The Sound of Bells

Home Page Next Previous Up

Simpson's first paper

This paper, as published in The Pall Mall Magazine, was quite a handsome affair, and included a number of pictures which have not been included below. They were: A Bell by Messrs. Taylor, of Loughborough; St Mary's Church, Stoke-by-Wayland (sic); Stoke-by-Wayland (sic) Church, Suffolk; Church of SS. Peter and Paul, Lavenham; Peal cast at Louvain for Church of Lower Beeding, Sussex; Church of S Picoue, Ghent; The Belfry of Ghent; and a bell from Mitford, with the following caption: This bell, which formerly hung in the church of Mitford, in Northumberland, is of unknown, but certainly of great antiquity. It is very interesting as marking a distinct stage in the evolution of the present form of bells. The chief points of interest are: (1) The height is EQUAL to the width - instead of four-fifths or less; and (2) The bell, instead of terminating in an OUTER cusp, terminates in an INNER one; so that, as seen in the illustration, it stands on its INNER rim.

The text of the paper is reproduced in its entirety below.


The Pall Mall Magazine

Vol. VII


pp. 183 - 194


The present writer has nothing to say on the ancient history of bells, nor will he attempt to make any addition to the many pretty things which have been said as to their sentimental power. It is simply to place on record certain facts which have come under his notice, during a course of observation extending over many years, and which he has reason to think would prove interesting to many of the readers of this Magazine.

We have the bells with us everywhere, and few people with musical ears have not, at one time or another, amused themselves, and (we will venture to suggest) puzzled themselves, in attempting to determine accurately the notes of their own church bells.

Many of us, also, have been struck by the apparent want of harmony in the famous carillons of Bruges and other Belgian towns; and some few have been at great expense to set up carillons of their own, and have been reluctantly driven to the vexatious conclusion that they are "painfully out of tune."

To all these we think we have that to say which will interest them. And we are not without hope that, through their influence, our bell founders and tuners may be roused to study their work more closely to try to understand better what was the purpose of the original designers of the present form of bell, and endeavour to fulfil that purpose more nearly than they have done in the past.

We begin by boldly asserting, as the result of a pretty wide experience, that there is probably not a single bell in England that is really "in tune with itself," and almost certainly not a single "peal" of bells that are properly in tune with each other.

We do not say that there are not many peals which are in excellent tune as to the most important note in each, and their general musical effect very pleasing. But eve do assert that the best of these might be much better; and, in the majority of cases, the irregularities we complain of are such as seriously to mar their musical effect, and such as ought to be, and might be, avoided by more intelligent founding, or (in most cases) rectified by more intelligent tuning after founding.

Now, this whole matter turns on the meaning of the expression "in tune with itself." Most people have an idea that every bell has one prominent, unmistakable note which characterises it, and as to whose pitch no two people with musical ears could differ. Thus, in the article on Bells in the latest edition of the Encyclopaedia Britannica, we read the following: "A good bell when struck yields one note, so that any person with an ear for music can say what it is."

If for "a good bell" we may read "a bell in good tune," this statement is true; but, as it stands, it condemns as "not good" some of the finest and best bells in the world.

One example, and that a notable one, will suffice at present to support us in this.

In the Times of July 20th, 1887, there is an account of the "inauguration" of the great bell "Gloriosa," made out of French cannon and hung in the Cathedral of Cologne. The account concludes thus: "The opinions of experts are divided as to whether the note which the bell sounds is C sharp or D."

We feel sure that many of our readers have felt a similar difficulty in determining the note of a familiar bell.

Now, to account for this, and to clear the way for further observations, we must understand what is the true "theory" of a bell, if we may be allowed the expression.

It would surely be unreasonable to suppose that the very peculiar form of bell which (with slight modifications) has been preserved for so many hundreds of years, both here and on the Continent, was adopted without the deliberate purpose of ensuring that the various tones and subtones of each bell should be in some fixed musical relation to each other.

What is that relation?

We make bold to suggest that it is this: Every true bell should give out, when fairly struck, a fundamental note or "tonic," its third, fifth, and octave above, and its octave below, thus sounding the full chord - do, mi, sol, do, with the bass do below.

This is the "theory" which was, we are satisfied, before the minds of the original designers of the present form of bell. Almost forgotten (if ever realised) by many of their successors, it is still recognised by some, and irresistibly forced upon the acceptance of those who, like the present writer, have made a study of the tones of bells as they are.

The following extract from the article on Founding of Bells in the Encyclopaedia Britannica, 5th Edition, 1815, is of importance as showing that such a theory was recognised in this country not so very long ago, though, it is true, there is no reference to the lowest note we have spoken of.

"The height of the bell in proportion to its diameter (is) as 12 to 15, or in the proportion of the fundamental sound to its third major: whence it follows (?) that the sound of a bell is principally composed of the sound of its extremity, or brim, as a fundamental - of the sound of the crown, which is an octave to it - and of that of the height, which is a third."

But now, to bring this paper within reasonable limits, we must dismiss all consideration of thirds and fifths, and confine our attention to the three more important notes - i.e., the tonic, its octave above, and its octave below. For convenience' sake, and for reasons which will appear further on, let us call the first of these the "fundamental," the second (or octave above) the "nominal," and the third (or octave below) by the name by which it is known in English foundries, the "hum-note."

If, then, a bell corresponded to its "theory," these three would sound the same note, in three consecutive octaves, and the bell would, so far, be "in tune with itself."

"But, alas! where shall we find such a bell? Whatever the cause may be - whether founders, in ignorance or indifference as to the importance of having these notes in accord, have, (1) for convenience of ringing, altered the original proportions of bells, or, (2) to obtain greater power, put more metal into them - certain it is that it is quite the exception to find a bell, which has any two of these notes in unison, and we have not yet met with one in which all three were in accord.

By far the commonest state of things is this: The "fundamental" is almost always the flattest of the three - irrespective, of course, of octave. The "hum-note" is almost always the sharpest, and the "nominal" generally between the two. Thus, if the nominal of a bell is C, the fundamental will probably be somewhere between C and B in the octave below, while the hum-note will probably be between C and C sharp in the octave below that. (It is not unusual for the hum-note to be much sharper than this.)

In support of this statement let us take a few examples. Take first the peal at Terling in Essex, which consists of five bells in the key of F sharp, by five different makers, and of various dates, covering a period of 240 years. This is an excellent example, as, from the variety of makers and dates, any general characteristics that we may observe cannot be considered as peculiarities of "time" or "foundry." And it has further this great value, that the tones of these bells have been carefully analysed by Lord Rayleigh, and tabulated in his most valuable paper "On the Tones of Bells," printed in the Philosophical Magazine for January 1890. An examination of these tables gives the following results.

1. In the first three bells, including the oldest and the newest, the tones follow just the rule which we have called the "common" one - i.e., they are nearly in octaves, but the fundamental is the flattest, the nominal sharper, and the hum-note the sharpest. In the fourth bell, the fundamental and the nominal are true octaves and the hum-note is sharper by a long semitone. In the tenor, the fundamental and the hum-note are true octaves, and the nominal a semitone flatter (very unusual).

We claim this peal as a powerful witness to the truth of our position - (1) that the fundamental, nominal, and hum-note were meant to be in octaves, and (2) that, as a matter of fact, it is the exception to find a bell in which any two of them are in accord.

2. Take next the little peal of six bells in the church of Fittleworth, Sussex. Of these, three are new, the fourth is about fifty years old, the fifth and sixth very ancient. In each of the six, without exception, the nominal is a quarter of a tone, more or less, sharper than the fundamental. And, in all but one, the hum-note is a trifle sharper than the nominal; the exception being the tenor, a very ancient bell, in which the hum-note and the nominal are true octaves. Can we help feeling that the general small defection from perfect octaves is an error from a design which is fulfilled in the exceptional case?

3. In the tower of Eastry in Kent, are five bells, the tenor about a ton in weight. In all these the common rule, as stated above, holds good, with the exception that, in the fourth bell, the nominal is a shade flatter than the fundamental - an unusual case, but still tending to confirm the theory that these notes were meant to be true octaves.

4. There is a very fine peal of six bells in the tower of Stoke-by-Nayland, in Suffolk. Every one of these, with possibly one exception, follows the same rule: the nominals, fundamentals, and hum-notes being nearly in octaves, the hum-notes being the sharpest, the fundamentals the flattest, and the nominals between the two. The possible exception is the treble, which is slightly cracked, and the hum-note not observable.

5. One of the finest peals of eight bells in the kingdom is that of Lavenham, in Suffolk. Here also every one, with one notable exception, follows the same general rule of slightly imperfect octaves between the three principal notes ; the errors in each bell being of the same kind - i.e., all the hum-notes a little sharper, and the fundamentals a little flatter, than their respective nominals. The one exception is the tenor, a marvellously fine-toned bell, which has, indeed, the reputation of being about the finest in England. In this, the fundamental and the nominal are apparently in perfect octave-a very noticeable fact, taken in connection with the reputation of this bell.

These are all instances of English bells. But, to guard against the supposition that these coincidences, or irregularities, are peculiar to our bells, we will mention, lastly, a peal of eight bells cast at Louvain about eight years ago, and placed in the tower of Lower Beeding, near Horsham, in Sussex. We had an opportunity of examining these bells on their arrival in this country. There was much to learn from them, which we may hereafter refer to. But with respect to the point now before us - viz., the relative positions of the three principal tones - there was nothing to distinguish them from an ordinary English peal. There was the usual approximation to octaves, but we only noticed one instance in which the relation was true. In the seventh bell, the fundamental was a true octave below the nominal. But it was evident that this bell had been greatly altered, and we have reason to feel sure that, originally, the fundamental had been flatter than the nominal, as in most English bells, and as was, and is, the case with the tenor bell by its side.

If now we have carried our readers with us so far, they will no longer be surprised at any difficulty they may have experienced in determining the note of any particular bell. For here we see that, in the majority of bells, we have three notes, very near to each other in pitch, though in different octaves, all struggling for the mastery, and each able - let us here say - under given circumstances, to assert its supremacy. Is it any wonder that even a skilled musician, if not learned in bell tones, should be in doubt as to the true note of a bell submitted to him? Is it so very surprising that "experts differed as to whether the note of the 'Gloriosa' bell was C sharp or D"?

We have not a doubt that the explanation of this is, that the sharper tone heard was the nominal, and the flatter the fundamental: this great bell following, in this respect, the common rule which we have so fully illustrated above.

"But" - we imagine our musical readers exclaiming - "if this be so, how is it that the sounds of bells are even tolerable? Any other instrument which gave forth simultaneously, e.g., C, with a rather flat C sharp above and a rather flat D below, would be unbearable, - and the succession of a series of bells of this imperfect character would surely produce nothing but a hideous noise."

So one might think; but it is not so. For reasons which satisfy the learned in acoustics, the difference in quality of these sets of tones is such that they do not interfere with one another, so as to give the sense of discord, which we should expect. The tones of nominals, fundamentals, and hum-notes, seem to move, as it were, in three separate spheres. And though any discord, between at least the nominal and fundamental, in a bell cannot but seriously detract from the purity and fullness of its tone, it does not produce "beats," nor affect the ear painfully. Consequently, there are thousands of bells, having their principal tones quite out of tune with each other, which, considered individually, are good and pleasing. They might be made much better, perhaps; but as long as they stand alone, no alteration is necessary in order to satisfy the ordinary musical listener. You may listen to whichever tone you like - sometimes to one, sometimes to another; all are pleasing, and no one interferes with another.

But when it comes to peals. and carillons, the case is different. In order to get any really musical effect, it is absolutely necessary, that some one, at least, of the sets of tones in the series of bells should be properly in tune with itself - i.e., all the nominals in tune with each other, or, similarly, all the fundamentals, or all the hum-notes. This, it is evident, is the very least that can be accepted. And, as a matter of fact, the choice is more restricted than this; for, important as are the hum-notes - far more important, as we are prepared to maintain, than either foreign or English founders seem to think - we are quite ready to allow that, unless all three sets are to be made to move in true octaves (which is devoutly to be wished, but is more than can at present be demanded), the "hum-note" set must be neglected, in comparison with either of the other two.

At this point, therefore - again to "lighten the ship" - we drop all reference to the hum-note till a more favourable time, and confine our attention to nominals and fundamentals.

We have then before us, let us suppose, a series of bells, direct from the foundry, all of the common character - i.e., with the fundamental in each bell more or less flatter than its nominal - and, as is sure to be the case, with neither its nominals nor its fundamentals quite in tune with each other. Now, what shall we do? Plainly, the one really satisfactory operation, if it could be done, would be to bring the nominal in each bell into unison with its own fundamental, and then to tune each bell, so rectified, to its neighbour. This would indeed be "something like" tuning, and we are prepared to maintain that, in all ordinary cases, it can be done.

But, as a matter of fact (speaking generally) no one attempts such a thing. What is done is this: the tuner (consciously or unconsciously) selects one of the two sets of notes (either the nominals or the fundamentals), puts them in proper tune with each other, and leaves the other set untuned, either to the other set, or to each other.

To be more explicit - and we now ask the attention of our readers, especially of those interested in carillons, to what we now state - "The Englishman tunes the nominals, and neglects the fundamentals; while the foreigner tunes the fundamentals, and (comparatively) neglects the nominals."

Now, this difference of practice is a very remarkable fact, which we have never seen referred to in any of the many treatises on bells, nor have we ever met with any one who seemed to be aware of it. Yet it is as certain as it is interesting and important.

1. That the Englishman tunes a peal by the (so-called) "nominals" is shown beyond question by Lord Rayleigh in the paper referred to. The Terling peal is pronounced by English bell-experts to be in the key of F sharp. Well, the upper series of notes in those bells is in that key; and indeed we call this note the "nominal," because the Englishman names each bell according to the pitch of this note in it. The fundamentals of this peal form no musical series at all, and evidently have not engaged the attention either of the founder or the tuner.

But any peal of English bells will prove the same point. If the bells are what an English tuner calls " in tune," you will find that it is the nominals that are in tune, and not the fundamentals. We have often tried to call the attention of professional tuners to the fundamental of the bell they were tuning, but they invariably treated it with indifference. And there is, further, this curious fact: that while a tuner always gave the nominal as the note of any Bell, he invariably gave the pitch an octave lower than it really was. We have, for many years past, lost no opportunity of calling the attention of founders and tuners to these things, but we know not with what effect.

2. That the foreigner, and in particular the Belgian, takes the fundamental as the principal note, and tunes the bells by it, is equally certain and unquestionable. The very fact that he calls it, as he does, the "fundamental" is almost conclusive. And no one can visit a Belgian foundry and engage in discussion about the bells without being convinced on this point. But we have clearer proof than this. The Belgian peal at Beeding was pronounced by the founders to be correctly in tune, and the seventh bell, in particular, they declared to be exactly in tune with the eighth. So it was, as to the fundamental; but, as regards the nominal, it was very flat, and the whole peal, generally painfully out of tune to the English ear.

Moreover, the internal form of foreign bells is such as to indicate how to sharpen them. But the effect of so doing is to sharpen the fundamental, without altering the nominal at all, or very partially.

We could adduce much more evidence on this point, and, on the whole, we feel safe in asserting what we have stated, as to the foreign method, to be true.

The foreigner is, indeed, quite aware of the existence of the nominal, and, to hear him talk, you might think that he brought both sets into unison. But, except perhaps in very large bells, he really does not do so, as any one with an ear may judge, from the inharmonious character of Belgian carillons, as well from such tangible examples as the Beeding bells.

One point more remains to be cleared up, before we can arrive at the practical conclusion we are seeking to reach. The foreigner tunes by the fundamental, the Englishman by the nominal: which is right?

A direct unqualified answer to this question is, as might be expected, impossible. Both are right in their way. But there is this difference: that while it never can be allowable to neglect the nominals, the fundamentals may be neglected in English ringing peals, and in the upper bells of carillons, not without some loss of purity and fulness of tone, but without painful injury to the harmonious effect. The reason for this distinction will appear from the following consideration.

When bells are struck at considerable intervals of time, most persons would be apt to take the fundamentals as the notes of the bells, on account of their full and persistent character. But the case is different when one bell follows another in rapid succession. At the instant of striking, the keen sound of the (higher) "nominal" is most perceptible; and, if followed immediately by another, there is no time for the "fundamental" to force itself into prominence, and so the car keeps following the nominals all through. This would be the case with an English peal, in which the bells follow one another very rapidly. And so also with the higher bells of a carillon, which take the "air." Consequently the nominals must be put into tune with each other in these cases. And it is the want of attention to this necessity which is the cause of the lack of harmony observable in foreign carillons.

But with the lower Bells of a carillon, the case is different. These strike, generally, at longer intervals, and the fundamental has time to assert itself, and to remain, so to say, master of the field. It therefore becomes necessary, in these bells, to tune the fundamentals also; and, of course, to bring them into unison with their respective nominals.

To sum up, then (omitting all reference to the hum-note) -

1. It is essential that all the nominals, throughout any peal or cannon, should be in tune with each other.

2. It is very desirable, in a carillon, that at least all the heavier bells (say, those above 7 cwt.) should have their fundamentals also brought into unison with their respective nominals.

3. It is best, in all cases, to bring the fundamental of each bell into true octave with its nominal, and then to tune the whole series of bells, so rectified, to each other.

Can this be done? We suggest, with some confidence, that in all ordinary cases, it can. But this is another question, which we are not concerned to deal with in this paper. Bell founders and tuners naturally do not care to take any more trouble than is necessary in order to satisfy the public. It has been our object to move the public to move the tuners. If we succeed in that, we feel sure that the tuners will find out how to satisfy the demands made upon them, so far as the nature of the case admits.

Nevertheless, if we thought that the publication of such knowledge as we possess - as to the possibilities and methods of Bell-tuning - would tend at all to the perfecting of the art in this country, we should be glad to return to the subject.



Last updated July 25, 2000. Site created by Bill Hibbert, Great Bookham, Surrey