The Cornelius Cabinet of Curiosity #4

One of the more delightful aspects of being a restorer, collector, and scholar of old flutes is the often peculiar way that items come to one's attention, and hopefully possession.

Every collector at some time comes across fragments which, fascinating as they might be, would be even more so if we could only see the complete instrument.

Crowley's music shop on MacCurtain St. was a Cork city institution for many, many years until it closed in 2013. From a traditional musician's, and indeed woodwind enthusiasts point of view, the story is enriched by the knowledge that the father of the owner was the renowned uilleann pipe maker Tadgh Crowley (1899-1952)

I got to know Mick Crowley in the early 1980s, initially I think, in connection with buying some African blackwood, and other timber that had survived from the old pipe making workshop, which had closed in 1966. Many of the pieces weren't suitable for flute making, but we struck up a relationship which lasted until his untimely death in 2010. Mick was a legend in Cork music circles, and you can read more about his shop and its famous customers here.

At sometime in the 80s, Mick showed me an old flute, which presumably had also come from the pipes workshop, and which would probably have ended up in the hands of a traditional player, were it not for a combination of it's unsuitability for trad music, and a scarcity of flute players in the Cork city of the time.

At that stage I was really only beginning my journey in the flute world, and all I could tell about the instrument was that it was a "frankenflute" made up of bits of different instruments. The body and the head, were that of Rudall & Rose #12...and before you all have heart attacks I have to tell you that it was an 1832 Boehm system, but of course a very early one of those as well.

The foot however is what I'm taking out of the Cabinet to show you today.

This is the front view, but you need to see it from the side to begin to appreciate what's going on here.

And what the dismantled keys look like.

  

It's with this side view that you begin to really see how odd this mechanism really is.

For the moment lets set the Eb key to one side, which leaves us with something which is basically mechanically similar to all 19th century pre-Boehm simple system foot joints. Two touches for the right little finger operate interacting keys, which close open standing keys for C# and C. The interaction of the keys ensures that when the player wants to play C, then the C# remains closed. In order to facilitate this essential action the touch of the C key must somehow "get around" or bypass that of the C#. This can be achieved in several ways, usually by bending the C key around or over the C#, as in these examples - on the right, a Potter from the 1790s with the 'up and over' system, and on the left Rudall & Rose #1376, with the by then standard "around the side" method 

In the foot under consideration though, the C# and C touches run parallel until they engage with the open standing pewters, and this leads to some very interesting, if bizarre, design features. The parallel nature of the touches essentially means that the C# and C holes cannot be in line, as they normally are. The obvious solution to this would seem to be to move the holes slightly to the side...one one way, and one the other, but this is not possible. Why? Because the bore of the foot is small and the C and C# holes are large, in fact just about the same as the diameter of the bore, so that moving the holes in the same plane ( at right angles to the radius of the bore) would mean that the holes wouldn't meet the bore without a little step to one side or the other. That's flute maker's talk, but here's a diagram that will hopefully make all of that clear. 

At A, you can see the walls of a hole that's almost the same diameter as the bore in the solid line, and then how a hole of the same diameter meets the bore, when the hole is shifted in a radial manner. ( the dotted lines )

At B, you can see how it's impossible to move the hole laterally and still line up with the bore.

The solution in this case was to move the holes, not in a flat plane, but around the circumference of the bore, ( as at A ) meaning that they now meet the bore without distortion. The following images attempt to show this.

Keys removed

Pewters in place

Fully assembled

This "twisting" of the mechanism is achieved by two means. 

Firstly the linkage between the touches and the pewters, being simply of the "peg in hole" type allows the plane of action between the two to be as variable as required.

Secondly, and much more complexly, the circumferential movement of the holes requires that the axles of the pewters are tilted away from a central line, one in each direction, and this is achieved by having a tall and a short pillar holding each axle. I hope the following image illustrates this.

The condition of the foot as it came to me was not mechanically good. The key touches were somewhat bent, but the major issue was that the axle rods were corroded in place. I eventually managed to remove the keys but at the cost of destroying the axles, so the assembly you see in these photos was achieved with push rods that I drew down from silver wire. As a result, the mechanism doesn't really work as I write this, but I've no reason to believe, that in it's original state it wouldn't have worked as well as any standard simple system foot joint...and this is as good a juncture as any to point out that the key work on this foot gives absolutely no mechanical advantage over the normal simple system "in line" C# and C keys.

Time now to look at the Eb key which fairly obviously seems to have a separate origin, and prompts me to reveal, that though the body discussed above was clearly stamped Rudall and Rose, the foot was clearly stamped as this image shows,

T. Prowse

Hannway St.

London

A close up of the key itself reveals several interesting features.

Remember that this foot was attached to a Rudall & Rose 1832 Boehm system, and here's the Eb key from one of those ( #18 in this case, which must have been made within weeks of #12 )

Note that the upper pillar of the Eb is in both cases an integral part of the ring, and that both our foot and that of the R & R 1832  are silver lined. ( R & R 1832 to the right)

I've spent many years now looking at this foot joint, and the more I look the more mysterious and inexplicable it becomes.

Does the stamp give up any clues?

Prowse worked in London from 1816-1868, so this is certainly contemporaneous with the body.  So could it possibly be a repurposed Prowse footjoint? That initially seemed a likely explanation, but had to be rejected due to what was pointed out above about the circumferential position of the C# and C holes. Secondly, the Eb is on the wrong side of the flute for a standard simple system foot.

It seems obvious though that the Eb key has been repurposed. Looking at the photo above you can see how the lower end of the Eb axle doesn't have its own post, but uses the tall post of the C# pewter, where a hole has simply been drilled to take the end of the Eb axle.

So the foot is clearly not part of the original flute, or at least not all of it is. The element that might be part of the original flute, is the Eb key, or at least part of it, as the touch appears to have been modified. 

There's also the possibility, given the stamp, that the Eb key could be from an 1832 Boehm system made by Prowse himself, which strongly resemble the R & R instruments, and in fact from what I can make out from the photo of a Prowse 1832 Boehm system in Robert Bigio's  "Rudall, Rose  & Carte-The Art of the Flute in Britain" (see Fig. 94, p. 91) this key may in fact resemble that of a Prowse 1832 Boehm system rather more than the R & R key pictured above.

Apart from the stamp, and the Eb key, there are certain other features which strongly implicate Prowse as the maker of this foot. Although the key work is very unusual, certain aspects of it replicate exactly Prowse's work as seen on other examples of his instruments...particularly the touches of the C and C# keys.

Here they are on the mystery foot

 and with the linkage separated...

Here's the same thing on a Prowse Nicholson

Improved #3929

                                                  

And then those same keys  dismantled

Compare how the interlinking of the C and C# touches compares on the two flutes, and I think you'll have to agree that they were made by the same hand. The way in which the C# touch (to the left) is shaped to fit under the C is absolutely characteristic of all the simple system flutes that I've seen by Thomas Prowse. 

The Nicholson Prowse is to the left, and our foot joint to the right.

             

There is no evidence that the touches were soldered onto some already existing key work, so the only conclusion that I can realistically come to, is that this foot was made in the Prowse workshop, presumably to replace the original foot of Rudall & Rose #12.

 But why would the owner of the flute choose to replace the original foot with something so unusual? Why not get R & R to simply make a new foot? The presence of what might be potentially the Eb key from a Prowse 1832 Boehm system, and the other elements of the keywork which strongly suggest the foot was made in the Prowse workshop still don't explain the extreme eccentricity of the construction.

It could be postulated that for some reason that the owner asked the Prowse shop to make a new foot, but then why repurpose an old key (the Eb), which I believe was not part of the original flute. And who designed the apparently unique system of operating the pewters, which as I've pointed out is of absolutely no mechanical advantage over the standard system.

I think the only conclusion to come to is that this was the flute of a wealthy amateur who had some sort of connection with the Prowse workshop.

How it ended up in Cork, in the workshop of a great uilleann pipe maker is a story that remains to be told.

 

The Cornelius Cabinet of Curiosity #3

 Rudall & Rose, who although in terms of 8 keyed flutes at any rate, produced a very recognisable style of flute over a period of almost 100 years, were also notable for undertaking to produce a very wide range of different flutes...if asked to do so.

Cabinet #2 was such an example, although perhaps, given that this was a very early example, it could be claimed that they hadn't developed their house style to the extent that they later achieved.

This flute though, is much later, and can be accurately dated, and although it has strong elements of the house style has also many idiosyncratic features.

This is Rudall & Rose #5631, address as stamped is:

                                            38 SOUTHHAMPTON ST.

                                                        STRAND

                                                        LONDON

Which rather conveniently dates it to 1847-50

Pictured like this the flute appears perfectly normal, but a view of the whole instrument shows the "curiosity" value.

As you can clearly see, this is a five keyed flute, with keys for Bb, G#, short F nat, Eb, and low C#, made at least 50 years after 8 keys had become the standard set up.

In fact this particular arrangement of keys occurs in some of the very earliest flutes with multiple keys from the 1760s and 70s, such as those by Potter Senior for example.

Why was this flute made, and for whom? Unfortunately, we'll never know.

What is remarkable is the condition of the flute. It was bought at auction in 2007, and received only the most superficial of restoration, pads, rethreaded joints, and and a light polish to the keys. This indicates to me that the flute has had very, very little use. It came in the original case, which was also almost pristine.

I've always believed that a good indicator that a flute has seen little use is the condition of the stamp on the upper middle joint, since it's just where one holds the flute to assemble and dissemble it. In this case it's absolutely crisp.

The condition of the flute leads me to believe that a very short time after it was acquired it was basically abandoned as a playing instrument, but had obviously been carefully stored.

Surprising though that it should have been "abandoned" in any sense as it's one of the finest R & Rs that I've ever played.

Here's the embouchure...

And a couple of details of the key work, showing the short F tilted for easy access by R3

and the lonely C# key...

However I've saved the most unusual thing to last.

I was so impressed by this flute and its playing capacity that I took measurements, and discovered what is the really curious thing about this instrument, which, in fairness will only mean something to another flute maker or restorer.

The bore of the cone bore flute, to give a rather simplistic account, is not generally a straight unvarying taper, certainly not in a flute of this period. Again to simplify matters, each joint follows a very broad pattern of starting at the wide end with a steep taper, followed by a less steep one.

Contrary to many people's idea, this variation in taper was not produced by having a profiled reamer which incorporated the two tapers in the one tool, but rather by using two reamers, the one with the less steep taper first, and then opening out the upper end with the reamer with the steeper taper. 

In this case however, although the lower joints follow the "standard" pattern, the upper middle most remarkably demonstrates the reverse, in that the steeper taper is in the lower end of the joint, as in B here...

Having thought about this for some considerable time, I even went back and recently re-measured the upper middle to make sure that I hadn't imagined it...but no - it is what it is.

How was such a bore made? And probably more importantly, why was such a bore made?

It must be understood that this was not simply a matter of using the reamers "the wrong way round", and although there may be ways of doing this using two separate reamers, it seems to me that in this case probably a profiled reamer was used.

Apart from this anomaly, the bore is a standard one for a R & R of the period.

All in all, definitely a good candidate for the Cabinet of Curiosity.

A Wonderful Confusion

The question of whether the material that a woodwind instrument is made from affects the sound and what might be called the playability of the instrument is a topic that appears with great regularity on fora as different as on-line discussions and the scientific press. This often takes the form of comparisons of the characteristics of instruments made from wood, metal, or plastic, but does indeed often also discuss the merits or otherwise of different wood types as well.

Those within the Irish wooden flute community will be familiar with arguments about boxwood versus blackwood versus cocus, and I think every musician's experience is that the material does indeed make a difference, but how to characterise that difference, and how to demonstrate scientifically that it actually exists is another matter.

As a flute maker, timber is by necessity one of my major concerns, and in fact any woodwind maker will spend a lot of their time sourcing, buying, seasoning and preparing it. This post however is not so much about the nature of wood as a material, but rather about how much, or in many cases how little, we as makers actually know about it. So to begin, a brief history lesson.

Broadly speaking, the first woodwind instruments were made from natural tubes, and some very sophisticated instruments, mainly in the East, still use natural tubes today. In the West however, tubes made by boring out solid wood began, even by the early medieval period to become standard. Right from the beginning there seems to have been a realisation that harder, denser wood gave a better sound, and woods of choice would have been fruit woods such as apple and pear, but also boxwood. Boxwood (Buxus sempervirens), is found in England,  but the makers preferred wood from further east. Rendall says:

The better qualities come from Italy and Spain, the best, the mottled Abassian boxwood, from the shores of the Black Sea. The English variety, while tougher, has a fatal propensity to warping.

In terms of the flute, it is around the time of the development of the baroque flute, towards the end of the 17th C., that we begin to see what are generally called exotic woods appear. It is no coincidence that this corresponds with the great expansion of European trade to the Americas and the Far East around this time, and I'm inclined to believe that these woods were not initially imported with woodwind making in mind, but rather for the furniture, treen, and even dyeing trades. In fact it is thought that much cocus and grenadilla was obtained as a side product of sugar importation. Rendall, speaking of wood used in the clarinet trade, says:

Cocuswood ( Brya ebenus ) was long a popular material with English players and makers. It has every good property. It is very hard, resonant, easily worked, durable, and its high resinous content makes it very resistant to moisture and atmospheric changes. The best qualities come from Jamaica - it was long exported as dunnage in sugar ships- but is now very scarce. ( Dunnage? loose wood, matting, or similar material used to keep a cargo in position in a ship's hold)

It is these exotic woods, and the high level of confusion about their source and even identity that I want to discuss in this post.

When starting to research this topic, perhaps not surprisingly since I'm dealing largely with 19th C. flutes, I began with a source from that period.

This book is a classic, and has often been reprinted ( the image above is from the 2nd ed.) and for anyone involved in craftmanship of any kind, particularly that concerned with the restoration of any type of object made in the 19th C. I highly recommend it.

Note that this first volume deals with the basic materials used in manufacture, and interestingly deals specifically with the the very woods under discussion.

On a first reading, I was forcibly struck by two things. Firstly there was no certainty about what the woods actually were, and secondly and even more surprisingly, no certainty as to where they came from. Remember the date of publication -1843 (2nd ed. above was 1846) places it at the zenith of wooden flute simple system manufacture. But read for yourselves... here's an entry of interest to us.

BLACK BOTANY BAY WOOD called also African Black-wood, is perhaps the hardest and also the most wasteful of all the woods; the billets are very knotty and crooked, and covered with a thick rind of the colour and hardness of boxwood; the section of the heartwood is very irregular, and mostly either indented from without, or hollow and unsound from within; many of the pieces have the irregular scrawling growth that is observed in the wood of the vine. The largest stem of Black Botany-Bay wood I have ever seen, measured transversely eleven inches the longest and seven and a half the shortest way, but it would only produce a circular block of five inches, and this is fully two or three times the ordinary size. The wood when fresh cut is of a bluish-black with dark grey streaks but soon changes to an intense jet black. Of the few sound pieces that are obtained, the largest maybe perhaps five inches, but the majority less than two inches diameter. It is most admirably suited to eccentric turning, as the wood is particularly hard, close, free from pores, but not destructive to the tools, from which when they are in proper condition it receives a brilliant polish. It is also considered to be particularly free from any matter that will cause rust, on which account is greatly esteemed for the handles of surgeons instruments. The exact locality of this wood has long been a matter of great uncertainty. It has been considered to be a species of African ebony but its character is quite different and peculiar; I have how ever recently heard from two independent sources that it comes from the Mauritius, or Isle of France. Col. Lloyd says the wood is there called Cocobolo Prieto; that is not the growth of Mauritius, but of Madagascar to the interior of which island Europeans are not admitted and that it is brought in the same vessels that bring over the bullocks for the supply of food. The stonemasons of the country used splinters of it as a pencil for marking the lines on their work. It makes a dark blue streak not readily washed off by rain.

I have only met with one specimen of this wood in the numerous collections I have searched, namely in Mr Fincham's: he assures me that his specimen grew in Botany Bay and was brought direct from thence with several others, by Captain Woodruff, R.N.  As I have recently purchased a large quantity imported from the Mauritius, it is probable that this wood, in common with many others, may have several localities. It would be very desirable for the amateur turner that the wood should be selected on the spot, and the better pieces alone sent, as a large proportion is scarcely worth the expensive shipment but the fine pieces exceed all other wood for eccentric turned works.

One thing about the above account, which makes me as a maker with forty years experience of African blackwood, very suspicious, is the description of the colour of the fresh wood. I've never seen fresh blackwood that was even remotely of that description. 

Admittedly, in 1843, what we now call African Blackwood was not in great demand by the woodwind industry, but cocus was at that stage the major wood in use for flute making...or was it. Holtzapffel ( whose name, curiously, translates as Applewood ) again...

COCOA WOOD It is really singular that the exact localities and the botanical name of the Cocoa wood that is so much used, should be uncertain: it appears to come from a country producing sugar, being often imported as dunnage, or the stowage upon which the sugar hogsheads are packed: it is also known as Brown Ebony, but the Amerimnum ebenus of Jamaica seems dissimilar.

(Here the account goes on to describe specimens, with their local but not Linnean names, from various wood collections which are similar to Cocus.)

He finishes with

The cocus wood of commerce is not easy to trace to any of the trees of the West Indies, the cocoa plum is Chryso balanus icaco which forms only a shrub; Cocoloba uvifera, or mangrove grape tree, grows large and yields a beautiful wood for cabinet work, but which is light and of a white color. In appearance and description it comes near to the Greenheart or Laurus chloroxylon which is also called Cogwood.

So far we have two types of cocus, apparently separated by their geographical origin, and one of them is also known as grenadille...but then...

GRENADILLO, Granillo, or Grenada Cocus, from the West Indies, is apparently a lighter description of the common cocoa or cocus-wood, but changes ultimately to as dark a colour, although more slowly. It is frequently imported without the sap. The tree yielding this has not been ascertained, the Bois de Grenadille of the French, is also called red ebony by their cabinet makers.

Is this a third wood or not? Interesting as well is the statement that basically it was not known what tree yielded this timber, and in fact no IDs are given to the other Cocus varieties or to the African Blackwood/Botany Bay wood. Yet despite Holtzapffel's assertion that the botanical name of cocus was uncertain, the species of trees that we now know as cocus and the other woods under discussion here had been known to science for some time. Cocus we now know as Brya ebenus, but that specific name was allocated in 1825, and that of African Blackwood, Dalbergia melanoxylon seven years later in 1832. Remember that Holtzappfel was published in 1843. Rockstro, much later in 1890, and even in the 1928 2nd edition doesn't use specific names either as a means of separating different wood types. It seems that botanists and craftsmen weren't having much of a conversation in the 19th century. Even more surprising is that John Jacob Holtzapffel was the elder brother of Jean Daniel Holtzapffel, a flute maker working in Strasbourg and Paris from 1812-1850. To be fair though, the first volume which deals with the woods was written by his brother Charles. I've only seen one flute by J. D. Holtzapffel, which was the original instrument of Breton virtuoso Jean Michel Veillon, and it was made from a black wood...probably ebony.

So there seems to be a pattern emerging regarding the most famous of the flute making woods. Most authors make the distinction between two types of cocus. That from Jamaica was considered the best, and that from other areas of the Caribbean not quite as good, and this type was sometimes called Grenadilla, or some variation of that name. Let's see what  Richard Sheppard Rockstro has to say.

The Cocus-wood of Jamaica gives a splendidly brilliant and powerful tone. This wood is extremely hard and resinous, and being therefore particularly non-absorbent, it retains its form under the influence of heat and moisture better than any other wood that has ever been tried, but it is prone to cracking, and owing to its great density, it interferes somewhat with the flexibility of the tone. It has an exceedingly handsome appearance when newly turned and polished, but it becomes dark, dull, and generally unsightly after being in use for a few years, and the application of French polish only defers the catastrophe for a little while, the ultimate result being worse than when the wood has only received its natural polish.

and again

Cuban and south American Cocus-wood or Grenadille. This material has for many years been employed for the manufacture of flutes. It is excellent for tone production, though its sound is scarcely so brilliant or so powerful as that of the Jamaica cocus or so sweet as that of box. Of all known woods it is no doubt the most suitable for flutes, and it is now almost exclusively used. It is nearly as non- absorbent as the Jamaica wood though less dense and not so liable to splitting but it is not by any means free from that risk, and it is not always permanent as regards its caliber, though a flute in my possession made it this wood by Messrs. Rudall Carte & co. in 1874 which has been in constant use ever since, is now even better than when it was new. The bore of this flute has remained quite perfect but the nature of the world having been somewhat mollified by age and use, the tone of the instrument has become more mellow and flexible without being less powerful than at first. I am bound to say that this is an exceptional case. Cocus-wood is found by some persons to produce serious irritation of the lip which necessitates the use of a silver or gold lip plate. 

Let's finish our look at cocus with what you might imagine to be from the horse's mouth, an account from a working, and reputed, flute maker, Ronnberg. 

Some flutes I make of granadilla (sic). You hear of cocoa flutes. There is no such wood used. It is all granadilla, the brown and the black. The wood comes from the West Indies, principally from Jamaica. The brown wood can be bought here for $45 or $50 a ton; the black is not in the market

It seems that the real problem here is not cocus, but grenadilla/granadillo, and is grenadilla in fact a type of cocus? To add to the confusion, at some point - and it will take more research to find out when - the woodwind trade began to also refer to African blackwood as grenadilla.

My next step in trying to clear all this up was to consult some of the many wood ID sites that can be found on line. One really interesting and useful one is Hobbit House run by Paul Hinds. Here's what he has to say about Grenadilla.

This is another one of those woods where I have considerable confusion, possibly because of similar names, possibly because of a plethora of species. In any event, I will attempt to get more information.
Uh, well ... OK, I HAVE attempted to get more information and what I have ended up with is a splitting headache. This name seems to be used for just about more unrelated woods than any other name I can find, and that's saying something, given the wild overuse of some common names.
The number of species using this name runs to the dozens (and from at least 6 or 8 different genera that I'm immediately aware of, and possibly quite a few more) and the number of alternate common names for various woods that use this name runs to at least 200.
I just don't know what to make of it; what's shown on this page is anything I find that the vendor chooses to list as granadillo.
One of the woods that is generally listed as granadillo is Dalbergia granadillo and while that species is also sometimes called cocobolo, "cocobolo" is generally used in the USA only with Dalbergia retusa and that's how I've handled it on this site.

If you go the Granadillo page on his site you'll see some great photos of many different woods that are grouped under this name. What really intrigued me when I first saw them were the images of one "Granadillo" which botanically is Platymiscium yucatanum ( right down at the bottom of the page). To my eye this looks incredibly like good quality cocus. Could it be that this wood is the Cuban/South American cocus of the earlier authors? It certainly has the right geographic distribution. Wouldn't it be fascinating to discover that it was used at least to some extent? There's only one way to find out.

You'll notice that there are both images and links to other images of the microscopic structure of the end grain of all these various timbers. This is really the definitive method of identification and differentiation, and is doable by someone with a basic knowledge of botany and a simple microscope. 

We could learn a lot, perhaps, by doing such work on some of the old flutes that pass through our hands, and since the process involves microscopically thin sections of wood, is not invasive, and could usefully be applied to even instruments in museums. 

Links, References, and Acknowledgments.

1/ I'm indebted to Michael Lynn for drawing my attention to the Ronnberg interview which was published in "Woodwinds in Early America" by Douglas Koeppe. Brother Francis Publishers.  Texas. 2015

2/Thanks to Paul Hinds for his permission to quote and link to his site. The relevant link is http://www.hobbithouseinc.com/personal/woodpics/granadillo.htm

3/Another useful wood ID site is https://www.wood-database.com/cocuswood/

4/ The Clarinet - Some Notes on its History and Construction. F.G. Rendall 3rd ed. London/Ernest Benn NewYork/W.W. Norton & Co. 1971.

5/ Turning and Mechanical Manipulation. Charles & John Jacob Holtzapffel. Published by the Author. London 1843

The Cornelius Cabinet of Curiosity #2

 The flute making firm, founded by the flute teacher George Rudall in 1821, to provide flutes for his pupils, quickly established a reputation as the finest of the London makers, which they maintained well into the 20th century.

Initially, in 1820 he had flutes made by Willis, who was among the finest makers in early 19thC. London, and were stamped George Rudall, and usually but not consistently "Willis Fecit" ( Latin for Willis made it, for those who didn't go to school in the 60s) on the foot.

By 1821 he formed an association with the Edinburgh maker John Mitchell Rose, and the first registered address is 11 Tavistock St. where they stayed till 1824. They worked at various addresses in the Covent Garden area, and after moving to possibly their most famous address, 15 Piazza, Covent Garden, they returned to 1 Tavistock St. in 1838.

How George Rudall and John Mitchell Rose came to establish the firm together is not known. Rose was from Edinburgh, Rudall a Londoner, and Rose, although several flutes exist apparently made by him, was far from being an established maker at the time. I strongly recommend the account of the Rose flutes in Robert Bigio's excellent tome "Rudall, Rose & Carte - The Art of the Flute in Britain"

.............

Although the vast majority of the 8 keyed flutes that Rudall & Rose, and later Rudall, Rose & Carte would make were of a very standard design, one which they were instrumental in establishing, they were also noted for very many flutes which broke that mould. Basically, it seems, they would make anything that a customer was prepared to pay for, and this has resulted in some very unusual flutes. I've come across several of them in my time, a few of which I mean to feature in the Cabinet of Curiosity series. This is the first of those.

This particular instrument was brought to my attention by a friend, who although he didn't own the flute was able to organise the photos which I've used in this blog. They were quickly taken and not intended for this purpose, unfortunately not showing some of the detail that I'd like to see, but will serve for the moment.

So first things first. The flute has small holes and is in cocus and silver, and has a standard early 19th century 8 keyed block mounted set up. The low C and C# keys are entirely missing. The blocks are lined in silver. It appears to be in rather good original condition, apart from those missing keys, with no cracks or other damage that can be seen in the photographs. The stamp is worn, but I'm told ( I haven't actually examined this flute myself) that apart from Rudall and Rose,  the address  is Tavistock St., and the figure 1 is somewhere in the #, that there isn't much else to be gleaned...well possibly something, although unlikely,  that we'll see later.

So what"s unusual about it?

Well it has no tuning slide, not unusual, in itself since I've seen several other R & Rs with this feature. The outstanding Irish player Catherine McEvoy has played for many years a Rudall & Rose with no tuning slide. I'm inclined to believe though, that these slideless heads were normally provided as an additional head along with a normal head with a slide. I'm basing this on a couple of originally cased R & Rs that I've seen with two such heads and the fact that similar examples can be found among the work of the other major makers of the period.

Secondly, you'll note that all the keys are pewter plugs, in the manner of Potter Patent flutes from around the turn of the18th century. Again I'm told that some other examples of this exist, if very rare.

Thirdly, the mounts. The sheathed head cap and end of foot are commonly seen, but although double rings are a feature of many American 19th C. flutes, I have  never seen triple ones as on this flute.

Finally, that long F key. 

From its general style and profile it has to be original, but of all the unusual features of this flute, this takes the biscuit. I think this is possibly the only example (although no doubt others more knowledgeable will rush to correct me) of a long F which doesn't have a kink or bend just above the axle block. The point of this feature is, of course to ensure the long F and long C don't hit and bend each other when the flute is taken apart. I wonder in this case if the angle of the straight key still allows this to happen?

So can we date this flute, from the information available? The general style, to me suggests earlier rather than later, so what can the address tell us. Rudall & Rose worked from two addresses in this street, no. 11 in 1821 (their first address), and no. 1 from 1838 - 47. On the basis of the scanty evidence I have at the moment, I'm edging towards 11 as being the more likely.

However, in the verbal report I had of the flute which led to the photos, the person who saw the flute said he read the address as 7 Tavistock St.

Now very likely this is a misreading of 1 Tavistock St. A random little mark on the left side of the top of a 1, could easily look very like a 7.

Who lived at 7, Tavistock St.?

George Rudall...and two at least of the very earliest known numbered instruments  (ref Bigio as above) from Rudall & Rose bear this address, could this be another?

Embouchures Again - The Cast Embouchure

 When talking of the embouchure in my post of  17/8/19 I was concentrating on the effect that Benade's ideas had on my own work. Now however I'd like to discuss what is perhaps at the heart of transverse flute making in a lot more detail.

I think anyone who has tried to make a flute of any kind, has been more disconcerted by the challenge of constructing a functional embouchure than any other aspect of the process. Most sets of flute plans or drawings that are available, will of course give the position of the embouchure, and its basic external dimensions and shape. It's what's happening "under the surface" as it were, that causes (or solves) the problems.

It doesn't take long for the aspiring maker to realise that the embouchure, even on early flutes, where it is generally circular, is not simply a parallel sided hole, and even given the difficulties of measuring the internal bore dimensions, it's manifestly more difficult to measure the embouchure in any meaningful way. A cursory examination will reveal that the first general principle is that the embouchure gets larger as it approaches the bore. The observant will also note that this expansion does not seem to be radially symmetrical, i.e. front, back, and sides are not the same.

One visual aid which can help examination of embouchures is a small mirror which fits loosely into the bore of the head, coupled with a light source of some sort which will also fit inside. This allows observation of the embouchure from the inside out.

Possibly the most useful means of visualising the shape of an embouchure is to take a cast of it. Normally considered as a means of documenting historical flutes (largely, if not entirely, now forbidden in museum collections ) it's also a very useful technique for the modern maker, if only to check that what they are trying to do, is actually what they're doing. 

If you have access to it, dental impression material is ideal for this purpose...just be sure to use some type of releasing agent ( Vaseline, or silicon grease both work well) 

 A brief historical survey of embouchures shows a move from small circular via oval, to parallelogram ( roughly speaking), increasing in size from the initially tiny baroque embouchures to the practical limit of a playable embouchure, which, it could be argued is not much greater than 12mm x 10mm. The common feature is that they all get wider approaching the bore, and that this is rarely if ever in a regular fashion. 

In the "Introduction to Flute Making" courses that I've taught for the Ballyfermot College of Further Education under the auspices of Na Piobairí Uilleann I've tried to break down the basic principles of embouchure design, and we did a series of experiments looking at some of the most basic parameters, such as size and depth. Using a delrin sleeve over a standard head tube ( see blog post 17/2/2016 Dante's Satan ) we drilled simple round parallel holes through the sleeve which could then be made to act as the embouchure. One advantage of this approach was that it was possible, using different barrels, for students to try these embouchures on their own flutes, and thus be able to attribute any differences to the change in the embouchure alone. Starting with size ( in this case diameter alone ) we drilled holes of 8, 10, 12, and 14 mm, observing what changed, as the diameter increased. Similarly we used sleeves of different external diameters, so as to vary chimney depth.

The first thing that this exercise will teach, is that it's possible to get a sound, sometimes a surprisingly good sound, from almost any size and shape of "embouchure". The second is that changing parameters in a direction which makes things better, only does so up to a point. To illustrate. A very small, say 6 mm embouchure, is almost impossible to play, but small increases in diameter lead to increasing facility and ease, until at about 14 mm, control becomes increasingly difficult. Of course it's impossible to separate the two basic parameters under discussion, because an embouchure is by its nature three dimensional, but at least we can vary one at a time and keep the other one constant.

The point of all this is to demonstrate the complexity of the embouchure even at this very basic level of the parameters of size and chimney depth.   

Consider that other parameters that need to be taken into account in properly functioning embouchure include shape, and not just on the basics: is it oval or square/rectangular etc. because for example there are long ovals, and almost circular ovals, close to pure rectangles and even occasional squares, there are these shapes with curved sides, almost blending into the oval, and so on. Then the undercut, as the way in which the embouchure expands towards the bore is called. This is usually different on the back, front and both sides, it can be straight or curved, curving more towards the outer surface or towards the bore, sometimes giving a slight "reverse" expansion. Another major consideration is how the different undercuts blend with each other. Extremely important are the edges where the embouchure meets the bore, and where it meets the surface. These can be sharp or rounded, and again they are different front, back, and sides.                               Finally, remember that all these parameters interact with each other with varying complexity. 

All this might lead one to wonder how on earth anyone can actually cut an embouchure, let alone learn to do it, but it bears consideration that in the account above, I'm deliberately emphasising the complexity, and intellectualising something that in practical terms is much simpler. Consider the art of reed making. Without the ability to make a reed, bagpipes, oboes, clarinets, bassoons, and many more would remain silent. Yet if a potential reed maker was to take the analytical, acoustical approach ( Benade, for one, has a lot to say about the acoustics of the reed) I'd suggest they would remain so. So how does the nascent flute maker approach this whole area?

In a similar way to that in which the current makers of simple system flutes began by copying historic flutes, they also copied the embouchures as best they could, given the caveat of mostly not knowing precisely what they were copying, but rather relying on what worked and what didn't. This is perhaps the major area in which it's essential that the flute maker is also a player, and not only that but a player in the tradition in which the flute is intended to be used. 

By whatever means, the contemporary makers learned to make embouchures that fitted into what might be called, to borrow a phrase from astronomy,  the Goldilocks Zone. The embouchure was somewhere in or around 10mm x 12mm, undercut a little more on the blowing side than the back, and perhaps less again on the sides, which might not be exactly equal in that sense. Other parameters, as mentioned above, would vary more according to the maker. I can only speak here for the  makers of modern simple system flutes, and not those who make historical copies for the early music enthusiast. They are much more restricted.

In my own work, I like everyone else began by copying as exactly as I could, but as I gained in confidence and moved away from copies I came up with an embouchure design which was in essence quite different physically and in concept from most other embouchures. Here's what I said about it in the "Irish Flute Player's Handbook" ( 2nd ed. 2008 )

"I pondered for many years about how to make a head which would incorporate the features that I wanted, such as a deep chimney, and an overcut, but which were easy enough to produce to allow me to use such a head as standard, and also were still basically of the standard rather than thinned head appearance"

 (I had just been discussing how a thinned head was allowing me to include certain embouchure design parameters, but was looking for something less expensive and time consuming for standard instruments)    

"Eventually in about 2000, I came up with a solution which has worked very well. The germ of the idea had been with me for a long time...there are drawings in my workshop notebook dating from early 1984 which state the basic idea. Normally the pilot hole for the embouchure is drilled in the centre of the head, i.e., it is centred on a radius of the head. This means that the initial chimney depth is equal on both the front and back of the embouchure. It occurred to me that if I shifted the axis of this pilot hole away from the centre in the direction of the side the player blows against ( in essence drilling the axis on a chord to the head circumference ) that two results would be achieved. Firstly the chimney is deeper than for a hole drilled on the centre, and secondly, when the normal undercut is made, the included angle is smaller, giving the same effect as an overcut. From the player's point of view, the effect is a noticeable improvement in response, similar to that achieved by a thinned head. The fact that the improvement is still obvious when features such as the pitch of the instrument or the shape of the embouchure are changed indicates to me that the embouchure parameters have been changed in a basic way, and this explains why an embouchure cut on the above basis works for a piccolo with an oval embouchure, or a low Bb with a square one."

Other aspects of this embouchure were a quite severe undercut on the blowing side in contrast with an almost vertical back and sides, the idea ( perhaps fanciful ) being to push the air forward against the edge where the vibration of the air reed was driving the sound production. The important thing was that it worked.

Because the embouchure was quite differently undercut on the across the flute axis, it necessitated cutting it "in reverse" for the left handed player...of which there are many in the Irish tradition.                                                                                      This led to several interesting observations. When I'm cutting an embouchure, I bring it as close as I can, by measurement and observation to what I want, and then I begin the final tweaking process, which involves playing the flute, and making the final adjustments to the embouchure, normally careful modifications to the edge profiles. So when it came to a left handed flute I had to make those final adjustments blowing the instrument in a way that is very counter-intuitive for me, but I managed to teach myself to do this. In fact what I would do, is to blow the embouchure even if left handed, right handedly at first, and then check the final cut by blowing it left handed. That I could get to perhaps 95% of where I wanted to be blowing the "wrong way", indicated to me that perhaps the degree of undercut was less important than claimed, and also called to mind the number of players who quite successfully played right handed flutes left handed. This included a few of my own customers who being left handed and having bought pre-owned right handed flutes, preferred to stay with the "wrong" embouchure when offered the choice.

Around this time as well, the letter that Benade had written discussing the embouchure cut in great detail became available and I began to give the ideas in it serious consideration. ( See post of the 17/8/19 ) The first thing that I began to realise was that my "own" embouchure design was almost exactly that of Benade's switched back to front...which perhaps to some extent accounted for the "playable in reverse anecdotes above. Secondly I began to wonder, if the volume of the space created by the embouchure was more crucial than had been realised, because of course this was one parameter which didn't change when the embouchure was reversed. Not being an acoustician, I also wondered if this had any relationship to the fact that the relationship between the air volume contained between the blades of a double reed and that of the bore appears to be absolutely crucial for double reed instruments of all sorts. ( See  Benade's The Fundamentals of Musical Acoustics )

Given how important the embouchure is to the flute, it's not surprising that flute makers have come up with both ways to repair damaged embouchures, and to replace them if required. 

This technique is called bushing, and basically consists of drilling a hole larger than the existing embouchure and filling this with a "bush" in which the new embouchure is then cut. In almost all circumstances, the bushing material is different to the original head material, being usually ivory, or horn. Apart from the decorative and aesthetic possibilities that this gives, it also reflected the fact that a small piece of wood would have little strength if worked on to any extent before it was inserted. 

Here's an example of a horn bushing on a George Rudall flute from c. 1820.

In this case the bushing tapers in as it approaches the bore in the opposite sense to the embouchure itself.

In this case the bushing tapers in as it approaches the bore in the opposite sense to the embouchure itself. You can see how this means that there is actually very little material in the bush itself. It's hard to make out if these bushes were largely finished before they were inserted, or inserted with a pilot bore worked in place.

It occurs to me that the way that the bush tapers would have prevented it being accidentally pushed into the bore in the finishing process Very many bushed embouchures were made in this fashion, although others where the bush itself was a parallel plug also exist. 

Note the turned grooves on the bush and the head, designed to give more grip to the adhesive, which was probably shellac.                                                                 

Unfortunately this bush is not in great condition as the horn has begun to denature and even delaminate slightly.

It may have to be replaced, but given the access that the detached bush gives to the embouchure cut, it should not be hard to do this accurately.

The ability to change an existing embouchure is a double edged sword.

Without resorting to bushing, a good maker can improve a badly cut embouchure, but the caveat is that this can only be done by removing material, and once gone too far, that can't be replaced.

I've seen several examples of embouchures that have been very professionally recut but which have resulted in a basically unplayable flute. This usually results from the desire to make a flute louder and more powerful, ( they were on flutes used in the Irish tradition where these qualities are prized) which a larger embouchure can do, but the temptation to pursue this course beyond its acoustical possibilities has obviously been too great with these examples.

The result of too big an embouchure, no matter how well made, is initially an apparent increase in power and volume, but the realisation soon dawns that it's only possible to maintain this for a very short period of play, which combined with a huge lack of flexibility can make a flute basically unplayable.

This is where the practice of bushing can be so valuable in retrieving such flutes to playability.

In my own work I have recently begun to experiment with bushed embouchures, not from the point of view of correcting or improving existing embouchures, but more in terms of consistency and simplicity of manufacture.

Of course, in the case of a new instrument, no advantage is gained by simply having to cut an embouchure in a separate piece of material which is then inserted, or inserting a piece and then cutting the embouchure.

The idea of having an embouchure cast in some material and which could then be inserted, leaving only a little tweaking to be done is something that had attracted me for some time. Of course, the idea is not original. The heads of Boehm system flutes are standardly made by soldering a cast chimney onto the head tube, and a lip plate onto that.

Here's a few images of a chimney and lip plate before assembly

The material to use was initially an issue. At various stages in my career I had thought of getting different components cast or moulded, but the problem was always one of proportion. These processes are only practical when many many thousands of units are required, when the cost falls to a few cent each. The answer was obviously to use a process where the production of a limited number of pieces didn't imply a prohibitive cost, and yet used a material which was at once both practical and aesthetically pleasing. So enter the silver cast embouchure.

The images below show the casting with the sprue still attached

There are other advantages.

1/ The main advantage of this system is consistency. Once the embouchure cut is determined it's reproducible within very close limits. 

2/ It's a very swift process inset the bush into the head and finish it.

3/ Making the flute left handed simply means turning the bush around in its "socket".

4/ A series of different cuts can be cast in this way, allowing a great degree of flexibility.

5/And of course the silver casting process is readily available at reasonable cost.

Finally, I don't think it looks to bad at all...

A flute of mine with such an embouchure...in fact this very one...will be available shortly via my Instagram account @hamiltonflutes

The Cornelius Cabinet of Curioisties #1

 This is the first of a short series of posts dealing with some of the more unusual things that I've come across in forty years of flute restoration. 

This first example arrived in a group of flutes that I bought at auction in 2007. It was described as "a rosewood flute, unstamped with nine keys on wooden blocks".      So far so normal. (except, of course, it was cocus...a very common auction house error)

A nine keyed flute is really nothing unusual. Very many keyed flutes have an extra touch to one key...very commonly a key for R1 ( first finger, which opens the Bb key from below.) Also common is another key for R1 which operates a trill key, often seen on German "nach Meyer" type flutes.

But when this flute arrived, this what I saw...

It took a moment or two to realise what I was looking at. The flute as you see it in the picture above is after restoration, but as soon as it was playable it didn't take long to figure out that the extra key didn't produce a note that existed in any known system of music, somewhere between F# and G. Otherwise it was a well made mid-19th century English flute, in fact a good bit better than average quality. It was a fine player, and I sold it sometime later, recouping costs to cover some of the other flutes bought at the same auction.

I puzzled over this key for some time, and the only function I can possibly see for it is to correct the notoriously flat F# that we find on many 8-keyed flutes.

Several other systems had been devised to do this.

This is a Tulou Système Perfectionné...

This might look a bit confusing at first. Hole number 6 is in this case covered with a plateau key, ( which is absent on most if not every other Tulou SP ). The touch under that opens the little key on the top of the flute just above the F# hole and is opened with R3 to correct the F#. This could be done on these small holed flutes by enlarging the F#, but the French ( and German ) predilection for small evenly sized holes ruled this out.

In fact this, although operating on exactly the same principle, is different to the more common set up...

This illustration is from Tulou's Methode, and you can see that the key is a long one on a cross tube mount, like a reversed long F but is still operated by R3.

Various other flutes have used a similar device.

To finish here's a neat example from a Hudson Pratten #185

You'll note that with this system, the ring key controls the key which sharpens the F# automatically, and with the Tulou, the same key is operated by a finger, R3 which is available without compromising other fingerings.

From this point of view, the 9th key on the first flute seems very counter-intuitive.

Was it meant to correct the F#, ( which it didn't, or at least very 

badly ) or is there a possibility that it was to vent some other note?

I'd welcome input from anyone who might be able to shed light on this.