The contents of this volume (hereafter MB 88) neatly capture the fashions in instrumental music in London during the period between the death of Purcell and the arrival of Handel. William Croft's admiration for both composers is a matter of record. In his noble setting of the burial service, which has been sung at almost every state funeral since 1722, Croft ‘endeavoured … to imitate that great Master and celebrated Composer [Purcell], whose name will for ever stand high in the Rank of Those, who have laboured to improve the English Style’ (in Croft, Preface to volume 1 of Musica sacra (London, 1724), 3–4); the fact that his setting incorporates Purcell's Thou knowest Lord (written for the funeral of Queen Mary) without any incongruity is testament to his success in this regard, as is the attribution to Purcell of Croft's C minor Ground for keyboard (zd221) in some sources. There is also evidence that after hearing Handel's ‘Utrecht’ Te Deum in 1713 Croft revised his Te Deum in D major (1709), a work that was originally modelled on Purcell's 1694 setting. (A Musica Britannica volume of Croft's Canticles and Anthems with Orchestra is currently in preparation.)

Croft would have first encountered Purcell's music regularly as a chorister in the Chapel Royal under John Blow. He probably joined the choir in the mid-1680s and was dismissed only in April 1699 at the age of twenty. During the 1690s Croft was a pupil of Blow, and copies of music by Blow and Purcell dating from this time survive in Croft's hand. Evidence of this English inheritance, however, is almost entirely absent from Croft's first published compositions, the three ‘Sonatas or Solos’ for violin, which were advertised in October 1699. These thoroughly Italianate works predate the publication of Corelli's Op. 5 sonatas (1700), and while it is possible that manuscript copies of at least some of Corelli's pieces may have been circulating in England, a number of factors suggest that Croft found a model for his works that was much closer to home.

Croft's three violin sonatas were paired in publication with ‘three for the Flute’ by an unnamed ‘Italian Mr’ who can almost certainly be identified as Gottfried Finger, the Moravian viol player and composer who had been resident in London since about 1687. Although at first it might appear that Finger's recorder sonatas were included simply to make up the numbers, the pairing of works for violin and recorder in a single publication was in fact a format that had previously proved commercially successful. (John Walsh certainly believed that the Croft/Finger publication would sell well, issuing a rival to John Young's edition within weeks.) The first English publication to use this format was Finger's VI Sonatas or Solo's, Three for a Violin & Three for a Flute with a Through Bass for ye Harpsychord (1690). These sonatas were evidently the fruits of Finger's travels in Italy, and in the dedication he stated that ‘the humour of them is principally Italian’. It appears that Finger's sonatas provided the model for Croft's 1699 publication, and also for Daniel Purcell's Six Sonatas of 1698. (The bass parts to Croft's three violin sonatas are also found (unattributed) in GB-Lbl Add. MS 31993 immediately after the three recorder sonatas from Finger's 1690 publication.) One device used by Finger and adopted by his English imitators is the prefiguring of the main theme in the basso continuo part at the start of movements; this device is also found in contemporary Italian vocal music and in English songs and cantatas of the period. The twenty-year-old Croft's grasp of the Italian idiom in these works is impressive, and, while his material is simple (and occasionally lapses into cliché), his handling of it is assured and effective. Much of the dynamism of the music derives from a strong treble–bass dialogue, including some invertible counterpoint (MB 88, No. 3b). Glimmers of the English tradition are perhaps discernible in the rhythmic displacement of motifs (as in the bass part of MB 88, No. 1b, bars 2–3 and 7), a section of triple time in a duple-time movement (MB 88, No. 2b) and occasionally a tunefulness reminiscent of Purcell (MB 88, Nos 2c and 3c, bars 14–16), but the one movement built on a ground bass – a feature so common in Purcell's works – sounds more Italian than English (MB 88, No. 1d).

There is also little evidence of Croft's English inheritance in his sonatas for two recorders without bass. Such works were immensely popular in England during the early years of the eighteenth century, especially among gentlemen amateurs; even complete operas were later arranged for one or two ‘flutes’ to cater for this curious phenomenon. Two editions of Croft's six sonatas were published almost simultaneously at the beginning of February 1704, John Young advertising his as ‘being the Original, and carefully corrected by the Author’; the sonatas were also re-engraved in 1706 for inclusion in a volume published by Estienne Roger in Amsterdam. As Diack Johnstone points out, ‘to compose music successfully for two treble instruments of the same kind without a bass to support them is a very considerable test of any composer's ingenuity and imagination (and particularly so with two instruments of such limited range and dynamic resource as the treble recorder)’ (xxv). Croft's six sonatas, which consist of three or four movements, demonstrate his ability to manipulate simple material to maximum effect and to incorporate contrapuntal artifice in an elegant and entirely natural manner (note, for example, the use of inversion in MB 88, no. 4d).

In the light of Croft's known admiration for Blow and Purcell, the near absence of elements of the English style in these published works is striking. While there is no reason to doubt that Croft's desire to master the Italian idiom was genuine, his trio sonatas tell a different story: they are clearly the result of a period of intense study of Purcell's works. Croft's trio sonatas survive only in manuscript and, considering the lack of commercial success of Purcell's 1697 set of sonatas, it is perhaps not surprising that they were not published during Croft's lifetime. The unique source (now MS 1262 in the Gerald Coke Handel Collection at the Foundling Museum, London) was copied by John Barker (1705–1781), a former Chapel Royal chorister under Croft who was also responsible for several important copies of Croft's work, notably his keyboard music; Diack Johnstone stumbled across this manuscript in 1977, and his presentation in a modern edition of these Purcellian sonatas alongside Croft's Italianate works provides a complete picture of Croft's compositional preoccupations and development for the very first time.

In the Sonata in F major (MB 88, no. 11) the general ground plan, the textures and some of the rhythmic and thematic motifs are all directly modelled on Purcell's ‘Golden Sonata’ (no. 9 of the 1697 set), while the last movement of the Sonata in B flat (MB 88, no. 12f) is a deliberate parody of another Purcell movement (the last movement of no. 8 in the 1697 set). The canzona-like movements vary in effectiveness: that in the Sonata in E minor (MB 88, no. 10b) suffers from a lack of harmonic variety and its looser counterpoint perhaps suggests that it was composed earlier than the highly accomplished (and musically effective) triple invertible counterpoint found in the Vivace of the Sonata in B flat (MB 88, no. 12e). The Adagio in F minor (MB 88, no. 11b) is a particularly effective essay in Purcellian style, complete with echoes of Dido's lament (at bars 30–37). By contrast, the Sonata in B minor is thoroughly Italianate and demonstrates again Croft's firm grasp of this style; its movements are linked by a common initial thematic cell, a technique found in Corelli's Op. 3.

Alongside these trio sonatas in the Gerald Coke manuscript are two further chamber works by Croft. The Sonata in F major for two recorders, two violins and continuo bears strong resemblances to a type of sonata developed during the 1690s in London by Finger and the German composer Johann Gottfried Keller. These works, some of which were published in collections in 1698 and 1699, were scored for two recorders and two oboes or violins with continuo (modern editions of four of them were published in the Nova Music series: in Nos 101 and 196 (Keller), edited by David Lasocki (London, 1978 and 1981) and No. 144 (Finger), edited by Peter Holman (London, 1980)). It appears that these pieces were performed by the chamber musicians (including the oboist Peter La Tour and the recorder player James Paisible) employed by Princess Anne (soon to be Queen Anne). Croft may have heard these sonatas at the concerts put on by Finger from at least 1693 in the York Buildings, off the Strand; Keller, who had established himself in London by 1695, may also have been involved, playing the harpsichord. In all these works, the pairs of instruments present material separately before engaging in close dialogue; they rarely play simultaneously. Croft's sonata is, however, much more tautly constructed and more economical in its use of material than Finger's looser works, and it is more thoroughly Italianate than Keller's, which incorporate French-style dance movements. Another significant point of difference between Croft's sonata and those of Finger and Keller is its three-movement plan (fast, slow, fast), which suggests the possible influence of Torelli. The compositional techniques, stylistic features and three-movement plan used in this sonata are also found in Croft's sonata for four violins and continuo MB 88, no. 14b. In addition, this work contains a greater proportion of writing for all four upper parts simultaneously, more textural variety and a bass part that participates more actively in the musical dialogue. The central movement, which is in an English style, begins with a particularly effective and satisfying passage in which the tessitura rises through a succession of suspensions (bars 1–11).

The date of composition of Croft's trio sonatas and the two five-part sonatas is not certain. The unique source was probably copied after Croft's death in 1727: John Barker signed and dated the manuscript in 1736 and his copying of music by Keller and Handel can be dated to the early 1730s, but the watermarks indicate that all three paper types were manufactured in the 1720s. Diack Johnstone draws a parallel between Croft's F major trio sonata and the first of Loeillet's Op. 1 sonatas (London, 1722), but it is perhaps more likely that Croft's works were composed before 1708, when his workload at the Chapel Royal and Westminster Abbey increased dramatically following the death of John Blow. Peter Holman has also argued that it is perhaps significant that all the other surviving music for two recorders, two oboes or violins and continuo can be dated within the years 1698–1704 (Preface to Godfrey Finger, Two Sonatas, Nova Music 144 (London, 1980)).

Croft's violin sonatas are also available in an edition by Ian Payne (Severinus Early Music Edition 4–6 (Sutton St Nicholas: Severinus Press, 1998)), who has edited the recorder sonatas that completed the original publication as well. Working from the same source material, Payne and Diack Johnstone make very similar editorial decisions, though the Musica Britannica volume contains a more detailed critical commentary. A useful context for the evaluation of Croft's sonatas is also provided by Payne's editions of the trio sonatas of William Williams (1700) and of two anonymous sonatas for four violins possibly copied in Italy by James Sherard around 1690 (Severinus Early Music Edition 7–9 and 14–15 (1998 and 1999)). Paul Rubardt's 1932 edition of the recorder sonatas (reprinted in 1954 as Hortus Musicus, volume 245) was based on the Roger re-engraving alone, which, while apparently incorporating accidentals added by hand to the original London editions, also introduced errors of its own. An apparent error in the source texts for the concluding bar of the Presto in the G major sonata (MB 88, no. 6b) is left uncorrected by Rubardt, while Diack Johnstone provides a more convincing solution.

The chief virtue of this volume lies in its provision of a modern edition of the manuscript sonatas; with their mix of Purcellian and Italian elements, these works give a much fuller picture of Croft's musical personality than the published works, which suggest a complete capitulation to the Italian style. Viewing Croft's chamber music as a whole, the Italianate style is nevertheless predominant, suggesting that Croft was drawn more to orderliness and elegance than to the dissonance, angularity and irregularity so typical of Blow and Purcell. Croft was clearly a very skilled assimilator of musical styles, but it would be a mistake to dismiss his chamber music as mere imitation; these works are effective and expressive in their own right and deserve to be more widely performed.