Establishing the size of the nineteenth-century harp industry, or indeed that of individual makers, is a challenge. The Erards’ London business accounts list 6862 harps sold between 1797 and 1917 but, with the exception of the Erat journal, which identifies instruments made between February 1821 and June 1824, no other company ledgers are known to survive. In order to calculate the number of harps made by others it is necessary to look elsewhere. Fortunately, most makers engraved serial numbers on their instruments, and surveying those in public and private ownership gives some clues. The single largest source of harp serial numbers are the Morley ledgers, which span 1891 to 1945 in 38 volumes. Joseph George Morley (and later his son John Sebastien) listed a number of harp types: single actions were entered as such, were abbreviated to ‘SA’ and occasionally entered as ‘simple’, from the French term for single action. Double action instruments were denoted ‘DA’ (referring mostly to Grecian harps but also sometimes gothic ones); the term ‘Grecian’ appears occasionally; and gothic harps were commonly described thus. Occasionally, interesting or unique instruments were given their own terms. For instance, whilst today Blazdell’s Elizabethan harp is mistakenly referred to as ‘gothic’, its original term was used in the ledgers: it may be that J.G. Morley learnt this term first hand during his time at Erard, perhaps even from a former Blazdell employee. French single action harps were commonly called ‘French’ or ‘Antique’, and the names of their makers rarely noted.
Excluding the resale of harps by Erard (for whom we already have a number) and those made by Morley (these dating from the twentieth century), the examination of a cross-section of eight of the 38 ledgers reveals 378 harps by 16 makers.[i] Some makers are underrepresented: Dizi (and Dodd-Dizi) for instance, numbered their instruments in hidden places; their serial numbers were rarely recorded. Whilst they appear in the ledger, it isn’t possible to say whether one harp is represented several times, making counting them impossible. That few Dodd-Dizi harps survive in public collections suggests that few were made. Indeed, more instruments inscribed Dodd, made in parallel to those inscribed Dodd-Dizi, survive today. Similarly, where Delveau harps were noted as having a ‘lyre pillar’, serial numbers are not noted; perhaps Delveau infrequently or never numbered this model. Fewer lyre-model instruments by Delveau survive when compared to those he made in imitation of Erard. What is immediately evident is that where some makers began numbering harps at one, others started at higher serial numbers or used more than one numbering system, perhaps to give an illusion of early success, to denote different models, or to hide business failure. The ledgers list ten single-action harps by Alexander Barry, numbered between 77 and 770 – no double actions are noted. As the numbers are broadly sequential and start low, we can assume that Barry made at least 770 harps. Only two harps, both double-actions, by Joseph Beasmore appear, their numbers, 422 and 424, suggesting that he started numbering higher, perhaps around 400, and that he made fewer harps. Beasmore, who made harps from c1835 until his death in 1841, probably made fewer than 100 instruments. Ten harps by Blazdell are listed, their numbers ranging from 791 to 4550. It is plausible that Blazdell’s serial numbers started at around 700 and that numbering was restarted at least once, perhaps corresponding to bankruptcy and business restarts. Only one harp was numbered under 1000, the next being 3050, and numbers then jump to nine instruments ranging from 4029 to 4550. From those with serial numbers of 4000 or higher (the others being outliers), one might deduce that Blazdell made about 500 instruments. Only two harps by Browne are noted, their numbers being 125 and 556. It’s unlikely that Browne could have made 556 harps during his estimated five years trading in London, before moving to New York. Based on Erat’s lowest annual output (40) twenty years earlier, we can estimate that Browne made a maximum of 200 instruments. Delveau is more straightforward than Brown: the Morley ledgers list 17 double-action instruments numbered from 26 to 2296. As the numbers are sequential and spaced, the total number of harps can be estimated at 2300. Dockree, a small maker, is represented by four harps, their numbers ranging from 40 to 611. His limited working period, and the scarcity of surviving instruments, suggests that he cannot have made 611 instruments; it is likely that he made no more than 200: perhaps he restarted numbering annually, moving up by 100 each year. Estimating the number of instruments made by Dodd is, like Delveau, fairly simple: 60 instruments, 10 single actions, 46 double-actions, a Grecian and two Gothics, and one harp that isn’t described, are noted in the ledgers, their serial numbers ranging from 16 to 922; Dodd would have made at least 930 instruments. Erat is also straightforward: 121 harps (35 single actions, 65 double actions, four identified as Grecian, eight Gothics, and nine not described) range in number from 320 to 4005; furthermore, that the Erat ledger records instruments with serial numbers as low as 12 confirms that numbering started at one. The Erats, therefore, made at least 4005 harps. Projected from serial numbers given in the Erat ledger, the business start and end dates, and datable instruments, estimated dates can be given to Erat’s harp by serial number (Table 1).
Table 1. Erat harps by serial number (1797-1858). Twenty-nine harps (one single, 11 double, 2 denoted Grecian, 13 Gothics, and two not described), ranging in number from 23 to 1136, indicate that Grosjean made a minimum of 1140. Haarnack’s serial numbers seem to follow three widely separated systems: of nine instruments (two single actions, a double action, and six gothics), one is numbered 400; six range from 723 to 929; and a pair, 4028 and 4030, are numerically distant. Again, the scarcity of harps by Haarnack tells us that it’s unlikely that they made over 4,000 harps. Although the company traded from c1860 until c1930, much of their business would have been in maintaining and restoring old instruments. It is likely that they made around 500 harps. Three harps by Mott, all double actions, were numbered 586, 589, and 590, suggesting that Mott, mainly a pianoforte maker, made few harps. His serial numbers perhaps started at 550. Harps by Mott are very rare today, and it is likely that he made fewer than 50. Rais is another minor maker whose instruments are rarely seen today. Only five instruments, four Gothics (nos. 1120, 1851, 1857), a double action, (no. 4399), and one not numbered, are recorded. That no instruments with serial numbers under 1000 are listed indicates that numbering started around there. Harp number 4399 is clearly an outlier: it is unlikely that Rais, like Mott, made more than 50 harps. Fifty-eight harps (three single-actions, 48 double actions, two Gothics, and 5 not described) by Schwieso (or companies in which Schwieso was a partner) are recorded, with serial numbers from 54 to 970. The sequential and contiguous nature of the serial numbers indicates that Schwieso, despite bankruptcies and at least three changes of partner, maintained one system. It is therefore reasonable to estimate that Schwieso made at least 1,000 harps. Twelve harps by Serquet (Schwieso’s former partner) are listed (six double-actions, five gothics, and one not described), numbered from 454 to 1136. Serquet may have started numbering at 400. If so, this would indicate that he made around 800 harps. Surprisingly, only 31 harps by Stumpff (four single actiond and 27 doubles) appear in the Morley ledgers. Stumpff traded from c1812 to 1847; a clue to the relatively small number of instruments is perhaps evident in the 1847 auction catalogue of his effects which, amongst many harp poles and bodies (more than Stumpff could use), were piano parts. It is plausible that whilst a maker in his own right, Stumpff was also making harp and piano parts for others. The serial numbers of Stumpff harps range from 88 to 1756, so it is plausible that Stumpff made 2,000 instruments. Several caveats must be invoked here. Firstly, without detailed company records, it isn’t possible to say precisely how many harps were made by each maker. Secondly, whilst they constitute the largest source of such data, the Morley ledgers only record the makes and serial numbers of instruments that passed through their business for repair or sale, and they are far from being a complete representation of makers and the number of instruments made. That said, by adding together the estimated numbers for each maker we reach the total 21,797. Taking into account unknown but small quantities of Dodd-Dizi and Delveau lyre harps, we can reasonably estimate that at least 22,000 harps were made in London during the nineteenth century. Makers can be sequenced thus, in order of the number of instruments made (Table 2).
Table 2. Estimated number of harps made by nineteenth-Century makers.
[i] For the purpose of this study, eight ledgers were chosen from across the early years of the company, these being for the years 1891, 1894, 1897, 1900, 1904, 1908, 1910, and an account summary covering the period 1908 to 1913.