¹ The most extensive work devoted to Elgar's recordings, including a full listing of matrix and catalogue numbers alongside correspondence and company records, is Moore, Jerrold Northrop, Elgar on Record (Oxford: Oxford University Press, 1974).Google Scholar A wide-ranging and up-to-date study of the subject may be found in Timothy Day, ‘Elgar and Recording’, in The Cambridge Companion to Elgar, ed. Grimley, Daniel and Rushton, Julian (Cambridge: Cambridge University Press, 2004): 184–94Google Scholar.

² For an account of Elgar as a conductor, supported by numerous eye-witness accounts, see Lloyd, Stephen, ‘Elgar as Conductor’, in An Elgar Companion, ed. Redwood, Christopher (Ashbourne, Derbys: Moorland/Sequoia, 1982): 291–306Google Scholar.

³ Recorded live at the Royal Albert Hall, London, with the Royal Albert Hall Orchestra on 26 February 1927. Published from test pressings in 1982 and subsequently issued on CD in Volume 1 of The Elgar Edition, EMI CDS 7 54560 2 (currently unavailable). See Moore, Elgar on Record, 65.

⁴ For an Elgar discography complete up to the early 1960s see Moore, , An Elgar Discography (London: British Institute of Recorded Sound, 1963)Google Scholar.

⁵ ‘The recordings of Edward Elgar (1857–1934). Authenticity and performance practice’, Early Music 12/4 (1984): 481–9.Google Scholar

⁶ Elgar's Complete Recordings 1914–25, GEMM CDS 9951–5; The Elgar Edition Volume 1, CDS 7 54560 2; Volume 2, CDS 7 54564 2; Volume 3, CDS 7 54568 2 (published 1992–93).

⁷ ‘Alternate’ usually means ‘rejected’, but not always: occasionally a side that was not included in an original issue might turn up in a later one or, perhaps, in an overseas issue. Moore briefly discusses the alternate takes in An Elgar Discography: 5, 31, 34–5. Many of the records were once in Moore's private collection, and he transferred some to cassette tape (from whence they came to me), but the location of many of the originals now is unknown, apart from a small number of test discs of the Cello Concerto, which are held in the Yale Collection of Historic Sound Recordings. As this collection is largely uncatalogued it is possible that more may be found there. One hopes that the location of the test pressings will be made known so that new transfers might be attempted; the present ones are poor. I presented a paper at the Dublin International Conference on Music Analysis (June 2005) entitled ‘The Vagaries and Vicissitudes of Recordings Adduced as Evidence of Performance Practice: A Revealing Case Study’, which considered the implications of the existence of seven surviving test sides (that is, seven different sides out of the total of eleven required for the symphony) of the First Symphony for our understanding of Elgar as a conductor and his approach to recording. It is hoped to publish a revised version of this paper soon.

⁸ It is coupled with the Prelude from Dream of Gerontius and Sospiri. The disc is most easily purchased from the company's website: www.eavb.co.uk/lp/indexcd.html.

⁹ An enquiry to Pavilion Records concerning their Pearl set of the acoustics elicited the following response from John Waite: ‘The Elgar boxed set, 9951–5, is still nominally in the catalogue, but stocks of the ’proper” boxed sets, with all attendant paper parts, has been out for some time, in the light of fading demand. The transfers were state of the art in their day and commercially it would certainly not be worth re-doing them. We could just possibly make up a few sets in an unorthodox way in that there would be the five discs and the one 16pp. booklet. But the discs would not be individually jewel-cased, with inlays, nor would there be the box.’ Email, 13 March 2006. I will make original transfers of some acoustic and electrical sides available as part of the Trinity College Dublin website: www.tcd.ie/Music (any help in locating records in good condition would be gratefully received). A disc devoted to Siegfried Sassoon's poetry and prose read by the author includes all of the acoustic Violin Concerto, though the movements are spread around the disc. Memorial Tablet CD41–008.

¹⁰ The list of works recorded only by the acoustic method includes Carillon, Polonia, Sea Pictures, The Fringes of the Fleet, extracts from The Sanguine Fan and The Starlight Express. See Moore, Elgar on Record.

¹¹ Day, , ‘Elgar and Recording’, 185.Google Scholar

¹² Moore, , Elgar on Record, 36–7.Google Scholar

¹³ Ibid., 39–40.

¹⁴ All information on the recordings is taken from Moore, Elgar on Record unless otherwise noted. Ibid., 32–6.

¹⁵ The wear test involved repeated playings of a disc in order to assess its durability on a typical domestic system. Many fine recordings were thrown away as a result of this procedure and others had to be dubbed to a second matrix in order to tame sound (often with unpleasant results). See ibid., 59–60.

¹⁶ Ibid., 59.

¹⁷ Day, , ‘Elgar and Recording’, 185.Google Scholar

¹⁸ This is a fertile area of theory. William Rothstein views this as taking place within a metrical scheme (see Phrase Rhythm in Tonal Music, New York: Schirmer, 1989)Google Scholar; others prefer to take alternative parameters into account, such as accent, motivic features, texture, and so on. In this article ‘hypermeter’ is taken to refer to groupings of bars such as phrases or sections of a work arising out of a combination of parameters such as phrase beginnings, accents, motivic features, and so on. The system is hierarchical so that at one level of a work there may be many hypermetric accents, but at the highest level there may only be two or three. At the level principally addressed here, hypermetric accents typically occur at the beginning of phrase groupings or sections, so this approach privileges just one of many potential levels. My use of ‘accent’ in the context of hypermeter generally refers to what Fred Lerdahl and Ray Jackendoff entitled ‘phenomenal accent’, meaning ‘any event at the musical surface that gives emphasis or stress to that moment in the musical flow’. Given that the points under consideration are generally the beginnings of phrase groups or sections, it is not always easy to distinguish between phenomenal accents and metrical accents. There is, however, within the Lerdahl–Jackendorf system a flexibility that is of great value in the study of performance, namely their recognition that ‘Phenomenal accent functions as a perceptual unit to metrical accent – that is, the moments of musical stress in the raw signal serve as ’cues” from which the listener attempts to extrapolate a regular pattern of metrical accents’. See Lerdahl, Fred and Jackendoff, Ray, ‘On the Theory of Grouping and Meter’, Musical Quarterly 67/4 (1981): 485Google Scholar.

¹⁹ Notes accompanying The Elgar Edition, GEMM CDS 9951–5.

²⁰ The Gramophone 35 (July 1957): 72.Google Scholar

²¹ RLS 708 (c. 1972) included the Variations, symphonies and other major works; RLS 713 (c. 1975) mopped up most of Elgar's other electrical recordings. Richard Osborne's lengthy and insightful review of RLS 713 appears in The Gramophone 52 (February 1975): 1480–91Google Scholar.

²² CEDAR Audio Limited has developed a number of tools for removing noise from various types of media. As the systems have developed so the level of refinement possible in the various denoising procedures has risen. For more information see www.cedar-audio.com.

²³ CDH 7 69789 (1989).

²⁴ I am indebted to Mark Obert-Thorn for this information.

²⁵ Recorded 18 March and 16 April 1929, matrices WAX4785–9/4, 4846–7, Columbia L2346–51.

²⁶ The Gramophone 50 (July 1972): 243.Google Scholar

²⁷ Ibid.

²⁸ Broadcast 4 March 2006 as part of BBC Radio 3's programme CD Review.

²⁹ New York Times (11 December 1966): D32.

³⁰ Moore, , Elgar on Record, 173.Google Scholar

³¹ This discussion has not addressed the question of whether the evidence of the score, genre and other evidence might allow a rhythmic distinction between the solo and orchestral parts, the one being more introspective than the other. It would also be of interest to study the two recordings of the Cello Concerto with Beatrice Harrison, a performer clearly attuned to Elgar's conducting (superficial impressions tend to the view that, while Menuhin was indeed remote from the Elgar style, Harrison was not).

³² EMI 7 64725 2.

³³ A session report by Edward Greenfield for the First Symphony recording describes Solti checking the tempo of the Scherzo with a pocket metronome. His main pulse is identical to Elgar's in this movement. The Gramophone (August 1972): 334Google Scholar.

³⁴ As bar numbers are not provided in published editions of this symphony, I have used cue numbers.

³⁵ I am most grateful to my research student Daniel Shanahan for his assistance in analysing tempo modification in the first movement.

³⁶ In his rough notes on the music Elgar stated that this ‘was to be considered (& labelled) as the principal theme’. Symphony No. 2, Elgar Complete Edition, ed. Anderson, Robert and Moore, Jerrold Northrop (Borough Green: Novello, 1984): viiGoogle Scholar.

³⁷ Simple microphone placement, which rarely amounted to more than three microphones placed in front of the strings, ensured that many recordings of the period are more truthfully balanced than post-war recordings, when 30 or 40 microphones were not unusual and the engineers could rebalance the entire orchestra if they wished.

³⁸ It would be interesting to conduct an experiment with a modern orchestra famed for its rich string textures, such as the London Symphony Orchestra, to find out if 1927 recording conditions and technology deprived the strings of some of their richness and brought the wind forward. My belief, until proven wrong by such an experiment, is that changing performance styles are of greater import than changing technologies.

³⁹ Moore, , An Elgar Discography, 5.Google Scholar

⁴⁰ As observed by the contralto Astra Desmond, quoted in Lloyd, ‘Elgar as Conductor’, 299.

⁴¹ Lloyd, , ‘Elgar as Conductor’, 299.Google Scholar