# A pendulum theorem

> An annotated edition of this paper on Fermat's Library, where scientists and readers explain its ideas, equations and history in the margins. Annotated edition: https://fermatslibrary.com/s/a-pendulum-theorem

- **Authors:** D. J. Acheson
- **Topic:** Physics
- **Original:** https://s3-us-west-1.amazonaws.com/fermatslibrary/uploads%2F1457913142137-vvikbgaowllf3iw6-a80063d7690f2ae8e63f87ca1b1e1217%2FA_Pendulum_Theorem.pdf
- **Pages:** 7
- **Annotations:** 8 comments in 8 threads
- **Featured:** Fermat's Library Journal Club
- **Annotated edition:** https://fermatslibrary.com/s/a-pendulum-theorem

## Annotations

### Annotation 1 (page 1)

Beside the passage: “BY D. J. ACHESON Jesus College, Oxford OX1 3DW, U.K.”

**Annotation** by Luís Batalha (2016-03-20):

> This paper was written by D.J. Acheson while he was trying to give a new twist to an old problem in dynamics, first studied by Daniel Bernoulli in 1738. He had considered a hanging chain of several linked pendulums, all suspended from one another, and discovered various different modes of oscillation.
>
> Acheson proved that it was possible to take all these linked pendulums, turn them upside-down, so that they are all precariously balanced on top of one another, and then stabilise them in that position by vibrating the pivot up and down.
>
> He started calling this gravity-defying experiment "Not quite the Indian rope trick" because of a famous magic trick performed in India during the 19th century. During this trick, a magician would hurl a rope into the air. The rope would stand erect, with no external support. His boy assistant would climb the rope and then descend.
>
> Although is theorem doesn't hold for the case $N\rightarrow \infty$ (a perfectly flexible string can be approximated by a series of infinite pendulums all linked together), thus not explaining the Indian rope trick it is still a surprising mathematical result that has a certain magic on its own!
> ![](http://www.stageillusion.com/news/wp-content/uploads/Indian-Rope-Trick.jpg)

### Annotation 2 (page 1)

Beside the passage: “Stephenson (1908a, b) showed that it is possible to stabilize a single rigid pendulum in its inverted, or upside-down, equilibrium position by subjecting the pivot to small vertical oscillations of suitably high frequency. He confirmed his theoretical predictions by a practical demonstration of the phenomenon. While …”

**Annotation** by Luís Batalha (2016-03-20):

> In 1908 Stephenson found that the upper vertical position of the pendulum might be stable when the driving frequency is fast. Yet no one could scientifically answer this unusual and counterintuitive phenomenon until almost fifty years later, a Russian physicist Pyotr Kapitza successfully analyzed the stability of this system by separating the motions into fast and slow, and introduced a new concept called Effective Potential (to read more about Kapitza's demonstration click [here](https://en.wikipedia.org/wiki/Kapitza%27s_pendulum))
>
> Once Kapitza found the effective potential, he then use it to analyze the motion and stability of the system.
> Systems always tend to move towards positions of minimum potential energy, so by finding the minimum
> values of the potential energy expression, we can find the stability points. In particular Kapitza found that the vertical position was stable.

### Annotation 3 (page 1)

Beside the passage: “t This paper was accepted as a rapid communication. Proc. R. Soc. Lond. A (1993) 443, 239-245 ? 1993 The Royal Society Printed in Great Britain 239”

**Annotation** by Julian Irwin (2016-03-23):

> This paper is from 1993! I had thought it was much older on my first read through.

### Annotation 4 (page 2)

Beside the passage: “? Figure 1. Typical region of linear stability for an upside-down N-pendulum, as given by (2.1). Here * = 0.450 g/Oax and co = 3.1432 ax/(min. The sketches indicate theoretical predictions of the behaviour just outside this stability region. Thus, if at a given frequency w the amplitude e is just too large, the system …”

**Annotation** by Luís Batalha (2016-03-20):

> As we can see in this video by Steve Mould the chain of pendulums can be balanced if the pivot is vibrated up and down by a small enough amount and at a high enough frequency.
>
> [![IMAGE ALT TEXT](https://slack-imgs.com/?url=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fgnn21smGVrQ%2Fhqdefault.jpg&width=400&height=300)](https://youtu.be/gnn21smGVrQ?t=56 "Inverted Pendulum")

### Annotation 5 (page 2)

Beside the passage: “Xi+? Xi,=0, i=l,...,N. (2.2)”

**Annotation** by Luís Batalha (2016-03-19):

> To prove this we are going to use the Lagrangian formulation and will start with the $N=2$ case (two coupled pendulums).
>
>
> ![](http://i.imgur.com/FPrkfpN.png?1)
>
>
> The Lagrangian $L$ of the system is given by $L=T-V$, where $T$ and $V$ are the kinetic energy and the potential energy of the system.
>
> $$
> T=\frac{1}{2}m_1(x_1^2+y_1^2)+\frac{1}{2}m_2(x_2^2+y_2^2) \\
>
> V= m_1gy_1+m_2gy_2
> $$
>
> For small angles ($\theta_i\approx 0 \rightarrow \cos(\theta_i)\approx 1$ and $\sin(\theta_i)\approx \theta_i$ ) we can write $T$ and $V$ as a function of the angles:
>
> $$
> T\approx \frac{1}{2}m_1l_1^2\dot \theta_1^2+\frac{1}{2}m_2l_1^2\dot \theta_1^2+\frac{1}{2}m_2l_2^2\dot \theta_2^2+m_2l_1l_2\dot \theta_1\dot \theta_2 \\
> V\approx \frac{1}{2}(m_1gl_1\theta_1^2+m_2gl_1\theta_1^2+m_2gl_2\theta_2^2)
> $$
>
> If we calculate the Lagrange equations $\frac{d}{dt}\left(\frac{\partial L}{\partial \dot \theta_i}\right)=\frac{\partial L}{\partial \theta_i}$ we end up with a system of coupled differential equations of the form:
>
> $$
> \left( \begin{array}{ccc}
> \frac{l_1}{g} & \frac{m_2l_2}{(m_1+m_2)g} \\
> \frac{l_1}{g} & \frac{l_2}{g} \\
>  \end{array} \right)\ \left( \begin{array}{ccc}
> \ddot \theta_1  \\
> \ddot \theta_2  \\
>  \end{array} \right)\ = - \left( \begin{array}{ccc}
> \theta_1  \\
> \theta_2  \\
>  \end{array} \right)\
> $$
>
> It is helpful to find a new basis in which the matrix $M$ is diagonal. The basis that diagonalizes $M$ is what we call the normal coordinates. In this basis we end up with a set of homogeneous second order linear differential equations, each of which corresponds to the motion of a simple harmonic oscillator (SHO).
>
>
> $$
> \left( \begin{array}{ccc}
> \frac{1}{w_1^2} & 0 \\
> 0 & \frac{1}{w_2^2} \\
>  \end{array} \right)\ \left( \begin{array}{ccc}
> \ddot X_1  \\
> \ddot X_2  \\
>  \end{array} \right)\ = - \left( \begin{array}{ccc}
> X_1  \\
> X_2  \\
>  \end{array} \right)\
> $$
>
> where $X_1$,$X_2$ are the normal coordinates and $w_1$,$w_2$ are the natural frequencies. Finally we get
>
> $$
> \ddot X_i+w_i^2 X_i = 0  \\ i=1,2
> $$
> The generalization for $N$ pendulums is a similar process that just involves calculating a more complex lagrangian and dealing with a bigger matrix, but in the end we always end up with a diagonalized matrix if we use the normal coordinates.

### Annotation 6 (page 2)

Beside the passage: “Now, each of the quantities ()2 will be simply proportional to g; this follows from the standard theory of small oscillations based on Lagrange's equations of motion (see, for example, Landau & Lifschitz 1976), given that the uniform gravitational field g is the sole source of potential energy in this problem. If we …”

**Annotation** by Luís Batalha (2016-03-20):

> To prove that the natural frequencies are proportional to $g$ we start again with the double pendulum and with the matrix:
>
> $$
> \left( \begin{array}{ccc}
> \frac{l_1}{g} & \frac{m_2l_2}{(m_1+m_2)g} \\
> \frac{l_1}{g} & \frac{l_2}{g} \\
>  \end{array} \right)\ \left( \begin{array}{ccc}
> \ddot \theta_1  \\
> \ddot \theta_2  \\
>  \end{array} \right)\ = - \left( \begin{array}{ccc}
> \theta_1  \\
> \theta_2  \\
>  \end{array} \right)\
> $$
>
> Since we expect oscillatory solutions let's start with the following ansatz:
>
> $$
> \theta_1 = A_1e^{iwt} \\
> \theta_2 = A_2e^{iwt}
> $$
>
> If we now substitute $\theta_1,\theta_2$ with the expressions
>
>
> $$
> \left( \begin{array}{ccc}
> \frac{l_1w^2}{g}+1 & \frac{m_2l_2w^2}{(m_1+m_2)g} \\
> \frac{l_1w^2}{g} & \frac{l_2w^2}{g}+1 \\
>  \end{array} \right)\ \left( \begin{array}{ccc}
> A_1  \\
> A_2  \\
>  \end{array} \right)\ =  0
> $$
>
> For this to be true for any amplitude, the matrix on the left must be singular i.e. must not be invertible, such that one cannot multiply both sides of the equation by the inverse, leaving the right matrix equal to zero. It follows that the determinant of the matrix must be equal to 0, so:
>
> $$
> \left(\frac{l_1w^2}{g}+1 \right)\left(\frac{l_2w^2}{g}+1 \right)-\frac{l_1w^2}{g}\frac{m_2l_2w^2}{(m_1+m_2)g}=0
> $$
>
> And finally:
>
> $$
> w_1=\frac{\sqrt{g/l}}{\sqrt{1-\sqrt m}} \\
> w_2=\frac{\sqrt{g/l}}{\sqrt{1+\sqrt m}}
> $$
> which means that $w_i^2 \propto g$

### Annotation 7 (page 3)

Beside the passage: “(2 1222/ 2 1 < e/g < 0.450, i= 1, ...,N,”

**Annotation** by Luís Batalha (2016-03-20):

> Near $\alpha = 0$, $ \sqrt{-2\alpha} <\beta <0.450$. Since $\alpha_i=\frac{-w_i^2}{w_0^2}$, $ \sqrt{2\frac{w_i^2}{w_0^2}} <\beta <0.450$

### Annotation 8 (page 6)

Beside the passage: “Suppose finally that L is fixed. Then, the larger the value of N, the smaller we must take e, and, consequently, the larger we must take to if stability is to be achieved. The limit N-* oo for fixed L corresponds to a perfectly flexible string, and was of major interest in the original investigations of Bernoulli and …”

**Annotation** by Luís Batalha (2016-03-20):

> Also the larger the value of $N$, the smaller the value of $l$ because $L$ is a constant, which means that it will push $w_0$ and $\epsilon$ out of the stability zone.

## Full text

Extracted from the PDF. [Annotation N] marks where each annotation above sits.

#### Page 1

A pendulum theoremt

BY D. [Annotation 1] J. ACHESON Jesus College, Oxford OX1 3DW, U.K.

We consider N linked pendulums which are inverted and balanced on top of one another, and establish a general theorem which shows how they may be stabilized by small vertical oscillations of the support.

1. Introduction

Stephenson (1908a, b) showed that it is possible to stabilize a single rigid pendulum in its [Annotation 2] inverted, or upside-down, equilibrium position by subjecting the pivot to small vertical oscillations of suitably high frequency. He confirmed his theoretical predictions by a practical demonstration of the phenomenon. While this is a well-known curiosity of classical mechanics it does not seem to be generally known that an inverted double, or even triple pendulum can be stabilized in the same way. This was, again, first predicted theoretically by Stephenson, in a comparatively overlooked paper of 1909, though the idea has reappeared in a number of subsequent studies (Lowenstern 1932; Hsu 1961; Kalmus 1970; Otterbein 1982; Leiber & Risken 1988). Here we present a simple but general theorem on the linear stability of an inverted N-pendulum of any kind. The generality of the theorem is achieved by relating the stability question to just two elementary properties of the system as a whole when it is in its non-inverted, or downward-hanging, state.

2. The stability of upside-down pendulums Theorem. Let N pendulums hang down, one from another, under gravity g, each having one degree of freedom, the uppermost being suspended from a pivot point O. Let O)max and wmin denote the largest and the smallest of the natural frequencies of small oscillation about this equilibrium state. Now turn the whole system upside-down. The resulting configuration of the pendulums can be stabilized (according to linear theory, at least) if we subject the pivot point 0 to vertical oscillations of suitable amplitude e and frequency (o. When (O2 > o)ax the stability criterion is

/V2g/oo tmin < e < 0.450g/2ma. (2.1)

Note. When several pendulums are involved, T2ax is typically much greater than 2i. The condition o2 > Omax is then necessary for the stability of the inverted state, so (2.1) then gives the whole stability region in the e-o)0 plane, as in figure 1.

t This paper was accepted as a rapid communication. Proc. R. Soc. Lond. A [Annotation 3] (1993) 443, 239-245 ? 1993 The Royal Society Printed in Great Britain 239

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM All use subject to JSTOR Terms and Conditions

#### Page 2

8*

? Figure 1. Typical region of linear [Annotation 4] stability for an upside-down N-pendulum, as given by (2.1). Here * = 0.450 g/Oax and co = 3.1432 ax/(min. The sketches indicate theoretical predictions of the behaviour just outside this stability region. Thus, if at a given frequency w the amplitude e is just too large, the system is unstable to rapidly growing buckling oscillations at frequency -o0. If, on the other hand, e is a little too small, the pendulums fall over comparatively slowly, keeping to the same side of the vertical.

Proof. Consider first small disturbances to the system about the original, i.e. noninverted equilibrium state, the pivot point 0 being fixed. Let the natural frequencies be oi, and let Xi be the corresponding normal coordinates, each of which will be some linear combination of the (small) angles which the pendulums make with the downward vertical. We then have

Xi+? [Annotation 5] Xi,=0, i=l,...,N. (2.2)

Now, each of the quantities ()2 will be simply proportional to g; this follows from the standard [Annotation 6] theory of small oscillations based on Lagrange's equations of motion (see, for example, Landau & Lifschitz 1976), given that the uniform gravitational field g is the sole source of potential energy in this problem. If we tackle the stability of the inverted state by changing the sign of g we then find that small disturbances are governed instead by

Xi-o Xi i=0, i==l,...,N. (2.3)

Suppose now that the pivot point 0 oscillates up and down so that its coordinate in the direction of the upward vertical is h = -ecoso0t. We may allow for this simply by replacing g in the above argument by apparent gravity g +h = g + e() cos o0 t. In this way we find that

X?--0~(? 1 + (eo)/g) coS ot)Xi = 0, i = 1,...,N.

On introducing the scaling T = (ot we therefore obtain N uncoupled Mathieu equations:

d2Xi/dT2 + (ci + ficosT)Xi = 0, i = 1,...,N, (2.4)

where i =-2/o2, /=-i =- /g. (2.5) Proc. R. Soc. Lond. A (1993)

D. J. Acheson 240

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM

All use subject to JSTOR Terms and Conditions

#### Page 3

11

1_ 01 ( 4 4 Figure 2. Stable (shaded) and unstable regions for the X= 0 solution of Mathieu's equation d2X/dT2 + (a + fP cos T) X = 0. The points (2.6) lie on the straight line OP, and for the inverted state to be stable they must all lie within the shaded region with a < 0, in the manner indicated (for N=2).

Now, the stability regions of Mathieu's equation are well known and indicated in figure 2. We achieve stability of our inverted system if, by choosing e and ow appropriately, we can place all the points

2(-/2, 6g), i = .. ,N (2.6)

in the shaded part of the region a < 0. To see that this can indeed be done it is simplest to pass directly to the highfrequency limit, i.e. 02 > 02ax so that w0)/02 is small for all i = 1 ...V,N. Now, when lal is small the stability boundary which passes through the origin is given asymptotically by a = _fl2 (see, for example, Jordan & Smith 1987, p. 257). Moreover, the upper boundary to the shaded region in a < 0 is known to meet the a = 0 axis at I0fI = 0.450. Thus all the points (2.6) will be in the stable region if

(2 1222/ 2 1 < [Annotation 7] e/g < 0.450, i= 1, ...,N,

i.e. if /2g/oi < e < 0.450g/0 (2.7)

for all i = 1,...,N. This is ensured by (2.1). While this completes the proof, it is worth considering a more general and geometric approach to the matter. All the points (2.6) lie on a line through the origin with slope -eo0/g. Imagine, then, that we attach an elastic band OP to the origin in figure 2, stretch it out, and mark points along it so that their distances from the origin are in the proportion o1' : : ...: (o. By choosing e and w0 we may vary the length of the band, and its slope - eO2/g, at will. Now, if )max and o)min are almost equal, so that the marked points are tightly clustered along the band, we may evidently steer those points into the stable region in figure 2 even if c1a is not particularly small, so that 02 is not large compared with 2ax. If, on the other hand, COmax is substantially greater than omin - as will typically be the case - we may still steer all the points into the stable region, but only by moving the band very close to the a = 0 axis in figure 2 and by shrinking it accordingly. The stability criterion (2.1) will then apply, and the region of stability in the e-oo plane will be as indicated in figure 1. Proc. R. Soc. Lond. A (1993)

A pendulum theorem 241

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM

All use subject to JSTOR Terms and Conditions

#### Page 4

D. J. Acheson

3. Examples (a) Stephenson's system inverted Stephenson's original double-pendulum analysis was for two identical rods of uniform density and equal length 1. An elementary exercise shows that the natural frequencies are then

Wmax,min = (3 +6//7)1 (g/l), (3.1) and (2.1) becomes

(1.653/o) (g/l) < e/l < 0.085. (3.2) The first of these inequalities may be written as

(1ew6)2 > 0.683gl, (3.3) which is Stephenson's original criterion, derived by a different method. His formula for the inverted triple pendulum appears, unfortunately, to be in error; the theorem of ?2 gives

(1.738/0o) (g/l) < 6l1 < 0.03 (3.4) as the high-frequency stability criterion for an inverted triple pendulum consisting of three equal heavy rods.

(b) The point-mass double pendulum Consider now a double pendulum consisting of two point masses ml, m2 and two light rods of equal length 1 (as sketched in figure 1). The normal modes of small oscillation about the lower, inherently stable equilibrium position (with O fixed) have frequencies

(g/1), (g/l) max = ) 1 and min = (3.5) (1- mi) (1 + mi)' where m = m2/(m +m2), (3.6) ml denoting the point mass which is attached directly to O. The high-frequency criterion (2.1) for the stability of the inverted equilibrium position then reduces to

V2( )i(0 < < 0.450(1-m). (3.7)

For comparison, we show in figure 3 the stability boundaries for this inverted double pendulum obtained by direct numerical integration of the equations of motion, linearized about the upward vertical. At high frequencies o0 the stable range of e/l is given well by (3.7), but if the upper mass m2 is much smaller than the lower mass m1 it is possible to stabilize the system with a comparatively small driving frequency o0 and (3.7) does not then apply. This case corresponds, of course, to Omax and tmin being almost equal (see (3.5) and the end of ?2); indeed, when m = 0 the two coincide and (3.7) reduces to the stability criterion for an inverted single pendulum. More typically, rather high frequencies w) are needed to give stability, and (2.1) then provides a good approximation to the whole stability region in the e-w0 plane. The case of equal masses (m = 0.5) is a good example: (1.85/(0o) (g/l) < e/1 < 0.132, (3.8) Proc. R. Soc. Lond. A (1993)

242

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM All use subject to JSTOR Terms and Conditions

#### Page 5

50 - 0.8

oo 40- 0.5 (gll)112

30 - 0.2

20 - 11

10-

I I I , I , 0.1 0.2 0.3 0.4 0.5 0.6 elI Figure 3. Computed linear stability boundaries for the inverted point-mass double pendulum for four values of m = m2/(ml+m2).

for the right-hand stability boundary in figure 3 is indeed almost a straight line parallel to the o0-axis, save at the very lowest frequencies giving stability. The larger the value of m, clearly, the smaller we have to take e and the higher we have to take o0, but there is in theory no upper limit to the mass ratio m2/m1 which can be stabilized by vertical oscillations of the support.

(c) An inverted N-pendulum We now apply the theorem of ?2 to a pendulum with N equal point masses connected by N equal light rods of total length L = Nl. The natural frequencies of oscillation about the lower equilibrium position are

i = (/l, i= 1,...,N, (3.9) where i denotes the ith zero of the Laguerre polynomial LN(x); this was shown by Daniel Bernoulli and by Euler in the late 1730s (see Cannon & Dostrovsky 1981). In the case N = 3 this gives 1 = 0.64(g/l)i, 2 =1.51(g/l)i, ()3=2.51(g/l), (3.10) and on taking the greatest and least of these we find that the criterion (2.1) becomes

(2.21/0l)(g/l)i < e/l < 0.07. (3.11) For comparison, figure 4 shows the stability boundary obtained by numerical integration of the linearized equations of motion. When N is large we can use the asymptotic expression $ X2/4N, where Xi denotes the ith zero of the Bessel function J0(x). With X1 2.4 and XN (N-) )n we then have

min- 1.2(glL) () *(N- 1) n(gL)H Omin -- Omax and therefore

1.18 g e 0.182 1.18(qie< .< 1)-82 (3.12) NO ) L 4(Nas the criterion for the stability of this particular inverted N-pendulum, when N is large. Proc. R. Soc. Lond. A (1993)

A pendulum theorem 243

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM

All use subject to JSTOR Terms and Conditions

#### Page 6

40-

30- \ S (g/1)1/2

20 -

10-

I I I I 0.05 0.1 0.15 0.2

eli Figure 4. Computed regions of linear stability for the upside-down N-pendulum, for N = 1,2,3, in the case when the various (point) masses are equal.

Suppose finally that L is fixed. Then, the larger the value of N, the smaller we must take [Annotation 8] e, and, consequently, the larger we must take to if stability is to be achieved. The limit N-* oo for fixed L corresponds to a perfectly flexible string, and was of major interest in the original investigations of Bernoulli and Euler. The region of stability of the inverted state vanishes altogether in this limit, so we cannot perform the Indian rope trick.

4. Concluding remarks

While the theorem of this paper concerns linear stability, numerical integrations of the full equations of motion have also been carried out, together with laboratory experiments (Acheson & Mullin 1993). These further studies show the stability of these inverted states to be remarkably robust, at least in the cases N = 1,2, 3. In particular, we have seen an inverted triple pendulum recover from very severe initial disturbances and gradually wobble back to the upward vertical.

I am grateful to Peter Clifford, Don Fowler, Janet Mills, Tom Mullin and John Roe for helpful discussions and encouragement.

References Acheson, D. J. & Mullin, T. 1993 On upside-down pendulums. Nature, Lond. (Submitted.) Cannon, J. T. & Dostrovsky, S. 1981 The evolution of dynamics: vibration theory from 1687 to 1742. Springer-Verlag. Hsu, C. S. 1961 On a restricted class of coupled Hill's equations and some applications. J. appl. Mech. 28, 551-556. Jordan, D. W. & Smith, P. 1987 Nonlinear ordinary differential equations, 2nd edn. Oxford: Clarendon Press. Kalmus, H. P. 1970 The inverted pendulum. Am. J. Phys. 38, 874-878. Landau, L. D. & Lifshitz, E. M. 1976 Mechanics, 3rd edn. Pergamon. Leiber, T. & Risken, H. 1988 Stability of parametrically excited dissipative systems. Phys. Lett. A 129, 214-218. Lowenstern, E. R. 1932 The stabilizing effect of imposed oscillations of high frequency on a dynamical system. Phil. Mag. 8, 458-486. Otterbein, S. 1982 Stabilisierung des n-Pendels und der Indische Seiltrick. Arch. ration. Mech. Analysis 78, 381-393. Proc. R. Soc. Lond. A (1993)

D. J. Acheson 244

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM

All use subject to JSTOR Terms and Conditions

#### Page 7

A pendulum theorem 245 Stephenson, A. 1908a On a new type of dynamical stability. Mem. Proc. Manch. Lit. Phil. Soc. 52 (8), 1-10. Stephenson, A. 1908b On induced stability. Phil. Mag. 15, 233-236. Stephenson, A. 1909 On induced stability. Phil. Mag. 17, 765-766.

Received 28 May 1993; accepted 16 July 1993

Proc. R. Soc. Lond. A (1993)

This content downloaded from 169.229.32.136 on Wed, 7 May 2014 15:58:31 PM All use subject to JSTOR Terms and Conditions

---

Source: [Fermat's Library](https://fermatslibrary.com/) — https://fermatslibrary.com/s/a-pendulum-theorem. More annotated papers: https://fermatslibrary.com/llms.txt
