Willmott and her Library

Ellen Willmott was born in 1858 to Frederick Willmott (1825–92), a solicitor, and Ellen Fell (d. 1898), a scion of the wealthy Tasker family. In 1875 the Willmott family moved from London to Warley Place, Essex, a large estate which comprised a house with two lodges, set in fifty acres of parkland (see Figure 1); the family also had a chateau at Tresserve, Aix-les-Bains, France, acquired in 1890.Footnote ¹⁷ Frederick Willmott died in 1892; Ellen inherited Warley, along with her sister Rose, who had married Robert Valentine Berkeley in 1891 and who moved to her husband’s home, Spetchley Park in Worcestershire, in 1897. Their mother died in 1898, leaving effects worth just over £134,000 — an extremely large sum, roughly the equivalent of £15 million today — the sole beneficiaries of which were Ellen and Rose.Footnote ¹⁸ With her mother dead and her sister having left home, Ellen was left at Warley alone, free to indulge in her apparent passion for purchasing books, both musical and botanical.

Figure 1. Warley Place, Essex, photographed by Ellen Willmott, date unknown. From a private collection; reproduced by permission.

Willmott’s earliest biographer, Audrey le Lièvre, was the first to point out that she was well connected musically: she was friends with Sir Frederick Bridge (1844–1924), organist of Westminster Abbey and a professor at the Royal College of Music and the University of London, Sir Charles Villiers Stanford (1852–1924), Professor of Music at the University of Cambridge, and Sir Walter Parratt (1841–1924), Heather Professor of Music at Oxford, Organist of St George’s, Windsor, and Master of the Queen’s Music — who le Lièvre reckoned ‘must have opened many doors’ for her.Footnote ¹⁹ But her papers shed light on an even larger cast of musicians and musicologists with whom Willmott corresponded, including Sir Hugh Allen (1869–1946), principal of the Royal College of Music, Sir John Stainer (1840–1901), Professor of Music at Oxford University, and, crucially, William Barclay Squire (1855–1927), who was for much of his life involved with the music department of the British Museum.

Willmott was also a capable musician herself. She was a lutenist, an ‘accomplished violinist’, and, according to one commentator, ‘an expert musician, playing the violin, viola, cello, organ and piano’.Footnote ²⁰ Her collection contained some twenty-eight instruments (most of them plectral or wind), some of which were valuable, including several Amati violins, and a harpsichord given to her by her father around 1870.Footnote ²¹ Willmott’s papers show that she regularly attended concerts and held musical evenings at home.Footnote ²² She was a member of the Folk-Song Society and the International Musical Society,Footnote ²³ and she also sang: she attended weekly rehearsals of the London Bach Choir and was ‘devoted to madrigal singing’.Footnote ²⁴ This would explain her interest in purchasing madrigal publications of the late sixteenth and early seventeenth centuries, the details of which are set out below.

By the time Willmott’s collection was auctioned off in 1935, it purportedly ran only to 575 items, divided in the auction catalogue into ‘The General Library’ (lots 1–117), ‘The Botanical Library’ (lots 118–381), and ‘The Musical Library’ (lots 382–575).Footnote ²⁵ With a library containing fewer than six hundred items, and a music section of fewer than two hundred, Willmott appears to have been the custodian of a carefully assembled bibliothèque choisie rather than an inveterate accumulator: her library was modest by the standards of some contemporary collectors whose hoards ran into the thousands.Footnote ²⁶ The 1935 Willmott auction catalogue makes clear on its title page that it was offering for sale items ‘selected from the library at Warley Place’, however; therefore it did not fully represent the true extent of her collection.Footnote ²⁷ Unfortunately no complete inventory for Willmott’s library survives, published or otherwise; nor are items she owned generally easy to track down because of her tendency not to mark her volumes in any way — Willmott’s books and manuscripts are not usually identifiable from any inscribed name or acquisition note; nor did she use bookplates.Footnote ²⁸ Consequently her items are seldom to be traced in subsequent sale catalogues, with only a few notable exceptions (mentioned below). A sense of the true extent of Willmott’s library may nevertheless be gleaned from the numerous invoices for books and manuscripts that survive in her papers, not all of which appeared listed in the 1935 auction, and also from an interim list of her books that was drawn up in a bid to tout certain items to London booksellers when Willmott hit financial hardship in the 1920s.

The majority of music collectors active in the late nineteenth and early twentieth centuries were regular attenders at book auctions — most of which were held in London, where the principal auctioneers and book dealers had their premises. Auctioneers’ copies of sale catalogues, marked up with the names of winning bidders and the prices they paid, reveal who bought what and for how much; in the case of music sales, these catalogues shed light on a relatively small but hardcore cast of purchasers who attended auctions regularly.Footnote ²⁹ Ouseley (who purchased ‘one rare item after another’), Marshall, and Cummings are names which crop up again and again in these annotated auction catalogues, all of whom were regular bidders at auctions;Footnote ³⁰ Willmott’s name, though, has not yet been found as a winning bidder in a single book sale held between 1900 and 1934, when, to judge from her papers, she was at her most active as a collector. She seems not to have attended any auctions herself, even though Warley Place is only fifteen miles from central London. Instead, Willmott’s modus operandi appear to have been those of an armchair purchaser: she perused sale catalogues, buying items at fixed advertised prices from specific dealers, or she had an agent bid on her behalf in the saleroom at auctions.Footnote ³¹ (Most of Willmott’s invoices record purchases from numbered catalogues, with fixed prices, rather than lots acquired at named auctions.)

Willmott’s various purchases are documented by her numerous invoices: these show that she had dealings with several booksellers, both at home and abroad. More than three hundred invoices survive from London-based firms, but there are also bills from dealers in Stratford-upon-Avon, Bath, and Derby, and some thirty invoices from booksellers in France, Switzerland, and Germany.Footnote ³² Most of the booksellers with whom Willmott had dealings appear to have supplied books of a general nature or on botanical subjects. Only a few (Frost, Lemencier, Rosenthal, and Tregaskis) supplied books on music; scores, recently published rather than antiquarian, were supplied chiefly by Augner, Breitkopf & Härtel, and Novello & Co. For a handful of dealers Willmott was a regular customer: nineteen invoices survive from Cumin & Masson of Lyon (1904–11); sixteen from Gregory George of Bath (1903–12); sixteen from Novello & Co., London (1908–22); and no fewer than sixty-six from the London firm of Wesley & Sons (1903–18).Footnote ³³ But it is the London bookseller Bernard Quaritch who supplied Willmott with the most items, including music. More than two hundred invoices and related correspondence survive in her papers from that firm, ranging from September 1899 to July 1912, shedding light on a thirteen-year relationship. Although only a small portion of these documents pertain to music purchases, they nevertheless bring into focus Willmott’s apparent musical interests, documenting her penchant for early modern music, especially madrigals.

Early Acquisitions, 1890–1904

Exactly when Willmott began purchasing books is uncertain. Audrey le Lièvre reckoned that the foundations of Willmott’s ‘magnificent library of rare books at Warley Place’ were laid in the 1870s, when she received for her sixteenth birthday a harpsichord which had purportedly belonged to Princess Amelia.Footnote ³⁴ Relatively few invoices survive from the end of the nineteenth century to prove this position (the earliest extant Willmott book invoice appears to be from Librairie Jullien of Geneva, dated 1897), although a letter to Willmott from the booksellers Ellis & Elvey, dated 19 April 1890, refers first to the cleaning and rebinding of a volume described as ‘Simpson’, and then to two volumes of music by the German composer Giovanni Girolamo Kapsperger (1580–1651), which Willmott had left with them ‘for enquiry’; the letter also tells us that she owned Book IV of his Villanelle and Book I of his Motetti. It has not yet been possible to find evidence for Willmott’s acquisition of these volumes, and worse is that some of her bills were recently disposed of because they were in very poor condition.Footnote ³⁵ If le Lièvre was able to substantiate book acquisitions in the 1870s and 1880s, then she presumably had access to documents which are today lost or inaccessible.Footnote ³⁶

This incomplete state of Willmott’s papers is illustrated by her various invoices from Bernard Quaritch. The earliest Quaritch invoice is dated 1 September 1899, for a single item: A Dictionary of English Booksellers, at a price of 1s. 8d. Next come nine further invoices, dated as far as 27 March 1902, for a total of £81 1s. 2d. Yet a statement issued to Willmott by Quaritch on 31 March 1902 records a balance of £861 10s. 9d. already owing on 31 December 1901 — about £91,000 today, even though no invoices can be found to explain how this balance was accumulated. The Willmott papers are therefore incomplete insofar as Quaritch invoices are concerned; consequently they do not fully represent the extent of her purchases. The same appears to be true of the paper trail for other sellers: if by le Lièvre’s reckoning Willmott’s collecting really did begin on or just after her sixteenth birthday in 1874, then some twenty-five years’ worth of acquisitions are today undocumented. The lion’s share of the invoices, insofar as they can be located, range in date roughly from 1897 to 1931, suggesting that Willmott’s purchases began about the time her sister Rose left home — perhaps boosted by the inheritance of her mother’s estate, which had come to Ellen and Rose in 1898.

The Julian Marshall Sale (1904)

Although she purchased some Amati stringed instruments of historical importance at the end of the nineteenth century,Footnote ³⁷ Willmott’s first major acquisitions of early modern music appear to have come in 1904, when she purchased fifteen lots from the third auction (of four) that dispersed the books and manuscripts of Julian Marshall (1836–1903), ‘the greatest of all British collectors’, whose library contained a ‘formidable array of early printed English music’.Footnote ³⁸ This third Marshall sale ran from 11 to 12 July 1904; it was handled by Sotheby, Wilkinson and Hodge.Footnote ³⁹ An invoice from Quaritch to Willmott dated 13 July that year lists her various purchases from that sale in sequence: except for lots 466, 580, and 589, all of her purchases were music (see Figure 2). Lots 1, 2, 373, 483, 484, and 495 were music volumes of the seventeenth century; lot 524 was a collection of songs published in about 1725; lots 22, 30, 376, 444, 445, 601, and 611 were madrigals published in the late sixteenth or early seventeenth century by composers including Thomas Bateson, John Bennett, George Kirbye, Thomas Morley, Thomas Watson, and John Wilbye. (See Table 1.)

Figure 2. Quaritch’s invoice to Ellen Willmott, 13 July 1904, following the Julian Marshall sale.

Table 1. Madrigal publications purchased by Willmott at the 1904 sale of Julian Marshall

Lot	Composer	Title
22	Thomas Bateson	The First Set of English Madrigales to 3, 4, 5 and 6 Voices (London, 1604), four volumes only (the Cantus, Altus, Bassus, and Sextus parts) and The Second Set of Madrigales (London, 1618), two volumes only (the Cantus and Bassus parts)
30	John Bennett	Madrigalls to Foure Voyces (London, 1599), three volumes only (Altus, Tenor, and Bassus); the Bassus was a MS copy in the hand of Dr Burney
376	George Kirbye	First Set of English Madrigalls to 4, 5 & 6 Voyces (London, 1597), five volumes only (Cantus primus, Cantus secondus, Tenor, and Bassus in one volume, Sextus loose and uncut)
444	Thomas Morley	The First Booke of Balletts to Five Voyces (London, 1595) five volumes complete, unbound, ‘2 leaves imperfect’
445	Thomas Morley	The First Booke of Balletts to Five Voyces (London, 1595) four volumes only (Cantus, Altus, Quintus, Bassus), unbound
484	John Playford	Select Ayres and Dialogues for One, Two, and Three Voyces (London, 1659), three volumes complete, unbound
487	John Playford	Choice Ayres, Songs, and Dialogues (London, 1675–84), five volumes complete, variously bound/unbound
495	Henry Purcell	Sonnatas of III Parts: Two Viollins and Basse: to the Organ or Harpsecord (London, 1683)
601	Thomas Watson	The First Sett of Italian Madrigalls Englished (London, 1590), Sextus and Bassus parts only, in one volume
611	John Wilbye	The First Set of English Madrigals to 3. 4. 5 and 6 Voices, Newly Composed (London, 1598), complete, and The Second Set of Madrigals to 3. 4. 5. and 6. Parts (London, 1609), five volumes of six, lacking Sextus part

This invoice is important for several reasons. First, it records the earliest of Willmott’s purchases that testify to her apparent affinity for early modern music — particularly madrigals. Second, it offers clear evidence that Quaritch was bidding at the auction as Willmott’s agent: Willmott had evidently selected items she wanted from the Marshall catalogue, so that Quaritch could bid at the sale on her behalf (a letter from Quaritch to Willmott dated 11 July 1904 confirms that he was ‘returning your marked catalogue which was given me by Messrs. Sotheby’ at the Marshall sale). Third, this invoice suggests that Willmott may have been interested in only the best available copies of early modern printed music. For instance, lots 444 and 445 appear combined in the invoice for a total price of £52 10s.; both are referred to simply as ‘Morley’, and are described as ‘the best copy made up from two lots’. Recourse to Marshall’s 1904 catalogue reveals that these were volumes belonging to Morley’s First Booke of Balletts to Five Voyces. The first (lot 444), although complete with all five partbooks present, described as a ‘fine copy’, had ‘2 leaves imperfect’; the second (lot 445), although it was similarly called a ‘fine copy’, lacked its Tenor partbook (see Figure 3).Footnote ⁴⁰ It would therefore seem that Quaritch combined them to make ‘the best copy made up from two lots’, as indicated in his invoice.

Figure 3. Lots 444 and 445 in the 1904 sale of Julian Marshall (annotated copy in the British Library, London, S. C. Sotheby 1268). Reproduced by permission.

Also telling is lot no. 183 in the Marshall invoice, a collection of autograph letters from the composer, dramatist, and novelist Charles Dibdin (1745–1814) and his contemporaries. This lot appears on the invoice without a price. The annotated catalogue for Marshall’s 1904 sale reveals that it went not to Quaritch but to the collector William Thomas Freemantle (1849–1931), for a price of £30.Footnote ⁴¹ Willmott, having marked up her copy of the Marshall auction catalogue, presumably identified not only lots of interest, but also maximum prices that she was willing to pay, including for the Dibdin letters (she already owned at least one letter in his hand).Footnote ⁴² It would seem that she was unwilling to pay as much as £30 for them, however, and that her maximum price, communicated to Quaritch in advance, was exceeded by Freemantle. This suggests that Willmott’s spending was not yet entirely unbridled.

Almost all of the music volumes purchased by Willmott via Quaritch at the Marshall sale were printed partbooks — a format in which each voice part is presented in its own volume, belonging to a set, rather than in score (partbook sets are therefore vulnerable to loss of one or more of their constituent volumes, which can render them useless for performance purposes). Scrutiny of the 1904 Marshall catalogue reveals that not all of the partbook sets Willmott purchased were complete. Lot 22 comprised two partbook sets: Bateson’s First Set of Madrigales (1604) and the same composer’s Second Set of Madrigales (1618). Each set ought to have had six separate volumes (Cantus, Altus, Tenor, Bassus, Quintus, Sextus), but the former offered only the Cantus, Altus, Bassus, and Sextus books — four volumes out of six; the latter offered only the Cantus and Bassus books — two out of six. Similarly, lot 445, Morley’s First Booke of Balletts to Five Voyces (1595), lacked its Tenor partbook, giving only four volumes out of five. Willmott’s madrigal purchases from the Marshall sale of 1904 are listed in Table 1.

Other items won at the same sale were acquired complete but were impractically bound. Lot 376, Kirbye’s First Set of English Madrigalls to 4, 5 & 6 Voyces (1597), and lot 601, Watson’s First Sett of Italian Madrigalls Englished (1590), were purchased as complete sets, so had all of their respective voice parts, but the partbooks of both publications were bound together in a single volume (see Figure 4). This would have required all four, five, or even six musicians to sing from different pages of a single composite volume simultaneously, making performance extremely difficult if not impossible. This suggests that Willmott was acquiring music for antiquarian rather than practical purposes — a point which receives further discussion below.

Figure 4. Impractically bound items (lots 376 and 601) purchased for Willmott at Julian Marshall’s 1904 sale (annotated copy in the British Library, London, S. C. Sotheby 1268). Reproduced by permission.

Quaritch’s invoice to Willmott for purchases made on her behalf at the 1904 Marshall sale (see Figure 2) is the only record we have to show that material from his collection entered hers. Julian Marshall’s bookplate has not yet been found in any material subsequently purchased by Willmott, even though it is ‘easy enough to encounter’ in other items he once owned (Marshall himself requested that his bookplates be retained by later owners).Footnote ⁴³ Since his bookplates do not appear in any of Willmott’s copies, it would seem that she had them removed in order to provide her own. Yet although Willmott was sent sketches and samples for the commission of a bookplate from the booksellers John and Edward Bumpus in 1907, a design seems not to have been settled upon. An invoice to Willmott from the Rembrandt Gallery, moreover, also dated 1907, lists the purchase of bookplates, but they do not appear to have been used. This makes tracing items from her collection difficult.

The Worshipful Company of Musicians Loan Exhibition (1904)

Willmott’s purchases from the 1904 Julian Marshall sale coincided with an exhibition of ‘beautiful old instruments, rare books, fine pictures, interesting manuscripts, chronological examples of music typography, and mementoes of musicians’, held at Fishmongers’ Hall, London, from 28 June to 16 July 1904 (see Figure 5).Footnote ⁴⁴ It was hosted by the Worshipful Company of Musicians to celebrate the three-hundredth anniversary of their royal charter. Although evidence of her acquisitions for the later years of the nineteenth century is thin on the ground, Willmott evidently had an established music collection in place by 1904, since, as well as lending twenty-eight instruments to the exhibition, she also lent several music manuscripts — including items in the handwriting of J. S. Bach, Beethoven, Brahms, Mendelssohn, Mozart, and Schubert.Footnote ⁴⁵ She also lent some autograph letters from composers including Crotch, Dibdin, Smart, and Walmsley.Footnote ⁴⁶ (Krummel pointed out that Willmott lent more music manuscripts than anybody else (twenty-one items), followed by Felix Moscheles (thirteen items)).Footnote ⁴⁷ By June 1904, then, Willmott had assembled a musical collection that included instruments, prints, manuscripts, and autograph letters, even though only a handful of music purchases are documented by her invoices by this point. Indeed, the fact that her music and instruments were included in the exhibition at all suggests that she was already known as a collector of some importance: the curators included material in the exhibition belonging not only to members of the Worshipful Company, but also to ‘other notable collectors’.Footnote ⁴⁸

Figure 5. Photograph of the Worshipful Company of Musicians Loan Exhibition (1904), Fishmongers’ Hall, London. Reproduced by permission of the Worshipful Company of Musicians.

Significantly, the Loan Exhibition of 1904 featured some forty-four early modern music items, most of which were printed partbooks containing madrigals. These were lent not by Willmott but by A. H. Littleton (1845–1914), W. H. Cummings (1831–1915), T. W. Taphouse (1838–1905), Sir August Manns (1825–1907), and G. E. P. Arkwright (1864–1944), among others.Footnote ⁴⁹ According to her documentable purchases, Willmott’s acquisition of early printed music did not begin until Marshall’s sale of July 1904, while the Worshipful Company’s Loan Exhibition was still in progress. Perhaps it was sight of the early printed material at this exhibition which induced Willmott’s future purchases, documented by the several items she bought from the third Marshall sale.

Further Purchases, 1905–08

Willmott’s next major haul of early modern music came in 1905, when she purchased numerous items from Quaritch’s catalogue no. 237, issued in February that year. This included a collection of ‘Elizabethan Madrigals and Part-Songs’, all of them by early modern composers, running to thirty-three items.Footnote ⁵⁰ An invoice from Quaritch to Willmott dated 6 March 1905 records that she purchased twenty-seven of the thirty-three madrigal prints listed, at a total cost of £559 17s. — roughly the equivalent to £63,000 today (see Figure 6).

Figure 6. Invoice from Bernard Quaritch to Ellen Willmott, dated 6 March 1905, recording various purchases from catalogue 237.

A few of the purchases from this catalogue were complete partbook sets. Item 2 was all five volumes of Byrd’s Psalmes, Sonets and Songs (1588), for £30; item 14 was all five volumes of Gibbons’s First Set of Madrigals and Mottets (1612), for £84; item 28 was all six volumes of Watson’s First Sett of Italian Madrigalls Englished (1590), for £72. Willmott’s other purchases, however, were incomplete sets, some with only two out of an original five or six volumes present. Item 1, for instance, was four volumes of six (Cantus, Altus, Bassus, Sextus) to Bateson’s Second Set of Madrigales (1618), bound together in a single volume, for £16; item 8 was five volumes of six (Cantus, Altus, Tenor, Bassus, Quintus) to Croce’s Musica sacra: To Sixe Voyces (1608), for £16; item 9 was two books of five (Cantus and Bassus) to Michael East’s Second Set of Madrigales to 3. 4. and 5. Parts (1606), for £10. Willmott also purchased several incomplete Byrd publications: item 3 was two volumes of six (Contratenor and Bassus) to his 1589 Songs of Sundrie Natures, for £10 10s.; item 4 was two volumes of six (Bassus and Sextus) to the 1610 edition of Songs of Sundrie Natures, also for £10 10s.; item 5 was four volumes of five (Superius, Medius, Contratenor, Tenor) to his Cantiones sacrae (1589), for £16; item 6 was two volumes of five from the second set of Cantiones sacrae (1591), for £6 10s.

In purchasing incomplete sets of partbooks, Willmott was presumably seeking to make good deficiencies in sets she already owned. This appears to be the case with item 1, the four volumes of six (Cantus, Altus, Bassus, Sextus) to Bateson’s Second Set of Madrigales (1618), bound together in a single volume and lacking its Tenor and Quintus books. Willmott had been able to purchase only the Cantus and Bassus parts to this publication at the Marshall sale of 1904;Footnote ⁵¹ her purchase from Quaritch’s catalogue 237 would therefore have given her the Altus and Sextus parts that were wanting, but duplicates of the Cantus and Bassus parts. The four volumes listed in catalogue 237 were also accompanied by a ‘transcript of the Tenor and Quintus parts from a copy which belonged to Edward Taylor, Gresham Music Lecturer, 1837–1863’ — essentially offering her a complete if not entirely authentic set.Footnote ⁵²

The same appears to be the case with item 28 — all six volumes to Watson’s First Sett of Italian Madrigalls Englished (1590), sold unbound. Willmott had already purchased the Sextus and Bassus volumes of this publication at the Marshall sale, bound in a single volume;Footnote ⁵³ her purchase of the same set from Quaritch would have given her a complete copy, even though she would have ended up with some redundant volumes. Her Byrd purchases from catalogue 237 are harder to justify, since no previous Byrd acquisitions are recorded in Willmott’s papers (none were purchased for her at the Marshall sale, even though some Byrd lots were available).Footnote ⁵⁴ Willmott may already have acquired some incomplete Byrd sets by the time catalogue 237 was issued, the purchase of which is unrecorded, or she was speculatively buying incomplete sets in the hope that the absent partbooks would eventually turn up for sale. (If the latter, then she did not succeed: the same Byrd sets are listed in her posthumous auction of 1935 as incomplete.)Footnote ⁵⁵

Willmott does, at least, seem to have been purchasing sets of particularly good quality from Quaritch’s catalogue 237, even if some were lacking their constituent partbooks. Item 7, five volumes of six to Byrd’s Psalmes, Songs, and Sonnets (1611), was unbound, but ‘in lovely state as though just issued’; item 9, two volumes of five of East’s Second Set of Madrigales to 3. 4. and 5. Parts (1606) was again unbound, but ‘in very fine state’.Footnote ⁵⁶ Item 14, all five volumes of Gibbons’s First Set of Madrigals and Mottets (1612), had uncut pages, but was ‘in beautiful condition’;Footnote ⁵⁷ item 28, all six volumes of Watson’s First Sett of Italian Madrigalls Englished (1590), sold unbound, was ‘in sound state’.Footnote ⁵⁸ Willmott’s acquisition of these sets of volumes, whether complete or incomplete, suggests that she was seeking the best possible copies she could find. This notion is supported by the volumes she did not buy from the Quaritch catalogue: items 15, 17, 18, and 29. Item 15 was all three volumes (Cantus, Altus, and Bassus) to the first edition of Morley’s Canzonets to Two and Three Voices (1593), but some leaves were missing, and various margins were described as defective; Willmott went instead for item 16, which offered the same publication but only the Altus partbook (which had only a defective last leaf).Footnote ⁵⁹ Presumably Willmott had already acquired the Cantus and Bassus volumes to this publication, or she intended to complete the set at a later date (if she did complete the set, then it did not feature in her posthumous auction, which lists only the third edition of this publication, issued in 1606, bound into a single volume).Footnote ⁶⁰ Lots 17, 18, and 29 were not purchased presumably because they offered volumes that Willmott had already purchased at the 1904 Marshall sale.Footnote ⁶¹ This suggests that Willmott was careful not to purchase material she already owned, unless seeking to complete a set that was missing one or more of its volumes.

Further evidence of Willmott’s preference for high-quality copies is offered by a correspondence card that accompanies Quaritch’s invoice of 6 March 1905, bearing the same date (see Figure 7). It shows that Willmott had sent some material to Quaritch ‘for comparison’. The volumes were drawn from seven partbook publications; six were acquired at the 1904 Marshall sale (all the sets snapped up at that sale, excluding those of Bennett and Kirbye), along with a Bassus partbook to Byrd’s Songs of Sundrie Natures, the purchase of which is not documented in Willmott’s papers. If Willmott was sending material ‘for comparison’ to Quaritch in advance of ordering items, then it would seem she was seeking to find the best copies available — ones which would complement her existing copies and offer an impression of uniformity. This same list shows that Willmott had also sent him her set of Morley’s First Booke of Balletts to Five Voyces (1595), which had been purchased complete from Marshall (lot 444 in his 1904 sale), but which had ‘2 leaves imperfect’; presumably she was seeking a perfect copy.

Figure 7. Correspondence from Bernard Quaritch to Ellen Willmott, dated 6 March 1905, detailing her property that had previously been sent for comparison.

The Willmott Manuscript

Willmott’s next major music spree came in 1908, when she made several purchases from Quaritch’s A Catalogue of Books on Music Offered for Sale by Bernard Quaritch, no. 270, published in November that year.Footnote ⁶⁹ An invoice to Willmott from Quaritch dated 2 January 1909 records nine purchases from that catalogue, varied in nature, at a total cost of £198 17s. — about £20,500 today (see Figure 8). The most expensive purchase, item 253, was two sets of madrigals: a complete set of Yonge’s Musica transalpina (both the first and second sets, 1588 and 1597), which went to Willmott for £84 — about £8,750 by today’s standards. The partbooks were bound together as ‘2 books of six pts. each, in 2 vols’.Footnote ⁷⁰ Willmott had already purchased four volumes of the first six-part set (1588) from Quaritch’s catalogue 237, at £20 (item 31), and therefore ended up with some duplicates. It was perhaps purchases such as this which caused le Lièvre to claim that Willmott accrued her library with ‘care and greed’.Footnote ⁷¹

Figure 8. Invoice from Bernard Quaritch to Ellen Willmott, dated 2 January 1909, recording her various purchases from catalogue 270.

The second most expensive purchase, however, was item 261: an ‘Elizabethan Music Book, MS on paper’, dated 1591, which went to Willmott for £52 10s. — roughly £5,300 today. This is the volume known today as the Willmott manuscript, named after Willmott herself. It is a lone contratenor partbook from a set originally of five volumes, containing sacred Latin-texted polyphony; it was one of several Tudor partbooks owned by the Norwich merchant John Sadler (d. 1592).Footnote ⁷² Willmott’s acquisition of a single partbook can hardly have been done in the name of practical use: a successful performance from this book alone, with only a single voice part present, would have been impossible. For some works represented in the partbook Willmott owned a printed edition, which could have supplied the missing voice parts;Footnote ⁷³ but attempts to sing directly from the Willmott manuscript itself, even if other voice parts were available elsewhere, would have been thwarted by the fact that several music-bearing leaves in this volume are now, and presumably were then, missing — something not reported in the description of the item in the catalogue (see Figure 9).Footnote ⁷⁴ Even if it were complete, there is no evidence to show that Willmott was herself conversant with mensural notation, so as to sing from the volume directly. Although she was involved with various musical groups, madrigal societies and choirs in the latter half of the nineteenth and the first half of the twentieth centuries typically sang from transcriptions, in modern notation, rather than facsimiles of originals.Footnote ⁷⁵

Figure 9. The description for the ‘Elizabethan Music Book’ in Quaritch’s catalogue 270, November 1908, advertised for a price of £52 10s.

Since a lone partbook from a set of five volumes cannot have been purchased with utility in mind, this raises the question of Willmott’s purchase motivations. Possibly she was interested in the volume’s contents: music aside, the book contains several illustrations, including one of a triple Tudor rose, dated 1591; therefore it might have appealed both to her botanical as well as her musical interests. (According to the sale description, the book’s depiction of a triple Tudor rose makes it ‘almost certain that royal ownership can be claimed for this remarkable volume’, but this cannot be proved, and seems unlikely.) Alternatively, Willmott might have been banking on other volumes from the same set making it onto the market, even though only one further partbook from this set is known to survive (a volume known today as the Braikenridge manuscript);Footnote ⁷⁶ the other volumes from the set, assuming they were ever produced, have never been definitively traced to any library or sale catalogue.Footnote ⁷⁷

It seems more likely, however, that Willmott was acquiring early modern material not primarily for practical use but for antiquarian purposes, as mentioned above. This would explain why she purchased partbook sets bound together as single composite volumes, which must have made performance from them a somewhat convoluted exercise, assuming it was ever attempted. Given that practical use seems not to have been Willmott’s principal motivation, it is perhaps unsurprising to find that material from her collection (insofar as it may be traced to the libraries in which it today resides) contains little in the way of performance annotations.Footnote ⁷⁸ With books and manuscripts owned by a musically literate individual, we typically find added performance directions, fingerings, bowing indications, or even the odd correction — signs that the owner engaged with or performed from them.Footnote ⁷⁹ But in material from Willmott’s collection there are none — at least none which can be said with any certainty to have come from her.

Loans

Apart from her loan of various items to the Worshipful Company of Musicians exhibition in 1904, Willmott occasionally lent other items from her collection. Her copy of Forbes’s Cantus, Songs and Fancies, to Three, Four, or Five Parts, Both Apt for Voices and Viols (1682) was lent to Sir Frederick Bridge for a lecture given at Gresham College in 1915 (Bridge had been Gresham Professor of Music since 1890, a position he combined with several others, including King Edward Professor at the University of London from 1903 onwards).Footnote ⁸⁰ A report of the lecture published in the Musical News of 1916 records that the ‘British Museum, the Royal College of Music and the late Dr Cummings possessed examples’ of this publication, printed in 1662, 1666, and 1682, adding that ‘the copy which Sir Frederick dealt with was lent to him by Miss Ellen Willmott; he was glad to return it and be free from the responsibility of so precious a book’.Footnote ⁸¹ Willmott owned the third edition of 1682, bound in blue morocco with gilt edges (this later appeared as lot no. 443 in the auction of 1935). Willmott’s source of acquisition for this volume has not yet been established: Quaritch listed two 1682 copies in January 1899, one of which was similarly bound in blue morocco with gilt edges and offered for £12 10s.; it was described as a ‘perfect copy’.Footnote ⁸² (Willmott’s copy, by the time it was sold in 1935, had ‘minute defects in [its] title’).Footnote ⁸³ Julian Marshall also had a copy which went as lot 224 in his 1904 sale, also bound in blue morocco with gilt edges, but with its last leaf ‘supplied in photo-lithography’; it was not acquired by Quaritch but by Ellis, for £11.Footnote ⁸⁴ Willmott also lent items to the Royal College of Music, discussed below.

Disposals

By the end of 1909, Willmott was experiencing financial difficulties. As well as the costs incurred by Warley Place, with its army of up to 104 gardeners, she had funded several overseas plant-collecting expeditions; she had also spent a considerable sum on antiques, cars, clothing, jewellery, and silverware.Footnote ⁸⁵ (Willmott’s first biographer, le Lièvre, appears to have hit the nail on the head when she wrote, ‘whenever she wanted something, she bought it’.)Footnote ⁸⁶ By 30 June 1905, the extent of Willmott’s balance owing to Quaritch was £1199 3s. 10d. — roughly £126,000 today.Footnote ⁸⁷ As well as her ongoing book acquisitions, she purchased a property in Boccanegra, Italy, for 12,950 lire in 1905; in 1906 she undertook a full redecoration of Warley Place, which included ‘extensive structural alterations in the music-room’ — work which cost £1131.Footnote ⁸⁸ To make matters worse, Tresserve, the family chateau near Aix-les-Bains (apparently filled with rare books from the bookseller Jullien), had burned down in 1907 and needed rebuilding.Footnote ⁸⁹ Willmott sought to sell her properties at Tresserve as well as Boccanegra, but without success.Footnote ⁹⁰

Willmott’s spending continued. An invoice from Quaritch dated 18 January 1909 asks for £151 for only three items (all of them botanical); a statement of her account dated 31 December 1909 records a balance owing of £713 1s. 6d. Although some returns made small reductions to her debt, a further statement of account dated 30 June 1910 still shows a substantial balance owing of £699 11s. 6d. This balance was further reduced, by returns rather than payments, to £599 14s. 6d. by 30 September 1910. Quaritch himself wrote to Willmott on 3 October 1910, beseeching her, ‘as I have several very heavy payments to make this month I shall be obliged if you could make it convenient to remit me a substantial cheque on account’.

No substantial cheque seems to have been forthcoming, for Willmott’s balance owing to Quaritch at the end of December 1910 was £626 12s. 10d. Her reaction was to return further items for credit against her account. Fifteen volumes of a botanical series were returned on 9 February 1911 for £63, although small purchases continued to be made; a statement of 30 June 1911 left a balance owing of £572 18s. 4d. A credit note issued on 2 September 1911 confirms that she had returned the two complete sets of Yonge’s Musica transalpina which she had purchased in 1909, even though Quaritch had written to her on 18 January 1909 persuading her to keep it (‘I do hope I can induce you to keep the Yonge Musica Transalpina, payment can of course be made entirely at your convenience, it is almost impossible to obtain another complete set’). Returned it was, however, and a credit note for £75 12s. (the original price, less commission) was issued on 2 September 1911. By 1 December 1911, her balance had been reduced to £389 21s. 2d.

In 1913, presumably in a bid to raise further funds, Willmott sold one of her Amati violins and a Stradivarius violin to the firm of William Ebsworth Hill of Bond Street, London — the same firm from which she had purchased her first Amati in December 1889.Footnote ⁹¹ She also tried to let out various satellite buildings on the Warley estate for £500 a year, and at Boccanegra, both unsuccessfully.Footnote ⁹² She did succeed in letting Tresserve, which had needed rebuilding following the fire of 1907 — work that had to be financed with a loan from the Credit Foncier de France.Footnote ⁹³ Evidence of some restraint in her spending comes in a letter dated 2 October 1918 from the London violin maker Arthur F. Hill. It requests that Willmott return a madrigal book (of which no further details are known) should she not wish to purchase it — a volume which had presumably been sent to her on approval (see Figure 10). Otherwise, Willmott’s spending seems to have been only slightly stunted in terms of the books purchased. She evidently didn’t cease searching for material, since a letter to her from Barclay Squire, dated 20 January 1917, states, ‘I have none of the autographs wanted’. (A further letter from Barclay Squire to Willmott, dated 3 March 1921, also mentions autographs.)

Figure 10. Letter from Arthur F. Hill to Ellen Willmott, dated 2 October 1918.

In 1918, Willmott sought to dispose of her complete set of Gibbons’s First Set of Madrigals and Mottets (1612), which she had purchased from Quaritch’s catalogue 237 in 1905 for £84, uncut but ‘in beautiful condition’. A letter from Ellis of 29 New Bond St dated 6 July 1918 states:

Whatever Willmott’s lowest price was is not known; no reply to this letter can be found. What is clear, however, is that these five Gibbons partbooks stayed with her until the end: they are listed in her Sotheby’s auction of 1935 (as lot 447).

Further attempts at disposal may also be traced. In 1920 Willmott sought to sell her two-manual harpsichord by Jacob Kirkman (1766), which had been purchased for her by her father around 1870 — thought once to have been the property of Princess Amelia. It appeared in a sale held by Sotheby’s on 12 November 1920, along with some furniture, but failed to sell, despite advertisement.Footnote ⁹⁴ Willmott also sought to sell some music. A letter to her from Barclay Squire, one of her long-term correspondents, dated 25 January 1921, explains that he had ‘written to Cockerell to say you will write to him about the Autographs’ (see Figure 11).Footnote ⁹⁵ This was Sir Sydney Carlyle Cockerell (1867–1962), then the director of the Fitzwilliam Museum, Cambridge. Presumably Willmott was hoping to sell some autographs to him, following an introduction from Barclay Squire; Cockerell was able to indulge in various purchases, ranging ‘from Greek vases to […] music autographs’ thanks to backing from the Friends of the Fitzwilliam.Footnote ⁹⁶

Figure 11. Letter to Ellen Willmott from William Barclay Squire, 25 January 1921.

Barclay Squire also adds, ‘as to the Elizabethan book I am told that the Museum would recommend the Trustees to give £10 for it. Could you let it go at this? Do, please, for it ought not to go to Leeds!’ As will be seen below, he was referring to the Willmott manuscript. Barclay Squire worked in the music department of the British Museum for thirty-five years until 1920, and was therefore well placed to intercede with potential prospective purchasers on Willmott’s behalf.Footnote ⁹⁷ The identity of the interested party in Leeds, whether institutional or individual, is not specified; possibly it was the Leeds-based music collector Frank Kidson (1855–1926), whose collecting focused on old English songs and dances, although he also owned various seventeenth- and eighteenth-century English publications as well as Thomas Morley’s Plaine and Easie Introduction to Practicall Musike (1597).Footnote ⁹⁸ No correspondence between Willmott and Kidson on the matter has been found to corroborate this possibility.

It has not been possible to locate Willmott’s reply to Barclay Squire (no letters from her survive in his papers), but his plea that the Willmott manuscript should not go to Leeds was apparently heeded.Footnote ⁹⁹ Documentation dated 1 April 1921 shows that a Mr Gilson of the British Museum was recommending that the trustees purchase from Willmott ‘a volume of 16th cent. vocal music, containing a part of pieces by Tallis, W. Byrde, J. Sadler, etc., dated 1591, with curious calligraphic ornament’; it was being ‘offered for £10’.Footnote ¹⁰⁰ The purchase of the book, apparently brokered by Barclay Squire, was approved by the Standing Committee of the British Museum on 9 April 2021 (see Figure 12).Footnote ¹⁰¹ A further report from Gilson, dated 25 April 1921, reports that the purchase was to be cancelled, however, as his recommendation ‘had been made under a misrepresentation. The book was submitted through a third person [Barclay Squire], and Mr Gilson believed the price to have been agreed, but this was not the case.’Footnote ¹⁰² A note of the cancellation was entered in the minutes of a British Museum Standing Committee meeting on 7 May 1921, when the trustees ‘authorised the withdrawal of the transaction’.Footnote ¹⁰³ Willmott presumably wanted more than £10 for the volume — not surprising given that she had paid more than £52 for it some ten years previously.

Figure 12. Excerpts from a ‘Report Respecting Offers for Purchase’, 1 April 1921, and a ‘Report from Mr Gilson to the Trustees of the British Museum’, 25 April 1921. Source: British Museum Records.

While Willmott was in touch with Barclay Squire about letting a partbook go to the British Museum, she was also communicating with Quaritch about disposing of her entire library. A letter from Quaritch to Willmott dated 24 February 1921 explains that he would be ‘quite prepared to buy a selection of your books, though at the present time it would not be possible for me to buy the whole’. He then suggests sending a Mr Craddock, who had ‘for many years been in charge of my scientific books’, to Warley to assess her library. (This suggests it was the botanical books that were being assayed rather than her music volumes.) On 3 March 1921, however, Barclay Squire wrote to Willmott, informing her that he had visited Quaritch’s, where he had run into Mr Dring, Quaritch’s stock manager. He reported:

Quaritch wrote again to Willmott on 20 July 1921, explaining that Mr Craddock was no longer with the company, and that a Mr Williams would instead make the inspection. Presumably the books to be inspected did not include music, for a ‘List of the Music, Books and Mss. belonging to Miss Willmott, Great Warley […] made from perusal and examination of the books’ was drawn up on 12 July 1921 by Barclay Squire (the document does not mention his name, but it is in his distinctive handwriting).Footnote ¹⁰⁴ This list extends to some fourteen pages, and provides a partial inventory of some but not all of the music in Willmott’s possession at the time. It includes a good deal of her printed material but excludes music manuscripts and autograph letters; consequently it did not list the Willmott manuscript or the Armstrong–Finch manuscript, for which Willmott is today known. Presumably she had wished to keep these volumes for herself, or she reckoned they would fetch a higher price if sold individually.

A further letter from Quaritch to Willmott, dated 28 July 1921, mentions that he had ‘not heard from Mr Barclay Squire regarding the books. Was he to communicate with me?’ Presumably Willmott had engaged Barclay Squire to lobby Quaritch into buying some of the music in her collection. Quaritch wrote next on 3 August 1921, confirming that he was enclosing ‘the list of books selected by my assistant yesterday, and I will make you an offer for them tomorrow’, but this presumably refers to botanical volumes. A further letter from Quaritch dated 4 August 1921 offers her £1250, although it does not mention music or Barclay Squire.

As well as seeking to dispose of some material herself, Willmott was occasionally approached by persons interested in purchasing items from her collection. One such request came from the artist William Westley Manning (1868–1954), who was, like Willmott, a music collector.Footnote ¹⁰⁵ He wrote to her on 27 January 1923, remembering ‘the interesting musical autographs which you lent to the exhibition at Fishmongers’ Hall’ and enquiring as to whether she ‘should feel disposed to part with a small Schubert’. Presumably she did not accede to his request, for three Schubert manuscripts featured in the eventual 1935 auction of her collection (lots 572–74). No further letters to Willmott from Manning have yet been found to document any further approaches; nor do any replies from Willmott appear to survive.

Wheldon & Wesley, 1924 and 1926

In 1923, Ellen Willmott turned to the firm of Wheldon & Wesley, London, to dispose of some books (the firm to which Mr Craddock had jumped ship from Quaritch).Footnote ¹⁰⁶ A duplicate copy of a letter, sent from Wheldon & Wesley to City and Midland Bank, dated 9 June 1923 and signed by Craddock, explains that the firm had been instructed by Miss Willmott to sell some of her library: ‘The first portion of her books to be dealt with by us were included in our Illustrated Catalogue. It has some biographical interest. Miss Willmott has realised up to the present from this portion, £1420. 2s. 6d.’—roughly £74,400 today. The letter then refers to forthcoming listings: ‘Part 2 of the Illustrated Catalogue is in the printer’s hands. It will include Miss Willmott’s beautiful collection of English Madrigals.’ Craddock also explains the basis on which the books were being sold: ‘We have made it a rule to advertise every book sold by us on commission for Miss Willmott […] we pay her the advertised price less 25%.’ (In 1924, Wheldon & Wesley re-issued her book Warley Garden in Spring and Summer, which they had first published in 1909.)

The music was listed in Wheldon & Wesley’s Illustrated Catalogue of Recent Purchases of Rare Books (1924), which featured forty-eight items of ‘Elizabethan Madrigals and other early English music’ (listed as nos. 136–84). Included were partbook sets by Bateson, Byrd, Gibbons, Morley, Watson, Wilbye, and Young; the manuscript of Purcell’s Violin Sonata was absent. The advertised prices in this catalogue are generally higher than those Willmott had herself paid. For instance, her four volumes from Bateson’s First Set of Madrigales (1604), listed as no. 136, were being sold for £38; the same four books had been purchased, along with two volumes of Bateson’s Second Set of Madrigales (1618), at the 1904 sale of Julian Marshall, for £17 10s. — roughly £2 18s. 4d. per volume. In the 1924 catalogue, four volumes were being sold for roughly £9 10s. per book. The same is true of no. 141, two volumes (Contratenor and Bassus) of Byrd’s Songs of Sundrie Natures (1589), which had been purchased by Willmott for £10 10s. from the Marshall sale, but which was listed for £24 — a price inflated by 120 per cent more than she had paid. Even her five volumes (of six) to Byrd’s Psalmes, Songs, and Sonnets (1611), listed as no. 145, which she had purchased unbound but ‘in lovely state as though just issued’ for £30, were listed for £42.Footnote ¹⁰⁷ No. 156 was Gibbons’s First Set of Madrigals and Mottets (1612), complete, which Willmott had acquired from Quaritch’s catalogue 237 in 1905, for £84. (Evidently it had not been sold by Ellis to their prospective purchaser in 1918.) In Wheldon & Wesley’s 1924 catalogue, it was being sold for £160 — almost twice what she had paid for it.

Only a few partbook sets were offered complete. No. 181 was Watson’s First Sett of Italian Madrigalls Englished (1590), complete and unbound, for £120. Willmott had acquired this set from the Marshall sale some twenty years previously for only £8 15s., and stood to make more than ten times what she had paid for it. (She had also acquired those volumes bound together as a single volume; clearly, by 1924 she had had them disaggregated into parts.) Also offered complete as no. 182 was Wilbye’s Second Set of Madrigals to 3. 4. 5. and 6. Parts (1609), unbound, for £70. Willmott had acquired five of the six volumes to this set in the 1904 Marshall sale, along with all six volumes to Wilbye’s First Set of English Madrigals to 3. 4. 5. and 6. Voices (1598), for a hammer price of £48 10s. — roughly the equivalent of £4 8s. 2d. per book. (The missing book to the second set was acquired by methods unknown.) At £70, each partbook was selling at roughly £11 13s. 4d., about three times what she had paid for them. (Any money received by Willmott, assuming the books sold, would have been subject to a deduction of 25%, for commission due to Wheldon & Wesley.)

The most important item offered for sale in the 1924 catalogue was the Willmott manuscript, which appeared as item no. 154 (see Figure 13). It was given more than a full page of description, as well as two plates (fols [16]^r and [38]^r). The catalogue claims, as Sotheby’s had in 1908, that the partbook has a royal provenance, noting that it was ‘doubtless used by the Queen [Elizabeth], the presence of the dated drawing of the triple Tudor rose indicating the royal ownership’. It also calls the volume ‘probably the finest MS. of its kind in existence’, listing it at £850 (roughly £44,700 today). Had it sold, it would have netted Willmott £637.50 after commission — more than ten times the £52 10s. she had paid in 1908.

Figure 13. Pages 60–61 from Wheldon and Wesley’s Illustrated Catalogue of Recent Purchases of Rare Books (Wheldon & Wesley, 1924).

Although the 1924 catalogue offered fewer than fifty items for sale, a brief report on it made it into the London Mercury and Bookman. This comments on some of the more notable items, including the Willmott manuscript (described as a ‘service book, folio, of 48 leaves, dated 1591, and written for Queen Elizabeth’); it also draws attention to the fact that some items are still ‘in the original wrappers and uncut’ — a guarantee of good condition, perhaps, and offering confirmation that the volumes were probably not acquired by her for practical use:

Messrs. Wheldon & Wesley, of 2, 3, & 4, Arthur Street, W.C.2, have sent me a copy of the second part of the new catalogue (New series, number 11). This contains only about fifty volumes, but they form a remarkably fine little collection of books of Elizabethan Madrigals and other early English music. Many of the greatest names are here. There is, for instance, at the price of £160, Orlando Gibbons’s First Set of Madrigals, quarto, 1612, in five parts (Cantus, Altus, Tenor, Bassus, Quintus). This book is in remarkable condition, since all the parts, save the Tenor, are in the original wrappers and uncut. Another book is T. Watson’s First Set of Madrigals, quarto, 1590 (six parts), for which £120 is asked; and there is J. Wilbye’s Second Set of Madrigals (six parts), quarto, 1609, for £70. Other famous musicians represented in this catalogue (which is illustrated) are Bateson, Bird, Farnaby, Morley, Playford and Purcell. A manuscript of great interest is a service book, folio, of 48 leaves, dated 1591, and written for Queen Elizabeth. The price of this is £850.Footnote ¹⁰⁸

Despite this positive report, it seems that the 1924 catalogue was not a success: a further catalogue was issued in 1926.Footnote ¹⁰⁹ This was a general catalogue which included a subsection of ‘Elizabethan Madrigals and Other Early English Music’; it listed thirty items, most of them printed partbook sets that were incomplete, but excluding some items that had appeared in the 1924 catalogue, including the Willmott manuscript. Prices were the same as those advertised previously (see Figure 14).

Figure 14. Pages 40–41 from Wheldon and Wesley’s Catalogue, no. 18, Rare and Valuable Books (Wheldon & Wesley, 1926).

As with the 1924 catalogue, the 1926 catalogue was subject to a brief review. An anonymous commentator reported in the Publishers’ Circular and Booksellers’ Record that

No librarian or collector can afford to overlook this catalogue, which includes a section, sure to create a furore amongst American collectors, devoted to Elizabethan Madrigals and other early English music. Since the Huth and Christie Miller dispersals no other collection to equal this has been offered for sale. It would require a special number of the P.C. [Publishers’ Circular] to deal with half the rarities in this catalogue.Footnote ¹¹⁰

Despite this recommendation, however, this sale seems not to have been a success either: although no correspondence has yet been located to shed light on the matter, all of the items listed appear in Willmott’s posthumous 1935 auction. Presumably the prices at which her items were advertised in both sales were deemed to be too high.

Final Years

According to one commentator, Ellen Willmott became a rather ‘sad and impoverished’ figure towards the end of her life, having become ‘increasingly eccentric and paranoid’; she ‘booby-trapped her estate to deter thieves’.Footnote ¹¹¹ (She was even said to keep a revolver in her handbag, for fear of being robbed.)Footnote ¹¹² Having squandered most of her fortune, she seems not to have given up on the idea of selling specific items in her library. A letter dated 1 August 1932 to Willmott from Sir Hugh Allen, director of the Royal College of Music from 1919 to 1937, reveals information about a loan — perhaps a permanent loan — of a ‘Choir Bk’ to the college (see Figure 15).Footnote ¹¹³ The letter does not reveal anything about this choirbook, but it may have been one of the three seventeenth-century ‘choir books’ listed in her posthumous auction.Footnote ¹¹⁴

Figure 15. Letter from Sir Hugh Allen to Ellen Willmott, dated 1 August 1932.

Willmott’s initiating correspondence with Allen cannot be found, but she had clearly written to him, claiming that some items she had lent to the College had not been returned. Allen states:

He then refers to a ‘Beethoven sketch’, which Willmott must have also claimed to have lent:

Allen then thanks her for her ‘kind loan to the College of the Elizabethan Bk’, but expresses regret if this loan has caused her ‘anxiety or trouble’, asking her to ‘let me know if have now found it and relieve me of any doubt as to any going astray from the time it was handed over to your agents in the matter’. Precisely what this Elizabethan book was is uncertain: it could have been one of the various printed madrigal partbooks she owned, although it would have perhaps been odd to lend a single volume rather than a complete set. Possibly this volume was the Willmott manuscript.

Willmott died on 27 September 1934, unmarried. Since she had no heirs, her entire estate, including her library, passed to her brother-in-law’s family (her sister Rose died in 1922). Willmott’s personal effects were assayed at £12,787 9s.; probate was granted on 4 December 1934 to her nephew, Robert George Wilmot Berkeley (1898–1969).Footnote ¹¹⁵ According to le Lièvre, it had at one time been Willmott’s intention to bequeath a portion of her library to the British Museum, but this did not happen; le Lièvre also alludes to the possibility that Willmott was minded to leave ‘a collection of rare madrigals to one of the madrigal societies in existence at that time’, but this does not appear to have happened either.Footnote ¹¹⁶ Nor did any items in which Quaritch had been interested seem to have left her collection before the posthumous auction of 1935.

Sales

Sotheby’s, April 1935

Willmott’s books and manuscripts were principally auctioned off by Sotheby’s, in a sale that ran for three days from 1 April 1935. The auction does not seem to have been particularly well advertised: only two advance notices of it appear to have been published. The first appeared in the Burlington Magazine for Connoisseurs in April 1935, just before the auction; it was very short, recording that only a ‘selection from the botanical, musical and general library, the property of Miss. E. A. Willmott’ was to be sold, ‘by order of the Executor’.Footnote ¹¹⁷ The second was published in the Saturday Review of Politics, Literature, Science and Art on 23 March 1935: this featured a brief article on the ‘botanical library of the late Miss Ellen Willmott’, but it also mentioned the ‘large collection of musical books and MSS’ in Willmott’s library, drawing attention to various early printed editions and Purcell’s violin sonata:

There is also a large collection of musical books and MSS in Miss Willmott’s library, notably 16th and 17th century madrigals, early printed editions of Purcell, Handel and others, musical manuscripts of Bach, Mendelssohn, Schubert, and Liszt, the sole existing MS. of Purcell’s only violin sonata, and autograph letters of many of the great composers.Footnote ¹¹⁸

The Sotheby’s catalogue itself offered a total of 575 lots for sale, divided in the auction catalogue into ‘The General Library’, ‘The Botanical Library’, and ‘The Musical Library’. As pointed out already, the catalogue makes clear on its title page that it was offering for sale items ‘selected from the library at Warley Place’ (see Figure 16). Presumably Sotheby’s selected only books which they believed to be particularly valuable — much in the same way that Quaritch had been interested only in a ‘selection’ of Willmott’s library in the 1920s. A preamble to the sale catalogue made mention in general terms of some of the more noteworthy items in her collection.Footnote ¹¹⁹

Figure 16. Inner title page of Sotheby & Co., Catalogue of Valuable Books and Manuscripts on Botany, Music and General Subjects.

Items in Willmott’s collection that were deemed to be the most important were allocated the most space in the auction catalogue. A few lots consequently have several substantial paragraphs of description, whereas other, less important items are given only a few lines. The most important lot descriptions also feature plates. The two lots which were evidently adjudged to be the most important of the music sale were the Willmott manuscript, the lone partbook once owned by John Sadler, and the Armstrong–Finch manuscript, containing the only extant contemporary copy of Purcell’s Violin Sonata.

The Willmott manuscript appeared as lot no. 537 (see Figure 17). Sotheby’s description made the same mistake as Quaritch in his catalogue 270 of 1908, since John Sadler is again named as a composer rather than as a copyist or an early owner, although a paragraph summarizing the various later owners named in the manuscript is provided — something which Quaritch’s previous listing had not included. The partbook is described as a ‘fine and important manuscript’, but the fact that some leaves are missing is also mentioned. When this catalogue was prepared in 1935, it was not yet known that another volume from the same set, the Braikenridge manuscript, was in private ownership near Bristol. Edmund Fellowes seems to have become aware of the Willmott manuscript in 1939, since he transcribed its contents that year; it was only in 1942 that the owner of the other extant volume initiated a correspondence with Fellowes, informing him of its existence, which caused Fellowes to realize the two volumes were part of the same set.Footnote ¹²⁰

Figure 17. Lot 537 in Sotheby’s Willmott sale (annotated copy in the British Library, London, S. C. Sotheby 1531).

The Armstrong–Finch manuscript went as lot 516 in the Willmott sale (see Figure 18). The catalogue description provided what was then the ‘best information available regarding both the volume in general and Purcell’s violin sonata in particular’, that is, until the manuscript recently resurfaced.Footnote ¹²¹ It includes important information on the state of the manuscript, its contents, and, importantly, its previous owners. Peter Holman has pointed out that some material described in the Sotheby’s catalogue is now missing from the manuscript: a ‘thematic index to the Violin Sonatas’ and an inscription, ‘Hunc librum scripsit Gulielmus Armstrong Londini, a.d. 1691’.Footnote ¹²² This makes the Willmott catalogue description an important witness to its prior state. (Some former listings for the same manuscript, in the sale catalogues of previous owners, were ‘rather summary’: for instance, that of 1814 did not mention any instruments, nor that the volume had come from Finch’s library.)Footnote ¹²³ Her copy of Forbes’s Cantus, in blue morocco leather, went to ‘Edwards’ for £10 10s. — a little less than the £12 10s. she had paid Quaritch for it in 1899, assuming that was her source of acquisition.

Figure 18. Lot 516 in Sotheby’s Willmott sale of 1935 (annotated copy in the British Library, London, S. C. Sotheby 1531).

Sotheby’s also listed some of Willmott’s more important musical instruments, in an instrument-specific sale held on 5 April 1935. The Kirkman harpsichord of 1766, with its apparently royal pedigree, which had failed to sell in November 1920, went to ‘Goodwin’ for £90, according to the annotated auctioneer’s catalogue. (It seems not to have left the family, however, since it was recently sold by the same family, via Sotheby’s, in a sale held on 11 December 2019; it fetched £56,250.)Footnote ¹²⁴ Some lots in this 1935 instrument sale attracted less than they were perhaps worth; one commentator, writing some nine years after the sale, records that the ‘fine collection [of instruments], formerly at Warley Place, Essex, in the possession of Miss Willmott, was dispersed piecemeal at her death’.Footnote ¹²⁵

Kemsley, June 1935

A further sale, involving instruments and also some music, was held on 4 June 1935 by Kemsley of Bishopsgate, London; an advert for this sale was placed in the Musical Times in June 1935. This refers to a three-manual pipe organ, a chamber organ, and ‘a number of antique musical instruments’; also mentioned is a ‘large collection of bound and unbound Operatic scores’ and ‘a small library of Musical Books’ (see Figure 19).Footnote ¹²⁶ It has not been possible to locate a copy of this catalogue, which had been available for purchase from the auctioneer. Whether any early music appeared listed in its pages is therefore uncertain.

Figure 19. An advertisement for the Kemsley sale, 4 June 1935, The Musical Times, 76.1108 (1935), p. 489.

Results and Subsequent Sales

The most successful item in the principal Willmott sale, handled by Sotheby’s, was lot 537, the Willmott manuscript. This attracted a winning bid of £155 (roughly £9,500 today) from Willmott’s brother-in-law, Robert Valentine Berkeley — not a bad price for an orphan partbook from a set originally of five volumes. Next was lot 550, Wilbye’s Second Set of Madrigals to 3. 4. 5. and 6. Parts (1609), offered complete, which went to Quaritch for £125 (Willmott had acquired the Cantus, Altus, Tenor, Quintus, and Bassus parts from the 1904 Marshall sale; the purchase of the Sextus book is not recorded). Morley’s First Booke of Balletts to Five Voyces (1595) was offered complete with all five partbooks present, thanks to Willmott’s purchase of two lots in the 1904 Marshall sale (even if she ended up with redundant partbooks); it went to ‘Maggs’ for £105. Lot 447, Gibbons’s First Set of Madrigals and Mottets (1612), with all its five volumes present, went to Harry Richardson Creswick (1902–88) for £102; he purchased it for Cambridge University Library.Footnote ¹²⁷ The prices fetched by these complete sets were high; Hyatt King has put on record that for ‘madrigals and the like, £20 was an average price for a complete set up to about 1918 when in the Littleton sale, Watson’s madrigals (1590) fetched £46’.Footnote ¹²⁸ By the time of Willmott’s 1935 sale, the average hammer price seems to have been subject to further inflation, since some complete sets, such as Wilbye’s Second Set of Madrigals to 3. 4. 5. and 6. Parts, went for as much as £125.

Some partbook sets, although they were offered complete, fared less well. Lot 548, Watson’s First Sett of Italian Madrigalls Englished (1590), with all six partbooks present, went to Quaritch, but for only £88 — possibly because the volumes were ‘not quite uniform in size’.Footnote ¹²⁹ Lot 494, Morley’s Canzonetts, or Little Short Songs to Three Voices (1606), had all three volumes present, but it went to Maggs for only £39, perhaps because all three of its books were bound in one single volume — even though the publication itself was described in the catalogue as ‘extremely rare’.Footnote ¹³⁰ Incomplete partbook sets naturally attracted lower bids than complete ones. Lot 493, Morley’s The Triumphs of Oriana (1601), went to Quaritch for only £8 — presumably because it offered only three volumes (Quintus, Tenor, and Bassus) of an original six. Willmott’s Bassus and Sextus volumes to East’s Third Set of Bookes (1610) went as lot 436 to Quaritch for £6, probably because four volumes were missing. Lot 551 offered volumes from three separate publications: the Altus book to Wilbye’s Second Set of Madrigals to 3. 4. 5. and 6. Parts (1609), lacking the title page and the next leaf; the Quintus book to Yonge’s Musica transalpina (1588), lacking the title page; and a manuscript copy of the Bassus part to Bennett’s Madrigalls to Foure Voyces (1599), in the hand of Dr Burney. All three were knocked down to the London bookseller Ellis for the bargain price of £1. These lower prices brought the average prices fetched by the early modern music in the sale down considerably. Hyatt King observed that up until about 1918, complete sets of madrigals generally fetched around £20 per set: although prices soared to as much as £290 per set in the Christie-Miller sale of 1919 (thanks to the bidding of the American dealer George D. Smith), ‘sanity returned’ in the 1920s; by the time of the Willmott sale, the average price of complete madrigal sets was closer to £40.Footnote ¹³¹ This observation, however, is only a rough average, and does not make clear that rarer items fetched higher prices or that incomplete or defective items attracted much lower prices. Some of the bidders at the principal Willmott sale of 1935 paid much more than £40 for complete sets, and some netted bargains.

The Armstrong–Finch manuscript, containing Purcell’s Violin Sonata, apparently sold to a ‘Marcus’ for £25 (Figure 18). For some years this was thought to be the antiquarian book dealer Hans Marcus, but he denied purchasing the volume when asked in 1972; Holman later suggested that it went unsold, and that it was bought in by the auctioneer because it failed to reach its reserve.Footnote ¹³² (It did, however, sell in a recent Sotheby’s auction for £87,500.)Footnote ¹³³ The music in Sotheby’s Willmott sale, including the autograph letters, raised £1931 8s — roughly £119,000 today.Footnote ¹³⁴

Even though the Sotheby’s auction was publicized with minimal fanfare, reports of the sale were published fairly widely. The most detailed summary appeared in the Publishers’ Circular and The Publisher and Bookseller on 27 April 1935. A preamble points out that ‘prices throughout the sale were excellent, and in many cases were far higher than was generally anticipated’.Footnote ¹³⁵ After the botanical books are dealt with, a section on music summarizes the most important lots of the sale, the prices they fetched, and even the names of their winning bidders. After giving some further results for ‘other works worthy of mention’, the author goes on to detail results for what they perceived to be the more important manuscript items of the sale, although both the Armstrong–Finch and the Willmott manuscripts were omitted:

The Willmott manuscript was mentioned in a summary of the 1935 Sotheby’s sale published in the London Mercury, a paragraph of which dealt with music. This also referred to other items sold, although the names of the winning bidders were not included:

Within a year of 1935, booksellers who had snapped up items at the principal Willmott auction were publishing their own catalogues that listed her stock. Quaritch, who, as we have seen, was already familiar with Willmott’s musical and other books from his dealings with her as her agent and from her attempts to sell him her library in the 1920s, was named as the winning bidder for some seventeen music lots. He seems not to have issued a music catalogue immediately, although he did so for Willmott’s botanical items,Footnote ¹³⁸ but he did list some of her stock in later, non-specialist catalogues. For instance, the Bassus and Sextus volumes to East’s Third Set of Bookes (1610), which had sold as lot 436 to Quaritch for only £6, were then listed in Quaritch’s catalogue no. 524 of 1936 for £20 — more than three times what he had paid for them (they were then purchased by the Folger Library).Footnote ¹³⁹ Similarly, Ward’s First Set of English Madrigals to 3. 4. 5. and 6. Parts (1613), which had been purchased by Willmott from the 1904 Marshall sale for £17 5s., went to Quaritch in her sale as lot no. 547, for only £9 9s. To whom Quaritch sold this volume, and for how much, is not known.Footnote ¹⁴⁰

In 1935, Ellis of Bond St issued a catalogue entitled Rare Old Music, Including Many Pieces of the 17th and 18th Centuries, Purchases from the Late Miss Willmott’s Collection, etc. (see Figure 20).Footnote ¹⁴¹ This listed a total of 124 items, even though Ellis’s name appears next to only twenty or so lots in the Sotheby’s catalogue for the Willmott auction of 1935. Exactly how much music in Ellis’s catalogue was Willmott’s is therefore uncertain; the catalogue does not mention her beyond the cover, and Ellis may have been listing stock purchased at Willmott’s Kemsley sale — the catalogue to which has not yet been located. Ellis, unsurprisingly, was also selling her music at a profit. Ellis acquired Watson’s First Sett of Italian Madrigalls Englished (1590), as lot 549 in her sale, for only £3; this appears in his catalogue for £12 12s. — more than four times the price. (The lone Altus partbook to Wilbye’s Second Set of Madrigals to 3. 4. 5. and 6. Parts (1609), which went to him for only £1, did not appear.) Similarly, Ellis’s purchase of Purcell’s Sonnatas of III Parts: Two Viollins and Basse: to the Organ or Harpsecord (1683), as lot no. 512 in Willmott’s sale, went to him for £18 10s., but was listed by him at a price of £30. Both Quaritch and Ellis stood to make handsome profits from reselling Willmott’s stock.

Figure 20. Ellis’s Rare Old Music, Including Many Pieces of the 17th and 18th Centuries, Purchases from the Late Miss Willmott’s Collection, etc.

The library of Ellen Willmott may not have been the largest assembled by an individual around the turn of the twentieth century; nor, to judge from its contents, was it the most varied. It did, however, contain various early modern music items, including printed partbooks and important manuscript items such as the Willmott manuscript and the Armstrong–Finch manuscript. Evidence for Willmott’s early acquisitions may be thin on the ground, but a substantial corpus of booksellers’ receipts, invoices, and other correspondence in her papers — which survive in unusually large quantity for an amateur collector — shed light on her methods of acquisition and expenditure. They show that the printed partbooks were acquired in haste, initially from the 1904 auction of Julian Marshall, and then at various stages from or through Quaritch — purchases which may have been motivated by her encounters with early prints and manuscripts at the 1904 Loan Exhibition of the Worshipful Company of Musicians, and was perhaps further driven by her involvement in madrigal societies and choirs which performed that music, albeit from modern transcriptions.

Information about items Willmott purchased from their respective sale or auction catalogues reveals that she was likely interested in only the best available copies (sending copies she owned to booksellers ‘for comparison’), even when partbooks were bound together in single volumes or where a volume offered only a single voice part. Composite volumes and books bearing lone voice parts may have been unsuited to practical use, but Willmott appears to have been collecting music primarily for antiquarian rather than practical purposes (even though at least one printed set of partbooks was dis-bound and disaggregated into separate parts while in Willmott’s custody, which may signal an ambition for at least some practical use, even if this ambition was not realized). Consequently none of her material traced so far exhibits any on-page markings to prove practical engagement with it, even though she was herself musical.

Willmott’s acquisition of early modern music, in a collecting stint that appears to have run for more than thirty years from its first recorded traces in 1898, is of significance to musicologists because it neatly coincides with the early music revival which burgeoned in England from the end of the nineteenth century onwards. In this revival, published writings introduced specific sources, their contents, and their composers; individuals such as R. R. Terry sought to transcribe from manuscript partbooks with the aim of liturgical performance; editions such as the seminal Tudor Church Music series published in the 1920s, which ‘played a pivotal role’ in the revival of sixteenth-century music, made that music accessible to the public.Footnote ¹⁴² The availability of early modern music at auction and sale, conversely, made the primary sources of that music available to connoisseurs and collectors — motivated by exhibitions and published editions, as well as liturgical and public performances.

Of greater significance, however, is what Willmott’s papers reveal about her various interactions with booksellers and auctioneers, and what these tell us about the music trade in the first few decades of the twentieth century. Coover has noted that ‘we know very little about the everyday working of the music market’.Footnote ¹⁴³ This situation is redressed somewhat by Willmott’s archive, which documents her purchasing activities and modus operandi. We can also trace her attempts to sell off items when she hit financial hardship in the 1910s and 1920s — attempts which seem to have fallen flat, presumably because Willmott had paid high prices for her items (as she was reminded by Barclay Squire) and was reluctant to let them go for less than she had paid. The eventual 1935 auction of Willmott’s library, handled principally by Sotheby’s, adds further to the picture. Some important manuscript items fetched high prices; some complete partbook sets also did well, while other complete sets fared less well, perhaps owing to their condition. Some items were had at knock-down prices: within a year of the principal Willmott auction, booksellers who had snapped up items were re-listing her stock in their own catalogues, with Quaritch and Ellis making handsome profits.

Today, Ellen Willmott is rarely identified as a former owner in items that were once part of her collection. (The main exceptions are the Willmott and Armstrong–Finch manuscripts.)Footnote ¹⁴⁴ But the various invoices and receipts in her archive do at least allow for her sources of acquisition to be tracked: the annotated auctioneer’s copy of Willmott’s 1935 catalogue at the British Library reveals the identities of the winning bidders and the prices they paid, and several items from that sale may then be found in subsequent catalogues. It is therefore possible for items in Willmott’s collection to be traced successively from owner to owner, revealing information on their custodial history. Willmott’s papers, unusually, offer us evidence of acquisition, attempted disposal, and then eventual dispersal; they offer us a window onto the economics of the music trade as items changed hands — driven perhaps by the early music revival, or by an item’s condition and completeness. They also build a compelling case study of a woman’s taste for early modern musical material, both printed and manuscript, driven apparently by an antiquarian impulse — a field in which Ellen Willmott stood with other unmarried women of her generation, including the instrument collector Margaret H. Glyn (1865–1946), the writer and editor Cecie Stainer (1867–1937), and the folk-song collector Lucy Broadwood (1858–1929), some of whom have received attention in recent work.Footnote ¹⁴⁵ Their activities, and the interconnections between them, invite further study.