The Hole in The Middle: Transverse Flute Bores in the Late 17th 

and Early 18th Centuries 

Ardal Powell

Paper read at the annual meeting of the American Musical Instrument Society, Elkhart

IN, 1994

ABSTRACT 

Though keys were added to the transverse flute at least as early as 1769, the primary

design concern of makers throughout the eighteenth century was the instrument’s bore.

Experiments to improve the flute’s evenness of tone, audibility in all registers, and

intonation were particularly vigorous in the first half of the century, but even as late as

1811 the brilliant inventor and famous keyed flute maker Heinrich Grenser still believed

that the extra keys were just a fad and the flute could be brought to perfection without

them.

The conical bore is the baroque flute’s essential feature, developed at the end of the

seventeenth century. Yet the bore has escaped systematic study because it is hidden inside

the instrument and can only be discovered by plotting a series of rather minute

measurements. It was almost never discussed in writing by contemporary authorities.

This presentation describes and comments on bores of surviving early-eighteenth-century

flutes, using graphs to illustrate the points of similarity and difference between them.

Ardal Powell is a 1993-4 Fellow of the National Endowment for the Humanities’

program for Independent Scholars, working on a project entitled “The Keyed Flute:

Technology and Musical Style in the Late Eighteenth Century.” For ten years he has

practised as a maker of historical flutes, as a partner in Folkers & Powell. He holds an

M.A. in English Literature from the University of Cambridge, and a Certificaat in

Baroque Flute Performance from the Royal Conservatory in the Hague, The Netherlands,

where his teacher was Barthold Kuijken. His translation of J.G. Tromlitz’s 1791 flute

treatise was published by Cambridge University Press in 1991.

[SLIDE 1—de la Barre and Musicians]

[TRANSPARENCY 1—Picart Hotteterre engraving]

The pictures you see on the screens are two of the icons of the baroque flute, a type of

instrument that emerged at the end of the seventeenth century, and first came to notice in

the hands of famous performers at the French court. This sort of picture is attractive to

look at, as are the instruments themselves with their ornaments turned in an elaborate and

characteristic style.

Still there’s only so much to be learned about musical instruments as sounding objects

from looking at pictures of their outsides. I’m here to take you on an introductory and

very preliminary tour of some of the earliest baroque flutes, avoiding their seductive

external appearance almost entirely, and instead concentrating on a number of their less

conspicuous aspects. I will be showing you snapshots of the flutes just to identify them.

But since the most distinctive thing about this group of instruments is what’s called the

“conical” bore I want to concentrate instead on rather unglamorous representations of the

empty space inside the instruments.

[SLIDE 2-Anon. Stuttgart]

[TRANSPARENCY 2]

Let me quickly explain what the graphs I’ll be showing you are meant to represent. The

left-hand section is the headjoint, the middle one is the middle joint, and the one on the

right is the foot. The bore of each flute was made with sets of short reamers, each of

which cuts a section of the profile a few inches long. The maker would start with a

cylindrically drilled tube, and pass the smallest reamer down to cut the sharp end of the

cone. The next reamer would go almost as far, but cut a larger diameter, and so on until

finally the largest reamer was used on the widest end of the tube.

The bore is measured with a set of gauges, which are nothing more than short sections of

Delrin rod with the ends rounded off, differing in length by a tenth of a millimeter. A

graduated stick holds the gauge at one end, while each gauge, beginning with the

smallest, is passed into the bore from its largest end, and the length coordinate read off

from the graduated stick. The coordinates are then plotted on a graph, where the vertical

axis is the diameter of the bore, and the horizontal one is the length along the tube. The

embouchure hole is on the left of the graph and the open end of the flute on the right. On

these graphs the vertical scale is exaggerated by a factor of ten so that you can make out

the variations more clearly.

If the bore is measured carefully in the way I described, it’s often possible to perceive the

profile of the reamers and figure out which parts of the bore were cut by which tool or

tools. On this graph of an Anonymous flute in a private collection in Stuttgart the marks

of several separate reamers are visible in the lower half of the middle section. It was

necessary to use such a large number or reamers over such a short length to make that

unusual concave shape you see in the bore, rather than the gradual taper you will see in

other instruments. In another part of the bore you can clearly see where two different

reamers were used to make an enlargement at a particular point. This one is just above

the third tonehole and would affect the octaves of G; D in the third octave; and the

relationship of G# with all these notes. It might affect other notes too—though the only

way to find out would be to try it.

You’ll sometimes see places where the bore contracts again after getting larger—this is

usually because a tenon has shrunk due to the pressure of a socket on the tube. This is a

hardly noticeable case—I’ll point out a more obvious example when we get to it.

[TRANSPARENCY 3-Handout]

On your handout, there’s a list of twenty-eight instruments of which I propose to look at

twelve in some sort of detail. It’s a bit dark to read it so here they are. All the flutes on

the list are alike in that they are in three sections, with six fingerholes and one key. I’ve

divided the flutes into three groups.

[SLIDE 3—Haka]

The first group contains two flutes: an Anonymous one, and one by Richard Haka, which

are not considered true baroque flutes. The reason they are disqualified is simply that

their bore is not tapered enough. The Haka flute has an overall taper of 1.8mm—that is,

the smallest point of its bore is 1.8mm smaller than the largest point. The Assisi flute is a

little more tapered, at 3.5mm. The taper is much more irregular than on the true baroque

flute, but we will not be detained by that point for now. We are usually told that the

principal innovations in flutemaking at the end of the seventeenth century took place in

France and consisted of three things: (1) three-joint construction, (2) the conical bore, and

(3) the addition of a key for D# or E♭. We can see from the two examples in Group I that

all three of these conditions existed, probably beforehand, and elsewhere than in France.

The late seventeenth-century changes to the flute, wherever they were first made,

involved only an increase in the bore’s taper. The non-qualifying Assisi flute has a taper

of 3.5mm starting at 19.2, or 18%; whereas in the qualifying groups II and III the

percentage taper, where we know it, is in the mid-20s or more.

The second group of instruments on the handout appears in this list just so they are not

left out, but we won’t look at them in any detail. They are excluded either because they

are lost or otherwise not extant, have been altered from their original state, are not what

they seem to be, or because I don’t know enough about them.

Group III contains the examples I’d like to discuss here, which were all probably made

within about 30 years of each other, between roughly 1690 and 1720, all except one by

makers of French origins.

Only five of the twelve play at a pitch we can be reasonably certain is the one they were

originally intended to work at. Five have had their embouchure holes interfered with—

this is an all too common misfortune to befall old flutes and is usually the result of

attempts to raise the pitch or to make the instrument seem louder. It’s almost always

possible to spot an altered embouchure, even when it has not been crudely done, because

of inconsistent workmanship or a different style of using the tools from the rest of the

instrument.

[SLIDE 4-Bressan]

[TRANSPARENCY 4-Denner]

If there’s enough light to see the back of the handout, there’s a league table of Group III

arranged according to bore taper. The flute with the shallowest taper in Group III is the

Miller collection’s Bressan. Unfortunately its middle section is generally agreed not to be

original, and we have no way of knowing what its bore profile was meant to be like. But

in most other cases the smallest diameter in the bore is in the footjoint, which is probably

original, and so we can at least estimate the overall taper. The ivory Denner flute in

Nuremburg has a C-foot joint as well as the D-foot shown on this graph. In the case of

this bore, the numbers on the handout don’t tell the whole story, because the headjoint

bore is almost a millimeter larger than the top of the middle section, and the D-footjoint

bore is rather narrow in comparison with the bottom of the middle section. If we counted

only the very shallow taper in the bore of the middle section, the Denner would actually

come at the bottom of the list, as it less tapered than any of those in Group III. It would

be an interesting digression to compare this graph with a graph of the Assisi flute (which

you remember is not a true baroque flute) to see the similarities, especially at the bottom

end of the middle section.

[SLIDE 5-Rippert St M]

[TRANSPARENCY 5]

The two flutes in Group III with bores most similar to one another are the two Ripperts.

Here’s a picture of the gorgeously ornamented one in St Moritz. If we lay graphs of the

two middle sections one on top of the other, we can see that the biggest difference

between them is about 0.3mm, at one point in the Glasgow flute where a reamer has been

used to enlarge the bore in the region of the 5th tonehole. This would affect the

relationship of F# and F natural and perhaps also G# in the second octave. The maker

could acieve almost the same effect by changing the undercutting of the fifth tonehole,

but that method would affect the first and second octaves and might not change the third

at all. There’s another difference of as much as 0.3 mm in the region of the second

tonehole, which might affect the octaves of B and the pitch of B♭ and A# in relation to

them.

Despite the 8mm difference in sounding length and the quite different footjoint bores,

these two instruments are exceptional among our sample in being so similar. On the

handout you’ll see that the overall taper of the bores in Group III, that is the difference

between the maximum and minimum measurements, varies over our sample from 6.6mm

to 4.2mm. Also the maximum bore size, which is usually found in the headjoint, can vary

by almost a millimetre and a half. That may not sound like much, but when you consider

that the sound and behavior of one of these flutes can be completely transformed by local

bore changes of only one or two tenths of a millimeter, fifteen or twenty-five times that

amount is a very large range.

[SLIDE 6-Hotteterre]

[TRANSPARENCY 6]

You’ll observe that the Graz Hotteterre flute, which is the best known of this type of

instrument and is usually thought to be “typical” of early French baroque flutes, is not

really typical in our sample as it is found at one extreme of the list. Its taper is quite

shallow, and its bore maximum extremely small. The headjoint has a tenon on the bottom

end which as you can see has shrunk dramatically—the flute in its present condition

exhibits intonation oddities that are at least partly a result of this shrinkage, caused by the

pressure of the ivory connector on the two thin wooden tenons which expanded as they

got wet with playing. There are only three flutes in Group III where the headjoint bore is

not pretty much straight as we saw on the Ripperts, and on all of these we have to allow

for a certain amount of shrinkage at the bottom end where two different materials are in

in a pressure fit with each other. I can explain that better by saying that sometimes the

bore shrinks, and sometimes it doesn’t—it can depend on the thickness and inherent

stability of the materials, how much and how carefully the instrument is played, how it is

kept, the weather, and suchlike imponderables. The only time we can be certain a

headjoint bore has not shrunk at any particular point is when it’s perfectly cylindrical.

[SLIDE 7-Leclerc]

[TRANSPARENCY 7]

At the other extreme of the taper league table is a flute by Leclerc in a private collection

in Brazil. (By the way the maker’s stamp is L-E-C-L-E-R and that’s why the final C

doesn’t appear on the graph.) Originally its headjoint bore was probably in the region of

20.2 all the way along: the contraction at the top end has no acoustical function since the

cork would stop the tube at about 55, and the slight contraction towards the headjoint

socket could also be from the demon shrinkage. Comparing the bore to that of the

Hotteterre flute as in the lower graph, you can see that there’s a huge difference of up to a

millimetere and a half all the way along. Another difference to note is the wallthickness

of the tube, which you can’t see on the graph but which is listed in two columns on

labelled WALLTHICKNESS AT EMBOUCHURE and WALLTHICKNESS AT

TONEHOLE 4 on the Handout; you have to tip your head over to your left shoulder to

read this I’m afraid. Roughly speaking, a flute with thicker walls produces a beefier tone

in the low register and better sounding flats and sharps than one with thin walls—all

other things being equal. The headjoint of the Leclerc flute is extremely thin at 4.3mm;

that of the Hotteterre flute is extremely thick at 5.6. However, the thickness of the rest of

the tube doesn’t follow the same proportions: at the fourth tonehole, a bit lower than the

middle of the middle section, the Hotteterre’s tube is less thick than the Leclerc’s, at 4.6

compared with 4.9. The instrument with the thickest tube where it matters, which is at

toneholes 2, 3 and 5 for notes like G#, F and B♭, is the Berlin Naust, which is otherwise

in the middle of the range for taper and overall bore size, while it is rather longer than the

average for its pitch.

[SLIDE 8-Naust StP]

[TRANSPARENCY 8]

This is a picture of another Naust flute, the one in St Petersburg—the original footjoint is

lost and here replaced with one from another instrument. The Berlin one looks quite

similar except that it has doesn’t have the ivory parts, which may have been added later

in any case. The graphs of the two Naust flutes are up to half a millimeter different.

There’s another notable thing about these two flutes: neither of the embouchure holes

seems to have been altered. Nevertheless, the embouchures differ in size by a significant

amount, and the wallthickness of the headjoints differ by half a millimetre. No matter

how similar the bores were, these factors would make the instruments behave quite

differently in practice. Also we have no way of knowing what the bore of the St

Petersburg footjoint was like—it might have been quite different from the Berlin one, and

could have made for some more notable differences in playing qualities. These two make

quite a contrast with the two highly congruent instruments by Rippert.

[SLIDE 9-Panon]

[TRANSPARENCY 9]

The remaining two flutes we have to consider are by makers about whom absolutely

nothing is known. This instrument stamped Panon is in a small museum in Toulouse.

Though it looks conventional enough on the outside, its bore is all over the place. The

original embouchure has been plugged up and a new one drilled right beside it. I’d guess

this was done fairly early in the instrument’s life, by someone who wanted to get the flute

to work a little better. As it is, the maker has achieved the extraordinary feat of designing

a flute whose first octave plays very flat at the bottom, and very sharp at the top, while

the second octave plays very sharp at the bottom and quite flat at the top, all with a tone

quality that could in all charity only be described as extremely feeble. If we compare the

wandering line of this bore profile with the confident and precise irregularities in the

instruments by Rippert and Naust, we can perhaps perceive the difference between

someone who was feeling his way and a skilled and confident maker.

[SLIDE 10-Chevalier]

[TRANSPARENCY 10]

The second mystery flute is the MFA’s Chevalier, which once belonged to Canon Galpin

and so has been in the literature of the flute’s history for most of this century. Its bore

taper is fairly shallow and very regular except for a bit of a blip in the region of the first

tonehole, which I would hazard a guess affects the D in all three octaves more than

anything else.

[SLIDE 1 again]

[TRANSPARENCY 1 again]

[OR: LIGHTS]

Many conclusions could be drawn from this brief survey, and there is room for much

more investigation and discussion. Already, though, the similarity of the Rippert

instruments, the distinctiveness of the Nausts, and the range of technical solutions in our

sample seems to suggest the wide variety of approaches among makers.

In this light, I’d like now to read you a few words from Anthony Baines’s 1957 book,

Woodwind Instruments and their History, which is still the textbook on this subject,

discussing the irregular bore of the baroque woodwinds, and of the flute in particular.

QUOTE “Such bore-construction naturally has some acoustic effect, but this is not a vital

one. . . although these bores successfully gave the musical results desired at the time, it is

hard to imagine that their irregular profiles were arrived at through deliberate experiment

in pursuit of musical objectives.” END QUOTE

We may pass over this passage’s inference that the musical results desired “at the time”

were of a low standard, as well as the notion that the woodwind-makers of the early 18C

designed their instruments in a state of ignorance or entirely by chance. A less logically

challenging assumption would that, as a class, the period’s makers fulfilled their

intentions and produced instruments as ideal for the music they were intended to play, if

for none other.

The bore graphs we have seen can support this proposition. They demonstrate that

instrument making left room for a distinctive personal style not only in the way

woodwind instruments looked, but also in the fundamental acoustical system that made

them work. Studying these technical aspects can sometimes enable us to posit a reason

why a maker made an expansion or a contraction at a particular point, and we can

occasionally get a feel for his skill in making these decisions. Yet the tireless

inventiveness of these makers and the depth of their understanding emerges only when

we stop to consider that everything about their art (or craft, or science, whichever you

want to call it) was variable: not just the features of the bore we’ve been looking at, but

the location, size, shape, wall thickness and undercutting of the embouchure and

toneholes, the materials and treatment of the tube itself, and finally the widely varying

technique and style (as evidenced by contemporary reports) among musicians whose job

it was to have the instruments make music when they played them.

Perhaps if we’d spent the last 20 minutes looking at just one of these graphs without the

opportunity to see the others we might take away the conclusion that Baines had rightly

interpreted its apparent randomness as evidence of a crude design and an ignorant

instrument-maker. One might well think, with all this variety, that you could put any old

bore in a 3-joint flute and it would work anyhow. But neither of these assumptions

survives contact with experience in the making and playing of early instruments. The

difference between a good design like the Berlin Naust and a poor one like the Panon is

evedent; anyone who has tried making even a slavish copy of the best original specimens

will confirm how difficult it is to match up to the musical standards—such as tone,

intonation, register balance, dynamic range—of the originals. Yet when we survey a

group of these instruments in the light of such experience—and study graphs of the

invisible hole in the middle—we see “deliberate experiment in pursuit of musical

objectives” going on before our very eyes.