Return to home page
  ^ documents ^ instrument ^ music ^ people ^ resources ^ shop ^ the book  


by Ardal Powell

Word count: 4000. Download size (text only): 30,000 bytes.

Author's note: This is an English version of "Die Eichentopf-Flöte: die älteste erhaltene vierteilige Traversflöte?", as it appeared in Tibia 1/95, 343-50.

Abstract: A unique transverse flute by Johann Heinrich Eichentopf (1678-1769) in the Bachmuseum, Leipzig, has a strong claim to be considered one of the earliest four-joint flutes to survive. Though the instrument has been altered in order to raise its pitch, this article describes a reconstruction of its essential features. A successful reconstruction seems to confirm that the sections have not been significantly shortened and the bore of the flute is entirely original. Characteristics of the bore at the central tenon-and-socket suggest that an important reason for the division of the three-joint flute into four in c.1720 was freedom of access to this region. It is further suggested that the division into four sections may have pre-dated the addition of corps de rechange by a short time.

Johann Heinrich Eichentopf (1678-1769) was a wind instrument- maker working in Leipzig from 1710 to 1749. Various suggestions exist of a link between him and J.S. Bach, including the fact that he delivered at least five instruments to the court musical establishment at Köthen, probably some time after Bach left service there.1

The list of surviving instruments by Eichentopf shows that his workshop had a wide range, producing all members of the oboe and bassoon families, recorders and brass.2 But only one traverso with the I.H. Eichentopf stamp is known.3 This is an ivory flute in the collection of the University of Leipzig, number 1244, located in the Bachmuseum, Thomaskirchof, Leipzig. It was examined and measured by Catherine Folkers and the author in May 1992 and March 1993. Our thanks are due Dr. Winfried Schrammek, the museum director, and museum staff for access to the instrument and permission to publish this report.

The museum catalogue states that the flute dates from about 1730 and no later than 1749, that the upper middle joint has been shortened by about 18mm, and that the tenons of that joint and of the lower middle joint (or heartpiece) have been altered.4 Despite the fact that its condition is not original, the instrument deserves attention because of its unusual proportions, its unique testimony to Eichentopf's work as a flutemaker, and of course its possible closeness to Bach. Furthermore as I will try to show, it has a strong claim to be among the earliest four-joint traversos in existence.


The flute has some unusual features. The very finely turned end- cap contains a screw-cork with its threaded spindle projecting through the cap. The screw-cap rotates on a tenon at the upper end of the headjoint, and this tenon has grooves in it. The mounts at the sockets are fitted on sleeves, so that an endwise view shows two concentric thicknesses of material. The workmanship of the ivory turning at the headjoint and heartpiece sockets, while it is competently executed, is not of the same quality as the very fine work of the original parts of the endcap, and of the end of the footjoint. In the workmanship of the footjoint it is quite easy to see the two different hands responsible for the flute's present condition: the mount at the socket, again fitted on a sleeve as far as the ring which holds the key, is not in proportion to the rest of the piece, and its ornamentation does not relate in size and shape to that of the outer ends of the flute. These discrepancies in the style of turning5 together with the shallow depth of the heartpiece socket suggest that the instrument has been shortened at the mounts, rather than--or as well as--at the tenons as the catalogue states, and that the work was done by a craftsman of less skill and artistry than the original maker.

In fact all the tenons connecting the pieces appear to have been shortened. On the heartpiece, not only the tenon but the actual sounding length at the bottom of the joint has been altered: the facing where the tenon begins is imperfectly executed, and since the sixth hole is thus closer to the footjoint, the key--which itself may well not be original--has been shortened to make room for the finger to cover the sixth hole. The sounding length of the upper middle joint gives no sign of having been changed.

An original embouchure, with its center at about 199.7mm from the present end of the headjoint socket, has been drilled out and plugged, and a new, oval hole made in the back side of the headjoint at 167.6mm. According to the catalogue, a silver shield, lost between 1903 and 1925, was set into the ivory to cover the plug.

Perhaps the most interesting feature of the flute is its unusual bore (see Figure 1). Traverso bores in the early eighteenth century generally have a more or less cylindrical headjoint (though conical headjoints both contracting and expanding are to be found), a generally contracting conical mid-section (or two mid-sections in the case of four-joint flutes), and a conical footjoint, usually expanding.

The overall taper of the Eichentopf's bore is very slight compared with most other flutes of the period. It is also unusual, though not unique, in having two prominent steps where the bore becomes larger, instead of gradually reducing in size in the normal way.6 The first of these is at the intersection between the bottom of the left-hand section and the top of the heartpiece, where there is a dramatic enlargement. The second step is at the beginning of the footjoint--whose bore continues with the same contracting conicity as the two upper joints, not flaring out again as is usual.


Almost certainly the instrument in its present, altered condition does not sound as Bach may have heard it. But curiosity about its original playing qualities made an attempt to reconstruct the instrument seem worthwhile.

The changes were evidently made with the intention of raising the flute's pitch to bring it up to the standard of a later age, an all too common fate of obsolete one-keyed flutes. The sounding length of the headjoint is made more than 30mm shorter by the new embouchure; the changes at the sockets also effectively shorten the tube. The range of possible original dimensions for each of the altered features is quite large, with a correspondingly wide range of influence on playing qualities, given the fact that alterations in bore, hole size, placement and undercutting measuring 0.1mm or less can have a transforming effect. Since no other traverso by Eichentopf is known, there are no direct comparisons to be made, and flutes having similar general proportions are extremely rare. But at least Eichentopf's working period, 1710-1749, helps us reduce the range of possibilities by ruling out any features typical of later flutes. The unusual length of the headjoint, seeming to derive from the proportioning of the three-joint flute, also invites comparison with models from the early part of the period.

The original embouchure must have been more or less exactly in the same place as the hole made for the plug which replaced it when the new embouchure was made. If they shared the same center it would be 199.7mm from the present end of the headjoint. But to judge from the outer profile the mount seems to have been shortened by about 8mm, and the depth of the socket increased to compensate. Thus the sounding length of a reconstructed headjoint would be 207.7mm. We would not expect to find an oval embouchure, like the one the flute has at present, on a traverso from the first half of the eighteenth century; we can be fairly certain the original was more or less round: not more than about 0.2mm larger from side to side than from front to back. Embouchure size rarely exceeds 8.9mm from front to back on well-preserved specimens from before 1750. And holes as small as 7.5mm in size are not uncommon on early examples. The most probable original size would be in the middle of this range, around 8.2mm, with diminishing likelihood for dimensions up to half a millimetre larger or smaller. The undercutting ought probably to match that of the flute's fingerholes in their unaltered state: the first hole could well be taken as an example since, as the hole closest to the embouchure, its pitch would rise more than any other hole's whan the flute was shortened and there would be no need to enlarge it, so it is probably in its original condition. The undercutting of the holes, generally speaking, is slight to medium, and quite concave, especially where the side walls join the front and back, with slightly more undercutting one one side than the other. The second hole has little undercutting on the bottom side, the third and fifth holes do appear to have been enlarged, though carefully, and the hole covered by the key is very little undercut. The size and undercutting of the fifth hole ought more or less to resemble the fourth; and the size and undercutting of the third hole can be determined--assuming for the moment that the bore has not been altered--from the tuning of A and G# in all three octaves, and of E'''.

It is inconceivable that the original endcap contained a screw- cork device. With the embouchure in its original position, there is barely room for the cork itself, much less any auxiliary machinery to move it about--and there is no sign that the top of the headjoint has been shortened. What we can learn about the invention and purpose of the screw-cork7 makes it improbable that the device was known to Eichentopf or used on a flute without alternative upper middle joints. The grooves in the tenon argue for thread windings which would have held the cap in place--this too quite inconsistent with its use as a screw device.

Though both the tenons of the upper middle joint seem to have been shortened, it does not appear that the section's sounding length has been reduced. The headjoint socket was probably deepened to give it strength after the joint was shortened, so perhaps the corresponding tenon, already 28.5mm long, has only lost a millimetre or two of its original length. In the heartpiece, by contrast, the large bore and small outside diameter leave a wall too thin to allow the socket to be given the extra depth, up to 10mm, one would expect. Thus the tenon that fits it is left a rather stubby 16.1mm long. The original heartpiece mount could have been up to 30mm long instead of the 18.5 it is; a first attempt at reconstruction showed that such a normal-length mount gave a scale whose low notes were not satisfactory. Further attempts suggested that the heartpiece mount had been shortened by only 8mm; restoring this much length gives the instrument a well-balanced scale and a socket 20.5mm deep. The comparative shortness of the resulting heartpiece mount (26.5mm) on the reconstruction is warranted by that of another apparently early ivory flute, the one by Scherer (Y14) in the collection of Nicholas Shackleton, Cambridge.8 The key, shortened to a fraction over 52mm, could not have been less than about 56.6 originally, and if it reached to the bottom edge of the 6th hole--its maximum practical length--the heartpiece must have been at least 4mm longer at the bottom end than it is now. So the total increase in length of the heartpiece would be about 12mm, to give a new sounding length of 140.9mm.

With the flute restored to something like its original length so that it will play at or slightly above the pitch it was designed for, let us now consider the bore. Is it in its original condition, or was it altered when the pitch was raised? Reamer marks at the top of the heartpiece or in the footjoint might indicate this, but there is no sign of irregularity: the bore is quite highly polished, and still shows an evenly-spaced pattern of tiny striations such as would be left by a burr or a small nick on the cutting edge of a reamer. Otherwise, the only indication that the bore is or is not in its original condition will be whether or not our reconstruction of the flute works as well as an instrument made with such evident skill and care ought to do.

One feature at least is an obvious alteration. The bore of the top tenon of the left- hand joint has been crudely enlarged with a knife or scraper. Here and in the places where tenons have been shortened, the bore might tend to expand or contract more or less in the same proportion as before. A "reconstructed" bore is shown in Figure 2.

Our model having the dimensions suggested here plays at 392--400 cps, depending on the player.9 The instrument has excellent intonation in all keys (except that the hole under the key produces D# rather than Eb), a good, even sound quality, and a quite exceptional facilty in the highest register. Clues that the instrument with its newly-restored dimensions works as intended are that F''', a note that on out-of-condition instruments is often difficult to obtain, is well in tune and easy to play; and A''', another note which sometimes seems reluctant to speak, can even be attacked piano and played with a crescendo or a diminuendo. From G' downwards, the first octave is strong and even in quality, so that even the unusually low tessitura of the G major trio-sonata BWV 1027 does not seem out of place. Our model's playing qualities certainly do not make it less suitable than any other pre-1750 instrument-type for any of the cantata parts or solo flute music of J.S. Bach.10 The unusual length of the headjoint, the shallow conicity of the bore, and the general appearance, sound and response of our model of the instrument are quite similar to the Brussels I.H. Rottenburgh (Conservatoire Museum No.2001, pitched at about a=400), copies of which some performers are using today to play Bach. However there are pronounced differences in tonehole spacing and bore profile, as well as in wall thickness--they are made of different materials--which distinguish the two instruments no matter how similar they may be in appearance11.


Whether or not this is considered an accurate method or a convincing reconstruction of the Eichentopf flute in its original condition, the model does at least confirm that the bore is generally unaltered and the original proportions must have been more or less as suggested here. With this in mind, it will be necessary to reconsider the catalogue's dating of the instrument at 1730 or later.

Quantz, who has been followed by all subsequent authorities, states that the three-joint flute was divided into four about 1720 to allow changes of pitch by means of corps de rechange.2 But if we consider flutes proportioned like the Eichentopf and I.H. Rottenburgh, this explanation does not seem entirely convincing. The upper middle joint of these flutes is already very short, while they play at the lower of two common chamber-music pitches.13 Upper middle joints more than a few millimetres shorter would upset the visual proportions of the instrument and bring the first tonehole uncomfortably close to the tenon: even when the Eichentopf flute's pitch was raised by a whole tone, the maker who did the work avoided the usual course of making the upper middle joint shorter. There is no indication that longer alternative sections existed--indeed, if they did, their pitch would be exceptionally low. It seems improbable, then, that this type of instrument, which retains the general proportions of the three-joint instrument and thus would seem to be very closely related, was developed to accommodate corps de rechange.

The only practical way to shorten the sounding length of such a flute by more than a few millimetres would be to recast the proportions of the instrument's top half so that the headjoint was shorter and the longest middle joint, playing at the present pitch or slightly below it, longer. Thus it would play at its present pitch with the longest of a set of middle joints, and shorter ones would give higher pitches. On surviving instruments made after about 1720, and on all known examples which actually have corps de rechange, precisely this change in proportions has occurred. The maker who raised the pitch of the Eichentopf followed the same prescription by drilling a new embouchure, effectively shortening the headjoint in relation to the rest of the flute.

So if Quantz was mistaken, and the three-joint flute was not divided into four to make corps de rechange possible, why was it done? The unusual bore of the Eichentopf flute provides a plausible, if somewhat surprising, answer to this question. Eichentopf, or the flutemaker who worked for him, clearly wished to make the bore in the region of the fourth and fifth holes larger, while leaving it undisturbed under the third hole. On a three-joint flute this would be impossible, because a reamer which cut a given diameter could not be used beyond a point in the tube that had a smaller diameter. The only effective method of making a bore like this would be to divide the three-joint flute's middle section in two, reverse the taper at the end of the first of the new pieces and begin the second with a larger bore, and this by itself is a perfectly valid reason, from an instrument-maker's point of view, for making the division. Such an innovation might have occurred to any maker, of course, but it might not have seemed like much of a novelty to one in a workshop specialising in oboes, which have two main sections joined by a tenon and socket in the middle.

We must also consider the possiblity that the bore of the Eichentopf flute arose as a result of the division of the three- joint instrument into four, not as its cause; that makers only took advantage of the opportunity to reverse the taper in the middle of the instrument and to create large steps in the bore after flutes in four joints (with corps de rechange) were already common. But this hypothesis would require the assumption that the three-joint flute was divided either for a reason we suspect to be unsound, or for some third reason which we do not know.

Surviving examples with Eichentopf/Rottenburgh proportions are extremely scarce. Though drawing conclusions from the random selection of instruments that have survived the centuries is risky, the rarity of this particular type might indicate that not many were made; that the idea of providing interchangeable middle sections to alter the pitch occured shortly after, and as a direct result of, the appearance of the first four-joint flutes. This would have provided an attractively simple explanation for the division of the three-joint flute into four, and the intermediate stage with its long headjoint and short upper middle joint would quickly have become obsolete.


Our experience with the Eichentopf flute suggests that it could be the unique survivor of a short-lived but important stage in the development of the instrument: the three-joint flute had been divided into four to give the maker access to the central part of the bore, but corps de rechange had not yet been supplied, at least by this maker. Considering only evidence in the instrument itself, which the construction of a successful model shows to be unaltered, there is reason to suppose that the Eichentopf traverso is among the very earliest four-joint flutes to survive to the present, probably dating from shortly before Bach's arrival in Leipzig.


1Bruce Haynes, `Bach's pitch standards: the woodwind perspective', Journal of the American Musical Instrument Society XI (1985), 101. A full survey of woodwind-making in Leipzig during the early eighteenth century is in Herbert Heyde, `Der Instrumentenbau in Leipzig zur Zeit Johann Sebastian Bachs', in 300 Jahre Johann Sebastian Bach ed. Ulrich Prinz, Tutzing: Hans Schneider, 1985

2Phillip T. Young, Twenty- Five Hundred Historical Woodwind Instruments, New York: Pendragon Press, 1982, s.v. "J.H. Eichentopf".

3Eichentopf may have made instruments sold by the Leipzig dealer Hirschstein (see Paul Hailperin, `Three oboes d'amore from the time of Bach', Galpin Society Journal XXVII (April 1975) p.26). Thus three other flutes, all in ivory, could be by Eichentopf: the Hirschstein No. 13 in the Musikhistoriska Museet, Stockholm; the Hirschstein flûte d'amour in the Dayton C. Miller Collection (DCM 1267) and a similar one listed by Sachs in Sammlung Alter Musikinstrumente bei der Staatlichen Hochschule für Musik zu Berlin, Berlin: Julius Bard, 1922, col. 256. The Miller instrument, while less well-made than the Leipzig flute, shares some of the distinctive bore characteristics noted below.

4Herbert Heyde, Flöten: Musikinstrumenten-Museum der Karl-Marx Universität Leipzig, Katalog, Band 1, Leipzig: VEB Deutscher Verlag für Musik, 1978, p.84 and the illustration in Table 9.

5One other traverso with mounts turned in this style is known: a boxwood instrument by Palanca, No. 851 in the Dayton C. Miller Collection. It is finely made and the style of turning, though unusual, is quite clearly deliberate.

6Some other instruments that have a large step between the upper middle joint and the heartpiece are: ivory Beukers, Haags Gemeentemuseum Ea 414 (1933); ivory Bizey, Paris Conservatoire 439; boxwood Schuchart belonging to Stephen Preston; boxwood Denner recently acquired by Konrad Hünteler. The first two instruments also have a step from the bottom of the heartpiece to the top of the foot. I am grateful to Rod Cameron for generously sharing data from his studies of these flutes.

7An overview of this subject is in Ardal Powell, "Science, Technology and the Art of Flutemaking in the Eighteenth Century," The Flutist Quarterly XIX.3 (Spring 1994), 33-42. [See technology.php3 in this directory]

8Illustrated in Phillip T. Young, `The Scherers of Butzbach', Galpin Society Journal XXXIX (September 1986), Plate VIII.

9The model made for the purposes of this study was in an artificial ivory material of cast polyester resin. Ivory was avoided for two reasons: it would have made transporting the model from New York to Leipzig for comparison with the original in March 1993 illegal under the Convention on International Trade in Endangered Species (CITES), and it would have provided stimulus to the ivory trade, though small, of a kind conscientious wind instrument-makers are presently trying to avoid.

10See Ardal Powell with David Lasocki, `Bach and the Flute', Early Music 23.1, 9- 29.

11Another flute with similar proportions, in ivory, by Johannes Scherer Junior (Butzbach: 1664-1772) or possibly Georg Henrich Scherer (Butzbach: 1703- 1778), is No. 153 in the Vleeshuis Museum, Antwerp, illustrated in [J. Lambrechts-Douillez], Catalogus van de Muziekinstrumenten uit de versameling van het Museum Vleeshuis, Antwerp: Ruckers Genootschap, 1981, p.63.

12Quantz, Versuch I.9. The earliest datable four-joint traverso known to me has unfortunately not survived to the present: an ivory instrument by Thomas Boekhout (Flanders: 1665-1715) was catalogued in 1922 by Sachs as No. 2678 in what is now the Staatliches Institut für Musikforschung in Berlin, but has since been lost.

13See Haynes, `Bach's Pitch Standards'.


  Go To Top Of Page Top of page



Copyright © 2000,