Here are the complete detailed notes to all the pieces in Volume 2, taken from Volume 2.

This was first published by Meissonnier in Paris in about 1817-22 with the title Troisième Fantaisie, at a time when Sor was living in London. If there ever was an English edition, however, no trace of it has yet been found. The Meissonnier edition bore the opus number 10, and it was doubtless called the “Troisième” (third) fantasia because Meissonnier counted the works now known as opp. 4 and 7 as the first two fantasias (see the notes to op. 4 above). (When Meissonnier came to publish op. 12, he called it “Quatrième Fantaisie” even though it seems that in England the work now known as op. 12 had been considered as the first fantasia. See the notes to op. 4.)

A slow introduction leads into a melodious theme, followed by two variations; then a section called “Majeur” which is rather an interlude than a true variation; then two more variations and a coda conclude this fine extended work.

*****

The sixth string should be tuned to F, a practice not unusual in Sor’s time (it was used, for example, in op. 1, no. 6).

Bar 11 of the introduction: the last two A flats in the bar are G in the original. Emended editorially.

The dots in the first bar of the theme doubtless apply also at bars 3, 5, etc.

__________

OPUS 11. DEUX THÈMES VARIÉS ET DOUZE MENUETS

This is something of a miscellaneous collection and first appeared in Paris in 1822 or shortly before, with the above title and opus number. Parts of it (and perhaps all) are certainly old, but the organization of so many separate pieces into a single publication is a product of the 1820s.

The theme of the first set of variations is almost the same as that of a “Thème Varié” which Meissonnier published separately with a different arrangement of variations and to which he gave the numbering op. 15(c). Minuet no. 12 was also published as op. 13, no. 4.

The twelve minuets are in an unusual style, with their abrupt changes of register, their sudden running passages, and their scordaturas. They are among the finest of Sor’s miniatures and have deservedly kept their popularity over many years. Some are certainly from Sor’s Spanish period, and it is quite possible that they all are: see Chapter 1 of my book Fernando Sor, Composer and Guitarist. Minuet no. 6 was certainly composed in Spain, having been published earlier by Castro in his Journal de Musique Etrangère (Paris, about 1810 or slightly earlier). Minuet no. 5 is found in a manuscript in the collection of Robert Spencer, London. Early versions of Minuets no. 3 (without attribution) and no. 11 are in MS 722-24[x] of the Biblioteca Histórica Municipal de Madrid (see Luis Briso de Montiano, Un fondo desconocido de música para guitarra, Madrid, 1995, p. 163). It seems likely, but not yet conclusively established, that all the minuets, and perhaps all of op. 11, are in fact music which was originally composed by Sor in Spain. The early versions of minuets nos. 3, 5, 6 and 11 offer particular points of interest and particular problems and deserve a special study which is not pursued in the present edition.

Minuet no. 3 is a special case. In a copy of an edition of op. 11 belonging to Robert Spencer, London, is a manuscript note against this minuet, possibly in the hand of José de Lira, reading: ‘En mai 1823 j’ai été logé à Girone chez M. le Marquis de Capmany dont le fils, Don Francisco Maria de Sabater y Camps, m’a assuré avoir composé le menuet suivant à Saragosse en 1803’ (‘In May 1823 I stayed at Gerona with the Marquis of Capmany whose son, Francisco Maria de Sabater y Camps, assured me that he had composed this minuet at Saragossa in 1803’). According to Ledhuy, Sor’s father had a composer friend called ‘le chevalier de Sabatea’, which may be simply a misprint for ‘Sabater’. Whoever composed the piece, the note does suggest an early date.

In some of these minuets the harmonies sometimes seem strange. But because of their special historical position, in the interests of historical accuracy they have not been changed here. An example is Minuet no. 8, bars 14 and 18, where the first G sharp in each case might be expected to be an A but has not been altered here. Sometimes the harmonies are particularly rich, for example in Minuet no. 12, bar 9. Performers may follow their own judgement about the performance of specific cases.

*****

In the first Thème Varié, Variation 1, bar 31, the third note is C in the original but to judge from the other variations, a D was probably intended and has been put here. In the first Thème Varié, Variation 6, last bar, the last two notes are marked “armonici”. The C must be an artificial harmonic. The E can optionally also be played as an artificial harmonic.

Minuet no. 3, bars 27, 32 and 33: the naturals are editorial on the analogy of bar 28 where a natural is present in the original.

Minuet no. 4, bar 14: the sharp on the G is not present in the original but is implied by the fingering. It returns also at bar 16.

Minuet no. 11, bar 21, first chord, the fingering 4 is 1 in the original. In the last theme and variations, bar 59, the fourth note in the bass is F in the original, changed here editorially.

__________

OPUS 12. FANTASIA This fantasia was first published without opus number in London by Clementi in 1815 or shortly after, a fact which is only now known, thanks to the recent discovery of a unique copy of the London edition. A fine extended work, it is in fact a set of variations and is an elaboration of the simpler set on the same theme which is today known as op. 3 and which dates from Sor’s Spanish period. Sor has added an introduction, coda, and three new variations. The new work is dedicated to Friedrich Kalkbrenner, the famous pianist who lived in London from 1814 to 1823, that is exactly the same period as Sor, and indeed the new work has dimensions more commonly associated with the piano music of that time than with the guitar music. Evidently this theme with variations was a favourite of Sor’s, and most probably this work would have been among the ‘Fantasia’s which he performed in concerts in London. Indeed, because it is an elaborate work, one may hazard a guess that if Sor’s lost Fantasia Concertante for guitar and strings is ever found, it may perhaps turn out to be a version of this fantasia. The theme was also used by Giuliani in his op. 112.

Variation 7 is a new discovery, because it is found only in the newly discovered Clementi edition which is the basis of this edition, and is not present in the later Meissonnier version which I printed in the earlier facsimile edition of Sor’s Complete Works for Guitar. In fact Variation 7 is a fine example of the ancient guitar and lute technique of campanelas. To us today it is extremely interesting because it demonstrates Sor’s links to the past history of the guitar in Spain, but perhaps for Meissonnier it was a vestige of the past and therefore was not modern and perhaps for that reason was cut. In any case I hope modern performers will now restore it to its place in this work.

Campanelas, by the way, are found also at the beginning of the Sérénade op. 37, a comparatively late work. Do they mean that the Sérénade has an older history? Or, more probably, did Sor continue to use this fine old technique even at that late date?

The original title in the Clementi edition was Fantaisie pour le [sic] Guitarre Espagnole.

* * * * *

The indication Lento ad lib towards the end is from the Clementi edition and is not present in Meissonnier.

Bar 18: performers may wish to add a tie from the first G to the second G. (There is no tie in the original).

Bar 199, last chord: the middle note is C in the original edition and is here replaced by B as it is in the Meissonnier edition.

__________

OPUS 13. SIX DIVERTIMENTOS

Another charming set of six short pieces, again first published in London, in 1819. As with the previous sets of divertimentos opp. 1, 2, and 8, this set begins with a short minuet introducing a waltz in the same key, followed by a longer piece in a different key. No. 4, “Cantabile”, is the same piece as op. 11, minuet no. 12. A long and delightful Andante Pastorale, and a march, conclude the set, which, once again, could with advantage be performed as a whole.

The original title in the London edition is Six Divertimentos for the Guitar, 4th Set. The London edition gave no opus number, which appeared only in later Meissonnier editions.

* * * * *

No. 3 bar 20: the natural on the B is added (not in the original).

In no. 3 the fingering in the original is seriously wrong in several places, that is to say simply misprinted or else impossible to play. Was it perhaps added by someone other than Sor? In any case, I have taken the exceptional step of removing all the fingering in this piece. Anyone who wishes to study it can do so in the Tecla facsimile edition.

No. 4 bar 15: the flat is added editorially. (It is present in the version of this piece in op. 11 minuet 12).

The set was praised by a reviewer for Ackermann’s Repository of Arts in 1820 by comparing it favourably with Sor’s vocal compositions, that is, especially his lyrical and successful Italian arietts with piano which were being published in London at this same period:

“Of the merit of the pieces themselves, we shall say quite enough, by stating that they are such as we were led to expect from our knowledge of Mr. Sor’s vocal compositions, of which we have, on several occasions, spoken in terms of the highest commendation …”

Here is the review in full:

Six Divertimentos for the Guitar, by Sor. 4th Set. Pr. 2s. 6d. If hitherto guitar-music has been considered by us as lying out of the sphere of our critical functions, we candidly plead in our defence, the opinion which our own experience had led us to entertain of the limited powers of the instrument; and the unimportant nature of the compositions which had come under our cognizance. It was with this bias that we recently chanced to hear Mr. Sor touch the guitar, and our previous prejudice instantly gave way to astonishment and admiration. We will not attempt to describe the sensations which the magic of his play excited within us, but our readers may form some idea of what we felt, when we state, that this gentleman executed, with the greatest precision, and with the deepest expression, scores of five and six distinct parts, nay, played fugues of the most complicated texture. How this was done, how it can be done, remains, at this moment, a matter of wonder to us. We have seen and heard; we therefore must believe: otherwise, we own, we should have staked a round sum, that two or three guitars had been actively at work at the same time.

This unrivalled perfection we find, upon inquiry, to be the result of a system at once simple and efficient, the fruit of matured experience, and of a diligent inquiry into the nature and the capabilities of the instrument. Without this, Mr. Sor’s excellence might astonish his contemporaries, like a passing meteor, whose reappearance in the horizon is a matter of chance. But, with his system, his art is in a manner perpetuated: some of his pupils already wield the lyre in a masterly manner, and its unassuming elegance daily gains ground in the higher circles. Thus the most ancient of all musical instruments, the invention of Mercury, the darling of Orpheus, seems once more reinstated in its pristine rights. Even music partakes of the spirit of legitimacy which characterizes our era.

Our readers may, in some degree, form a conception of the capabilities of the guitar, by an inspection of the six divertimentos which are the subject of the present article. They are generally written in three distinct parts, and in keys of very different kinds: the harmony is full, and modulation appears in no instance to be pressed or confined by the limits which we ourselves had erroneously considered as dictated by the nature of the instrument. Of the merit of the pieces themselves, we shall say quite enough, by stating that they are such as we were led to expect from our knowledge of Mr. Sor’s vocal compositions, of which we have, on several occasions, spoken in terms of the highest commendation; and a further specimen of which is now in our portfolio, to be brought before our readers in this month’s Critique, if time permits, but at all events in our next Number.

(Ackermann’s Repository of Arts, Literature, Fashions, Manufactures, &c., February 1820)

__________

OPUS 14. SONATA PRIMA (“Grand Solo”)

This one-movement sonata, perhaps Sor’s finest extended work, remains still something of a mystery in its origins. Grandiose, imaginative, using the resources of the instrument to the full, it certainly dates from Sor’s Spanish period and was first published as “Sonata Prima” by Salvador Castro de Gistau in Paris in 1810 or shortly before, without opus number. As a one-movement sonata, it has affinities with the keyboard sonatas of Domenico Scarlatti. In about 1822, Meissonnier published a version of it which he called “Grand Solo”, a name which does not necessarily have Sor’s authority, and in which the opus number 14 was added. There is no way of knowing whether or not either the Castro or the Meissonnier versions had Sor’s approval. It should be emphasized that, for the present, neither edition can be established as having full authority for a definitive text. It may be that a text faithful to the original version of the work—composed, as we know, in Spain—is yet to be found in Spain itself.

Recently a text has come to light which may refer to this work. It is the diary of a Barcelona gentleman of Sor’s time, Rafael de Amat y de Cortada y Sentjust (in Catalan: Rafael d’Amat i de Cortada i Sentjust), fifth Barón de Maldá. This diary has been studied by Josep Maria Mangado in his recent book La Guitarra en Cataluña (London, Tecla, 1998), from which I take these references, and I have discussed it in relation to Sor in my recent edition of More Seguidillas of Sor (Tecla, 1999). Baron de Maldá’s diary mentions Sor’s boleros, but on one occasion it tells of how Sor played the guitar alone. It is the earliest known reference to a solo performance by Sor. The occasion was May 7th 1802, in the Barcelona house of the Marqués de Castellbell:

…avisats item tots los senyors tertulians de berenars a fora, en est deliciós temps de primavera i de pau general…, per gaudir i menjar d’un com espècie d’ambigú en taula parada, és dir, berenar i sopar a un temps… a tres quarts de nou de la nit, que la prima fóra acabada en quant a la del gust de la llengua, i entraria l’harmonia als oídos, ab aquella com lira d’Orfeu, los dolços arpegios de la guitarra de Fernando Sors. Est encara no hi era, sí que tot just aquella “amada” i alegre societat, ab lo doctor Anton Bardolet unit a tot aquell bell aparato limpio de la taula, en l’estrado immediat al saló …, ab tot que m’istaren que els fes companyia; mes, com nondum venit hora mea, no en vaig tastar res, esperant a sopar a ma hora en casa, ab tot que molt alló temptava a hom a provar, especialment d’aquells gelats…

Luego de ser fora lo tal ambigú, com escena de teatro, que canvien los bastidors, nos en passàrem tots a l’estrado, después del de l’alcova principal de casa Castellbell, i, enrotllant tots al Fernando Sors, sentats en cadires, oírem sa guitarra—luego de ben trempada—que ens tocà un de sos entusiasmes de música, ab aquella dolçura i destresa de dits, que ens paragué oir un pianofort, en la varietat de punts, ja desmaiats i ja forts, ab certes carretilles que feia, sens discrepar en punt algun de sa ben trempada guitarra en la tocata que ens tocà al principi, ab moltes varietats i estranyeses de música; de seguida cantà algun bolero—que d’estos ne puc ser catedràtic.

(“The company was invited to supper outdoors, in this delicious springtime and peacefulness, to enjoy some refreshment, that is, to have supper and dinner at the same time.. . At a quarter to ten in the evening, when the first part was finished as far as the tastebuds were concerned, harmony could enter into our ears, as with the lyre of Orpheus, with the sweet arpeggios of the guitar of Fernando Sors. He had not yet arrived, and all that agreeable and happy company, with Doctor Anton Bardolet in charge of all the fine equipment of the table in the drawing room next to the salon … asked me to join them; but as nondum venit hora mea [my hour is not yet come], I didn’t want to eat anything as I expected to dine at home, but I was much tempted to taste, especially the ice creams…

When the refreshment was over, we changed the scene as in a play, and we all went to the drawing room beyond the main room of the Castellbell house, and all gathering round Fernando Sors, seated in chairs, we listened to his guitar, after he had well tuned it, on which he played one of his inspired pieces of music, with such sweetness and dexterity of the fingers that it seemed to us that we were listening to a pianoforte in the variety of expression, sometimes soft and sometimes loud, with certain scales that he performed, never missing one note on his well-tuned guitar in the toccata which he played to us first, with many variations and musical modulations; then he sang a bolero or two, in which he is the champion.”)

What was the “toccata” which Sor played on May 7th 1802? The description certainly could well fit the work which we now know as his op. 14.

In this edition I give both Castro’s and Meissonnier’s versions. The two are very different. Castro’s version dates from when Sor was still in Spain, and its idiom speaks to us in many details of that earlier time when later musical and notational conventions did not apply there. Players of Castro’s version can use this edition, and they may also find it of interest to consult the facsimile in volume 2 of the Tecla facsimile edition. Meissonnier’s version has many changes of detail which there is no reason to suppose were necessarily authorised by Sor.

Another edition of this work published by Porro in Paris within the period about 1807-1815 is mentioned by Mario Torta in his article “Le sonate per chitarra di Fernando Sor”, Guitart, January/March 2001. However, as Mr. Torta has unfortunately not given the location of a copy, it has not been possible to see a copy in order to take account of it in this edition. Although one cannot be sure without examining it, it appears from Mr. Torta’s description that this Porro edition may well date from after Castro and possibly derive from Castro’s text.

Opus 14, Castro’s version:

Bar 43, in the first and second groups, the original has an extra B an octave above the bass B, here removed.

58 in the original the bass notes are E, which are hard to play. Meissonnier’s version puts the first E up an octave and omits the second. 139-140 the natural signs are following Meissonnier.

Opus 14, Meissonnier’s version:

26 The unusual notation, unchanged here, perhaps means that the first upper D is played open on the fourth string, and the next D and C sharp on the fifth string.

OPUS 15(a). FOLIES D’ESPAGNE AND MINUET

One of Sor’s most delightful early works, first published in Paris by Meissonnier in about 1822-25 with the title Les Folies d’Espagne, variées, et un Menuet.

This is one of my favourite pieces by Sor. A theme and four variations in E minor, it is followed in the first editions by a minuet which may or may not itself be of an earlier date. It shows a natural musical gift, lyricism, and strength: surely one of the most attractive guitar pieces of this period.

In the minuet, the D.C. al Fine is editorial.

__________

OPUS 15(b). SONATA

The sonata in C now known as op. 15 again consists of a single movement, an allegretto. It is an uncompromising work, developing its ideas to the full and concentrating on musical values rather than on what the guitar can easily do. As a one-movement sonata, it again has affinities with the keyboard sonatas of Domenico Scarlatti.

Like the first sonata [opus 14], it was published by Castro in 1810 or shortly before and therefore dates from Sor’s Spanish period. Castro gave it no opus number, and called it “Sonata Seconda”, having called the first sonata [op. 14] “Sonata Prima”. It was also later published by Meissonnier in about 1817-22 with the opus number 15.

The edition presented in this present volume of the New Complete Works of Sor is a new text, in which some textual problems of transmission have been studied and dealt with. Here are some details, for those interested, about the way in which this present edition has been established.

An examination of the structure of this sonata in its earliest editions shows that the earliest known source, which was published by Castro, already contained what appear to be corruptions in textual transmission, and that the later Meissonnier edition, which has served for most performances and modern editions of this work for the past nearly two hundred years, contains the same corruptions and adds yet more of them. It is hard to believe that the composer had any part in putting out either edition, and especially hard to believe that the Meissonnier one, with its apparent extra corruptions, has any special claim to validity, so in this present edition I have decided to use only the earliest known source, namely the one published by Castro, and not to use the Meissonnier one. Anyone who wishes to consult the Meissonnier version may do so in the Tecla facsimile edition which is available.

The problem in the Castro version is that certain sections have only seven bars where one would expect eight, or six bars where one might expect eight. Perhaps certain bars which were present in Castro’s source (or at an earlier stage) were simply omitted by a copyist. However that may be, in this edition I have added some bars in order to make a version in which those apparently missing parts of the structure are restored. The new edition presented here is a strict edition of the Castro version, to which I have added a total of five hypothetically reconstructed bars. Of course anyone who doesn’t like them can omit them. They are those bars which are numbered 69, 86, 87, 167, and 168 in this edition. For those interested, here is an account of what has been done.

In the Castro version the piece begins in normal eight-bar phrases. But then on page 9 of the Castro version (see the Tecla facsimile edition), the section which begins at line 3 last bar and ends at line 5 bar 2 has only seven bars where one would expect eight. This is so unlikely that I have added a hypothetically reconstructed bar (bar 69 in this edition). Anyone who wishes to do so can simply omit it.

A little further down, the section beginning on page 9 line 7 bar 1 has only six bars where one would expect eight. Here I have added two bars (bars 86 and 87 in this edition), which are simply a repeat of bars 84 and 85, in order to make an eight-bar section. The player can omit them if he or she wants.

On page 10 of the Castro edition, in line 6 last bar to line 7 bar 3 there are only four bars where one might expect eight. I have made no changes, but the player who wishes to can simply repeat these four bars (bars 121 to 124 in this edition) to give an eight-bar section.

On page 11 of the Castro edition, the section from line 5 bar 1 to line 6 bar 1 has only six bars where one would expect eight. Here I have added two bars (bars 167 and 168 in this edition), which are simply a repeat of bars 165 and 166, in order to make an eight-bar section. The player can omit them if he or she wants.

At the very end of the piece, the last section (from bar 181 on) has ten bars where perhaps there might originally have been twelve. I have not made any changes here, but the adventurous player could repeat bars 188 to 189.

These features, especially the section which has seven bars, seem to indicate textual corruption either made by Castro’s engraver or else already existing in the manuscript which served as his source.

If one then compares the later Meissonnier edition, one finds all these problems, but in addition there are also yet more problems which are not in Castro. Thus, the Meissonnier edition omits the bars numbered 73 to 76 in this edition [Castro page 9, lines 5-6], which cannot be right. It is hard to imagine that such a change was made by Sor or with his approval. Also, Meissonnier omits four more bars (52-55 in this edition) which are needed to make an eight-bar phrase.

Ever since Sor’s day the Meissonnier version has been the version most used for the performance of this famous piece.

The piece is also found, entitled “Sonata di Sor”, in another publication called Collection de morceaux choisis des meilleurs auteurs espagnols et italiens pour guitare ou lyre, rédigée par un Espagnol, [1er] Cahier, Paris, no name of publisher or date (c.1815?). But this shares the faults of Meissonnier and indeed may have served as Meissonnier’s source.

* * * * *

There is a difference in tempo indication: Castro has Allegretto, but Meissonnier changes it to Allegro moderato.

The double stems on the first notes probably indicate that the notes are to be played on two strings and that the lower string is struck with the thumb.

The diagonal lines through the chords at bars 1, 9, etc., perhaps indicate that those chords are to be played rasgueado.

The repeat sign at bar 88 is from the original edition. The original edition also has another repeat sign at the beginning of bar 89, but this appears to be a mistake and has here been removed.

__________

OPUS 15(c). THÈME VARIÉ

This uses the same theme as the first set of variations in op. 11. First published in Paris in about 1822.

__________

AIR VARIÉ

An otherwise unknown work published only by Castro in 1810 or shortly before.

In bar 105, the first note of the last group was G in the original, here replaced by A which seems more likely.

__________

THEMA VARIÉ

This set of variations, published by Castro in 1810 or shortly before, uses the same theme as op. 20.

__________

FOUR MINUETS Published by Castro in 1810 or shortly before. The second minuet is the same as op. 11, no 6. These four delightful minuets are indubitably from Sor’s Spanish period.

No. 1: at bars 9 and 11, the notation is probably conventional, with triplets intended to be played in both voices.

__________

MARCH FROM “CENDRILLON” Cendrillon was a celebrated ballet with music by Sor, first performed in London in 1822, then in Paris and Moscow. Mercandotti, Noblet, and Félicité Hullin all danced the title role. This arrangement of the March from it, for guitar, was published in Paris in about 1823-25.

In some modern editions, this work is called “op. 15 (c)”, but there is no known documentary authority for this. It does, however, date from the same period of publication as opp. 15 (a), (b), and (c), and for that reason has been placed here in this present volume.

51, second chord: the B in this chord is editorial. The original had D sharp, which cannot be correct.

__________

LA CANDEUR

An eloquent work, published in the Encyclopédie Pittoresque de la Musique (Paris, 1835).

__________

OPUS 16. FANTASIA ON PAISIELLO’S “NEL COR PIÙ NON MI SENTO”

An extended set of variations using many of the different techniques possible on the guitar. First published in Paris by Meissonnier in 1823 with the title Cinquième Fantaisie pour la Guitare avec des Variations sur l’Air de Paisiello Nel cor più non mi sento.

Variation 4, bars 90 and 102: the C in each of the last chords is an artificial harmonic. Sor writes in his method about the harmonics in this variation as follows:

Quelquefois je profite aussi du procédé du violon dont j’ai parlé, mais très rarement, et avec de grandes précautions. Dans une de mes variations sur le thème de Paisiello, Nel cor più non mi sento, je termine les deux reprises par les deux sixtes en sons harmoniques, exemple soixante-dix-neuvième; on voit que, pour la dernière, je fixe le premier doigt sur la cinquième corde à la troisième touche pour bien déterminer l’ut; la touche devient le sillet, et si, à cinq touches, je devais produire la double octave de la corde à vide la, je produirais, à l’octave (devenue quinte), la double octave d’ut; je produis avec le second doigt la double octave de mi, sur la cinquième touche avec la sixième corde, et les deux notes réussissent très distinctement; mais aussi cette variation est plus lente que les autres. Je ne dis pas qu’avec beaucoup d’exercice on ne puisse parvenir, par les moyens indiqués, à jouer en sons harmoniques avec plus de vitesse que moi; en ce cas, je me féliciterai toujours d’être surpassé, car ce sera une preuve de l’utilité de mes recherches.

(“Sometimes I take advantage likewise of the violin-method already mentioned, but very rarely, and with great caution. In one of my variations on Paisiello’s theme, Nel cor più non mi sento (Hope told a flattering tale), I terminate both strains by two sixths in harmonic sounds, example 79 [in the illustrations to the method]. It is seen that, for the latter, I fix the first finger on the fifth string, at the third fret, to determine C properly: this fret becomes the nut, and if, at five frets’ distance I should produce the double octave above the open string A, I should produce at the octave (become fifth), the double octave of C. I produce the double octave of E with the second finger, on the fifth fret with the sixth string, and both notes sound very distinctly: this variation, however, is slower than the others. I do not say that, with a great deal of practice, success might not attend the methods mentioned for playing harmonics, with greater rapidity than I can. In this case, I shall always rejoice at being surpassed, for it will be a proof of the utility of my researches.”)

(Originally in French in Sor’s Méthode pour la Guitare, Paris, 1830. English translation by Arnold Merrick published as Method for the Spanish Guitar in London in 1832, available in reprint form from Tecla.)

In other words, if I interpret Sor correctly, stop the fifth string at the third fret, then play harmonics on both the fifth and sixth strings at the same time, touching the fifth string at the eighth fret and the sixth string at the fifth fret. At bar 84 stop both strings at the third fret and touch both strings at the eighth fret.

In variation 7 the stems pointing down in the first bar indicate that those notes are to be played with the right hand thumb (and by inference, all the similar notes in this variation).

A charming reminiscence of a performance of this work may be found in the catalogue of the Tiscornia collection of guitar music, Catálogo Breve, which was published in Buenos Aires in 1948. In the preface of that 1948 book, Jorge M. Furt remembers how his late friend Eleuterio F. Tiscornia used to play this piece on his Torres guitar:

Me detengo ante este papel blanco y me lo veo, cerrando los ojos, en una de nuestras tantas noches, diciendo en su Torres ilustre, sonora como un milagro de bajos cálidos, de clarísimos altos, la música de Coste o de Sor en una lección expresiva que le escucho patente, más allá de su muerte y de mi vida, como pura insuperada intención armónica. Todo oídos en este silencio puedo seguir, como seguía el muchacho lejano, el fraseo limpio y preciso de la Fantasía de Sor sobre tema de Paisiello, todas sus variaciones, solas el respiro de las pausas. (Una tarde, bajando al volver de Fiesole, me entré en la Villa Albrizzi y le corté una ramita de uno de esos viejos laureles junto al camino que Paisiello recorría para su vivienda, y se la mandé a Buenos Aires en memoria de una noche en que, disuelto el último compás de la Coda, no supimos ni él ni yo por unos minutos atinar palabra, quebrar el silencio, en el cuarto lleno de libros donde duraba, suspensa, la melodía: ‘nel cor più non mi sento…’)

(“I sit before this white paper, closing my eyes, and I see him on one of so many of our nights, interpreting the music of Coste or of Sor on his fine Torres, miraculously sonorous in its warm basses, its very clear trebles, in an expressive reading which I can clearly listen to beyond his death and even beyond my own life, as though for a pure and insuperable harmonic intention. All ears in this silence, I can still follow, as that far-off lad used to follow, the clear and precise phrasing of the Fantasia of Sor on a theme of Paisiello, all its variations, and apart, the breath of its silences. (One evening, as I went down coming back from Fiesole, I went into the Villa Albrizzi and I cut for him a small branch of one of those ancient laurel trees by the side of the road which Paisiello used to walk along on the way to his house, and I sent it to him in Buenos Aires in memory of a night in which, the last bar of the Coda having died away, for several minutes neither he nor I knew how to find a word to say, to break the silence, in the room full of books in which there remained, suspended, the melody: ‘nel cor più non mi sento…’).”)