Skip to Photo Index

The Curious Case of the “W” Bows

WE Hill and Sons bow makers

From the outside, selling bows may look rather easier than falling off a log, but the reality is that every buyer has their own unique set of preferences.

There are the thoroughly laudable demands of musicians that a bow be stiff but not too stiff, neither tip heavy nor tip light, grippy but not gritty etc; there are the expectations of collectors that a bow be unused yet of proven musical value, that the lapping be original tinsel yet the weight conform to modern norms; there are the strange and unaccountable prejudices that attach to certain makers, and there’s clear preference for bows by makers whose names are easy to pronounce…

Once these various hurdles have been leapt over and a buyer decides they like a bow, the dealer must now contend with an army of people, hidden up until this point but gathering in the wings for the express purpose of offering second opinions. Rival dealers, trusted restorers, desk partners and quartet colleagues, spouses… all will have opinions. Even close friends of the buyer who claim to be tone deaf will have their say.

At some point I started to feel that my strike rate was poor, and that I was showing an awful lot of bows before I succeeded in selling one. I decided to conduct a straw poll of colleagues, just to see if my experience was normal, or whether I was attracting a particularly demanding clientele. Far from it, it turned out – it would appear that amongst all my friends and colleagues, for every bow sold around 10 will have been taken out on trial and rejected.

So it was a pleasant surprise to be phoned up the day after showing a lovely gold Hill to a client and to be informed that the bow had completely hit the spot, and that the funds should already be in my account. I congratulated myself on having shown a bow that was in near perfect condition, of the right period, and which played excellently. Surely after all these years I was finally getting the hang of this!

A week later I got another call, this time to inform me that the bow was to be returned, and that the client would please like their money back.
Of course, I assented. What seems to be the problem with the bow?
Oh I showed it to XXX who does all my rehairs and he pointed out that it’s a W.
A W? What’s that?
It’s got a tiny W branded onto the back of the head. XXX says that means ‘weak’ – apparently in the 30s Hills made lots of bows with the grain set the wrong way. They all broke so they were recalled and then pinned. This really isn’t the sort of investment grade bow I was looking for…

Well, hell hath no fury like a violin dealer whose sale has been killed. The bow I had shown the client had no head break, I had been over it with a UV torch and knew it to be in perfect condition. I determined there and then that the whole W thing was so much nonsense. I had never heard of XXX who must surely be an opinionated fraud, I had never heard of W bows, or indeed of a Hills’ recall. I did have some dim recollection involving Hill bows with original pins in the head, but I assumed this was some belt and braces innovation that Hills had tried. The Hill workshop had always struck me as a kind of engineering workshop that happened to make bows, and I had filed this piece of information away as being commercially unimportant, rather like the original splines to be found on the heads of some very early French bows.

Nonetheless, I went and picked up the bow, and sure enough, almost invisible to the naked eye, there was a tiny W stamped into the back of the head just above the face plate. And on returning to the shop, I was dismayed to go through our Hill bows and find another ‘W’.

WE Hill and Sons bow makers

Yet neither had a head break, and both were wholly intact after almost 100 years. It was conceivable that in both bows the grain was running more down the head than across it – but what was the big deal there? Bowmakers regularly discuss and disagree on which grain orientation gives the better spring in a bow stick, and some of the greatest French makers produced very successful bows with slab cut blanks.

Slab-cut head vs quartersawn head
WE Hill and Sons bow makers

The assumption that a slab-cut blank would lead to a tendency for the head to shear off, and that it would therefore benefit from a hidden pin, rests on the premise that wood fibres are harder to separate along the orientation of the annual rings than along the radius of a log. And that the natural tendency of wood to split along a radius (shown as a dotted line in the diagram above) requires a bow maker to orient that line of weakness down the head rather than across it where it’s very narrow.

But is this in fact true? Is wood stronger in one orientation than another?

This was an easy question for me to answer. Before getting involved in violins and bows, I ran a business as a cabinet-maker and carpenter, making wooden objects varying in size from jewel boxes to large house extensions, all from trees acquired from the estates around my home in the Scottish Borders. We took wind-blown trees or other trees that needed to come down and processed them in situ on a mobile sawmill. All the branch wood or lower grade stuff was converted into firewood, and I spent much of every winter hacking logs. To say that I have an intimate understanding of the splitting qualities of timber would be something of an understatement.

In most wood species, there is a slight difference in the ease of splitting on the radius of the log or across it. When the wood is wet, the difference is marginal – we tend to use the axe on a radius or clean across the diameter of the log simply because that allows for more efficient working. Also the wood splits more cleanly which gives a nice feeling. When the wood is dry, splitting becomes much harder generally as the fibres have hardened and fused, but splitting along the radius seems significantly easier. However, this is more because for a round log, the process of drying out creates checks or splits in the timber that follow the radius. And this in turn is because of differential shrinkage – wood expands and contracts according to its moisture content, and the rate of movement is around twice as great along a growth ring as along the radius.

Differential shrinkage
WE Hill and Sons bow makers

In this diagram A represents a line travelling along the growth rings of a log, B represents a line travelling at right angles to it. As a rule of thumb, wood expands and contracts twice as much on the A axis as it does on the B axis.

Obviously the point where the wood experiences the greatest change in moisture, and therefore the greatest movement is in the initial seasoning, from ‘green’ or living wood to dry wood, where its moisture content is the same as that of the air around it. If this seasoning process takes place in dimensioned timber or billets, much of the stress is relieved, but if the wood is left in a log it results in big splits, always along the B axis.

But back to Hills…
My indignation at the rejection of my unusually fine bow was turning into an obsession. So, like everyone suffering from a well-developed sense of injustice, I headed for the internet in order to corroborate my unfounded beliefs. There was a surprising amount of discussion about Hill bows with original pins … for example from a discussion on Violinist.com:

“There’s an interesting story about Hills that they once ran out a series of bows with the grain of the wood running the wrong way (slabbed to the side rather than quartered). In an attempt to reinforce them, they pinned every one they could get their hands on. Subsequently, the only ones that broke were the ones with pins.”

Or an article about Bill Watson by Tim Ingles :

“From our first meeting Bill was keen to talk about Hill bows and to share his knowledge – he taught me about the maker’s marks… the introduction of date stamping the bows, the inverted W on the back of the head on a bow which had been ‘cut wrong’ and subsequently pinned, the changes in handle design in the 1920s etc…”

There were references on a couple of dealer websites (some dealers priced these W bows lower, some not), but all with contradictory information – did these pinned bows break or didn’t they, how would bows have been recalled, did W really mean ‘weak’? It seemed that the same illogical tale was popping up in various places, so there must be a source…

Derek Wilson was the person who managed to shed the most light. As co-author of The Hill Bow Makers and one of the last generation of Hill makers (he worked at Hills between 1978 and 1985) he can be regarded as the horse’s mouth. It turned out he had also had some discussion about these bows with Bill Watson:

I wrote a little about this in my book under Frank Napier’s biography. There are different opinions about these bows… Bill Watson said it was a mistake to do the pins and devalued them a lot.

I think that it’s all part of the history and they got as many bows out of the log as possible at that time … of course it’s better not to have a pinned one, but as I’ve only seen two or three that have broken in the last 40 years, either the pins did the job well, or they were not necessary.

It’s fair enough for Bill to have had a different point of view, because the bows that were weak may have broken during his time, which has now left the bows that don’t have a problem. I would personally only devalue by a minimal amount…

I don’t think I’ve ever heard of the ‘w’ meaning weak or that they were ‘recalled’.

Derek’s biography of Frank Napier further fleshes out the picture. From The Hill Bow Makers, page 31:

Napier also tried to make more efficient use of raw materials, calculating that more sticks could be cut from a log if the planks were cut ‘on the slab’ in simple, stacked layers. Traditionally the wood was cut ‘on the quarter’ leaving the annual rings always to cross the head, but this yielded fewer sticks per log. In the 1920s many sticks at Hills were cut ‘on the slab’ before the results had been properly assessed. Some bows began to be returned to the shop with broken heads. To prevent further failures the workshop manager, with the agreement of the Hill brothers, introduced a strengthening wooden dowel to secure the head repair. These bows were marked “W” on the back of the head (and usually also had a punched dot beside the match mark and date stamp on the handle). Others had a brass dowel inserted and were marked “N”. Retford and Bultitude disapproved of this ‘joiner’s solution’, though it seems to have been effective, since few bows with dowels were found to fail. When it was pointed out that French bow makers had nearly always used wood cut ‘on the quarter’, which did not suffer this weakness, the system of slab cutting was abandoned.

There’s also a further footnote clarifying that in Bill Watson’s opinion the practice of fitting dowels in slab-cut bows continued into the 1930s. This is borne out by the image below, which shows the handle of a “W” bow with a date stamp of 1931 and a dot or awl mark indicating that the bow was doweled.

WE Hill and Sons bow makers

Here is a “W” bow from the 1920s, too early to have a date stamp, but with a clear awl mark.

WE Hill and Sons bow makers

So the “W” denotes a wooden dowel, presumably inserted in the head through the head mortise and stopping short of the top of the stick. This dowel was to serve as a reinforcement for a bow which might break, rather than as a repair to a bow which had already broken. The dowel’s usefulness was debated at the time and remains unproven. 90 years on, it can be viewed either as a fault or as an interesting historical feature.

One thing that we can’t say with any certainty is whether any given bow has fared better for being pinned than it would have fared without a pin. It also seems clear that a W bow that hasn’t broken isn’t going to break – or at least, it’s no more likely to break than any other bow.

To me, Derek’s account seemed primarily to be a story of workshop politics. Frank Napier, with his rather industrial outlook and his penchant for efficiency, must have questioned the practice of quarter-sawing Pernambuco logs and decided to try plain-sawing them – a reasonable enough idea since it saves a bit of wood and it saves a lot of time… even without knowing of the existence of historic bows cut on the slab, it must have seemed a rational course of action.

Without going into detail about the difference between true quarter-sawing and partial quarter-sawing, here’s an illustration of the two methods at their most simple.

Quarter-sawn vs plain sawn
WE Hill and Sons bow makers

We can see in this diagram that plain sawing, where we mill down the log in regular lines, is much easier – it also results in wider boards and seems to maximise the available timber since there are no nasty little wedge shaped pieces left over. But much of the resulting wood, specifically the boards cut from the outer part of the tree, has the growth rings travelling along the width of the board rather than bisecting its width. Quarter-sawing is finicky, it results in narrower or less standardized boards, and there’s a certain amount of wastage – or at least, there’s a definite ratio of time to wastage that has to be considered.

Napier doubtless made these calculations, and determined that from the perspective of a Henry Ford style production line, quarter-sawing was inefficient.

We know that the Hill shop was a cantankerous place, and that relations between the workshop manager and the makers were often strained – it’s easy to imagine a workshop manager taking exception to Napier’s fancy new ideas, being alarmed by failures or returns, and insisting on a belt & braces procedure which would reinforce the workshop hierarchy as much as it reinforced the bows.

As an aside, the myth of a ‘recall’ seems to be a misunderstanding of the fact that some of these slab cut bows were returned because of head breaks. It’s hard to imagine how Hills would have managed a recall in the 1920s, or that it would have been thought necessary for what was at the time an inexpensive accessory.

At some point in my indignant phase (which lasted quite a while), a colleague who also had a Hill “W” bow pointed me in the direction of Richard Sadler’s book Arthur Bultitude and the Hill Tradition. Reading this was a revelation, though not for the reasons the author intended!

Arthur Bultitude’s early grounding in bow work coincided with William C. Retford achieving the first glimmerings of insight into a serious problem that was baffling the Hill Bow Shop during the mid-1920s. Heads had been “dropping off” Hill bows for no apparent reason and returned as faulty. No one at Hill understood why this was happening, cryptic utterances on wood being “cut the wrong way”, relayed by Charles François Langonet on return from holidaying in his native France, were met with blank stares.

In an effort to pinpoint the problem, William C: Retford commandeered Bultitude to assist him with a study of some two thousand bows. Bultitude recalled that Retford and himself received some prompting in their deliberations from “an old boy in Hanwell who made ladders and wheelbarrows – he was about a hundred! He would sit in the front garden and talk to people – I think he was beyond working when I met him…”. In the course of conversation, this wise old ancient told Bultitude that the wooden rungs for scaling ladders were always made from wood that had been “cut on the quarter” for added strength.

Bultitude duly relayed this conversation back to Retford. He later felt that it may have given Retford some idea of where to start looking, when they began this period of extensive practical examination.

A conclusion was reached that the first-rate bow makers had indeed made their bows from wood cut ‘on the quarter’. Hill had, earlier in the 1920s, switched to using wood that was ‘slab cut’, on the basis that substantially more bow sticks could be obtained from each Pernambuco log sawn in this manner.

Frank Napier, whose idea this may have been, apparently still felt that the strongest – ‘stiffest’ – sticks were those fashioned from ‘slab cut’ wood. However, that research conducted by Retford and Bultitude meant Hill bows were henceforth made only from wood cut “on the quarter”; a knotty problem had been solved.

The strangest thing we learn from this account is that no-one in the Hill shop at the time knew anything about the properties of wood. Nothing illustrates this more perfectly than the concept of ‘quartersawn ladder rungs’, since these are of course cylindrical and are by definition both quarter-sawn and slab cut. That Retford and Bultitude should be reliant on an old boy in Hanwell to explain the nature of wood, albeit in the most half-arsed and surreal manner, is almost unbelievable. Or that 50 years later, when talking to Richard Sadler, Bultitude should reveal his continued ignorance so disarmingly… it’s all a bit reminiscent of Eric Newby climbing in the Death Zone of the Nuristani Himalayas in a tweed jacket and a deerstalker.

It’s also quite telling that Langonet, a violin-maker descended from a line of violin-makers, seems dismissive of his colleagues in the bow department, and equally telling that he did have some understanding of wood. Looking at the biographies of the Hill bow makers, it’s very noticeable that at this time, no-one in the workshop had had any kind of apprenticeship in bow-making. All had come from apprenticeships in case-making or general dogsbodying, very few even had carpentry skills. Yeoman came from the Hill case-making shop, William Napier had been making packing cases for Broadwood pianos, William C. Retford was a country boy with some basic joinery skills, Johnston, a piano-maker…

In a way, they were re-inventing bow-making along industrial principles without really understanding the materials they were using. It’s also clear that their understanding of historic bow-making was minimal, and that these failures of design gave Retford and Bultitude the bright idea of looking at some French bows, as if for the first time.

To get some kind of alternative perspective on this, I asked Pierre Guillaume how he treated the question of slab cut versus quartered Pernambuco. Pierre trained in Mirecourt, and is steeped in and representative of a tradition that goes back to Tourte and beyond. He explained that while quarter-sawing was an ideal, he would use any blank, even one cut on the slab. The simple way to test the likely strength of the head was to give it a firm twist before starting to make the bow. If the head fails you put the stick in the bin, if it doesn’t then you make a bow. And if the head doesn’t snap off before you start work, it won’t snap off later.

This very pragmatic approach could have saved Hills a great deal of grief – what a pity they didn’t have any bow-makers to advise them.

My own theory is that slab-cut bows are only compromised if there are undetected drying checks in the wood. The area just below the head is vulnerable, and there’s a good deal of twisting force applied to it in the process of playing. Any failure of the fibres would weaken an already precarious structure. But equally, without such flaws, a slab-cut bow should be as strong as a quarter-sawn bow. And a pin or dowel will make no difference.

So the case of the “W” bow reveals a larger truth about Hill bows. The only maker associated with Hills since James Tubbs who had any contact with traditional bow-makers was Sam Allen, who seems to have learnt a lot from Emile Bazin. But from then on, a procession of bright or willing young lads who were good with their hands were given the task of reverse engineering the bow! The most remarkable aspect of this is how well they succeeded given their staggering ignorance when it came to the properties of wood.

The way that the Hill shop went about bow-making seems a long way from the sort of traditional apprenticeship that French makers underwent. And obviously some well-established norms or practices of the kind usually handed down through the generations in the apprentice system simply escaped them. However, they did apply a new rigour to the manufacturing process and to quality control, and this is why Hill bows have enjoyed such incredible success. It’s almost unheard of for a Hill stick to warp or go out of winding, the mounts never fall apart, and the weight and balance of a Hill bow tends always to be optimal.

In a way these “W” bows were an inevitable and brief glitch in the development of an otherwise excellent system. It turns out they have a lot to tell us about the Hill shop between the wars, and about the history and evolution of the 20th century English bow. For me they are Hill bows like all others, and if they are still intact in the 2020s then I see no reason to devalue them or to think less of them.

WE Hill and Sons bow makers

Article by Martin Swan, January 2023


Photo Index: WE Hill & Sons