive with the Shaker work ethic. Some of those cracked antiques were most likely built by those who didn’t know any better. Not only that, but they didn’t have to deal with forced-air heat. Why would you build a piece knowing full well it was going to crack? I like to think that we learn from our mistakes, as well as the mistakes of others, and hence improve. I don’t think that any craftsman working for a living 100 or 200 years ago would consciously build in a fashion that he knew full well would fail. That’s not my idea of old-fashioned craftsmanship – but as I stated, everyone is entitled to set their own standards and opinions.
Figure 6.12. Three 25 mm (1″) long lengths cut sequentially from an English oak plank were prepared and tested. Each piece started the test weighing 73± grammes, and were approximately 177 mm wide by 21.5 mm thick. Piece A was kept near a radiator for three days prior to further experiments. The weight recorded after this additional drying was 69 grammes. Subsequent microwave oven drying to 0 percent MC and a final weight of 65 grammes shows the piece started the test at 6.5 percent MC. After oven drying it measured 171 mm wide x 21 mm thick. B, the middle section of the three cut pieces, was placed outside but sheltered from rain for three days after which it, too, was oven dried to 0 percent MC. Prior to drying it weighed 76 grammes with a dry weight of 66 grammes. This equates to just over 15 percent MC, indicating its MC rose probably 7 percentage points over the three days. Dimensions of the piece just before oven drying were 177 mm wide by 21.5 mm thick. C was weighed and measured after three days of soaking in water. Its weight after soaking was 92 grammes and measured 184 mm wide by 22 mm thick. Subsequent oven drying to 0 percent MC and a final weight of 67 grammes show MC of this piece was 37.5 percent after soaking. Dividing the narrowest width (A) by the widest width (C), i.e., 171 / 184 = 0.93 or ~7 percent shrinkage from 30 percent MC, or greater. Similarly dividing the narrowest thickness by the greatest thickness, i.e., 21 / 22 = 0.955 or 4.5 percent shrinkage. The photograph was taken after A was oven dried but before pieces B and C were dried. The dimensions are very approximate as only a steel rule took the measurements. Although measurements are only approximate it’s interesting to note shrinkage factors for the tangential shrinkage (~7 percent) and radial shrinkage (4.5 percent) for this piece of oak are quite close to European oak numbers provided in various wood movement tables, i.e., 8.9 percent tangentially and 5.3 percent radially.
This is an excerpt from “Cut & Dried” by Richard Jones.
Oven drying in a microwave oven takes between 20 and 45 minutes. The average time is 30 minutes. It saves a great deal of time compared to drying wood in a regular oven. It does, however, require care and attention to details. Poor methodology and mistakes in the procedure usually lead to problems and failure.
You will need to be able to weigh the wood samples. I find electronic postal scales purchased at a reasonable cost from an office supplier work well enough for my needs. If you require more accuracy, more expensive scales are required. My scales provide readings in 1 gramme divisions from zero up to a maximum of 2,200 grammes, and the machine can be set to give readings in either grammes or ounces.
To dry the wood I use a turntable-type microwave oven with several power settings. The only two settings I use are the very lowest setting and the next higher setting which is “defrost” – your oven is likely to have a different configuration. But whatever marked settings are available, restrict yourself to the lowest one or two power levels. As the wood is heated, moisture evaporates from all exposed surfaces, including the bottom face resting on the turntable; three to five paper kitchen towels laid under the wood absorb and dissipate the condensed moisture drawn downward from the wood. If you’re testing several samples, make sure they don’t touch each other because this can concentrate the energy and can lead to smoking and possibly fire.
If the wood starts to smoke during the drying procedure the sample is ruined and you need to start again with a new sample. Smoking during the cooking means you have burnt away some of the wood volume, so weight measurements taken thereafter are inaccurate. This is why I mostly restrict myself to the lowest power setting and short bursts of heat. The second lowest power setting, defrost on my microwave oven, is seldom used, but I do sometimes use it for the initial drying cycle of very wet wood.
The ideal wood sample is the same as described in section 6.6, i.e., a full thickness and width piece taken at least 400 mm in from the board’s end, approximately 25 to 32 mm (1″ to 1-1/4″) long. Weigh your sample and make a note of this. If the sample is already partially dried, e.g., about 25 percent MC to 15 percent MC, cook the wood at the lowest oven setting for between one and a half and two minutes in the first cycle.
If you know the wood is already below 10 percent MC, I recommend you cook it at the lowest setting of the oven for no more than 45 or 60 seconds to start with.
When wood is definitely very wet, 30 percent MC or above, the first cooking should last no more than between one and a half and three minutes with the oven at the second lowest setting. Even in this circumstance I prefer to use the lowest oven setting. It takes a few minutes longer to dry the wood but is preferable to starting again because of a burnt sample.
After the first cycle, weigh the sample or samples again to form an impression of how quickly the wood loses weight, i.e. loses water. Let the sample rest for a minute or so and re-cook it for between 45 and 60 seconds and re-weigh.
Continue with this routine until you can’t measure any weight change, i.e., less than 0.1 of a gramme variation if you are using highly accurate scales. My scales read only to the nearest gramme, so I stop cooking when five or six low-weight readings are recorded.
When this point is reached, use the formula provided earlier, i.e., MC percent = ((WW – ODW) / ODW) x 100, where WW is wet weight of the sample, and ODW represents the wood sample’s oven dry weight.
The following cautions are important: Do not use the microwave oven’s high power settings. The internal heat built up in the wood needs to dissipate, and high settings cause rapid heat build up, smoke and even fire.
The more wood tested in one go, the more time is required to complete the job. This is useful because after the initial heating of a large batch you can rotate from one sample to the next in the oven with short bursts of cooking for each piece. This gives each sample a break between heating cycles, thus reducing the chance of overheating any one piece.
I generally find kiln-dried wood samples react differently to cooking than green or air-dried samples. It’s best not to mix samples of very different moisture contents and different wood species during the test, but it’s possible if you proceed with care.
Being sure the wood sample or samples is, or are, truly oven dry requires patience and careful weighing using accurate scales. It’s better, and safer, to use several short cycles in the oven at low settings than it is to try and rush the job using a higher setting for extended times. The latter strategy usually results in burning the wood and failure.
In closing, these final, following warnings probably seem obvious, but they’re worth mentioning. Removing cooked wood from the oven requires care. It’s usually quite hot, and can and does burn skin – you probably don’t need to ask how I know that! Use an oven glove or heavy leather work gloves. Also, be aware that at the end of testing, and unknown to you, wood might have charred on the inside: It can smoulder and burn and, if placed in a rubbish bin, could start a fire. Careful disposal is essential. The safest thing you can do is put the cooked wood in water when you’ve finished drying it to ensure it doesn’t burst into flames later – it can happen.
The construction of a drawer seems a straightforward and fairly obvious piece of work; there is an accepted way of making it which experience has proved reliable. Yet drawers were not always made in this way, and it is extremely interesting to see how woodworkers of past ages solved the problem of making and sliding them.
It was not until the 17th century that drawers in furniture were used to any extent. The chest of drawers was entirely unknown, and it is with something of astonishment that one comes to realise that through all the centuries previously men had been content to bundle out the entire contents of a chest in order to reach something at the bottom. When the idea did come, however, its advantages were quickly realised, and from the middle of the century the chest of drawers became established and has remained popular ever since.
Most early drawers were supported by the rather curious method shown in Fig. 1. A groove was worked along the side, this fitting over a runner fixed to the carcase side. It seems a little strange that the method should have been adopted because it must have meant more bother than putting runners between the drawers; furthermore the grooves rather weakened the sides. Still, they used thicker sides then than we think necessary to-day.
FIG. 2. SAME DATE AS ABOVE. Here the worker has used crude dovetails instead of the lapped joint.
The actual construction is really very crude. The front is rebated at ends and bottom, and the sides glued and nailed on. So also is the bottom, which fits in the rebate at the front and is merely butted beneath the sides. The entire weight of the contents is taken by the nails holding the bottom. Occasionally one comes across a piece made by a man who had ventured into the mysteries of dovetailing. In Fig. 2, for instance, is a drawer which is lap-dovetailed at the front and has a through dovetail at the back. The bottom is attached in the same way as before, but since this actually rests upon the runners there is no strain on the rails. The weakness, of course, is that as the wood wears away the nails are left in projection and so score the runners.
FIG. 3. SECOND HALF 17th CENTURY. The front is of deal veneered with walnut. Sides, back, and bottom are oak.
Both the foregoing examples are of oak furniture. During the second half of the century walnut gradually superseded oak, though it was mostly used in veneer form. Oak remained the chief wood for linings, however, and thus it is in the drawer in Fig. 3—the sides, back and bottom are of oak. In some instances oak was used for the groundwork of the front also, but it was soon realised that it was not the ideal wood for veneering. It was too coarse in the grain, and, owing to the presence of the medullary rays, which were harder than the rest of the wood, marks were liable to show through to the surface owing to unequal shrinkage. Consequently deal was mainly used for the fronts as in Fig. 3. Dovetails are used here, but they are of a very coarse type and run right through, a poor way of doing the job because the exposed end grain at the front offers a poor grip for the veneer. Furthermore, the joint eventually shows through at the front, owing to the front shrinking and leaving the dovetails standing up. The example is interesting, however, in that the sides as well as the front are rebated to hold the bottom.
FIG. 4. EARLY 18TH CENTURY. This shows the first attempt at cutting really neat dovetails.
As men’s skill increased they began to make altogether neater dovetails, and in the next example in Fig. 4, which dates from the early 18th century, an altogether more refined construction is apparent. In fact it is really the beginning of the modern way of drawer making. Note how the dovetails are lapped at the front, and how narrow the pins are between the dovetails. They run almost to a point. A rebate is still worked in the sides to hold the bottom, but the interesting point is that the maker has realised the desirability of raising the bottom slightly to prevent it from sagging and rubbing on the drawer rail. He has accomplished this by making the rebate extra deep and fitting a slip beneath. The advantage is two-fold, for, apart from raising the bottom, the slip gives a wider bearing surface and so reduces wear.
The practical reader will realise that the front is rebated to hold the bottom, this being evident from the half-dovetail cut at the bottom, the purpose of which, of course, is to conceal the rebate. At the back the bottom passes beneath the back.
Note incidentally that the grain of the bottom runs from back to front as in the previous examples.
In this particular example a cocked bead is fitted around the front, rebates being cut all round to accommodate it. As a matter of passing interest we may note that later in the century the rebate was worked at sides and bottom only. At the top the bead extended the full width so that no joint was visible. The small sketch at A, Fig. 4, shows how a slip of walnut was let into a rebate when a projecting thumb moulding was needed. The veneer at the front concealed the joint.
FIG. 5. SECOND HALF 18TH CENTURY. This is in mahogany and is similar to Fig. 4 but bottom is grooved in.
Fig 5 dates from the late 18th century, and here the bottom is fitted into grooves in sides and front. Otherwise the construction is similar to the previous example. We may consider here why the groove was used in place of the rebate. It has been pointed out that in previous example’s the grain of the bottom ran from back to front, and the reason for this was that, since most drawers were wider than they were deep, the grain ran across the shortest distance and was therefore stronger. This meant, however, that the bottom was most liable to split owing to the great width running across the grain. By grooving the sides and allowing the grain to run from side to side, there was no need to fix the bottom except at the back. It was thus free to shrink without being liable to split. In any case, there was less distance across the grain to shrink. To prevent any sagging in wide drawers a centre muntin of stouter stuff could be fixed. The likelihood of this being the reason for the change is shown by the fact that in nearly all drawers in which the grain of the bottom runs from back to front and is fixed rigidly the joints have opened.
FIG. 6. 19TH CENT. Alternative drawer slips.
There was, however, one weakness in the grooved sides. They were weakened by being cut practically half-way through, and the only bearing surface was that of the drawer side thickness. This accounts for the introduction in the early years of the 19th century of the drawer bottom slip moulding, A, Fig. 6. The side was not weakened and the bearing surface was approximately doubled. The alternative form was introduced later.
