Category: Roman Workbenches

Today I turned the first four legs for the low Roman workbench but I left my lathe set up because I might need to turn four more tomorrow.

While most images I’ve come across of Roman workbenches show them made with four legs and with staked construction, there is one image from Herculaneum that looks a lot like there are eight legs. The image below is actually an 18th-century drawing of a painting from that doomed city that has deteriorated – all we have left is this image (plus a host of wrongly interpreted images).

Suzanne Ellison and I are fairly convinced this image is accurate for a variety of reasons that I’ll discuss in the forthcoming book. But even if the 18th-century image is accurate, a few experts have suggested that the Roman image it is based on is inaccurate.

It’s a bit of a hall of mirrors to discuss. But the bottom line is that some people think that the Herculaneum image actually depicts a work surface on top of two four-legged sawhorses. I tend to disagree. But there is no way to settle the issue.

So what I’m planning on doing is making the bench with four legs, like this image from Pompeii.

Then I’m going to work with the bench and see if I think it needs four more legs to be robust enough for operations such as hand mortising.

So stay tuned.

These four legs feature 4”-long tenons that are 1-3/4” in diameter. I roughed them out on the lathe and then shaved and scraped them to the finished shape. I like the process and the finished texture.

— Christopher Schwarz

I only want to work each surface of this benchtop once. The slab weighs enough (approximately 115 pounds) that I have to struggle with it to get it in and out of the vise and onto the benchtop.

So every move with the slab is planned with care so I don’t end up injured or (at best) embarrassed at having to ask a friend to help me get the benchtop off the floor.

Today I dressed the two broad faces of the benchtop: the underside and the benchtop itself. Both have to be fairly flat and free of twist in my experience. Of course, this slab turned out to be a weird one. Typically the bark side of a slab will be concave across its width, and the heart side will be convex. This slab was reversed.

So I started on the convex face. Normally when I dress a rough convex face I remove the hump in the middle using with-the-grain strokes with my jack. But because I had a lot of wood to remove (about a quarter of a thumb), I used a different tactic. I traversed the hump alone at first and stayed away from the long edges of the benchtop. Traversing allowed me to take a bigger bite with the jack plane.

Once the hump was gone, I checked the top for twist. It was indeed twisted. So I used my jack plane to work away the two high corners, which were diagonal from one another.

When the underside was flat (according to the winding sticks), I dressed the entire underside of the top with the jack to leave a consistent and tidy (if scalloped) surface.

Then I flipped the benchtop over to work on the concave side.

Because I knew this surface was also twisted, I began working away the two high corners straightaway. After bringing all the corners into the same plane, I dressed the surface with a jack. I’ll probably dress it with the jointer in the morning and leave it like that until after assembly – that’s when I’ll clean up the leg joints protruding through the benchtop and tooth the surface with a toothing plane.

I also managed to rough out the legs on the band saw today and hope to turn them on the lathe tomorrow.

If this sounds like a lot of work, it’s not. I have logged only two full hours of shop time. How much more time I’ll log will be determined by whether I decide to build an opossum or an arachnid.

— Christopher Schwarz

This afternoon I got a good start on my first Roman workbench – a knee-high bench with almost no workholding, aside from holes for pegs or holdfasts.

I’m building it using a red oak top from Will Myers, who dried the slab in his homemade kiln in North Carolina. The legs are some white oak stock that is sold at the lumberyard for making rustic mantles. (I was going to instead use some firewood I have in my shop, but that firewood is actually going into two upcoming commissioned chairs.)

The real fun part of the project is the measurement system. Thanks to Brendan Bernhardt Gaffney at burn-heart.com, I have a Roman ruler to guide me as I design and build these two workbenches. I’m using his Cubitus Ruler, which combines several Roman systems onto one pretty stick.

So here is the cutting list for this first Roman workbench:

1 benchtop, measuring 3.4 thumbs x 14 thumbs x 4.9 cubits (or 87.8 thumbs)
4 legs, measuring 2.2 thumbs x 2.2 thumbs x 1.25 cubits (or 21.3 thumbs)

Before you do the math, just think of the cubit as the distance from your elbow to the tip of your middle finger. And the thumb as the length of the second segment of your thumb. That’s accurate enough.

Today I dressed the front edge of benchtop with my jointer plane, making sure it was square to the benchtop (the benchtop is the heart side of the slab, FYI). Then I marked the final width of the benchtop using a large square – my panel gauge is in my other shop.

That’s when I found that I had to remove almost 1/2 thumb of wood in places to make the front edge and back edge parallel.

I looked for my hatchet. Dangit. It’s also in my other shop.

So I decided to traverse the edge with my jack plane. After marking the final width of the benchtop, I use my jack to create a chamfer on the corner that touched the line that represented the final width of the benchtop. The chamfer acted as gauge – as the chamfer disappeared I knew I was closer to my finished width. It also protected the corner from spelching during the traversing.

This dodge worked surprisingly well.

Tomorrow I’ll dress the benchtop and start shaping the legs.

— Christopher Schwarz

P.S. The book “Roman Workbenches” is unlikely to have any photos because we are printing it via letterpress, so I’m not sure why I’m documenting every step. Old habits die hard, I suppose.