TThhee WWoorrkkiinngg MMeetthhooddss ooff GGuuaarrnneerrii ddeell GGeessùù aanndd tthheeiirrIInnfflluueennccee uuppoonn hhiiss SSttyylliissttiicc

DDeevveellooppmmeennttText and Illustrations by Roger Graham Hargrave

PPlleeaassee ttaakkee ttiimmee ttoo rreeaadd tthhiiss wwaarrnniinngg!!

Although the greatest care has been taken while compiling this site it almost certainly contains many mistakes. As such its contents shouldbe treated with extreme caution. Neither I nor my fellow contributors can accept responsibility for any losses resulting from information or opin-ions, new or old, which are reproduced here. Some of the ideas and information have already been superseded by subsequent research and de-velopment. (I have attempted to included a bibliography for further information on such pieces) In spite of this I believe that these articles arestill of considerable use. For copyright or other practical reasons it has not been possible to reproduce all the illustrations. I have included thetext for the series of posters that I created for the Strad magazine. While these posters are all still available, with one exception, they have beenreproduced without the original accompanying text.

99

TThhee MMoouulldd aanndd tthhee RRiibb SSttrruuccttuurree

The elegance and purity of the violin form exer-cises such fascination that it has been the subject ofenquiry for more than two centuries; questions aboutits design have generated almost as much interest asthe subject of Cremonese varnish and its composi-tion. Although several eminent studies have demon-strated the regular use of mathematics in the processof early instrument construction, recognising thatmathematics was applied is considerably easier thanunderstanding how it was applied. If they ever ex-isted as such, the exact mathematical formulae usedto create the first Cremonese violins are unlikely tobe rediscovered.

The violin family was developed more than fourhundred years ago, and it was already two hundredyears old when Guarneri Del Gesù began making in-struments. Although it is probable that he wouldhave known about any existing mathematical for-mulae, he may never have been required to use them.Unlike the Amatis (and, possibly, Stradivari), Del Gesùalmost certainly made instruments based upon al-ready existing designs. Indeed, his supposed varietyof original designs is assuredly an illusion. If his vio-lins are unique it is because of his free-ranging use oftools and materials, and above all his creative fantasy;

they were not, in their basic form, the result of anyinnovative mathematical composition.

The outline of a violin is only one element of itscomplex design. It is still not known which was es-tablished first, the mould around which the violinwas constructed (which represents the chamber ofair inside the instrument) or the complete violin,from which the mould was then derived. However,because it is the starting-point for the process of con-struction, the mould has been the major preoccupa-tion of design theorists. The use of an inside mouldwas central to the Amati system, and accordingly itbecame the cornerstone of all Cremonese construc-tion. In the seventeenth century, with the exceptionof Jacob Stainer it is hard to find evidence of the con-sistent use of such a mould by makers outside Cre-mona.7 Whether later Cremonese makers createdtheir moulds mathematically or simply copied oradapted existing moulds, they were still effectivelyworking within the Amati tradition.

The contents of the Museo Stradivariano indicatethat Antonio Stradivari developed and used a com-paratively large number of different moulds duringhis long working life. But this does not mean that all,or any, were mathematically devised. Pollens8 hasdemonstrated that when the twelve surviving Stradi-vari violin moulds are superimposed on each other,

they fall into several groups.9 Stradivari appears tohave retained certain sections of particular moulds(for example, the top and centre bouts), while hemodified the remainder (perhaps the lower bouts);as a result, the differences between some of hismoulds are remarkably small (figure 5).

This implies a gradual adjustment to the lines ofthe form, rather than a fresh mathematically calcu-lated construction for each subsequent development.

The variety to be found among Del Gesù’s violinsgives the initial impression that he too must havecreated a number of moulds. Yet this was probablynot the case. By the 1730s Cremona had a rich, well-established tradition of lutherie and there was noneed for each maker to produce new designs. Thetrend appears to have been towards copying or mod-ifying existing ones. A personal style could beachieved by adjusting a line here or there or bychanging the details of the scroll, soundholes oredgework; it is not unreasonable to assume that thisis what Del Gesù did. Although Stradivari would thenappear to stand alone in his radical reworking of theAmati mould, it could be argued with some force thateven he was simply adapting rather than innovat-ing.10 There may have been some underlying mathe-

matical formulae in Cremona, but well before the endof the eighteenth century the violin maker GiovanniAntonio Marchi commented that his contemporarieswere merely tracing earlier violins in order to arriveat their own models.11

Count Cozio Di Salabue, in his extensive notesabout the classical Italian violin makers, says of DelGesù, “The violins that he built from about 1731 until1743, though ordinarily he retained the same form,they are generally quite badly worked.”12 Disregard-ing his verdict on the craftsmanship, what is inter-esting is that Count Cozio was clearly of the opinionthat Del Gesù used a single mould. To some extentthis notion can be confirmed by superimposing theapparently wildly different outlines of Guarneri vio-lins made during this period. At no point do theseoutlines encroach upon the single mould which canbe marked out within them.

The exceptions to this rule are the violins of thetransitional period, which include the “Dancla”. Withits shorter C bouts and the lower placement of theupper corners, the “Dancla” was unquestionably con-structed on a different mould from every instrumentknown or believed to have been made by Del Gesùafter 1731. As might be expected, its outline matchesseveral outlines of Del Gesù’s father; in particular, itis virtually identical to one dating from 1705. TheseGiuseppe Filius Andreæ outlines are in turn derivedfrom Del Gesù’s grandfather, Andrea Guarneri, whoclearly obtained them from the Amatis. In fact the“Dancla” outline matches several Brothers Amati vi-olins and even an Andrea Amati remarkably well, andalso fits quite well around the “S” (ms. 39) mould atthe Museo Stradivariano. In comparing these out-lines, the positioning of the corner blocks and thesize of the overhangs could not be taken into ac-count. Nevertheless, there can be little doubt aboutthe original source of the design.

It is possible that after 1731, Del Gesù was usingtwo very similar moulds. If this was indeed the case,the first was preferred up until about 1738/9, al-though the “Stretton”, “Kreisler”, “King” and“Joachim” are among the exceptions which may havebeen built on a slightly wider mould. After about1740, all the instruments appear to have been con-structed upon the wider model. This could well bebased upon the “grand pattern” of the Amatis, withwhich it corresponds very well. On the other hand,in almost every case, the violins which appear tohave been constructed upon a larger mould also havelarger overhangs; to complicate matters there are

Figure 5. Several Stradivari mould outlines superimposed upon each other as described by Pollens(see footnote 10).

1100

1111one or two instruments which lie between both cat-egories, in particular the “Ysaÿe”. The idea that suchinstruments as the “Diable”, the “Kemp” and the “OleBull” share the same basic design and were con-structed upon the same or very similar moulds,seems at first glance virtually insupportable. How-ever, the overhang variations may well be enough toaccount for most of the apparent differences in bothlength and width.13 In the final analysis, it may be im-possible to prove whether Del Gesù used one, two,three or more moulds, but in any case they wouldhave been extremely close in size and shape.

All the violin moulds in the Museo Stradivarianohave certain features in common (figure 6).

They are Xat boards, mainly of slab-cut walnut, al-though willow and poplar are also to be found. Theseboards range in thickness from 13 to 15 mm, and DelGesù’s moulds must have been much the same in thisrespect: A thicker mould would not have allowedboth sets of linings to be fitted, whereas a thinnermould would have been too flexible. Each mould hassix recesses, or mortises, into which the two endblocks and the four corner blocks were set and ontowhich the ribs were eventually glued. These have ob-

viously been subject to extensive wear and severalhave been repaired. The interior work of Del Gesù’sviolins, the size, shape and position of the blocks andlinings all indicate that he used a mould which wassimilar in concept to those in the Stradivari museum.Unfortunately, reliable measurements of Del Gesù’sblocks are difficult to obtain and for the time being itis only possible to make general observations abouttheir size, shape and position in relation to the sur-viving Stradivari moulds.14

Although similar in depth, Del Gesù’s neck blockmortises were probably somewhat shorter than thoseof the Museo Stradivariano moulds: His rare surviv-ing neck blocks seem to be about 50 mm across,15considerably less than the neck mortises in theStradivari moulds, which range from about 55 to 66mm. Del Gesù's end-pin blocks are likewise narrower

– about 43 mm compared to the 45 to 50 mm foundon Stradivari instruments and on the existingmoulds. Most of the latter have small cut-outs in thetop and bottom block mortises to facilitate separa-tion of the mould from the rib structure. Measuredinside the instruments, there is little variance be-tween the corner blocks of Del Gesù and Stradivari:Both are about 25 mm wide and closely match themould mortises.

Each of the Stradivari moulds has ten 8 mm to 9mm round holes set at strategically important points.These were for locating the binding sticks, used toclamp the ribs to the blocks. The technique is fully

Figure 6. Reproduction of a Stradivari violin mould from the Museo Stradivariano.

Figure 7. Method used to bind the ribs to the mould as used by Stradivari.

1122described by Pollens and Sacconi and further sup-ported by the surviving sticks and counter blocks inthe Museo Stradivariano (figure 7).16Without an au-thenticated Del Gesù mould, it is virtually impossibleto establish the exact method he used to bind the ribsto the mould and blocks. However, the possibility thatsome of the museum moulds were not of Stradivari’smaking suggests that this method of clamping was inwidespread if not universal use among Cremonesemakers.17

Most of the moulds have incised markings, includ-ing a centre line, lines showing the position of thecorner blocks and a compass point with two shortarcs of a circle indicating the block heights (figure 6).In almost all cases, Del Gesù’s end-pin and cornerblocks were finalized to a height of about 32 mm; theneck block was lower, usually about 30 mm. Thesemeasurements are generally consistent with those ofStradivari but there are several interesting excep-tions, including the “Cannon”, the “Carrodus” andthe “Leduc”, where all the block heights are in-creased by at least 1 mm. Interestingly, in CountCozio Di Salabue’s original notes on a fourth violin,the “Vieuxtemps”, the measurements are given as 33mm and 31 mm, making it consistent with thesethree. Assuming the Count was correct, the ribheights must have been reduced subsequently. (Seemeasurements of the “Vieuxtemps”, p. 95.)

Previous Cremonese makers including Stradivariusually made their blocks and linings of willow.18However, following the example of his father’s laterworks, Del Gesù preferred spruce. Both blocks andlinings were probably fashioned from split wood,since although they often give the appearance ofhaving been cut back in haste, no seriously unevensplitting has occurred. Otherwise, Del Gesù paid scantattention to the course of the annual rings which canbe found running in all directions (figure 8).

His block wood is generally coarser than the bellywood and it may be that he was economising by usingoVcuts, since there is no doubt that willow linings areeasier to bend. The particular shape of Cremonesecorner block mortises meant that the blocks onlyneeded to be squared on two sides and this wouldhave been particularly useful whenever oVcuts werebeing used.

Once the blocks were glued in position, they weremarked out for cutting. Stradivari achieved this withthe aid of small, individual templates (figure 6); re-markably, several sets of these templates have sur-vived the ravages of time.19 Both Pollens and Sacconidescribe the process he used.20 Individual templateswere not only more straightforward to make than thealternative full- or half-body templates; as mouldsbecame worn, warped or damaged, especially in themortise areas, they were both more easily adjustedand more compatible with badly aligned blocks. Itwould seem that Stradivari used only two cornerblock templates, one for the top corners and one forthe lower corners, and he simply transposed thesefrom side to side. Although it is unlikely that thecurves of these templates were altered, even rela-tively minor variations in the positioning of theblocks (as a result of a worn or twisted mould) wouldhave contributed to the type of variation in cornershapes which we associate with all Cremonese work,including that of Nicolò Amati and Stradivari. It maybe significant that the “PG” mould21 has four indi-vidual templates. This mould has been heavily re-paired in the mortise areas, and as a result the cornerblock mortises on one side are slightly deeper than

Figure 8. Reproduction of the four corner blocks and the end block of “Soil” del Gesù, showing the direction and width of the year rings.

1133those on the other. Two extra corner block templateswere made to Wt the reshaped mortise holes, but thecurves have been carefully matched so that the cor-ners theoretically remain the same.

One of Del Gesù’s consistent peculiarities is his se-lection of rib wood. It was usually fine grown maple,cut on the quarter. Unlike the Amatis, he appearsnever to have used slab-cut ribs, even when usingslab-cut backs; centuries of experience had probablytaught the later Cremonese masters that these wereprone to cracking and warping.22 Del Gesù’s ribs sel-dom matched the back wood and only rarely werethey cut from the same billet. Even where the matchappears perfect, as with the “Leduc”, he generallyfailed to align the slope of the figure in the same di-rection all round the instrument.23 Often the ribswere made up from unmatched pieces: Plain woodwas used in conjunction with highly figured wood, ornarrow and wider Xames were mixed. In particular,the lower rib(s) were often completely different fromthe others.

Del Gesù thicknessed his ribs with a coarse-toothed plane iron. Unlike Stradivari, he made no at-tempt to remove the marks left by the toothed ironon the inside of the ribs, and they are usually clearlyvisible through the soundholes. The ribs were fin-ished with a scraper on the outside only. However,even on his early instruments (see the “Kreisler”photographs, pp. 24-27, volume I) the remains oftooth plane markings can sometimes be seen on theoutside of the ribs, beneath the varnish.

As a rule, Del Gesù thicknessed his ribs fairly evenly to about 1 mm. Exceptionally, as in the case ofPaganini’s “Cannon”, they average 1.5 mm. Occasion -ally, as with the “Soil” of 1733, in the immediate cor-ner block gluing area they are reduced to as little as0.3 mm. Thinning the rib ends in this way was onlyviable because the block itself provided a stable back-ing. This may have been done to make the curves ofthe centre bouts easier to bend. From the beginning,Del Gesù’s centre bout curves were never as tight asthose of Stradivari; in fact, because of these moreopen curves, it may even have been possible for himto bend them without the aid of heat.24 There is someevidence which suggests that Del Gesù was usingfresh wood;25 this too would have eased the bendingprocess. In fact, the combination of shallow curves,thinner ribs, and fresh wood must have made the ribbending process easier for Del Gesù than it was formost Cremonese makers. Nevertheless, it appears tohave been an onerous task for him. Numerous creases

and cracks occur in the tighter curves of his rib struc-tures, the “Heifetz” being an excellent example. Gen-erally, the ribs are quite buckled along the Xame.Although this may be due to the use of fresh wood, inmany cases it looks very much as if the ribs were bentin a series of small creases rather than in a smoothcurve.

Following the marking process, the blocks were cutto shape, beginning with the centre bout curves.26The centre bout ribs were bent and glued in place,and the outer curves of the block to which the upperand lower ribs would be attached were cut. Del Gesùhabitually cut the points of his corner blocks consid-erably shorter than those of Nicolò Amati and Stradi-vari, producing in his earlier violins rather smallcorners.27 In these earlier works, the corner blockswere clearly set out and cut with care. Judging by DelGesù’s attitude to other tasks, it seems unlikely thathe worked without reference to some form of tem-plate even in his later years;28 however, it can be as-sumed that his swift cutting of the blocks resulted inrib corners which were neither square nor true to thecurve of the template (figure 9). On Del Gesù’s violinsit is not unusual to find the four corner blocks cutwith different curves, running in different directionsand finishing either shorter or longer than eachother. Furthermore, the rib mitres often lean at var-ious angles, a feature which can be extreme in laterworks. As the Hill brothers observed, “the cornerblocks were not left true by the gouge – the only toolhe made use of – nor were the sides accuratelybent.”29 Small wonder that as a result, all eight cor-ners (back and belly) are frequently dissimilar.

Figure 9. Possible curve and length variations of the corner blocks, of which del Gesù used several combinations.

Whenever the blocks were cut at different angles,it caused the ribs to twist slightly on the mould. Thisin turn altered the curves outside the immediate areaof the corner, creating back and belly outlines whichwere different from each other and from the moulditself. On the “Lord Wilton”, the centre bout ribs areclearly not square; they taper inwards towards thebelly side, and the belly is considerably narroweracross the centre bouts as a consequence. This taperis more marked on the treble side, where the centrebout of the back is much straighter, and the belly out-line is deeply curved in compensation (figure 10). Theexplanation is almost certainly that the ribs twistedon the mould as a result of the corner blocks not hav-ing been finished square to it, and in fact this detailis apparent from the rib corners. Such discrepanciescould have repercussions far beyond simply alteringthe outlines: In the long run, the disposition of thesoundholes was also affected. As the Hills acknowl-edge, “We cannot say the master was over particularin making his sides conform quite accurately to themould. Approximately correct was in all cases suffi-cient unto the day!”30 This approximation is theprime cause of any variation in the shape of the out-

lines, although other contributory factors willemerge in the course of our discussion.

Any changes to the two endblocks were also capa-ble of modifying the upper and lower bouts signifi-cantly. This can be observed on some of Del Gesù’slater works, where the top block is occasionallyrather pointed, as on the “Heifetz” (1741), or ex-tremely Xat, as on the “Lord Wilton” (1742). By fin-ishing the end blocks proud of the mould, the ribstructure could have been lengthened significantly.This process may or may not have been carried outdeliberately. On the “Vieuxtemps” of 1741, the posi-tion of the neck block appears to have been extendedconsiderably, resulting in what for Del Gesù is alonger than usual stop length and upper bouts.

The extension of the “Vieuxtemps” may have beencarried out on the normal sized Del Gesù mould, inwhich case the neck block would have been cut proudof it (figure 11). The ensuing gaps between the ribsand the mould could have been filled out in advance,possibly by the addition of paper or card strips fixedto the edges. Equally, the ribs may simply have beenstretched across these gaps. Another alternativewould have been to elongate the ribs after removingthem from the mould and before finalising the backoutline, and, as shall be demonstrated, this seems tobe the most likely explanation.

Interior of “Soil” del Gesù, 1733, showing typical form of delGesù’s blocks and linings, middle bout linings quickly mortiseddeep into corner blocks.

Figure 10. Cutting the corner blocks at various angles led tovariations in the back and belly outlines of del Gesù’s instruments.

1144

The “Fountaine” pochette of 174031 probably hadthe opposite treatment. Here the lower block appearsto have been extended, producing distorted lines inthe lower bouts which suggest that in this case thegaps between the ribs and the mould were notpacked out. The “Fountaine” also has a longer bodythan the more conventional looking 1735 pochette,the “Chardon”,32which was presumably made on thesame mould. As has been mentioned, Del Gesù’s cor-ner blocks were invariably kept short and stubby.Until about 1742 the rib mitres were also short, withthe top and bottom ribs only barely overlapping thecentre bout ribs (figure 12)

.

After this period the overlapping ribs becamelonger; this is well illustrated by the “Leduc”, wherethe upper and lower ribs overlap the centre bout ribsby a considerable margin. Occasionally, the centralrib was hardly feathered at the end, and both rib endssimply came together to form a thick wedge. Rib

mitres of this nature may have been the result of twodistinct trends in Del Gesù’s work: Firstly, the cornerblocks were being finished even shorter and withflatter curves, and secondly, the back and belly cor-ners were becoming increasingly longer and morefragile. The vulnerability of these longer back andbelly corners may have inspired Del Gesù to providemore support through extended and somewhatthicker rib mitres. Almost uniquely among Cre-monese makers, Del Gesù blackened the ends of therib mitres.33

It seems to have been the accepted practice in Cre-mona to form the upper ribs from one continuouspiece. This holds true for all of Del Gesù’s instru-ments, with the possible exception of the “Ko-rtschak”. Although virtually all existing instrumentshave had their upper rib cut through during theprocess of mortising replacement necks, the grainand figure of the upper rib wood always appears tobe continuous across the neck root. The reason forthis practice is not hard to fathom. The neck was fas-tened to the block with the aid of several nails, whichwere driven through the block into the neck root.Even using pre-drilled holes there was some dangerof splitting the block, a calamitous occurrence at thisstage of the construction; the reinforcement pro-vided by the continuous upper rib running across theblock reduced this risk considerably. Del Gesù’s useof (possibly) up to five nails would have made hiscoarse spruce neck blocks particularly vulnerable tosplitting. Cremonese lower ribs were also generallyof one piece, a fact which can easily be demonstratedand which supports the theory that the upper ribswere made in the same way. The use of upper andlower one-piece ribs avoided the need for a carefuljoint at the middle of the lower block. A one-pieceback is long enough for a one-piece top rib to be cutfrom the same billet, creating a perfect match. How-ever, one-piece bottom ribs are much longer than theback of a violin, and these could not have been cutfrom the back wood unless the billet was longer thannecessary.’34 This may be why many Del Gesù violinshave a one-piece bottom rib which matches neitherthe other ribs nor the back, but does match the bot-tom ribs of several other violins. It would seem thathe simply cut several bottom ribs from a suitablylong piece and, regardless of their matching quali-ties, used them as bottom ribs only. Generally, when-ever Del Gesù used a one-piece bottom rib, the centreof the instrument is marked by a tiny knife cut on theedge of the rib, where it comes into contact with theback plate. On the “Soil” of 1733 this tiny nick on theribs lines up with the remains of a centre line scribed

Figure 12. Variations in the rib mitres at the corners. a) The earlier shorter version. b) The extended outer rib of the later longer corners.

Figure 11. How the upper ribs of the “Vieuxtemps” may have been extended by packing out the mould.

1155

1166into the overhang of the one-piece back. Such detailscan also be found on instruments by Stradivari andAmati.

The advantage of a one-piece bottom rib is that itavoids the need for a careful centre joint. The greatdisadvantage of both one-piece top and bottom ribsis that they are more difficult to bend with precision,and they compound the inaccuracies of badly cutblocks. Del Gesù obviously had problems in both re-spects. Although the first corner of a one-piece rib isrelatively easy to bend, as is the gentle curve aroundthe upper and lower bouts, the second corner is lessso. It must be bent against the prepared curve of thebouts and at exactly the right place. On several laterinstruments, Del Gesù plainly had difficulties inachieving this: When the ribs were fastened to theblocks, a small swelling appeared where they couldnot be pressed home against the mould, and thiseventually transferred itself to the outline. This mayprovide some explanation for the fact that when lateoutlines are superimposed upon earlier outlines, theyappear to spring outward at the blocks in a fairly ran-dom manner. To some extent this is true of the “LordWilton”, the “Carrodus” and the “Cannon”.

Especially after about 1740, Del Gesù’s apparentlyerratic preparation of the corner blocks resulted inconsiderable twisting of the one-piece ribs on themould, leading in turn to even greater variations inback and belly outlines. He may have been forced tocut some lower ribs to compensate for such twisting,but this was probably a rare occurrence. Wheneverhe appears to have used a two-piece bottom rib, it ismore likely to be the result of later repair work. The“Cannon” and the “Leduc” are rare examples of gen-uine two-piece bottom ribs. All the “Leduc” ribs alsomatch the back of the instrument, an uncommon fea-ture, and it can be assumed that Del Gesù cut themfrom the back wedge which was otherwise too shortfor a one-piece bottom rib.

With the ribs bent and glued in position, Del Gesùfitted the linings, which were cleaved from straight-grained spruce. In common with those of Stradivari,they were probably about 2 mm ¥ 7.5 mm in section.Employing the Amati system, he mortised the centrebout linings deep into the corner blocks (figure 13).Often the linings were brutally spliced into crudelycut oversized mortises. Del Gesù did what was essen-tial with the minimum of fuss; his rib structures maybe fairly described as stable but hastily made. In hisfinal years, he appears to have chosen an evenquicker method of mortising the centre bout linings.

This was a technique favoured by Carlo Bergonziwhereby, instead of cutting an oblong slot, two rapidknife cuts were used to prepare the block and oneswift cut to prepare the lining (figure 14). The liningsfor the “Ole Bull” were inserted in this way.

Until recently, the deep mortising system em-ployed by Del Gesù and his contemporaries wassomething of a puzzle. However, when it was realisedthat the classical makers glued the back and front lin-ings in place before releasing the ribs from themould, the reason for both the deep mortises and thepeculiar shape of the corner blocks became apparent(figure 15). Using this method the ribs are sprungrather than slipped off the mould, and the wholeprocess of disengaging the two can place consider-able stress on the structure. Mortising the liningsdeep into the corner blocks helps prevent accidentaldamage during this delicate procedure. With bothsets of linings glued in place, the rib structure is far

Figure 14. SimpliWed drawing of the method of mortising the centre bout linings, favoured by Carlo Bergonzi, which del Gesù used on the “Ole Bull”.

Figure 13. SimpliWed drawing of the Amati system of mortising the centre bout linings into the corner blocks.

more stable both before and after being releasedfrom the mould. There could be no possibility of fur-ther inaccuracies being built into the structure as aresult of forcing the second set of linings into placeon the fragile ribs, unsupported by the mould.

Once the rib structure with its strengthening lin-ings had been completed, but before the mould wasremoved, several important operations were carriedout. In each case these operations were aided by theextra stability which the mould imparted to the ribstructure. Initially, Del Gesù planed Xat the side ofthe ribs which would come into contact with theback, and the rib heights were established at approx-imately 32 mm, exceptionally 33 mm. It is generallyaccepted that the Cremonese ribs were tapered insome way, with the end-pin block being higher thanthe neck block. Where this taper begins and ends,and whether or not it was taken from the back or thebelly side, is difficult to establish, especially whenmore than two hundred years of damage and distor-tion have clouded the evidence. In Del Gesù’s case,his characteristic inconsistency makes even well-pre-served examples difficult to evaluate.

The available data indicates that Stradivari taperedthe ribs on the belly side. The “ex Regnier” Stradivariviolin of 1722 has the remains of a scribe line on theupper bass rib, indicating the final height of theupper block (figure 16).

This scribe line runs parallel to the back and ribjoint, from which point it was obviously struck witha marking gauge. The rib measurements, and theangle which the scribe line makes between the neckroot and the top edge of the rib, leave little doubt thatin this case the taper ran between the neck block andthe upper corner block only. This observation is sup-ported by the measurements of a large number ofwell-preserved Stradivari violins.

In spite of the fact that Del Gesù’s rib structureswere not always finished with Stradivarian accuracy,the available measurements indicate that he also ta-pered them from the top corner blocks to the neckblock. The reasons for this taper are unclear – andunless any definitive documentary evidence is un-covered, will probably remain so – but the result isthat the belly is bent downwards from the upper cor-ners. This certainly imparts some stress to the belly,which may provide some acoustical or structural ad-vantage.

Having tapered the ribs, Del Gesù again turned hisattention to the linings. These he trimmed back witha knife; numerous tiny cuts on the insides of the ribssuggest that they were cut back with some speed (figure 17).

The edges of the moulds in the Museo Stradivari-ano are also heavily scarred with knife cuts, revealing

Figure 15. The open form of the Cremonese corner blockmortise, with its angle of more than 90°, making it easier to spring the block oV the mould.

Figure 17. SimpliWed drawing showing the position of the knife marks on a set of del Gesù ribs.

Figure 16. SimpliWed drawing showing the position of the scribe line on the 1722 “Ex Regnier” Stradivari.

1177

1188that the linings must have been cut to shape whilethe mould was still in place.35 This was only logical:With the rib structure still rigidly fixed on the mould,the job would be that much easier to accomplish. Be-sides, with both sets of linings glued in place, shapingthe linings before removing the mould made the taskof removal itself considerably easier.36 Stradivarishaped his linings with some care, certainly morethan the Amatis, but Del Gesù simply sliced themback – an act which confirms that the linings weremade from split wood. Although he made some at-tempt to take off the roughest edges with either arasp or some flexible abrasive,37 he probably rea-soned that they were of little aesthetic importance.With the exception of a few minor details, the ribstructure was now complete and ready to be pre-sented to the back, in order to mark the outline.