This article was originally published in The Ringing World of 17 August 2018 and has been updated to include more Taylor bells. The conclusions are unchanged.
Taylors claim to be the founder of the world’s finest sounding bells, and rightly credit for the rediscovery of true-harmonic tuning in the late 19th century. The sound of Taylor bells has changed considerably over the last two centuries, as an investigation into 1,838 bell recordings in my collection shows. The history includes interventions by outsiders, changes to reflect a changing market, effects caused by the mechanics of bell production, and a recent return by Taylors to their historic bell designs of a century ago.
The sound of a bell when struck is made up of many different frequencies or partials. Some of them have names: the nominal determines the note we hear (hence nominal – the note name of the bell). The hum is about two octaves lower, and the prime about an octave lower than the nominal. The tierce is an approximate minor third above the strike pitch, the dominant note heard by the human ear.
The tierce is very important to the sound of a bell. It is one of the loudest partials, and the interval of a minor third to the strike pitch gives bells their haunting, melancholy sound - bells with major third tierces sound quite different. The interval between the tierce and the nominal doesn’t change much when a bell is tuned unless the tuning is heavy – the interval is set by the overall design of the bell. Unless the shape of the bell is specially designed to control the tierce interval, the thicker the bell, the sharper the tierce against the nominal.
There is no correct tuning for the minor third interval. Instruments were traditionally tuned in a temperament such as meantone, where the minor third is comparatively sharp. A disadvantage of meantone is that modulating into different keys can be musically unpleasant. In just tuning, the frequency ratios for all intervals involve small whole numbers. The minor third is equally sharp, and key changes are even more unpleasant. However, for a bell tuner just tuning allows easy calculation of frequencies due to the simple arithmetic. In equal temperament, all the semitones are the same size, so that all key changes sound the same. Minor thirds in equal temperament are flatter than meantone or just tuning by a fifth of a semitone.
The tierce has proved an illuminating way to investigate the history of Taylor bells and their tuning. It is by no means the only partial that matters to the sound of a bell, but changes in tierce tuning can be used to show changes in design over time.
The chart shows how the tierces of Taylor bells have changed since the late 1700s, and marks some significant phases of Taylor history. The plot is a running average over 5 years. The 1,836 recordings are of bells of all sizes and include ringing peals, carillons, clock chimes and individual bells. Most are from the UK but there are also bells from Australia, Belgium, Holland, Ireland, Malta, Pakistan and the US. From 1870 to the start of the second war there are between 100 and 200 bells in every decade, and all other decades from 1860 have between 50 and 100 bells.
The chart shows the interval in cents between the tierce and the nominal. Higher points on the chart are sharper, lower are flatter. As explained above, the interval between nominal and tierce doesn’t change much on tuning, so whether a peal has been retuned doesn’t affect the analysis, and the chart gives a picture of design changes over the years.
In this period, Taylors were casting in multiple foundries: St Neots 1784 – 1821, Oxford 1821 – 1854, Buckland Brewer 1825 – 1834 and the old foundry in Loughborough 1839 – 1860. The jagged nature of the graph prior to 1850 reflects both a small number of recordings and the itinerant nature of the work. Ringing peals from this period include Swaffham Bulbeck (1820), Halwill (1826) and Eye (1831).
In 1848, the first contact occurred with Edmund Beckett, later Lord Grimthorpe, the clock designer and architect famous for designing the clock and bells for the Palace of Westminster. Grimthorpe provided Taylors with a steady stream of work for many years, but he had strong views on bell design, advocating bells which were very thick and heavy for their note. The very sharp tierces from 1848 onwards are due to the thickness of the bells. Unfortunately Grimthorpe was wrong about bell design, and bells from this period are not noted for their musical qualities. Bells included in the analysis from this period are Oxford St Giles (1850), Halifax Town Hall (1852), Dunham Massey (1854), Horninglow (1862), Gainford (1865), Madresfield (1867), Inkberrow (1868), Mirfield (1869, some bells now replaced), Worcester Cathedral and Healey (1870) and Bradford Town Hall (1873). The Worcester Cathedral bells were replaced in 1928 but I have the tuning figures taken before the bells were broken up. Many of Grimthorpe’s heavy peals have been replaced, with no tears shed at their passing.
Eventually Taylors rejected the Grimthorpe bell designs, and reduced the weight or deepened the tenor note of their rings, leading to flatter tierces. It is said that Grimthorpe specified the bells for St Pauls Cathedral as 3 tons in D-flat, but Taylors cast them in B-flat instead, claiming to have mistaken the instruction. Their ability to design and cast bells had advanced considerably, and for the next 20 years they produced a number of classic and well-loved old-style rings: Hoar Cross and Derby St Luke (1875), St Pauls (1878), Kingston Dorset (1878, since retuned), Low Elswick and Stockton Heath (1880, both since retuned), Bolton All Souls (1881), Melbourne (1882/7), Ryde (1886), Poynton (1887), Lymm and Ospringe (1891), Newcastle Cathedral and Imperial Institute (both 1892), Aberavon (1893), Curdridge (1894), Newcastle St Matthew and Cambridge OLEM (both 1895).
The plot of tierces jumps around a little – but compared with the previous 75 years displays more consistency. The bell shape they were using gave a tierce mid-way between just tuning and equal temperament.
In a true-harmonic bell, three low partials (hum, prime and nominal) are in perfect octaves. In an old-style bell, the hum (lowest partial) is generally sharp, and the prime (the next highest partial, around the strike pitch) generally flat. Continental founders in past centuries, in particular the Hemony brothers, had known how to produce true-harmonic bells but the knowledge died with them.
In 1894, Canon Arthur Simpson, a clergyman from Fittleworth in Sussex, wrote to Taylors with new ideas on bell acoustics, and went on to write two papers in 1895 and 1896 which among other things explain the importance of true-harmonic tuning to the sound of bells. This innovation is often called ‘Simpson tuning’, and he is widely credited with having rediscovered the technique. One of the Fittleworth bells, and also one from Simpson’s previous parish in Bexhill, had numerous grooves inside the soundbow when inspected by Elphick in the 1930s / 1950s which may indicate tuning experiments.
However, Taylors were already working towards true-harmonic tuning prior to Simpson’s first contact. A plot of the hums of Taylor bells cast between 1885 and 1895 shows that they were changing their bell designs to flatten this partial to the double-octave, and the 41cwt bell at Richard’s Castle of 1892 is true-harmonic to within normal tuning errors.
The new bell shapes needed for true-harmonic partials meant that the strickles or crooks used to mould the bells had to be remade. In 1894 Taylors purchased a new set of tuning forks, and in 1895 installed a tuning machine. By 1896 they had solved the problems in bell design and tuning and for the first time were able consistently to produce bells with hum, prime and nominal in perfect octaves. Transitional rings like Merthyr Tydfil (back seven 1893) show the progress. They were cast before Simpson’s first contact, but retuned in 1896, and have octave hums but flat primes.
It is clear from the plot of tierces that Taylors achieved an astonishing consistency in tierce tuning in this 25 year period, not achieved before or since. The analysis includes many well-loved peals from this period, including Norton and Brewood (1896), Reddish front six and Towcester (1897), New Mills (1898/1901), Hallow (1900), Bridlington (1902), Lahore, Newton le Willows and Redcliffe (all 1903), Taxal (1904), the fine bells in the notorious tower at Leamington St Peter (1905), Bromham (1907), Caerphilly (1910), Middleham (1911), Henfield (1913), Chagford and Nether Stowey (1914), St Neots (1919), Inveraray (1920), and Skipton and Westbury (1921).
However, what is hidden by the averaging in the plot is that the tierce tuning varies across the peal in these rings. The tierce in the tenors is sharp, a little sharper than just tuning, whereas in the trebles it is flat, a little flatter than equal. This trend is consistent across 386 bells in dozens of peals. The reason is the difference in scale to give heavier trebles (as is always done for ringing peals).
During this period, Taylors were also changing the nominal or strike pitch tuning of their bells. In the ringing peals I have investigated, up to 1910 nominals were generally tuned just – with flattened thirds, sixths and sevenths in the musical scale. It is said that the demonstration carillon at the foundry, created in 1904/5, had nominals tuned in equal temperament, to allow music to be played in any key. All peals I have investigated from 1911 onwards have nominals in equal temperament, but still have just tierces in the tenors. It is clear from the day-book entries for Inveraray, kindly provided by George Dawson, that Taylors were attempting to get equal tempered tierces in all the bells but didn’t have the bell shape to do it.
For the carillon market, where very accurate tuning is required, the combination of equal-tempered nominals and just tierces is not satisfactory. The tierce plot shows that from the early 1920s Taylors changed their bell shape to give the flatter tierces. This coincides with a major increase in their carillon business. The earliest carillons for which I have recordings (Charthouse School, 1921-3 and Loughborough, 1923) have tierces that on average are equal-tempered, though some are sharp and some are flat.
What was lost in this change was the consistency of tierce tuning across all their output that had had been achieved in the previous 25 years. I suspect this is due to two things. First, all the strickles may not have been remade at once - Whitechapel converted their strickles slowly over a period of time when they moved to true-harmonic tuning. Secondly, the recordings I have in this period are a mixture of ringing peals and carillons and they may have continued to use the old strickles for some change-ringing bells.
Recordings investigated from the 1920s while this change was in process in addition to the Charterhouse and Loughborough carillons include Sutton on Trent (1922), Fairfield (1923), Swindon and Melksham (1924), Tavistock, Belper and Norbury (1925), Worcester (1928/9, replacements for the Grimthorpe bells), Bideford (1929), and others.
There is a significant trend in the chart during this period of flattening of the tierces, especially from the late 1960s onwards. When I first observed this, I guessed that it might be due to strickle wear, and this was confirmed in a discussion with Taylors a number of years ago. When the cope and core moulds are produced for a bell, the metal strickles are abraded by the moulding sand. Due to wear, the shape of the bell as cast changes slightly, and this lowers the tierce. The picture shows how a strickle is used to shape the moulding sand.
Bells included in this portion of the graph include Boston (1932), Albury (1935), Shalford (1936), Winchester (1937), Manchester Town Hall ringing peal and carillon (1937, replacements for another Grimthorpe peal), the heavy chime at Bradwell (1938), Doncaster and St Brides Major (1939), Valetta St Dominic (1948), Evesham (1951), Little Chart and Cambridge St Andrew (1956), Wolvercote (1957), Farnham (1959), Tewkesbury and Northallerton (1962), Newcastle carillon (1963), East Retford (1968), Houston (1971), Chester (1973), Durham St Oswald (1976/7), New Castle Delaware (1981), Tunstall (1995), Dorking (1998), and some new bells plus trial bells for the carillon at Brielle (1999) cast to Hemony profiles with meantone tierces. All of these apart from the Brielle carillon have equal-tempered nominals.
Between 2000 and 2015, work was clearly done to remake the worn strickles and the tierces sharpen. Bells analysed in this period include Chilcompton (2000 – 2008), New York (2006), Cambridge Great St Mary (2009), Worcester Old St Martin and Cornhill (2011), and the new bells at Guildford St Nicolas (2013).
The most recent Taylor peals investigated are Bampton, Godalming and Ypres (2017) and Grappenhall (2019. In these peals, the tierces are very much sharper. I understand that Taylors are now once again casting bells using their designs from the early 1900s.
It was once suggested to me, perhaps not very seriously, that Taylor bells from the early 20th century sounded good because they were tuned with forks, not electronic equipment. This investigation shows instead that the particular bell shapes used creates the special sound of peals from this period.
As ever I must thank the tower captains who have allowed me to record their bells, and the enthusiasts who post bell videos to YouTube. Thanks also to Will Bosworth who helped edit this article.
On Bell Tones: Canon A B Simpson: 1895 and 1896, available on this website: first paper and second paper.
A Rudimentary Treatise on Clocks, Watches and Bells: Edmund Beckett, Lord Grimthorpe: 1903 available at the Gutenberg Project.
Two Hundred Years of History: Paul Taylor: 1960 available on this website: Taylor history.
Master of my art: The Taylor bellfoundries 1784-1987, Trevor Jennings, John Taylor and Company (Bellfounders) Ltd, 1987, ISBN 978-0951198810
Bells and Bellfounding: A History, Church Bells, Carillons, John Taylor & Co, Bellfounders, Loughborough, England, Michael J. Milsom 2017, ISBN 978-1547239153
Great Bookham, Surrey
Last updated November 15, 2019. Site created by Bill Hibbert, Great Bookham, Surrey