You might look at this elegant five-rung side chair, and declare that you'd seen it before on this website. At first sight, it seems identical to the side chair (No 4a) illustrated on the Ladderbacks page that is, essentially, a full-size armchair without the arms. But this chair is different. The dimensions of its seat is about 90% of the size of the seat on the armchair. This makes the back narrower, so the rungs of the back are scaled down accordingly and, because the overall height is the same, this chair appears taller and, arguably, even more graceful.

One of the two bills of sale that have survived from Clissett's workshop details two types of armchair. The "Ordinary size, Low back, Rush-seat Chairs, without Arms" is almost certainly the small three-rung side chair, No 6 on the Ladderbacks page. The others are "High back, Ash Rush-seat Chairs without Arms, second size". Until now, I had no idea what was meant by "second size". Now, it seems entirely possible that this new chair is the second size, while the full size side chair is likely to be the first size.

Another little mystery potentially solved.

I've long puzzled over how Philip Clissett bent the curved parts of his chairs. Modern woodworkers would tell me that he must have steamed or boiled the parts before bending them in some sort of jig (you can watch Mike Abbott do this in a Youtube video) . They would then have been retained in a similar jig while they dried out.

We have two quite detailed accounts of how Clissett made his chairs. Both are incomplete, and only one comes from a known source – an interview with Clissett's grandson, Will, who made chairs alongside his grandfather for nearly twenty years. Neither account mentions steaming or boiling. Chair legs were simply bent, apparently cold, in a jig.

It was green woodworking guru Mike Abbott who gave me a clue to what might have been going on. He asked me why there was such a difference between the back legs of Clissett's chairs and those produced by Edward Gardiner and his followers, Neville and Lawrence Neal – who make copies of Clissett's chairs. In contrast to the curve of Clissett's, the legs on the other chairs have a distinct kink. You can watch Lawrence bending chair legs in a short video of him at work. This is instructive, because the jig that Lawrence uses is more or less identical to Clissett's, and should produce the same results – but doesn't.

You will see that Lawrence boils the legs, then bends them over a fulcrum in the jig. My thought was that the hot, relatively soft wood bends at the fulcrum resulting in this kink with quite straight timber above and below. On the other hand, if Clissett bent chair legs cold, then I could imagine the stiffer timber resulting in a relatively smooth curve over the fulcrum. In one of the accounts of his methods, there is a graphic description of how the jig containing eight bent legs was placed in a cooling bread oven for twenty to thirty minutes, then left in a warm place to dry out. I believe the high temperature in the oven would result in the same softening of the timber that is caused by steaming or boiling.

There is no question that the oven trick will set bent green wood. You can try it with small stuff cut from the hedgerow, and it works convincingly. The big question is, can you bend a chair leg (1¼ inches thick) without heating it first?

Mike Abbott was doubtful. But he kindly, and very open-mindedly, gave me a chance to find out. He made a leg slightly thinner than he normally does, adding 10% to the usual diameter of a Clissett – then gave it to me to try to bend. To his surprise, and mine, it went very easily. He then tried it with one of his slightly thicker pieces – again, no problem.

There are several books out there that deal with bending wood, and some of them refer to cold bending. All seem to consider it of limited use, mainly because of the tendency for the wood to regain its original shape soon after release from pressure. But none seem to consider the use of heat after bending. So maybe Clissett was using a method that has since been lost.

Cold bending would be of no use for the dramatic curves of Windsor chairs but, for gentler bends there is clearly some potential – if a woodworker has access to a suitable source of dry heat. Clissett was, of course, using heat that would otherwise have gone to waste, so keeping his costs to a minimum. We still can't be certain that Clissett didn't steam or boil chair parts before bending, but we now know that he could have done it without – and if he didn't need to steam or boil, why would he have done so?

You can read Mike Abbott's account of this experiment, with photographs, on his blog.