On Foucault's Pendulum

Author(s): A. S. Chessin
Source: American Journal of Mathematics, Vol. 17, No. 1 (Jan., 1895), pp. 81-88
Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/2369710
Accessed: 05/10/2010 07:24

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=jhup.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to
American Journal of Mathematics.

http://www.jstor.org
 On Foucault's Pendulum.
BY A. S. CHESSIN.

In this paper we shall consider the motion of a physical pendulum on the

surface of the earth, taking into account the rotation of the earth about its axis.
The initial velocity of the pendulum relatively to the earth will be supposed
equal to zero, as in the famous experiment of Foucault with his practically
mathematical pendulum. The name of " Foucault's pendulum " is therefore
retained, although oscillations of any finite amplitude will be considered.*
y(East) Let 0 be a point
_P on the surface of the
(North)X , 0) X(5outh) the meri-
(North) X X (South) earth; OXZ
dian of this point; 0&)
the direction of the axis
( West Y'
) ~ - 1/ ~ < >1 ,> / of the earth towards the
Q\ 90 \ /\( / south. Then L XOG is
\ {C wx (South) the latitude X of the
place, OX being the in-
z @ tersection of the meri-
dian plane with the
N/r horizon XOY, directed
again towards the south; 0 Y in the plane of the horizon 900 from OX and
towards the east; OZ in the direction of the force of gravity. Farther, let
ON be a physical pendulum, 0 the angle its axis makes with OZ, the pen-
dulum being a body of rotation. Let OP, 0Q, ON be a system of three
principal axes of inertia about the point 0, the axis ON coinciding with the
geometrical axis of the pendulumn,OQ under 900 to ON in the plane OANZ

*As is well known, only very small oscillations were given to Foucault's pendulum-
11
82 CHESSIN: On Foucault's Pendulum.

(the plane which would be the plane of oscillations of the pendulum but
for the disturbance due to the rotation of the earth); OP, perpendicular
to this plane in the sense shown on the figure. Let farther C be the centre of
inertia of the pendulum and 0 C = 1; M be the mass of the body; A, A, C be
the three prilncipal mioments of inertia of the pendulum about the point 0;
l, y2. r3 the angles which the axes OP, OQ, ON make with the direction of
the axis (c)) of the eartb. Finally, let 0, cp, 4 be the three Eulerian angles
which define the position of the pendulum relative to the earth, i. e. 0 L NOZ;
ci = L POX; A -angle of a determinate plane passing through ON and fixed
in the body, with the plane OZN. Then we shall have
cos = cosX cos ci,
cos y2= sin I sin 0 - cos X cos 0 sin cp,
cos y3=sin A cosO + cos sin 0 sinc. (1)
We shall make use of Bour's differential equations for the relative motion
d _ aT2 _ a(U+ )
k@2b
I
__

dt aqk i)qk
where T2iS the kinetic energy of the absolute rotation, U + K is the potential of
the force of gravity (and not of attraction) * in the case of the earth. Hence
Ul+ K= Mgl cos 0, (2)
T2 = {A
I [(0O+ X cos y1)2+ (<p'sin 0 + cacos )2] + C (r + cos 73)2}, (3)

where r-=' + P'cos 0,

and Ga is the angular velocity of the rotation of the earth. Bour's equation
gives first (q,=4'),

as neither T2nor U+K contains the variable A, and thus we have a first integral
r+ a cosY.3 =4 + 4 cos 0 + cos Y3= C1.

We have supposed the initial velocity of the pendulum equal to zero, hence
=l C0 730 and
aO
r+o cosy3= 6cs73o, (5)

* U is the potential of the force of attraction.

 CHESSIN: On Foucault's Pendulum. 83

where COsYsomeans the substitution of the initial values 00, c)6 into the expres-
sion of cos73
Next take qk = (aT2> -
T2 =
dGt
\apJ aq
or

gd [A (qpsin 0 + cacos 72) sin 0 + Ccacos y30 cos 0]

- AGIcos X sin 0 [0' + qi cos 0 cosq],

leaving off the terms of the order of Ca2.* After a slight transformation this
equation becomes

d- [A sin2 0 (cli + casin A) + Cabcos 730 cos 0] -2Ac cos x sin2 0 sin .pol. (6)

If we neglect terms of the order of 02 and higher, we may substitute in the

right-hand member of (6) the value of sin20 sin p.0' calculated in the supposition
of the immobility of the earth, because this expression has the factor ca. Let a
be the angle formed by the axis of the pendulum with OZ in this hypothesis.
Then we may instead of sin20 sin q .' in (6) substitute sin pOsin2a. a'. Inte-
grating the equation after that and putting for brevity

f1 (0) = sin2 00-_sin2 0 + cOs00 (cos 00 - cOs 0), (7)1

A (0) = sin 00(cos 00- cos 0), (7)2

A (a)= 00- sin 00cos 00- ( - sini cos a) (7)3

we shall find
fi (0) sin x l
sin 0. =' + f2 (0) +fA ()] cos X sin(o (8)
sin 0
This value of sin 0. qI we now substitute in the third and last integral, that of
the kinetic energy in the relative motion:
A (0s2 + Sin2 0. qp2) = 2Mgl (cos 0 - cos 0,). (9)
Now, the expression (8) shows that sin20. qp2involves the factor 2a. But this term
cannot be neglected. This exception is due to the fact that in the right-hand

* In this problem, like in all other problems of motions on the surface of the earth, it is useless to
keep terms of the order of W2. if the force of gravity is considered as constant.
84 CHESSIN:On Foucault's Pendulum.

meniber of (8) sin 0 appears in the denominator. We do not know a priori what
the mninimumvalue of 0 may be. It may perhaps be zero, like the milnimum
value of 3 (i. e. like in the ordinary pendulum), in which case the right-hand
member of (8) would become infinite; or it may be of the order of , in which case
the same expression would have a finite value. It follows from this remark that
the term fir20i (0) sin X + [f2 (0) + fA ()] cos X sin po}2 cannot be neglected

a priori whatever be the desired approximation. There are, however, some terms
in the expression (8) which may be dropped, namely, - sin2 0 in (7)1,
- sin 5 cos ,) in (7)3. This is obvious. Hence, we must substitute for
sink0.q(I into the left-hand member of (9) the expression Xjfi(20) , where

f (0) = sin x [sin2o0 + C Cos o0(cos 0 - cos 0)]

+ cos x sinq0 [?a sin 00(cos 00-cos0) +00-sin0o cos0o] . (10)

We then obtain the equation

(d cos 0 2Mgl sin2 0 (cos 0-cos 0o)-2 f2(0)X (11)
)t A
the integration of which gives
cos 0 _c Oosnsn2 (t + T) + (1- E2) cn2i (t + T),
where N/Mgl 1
A'
/Af2 dP |p (12)

k-sin 002 j
2'

e= f(o) (13)

We may write cos e instead of (1 -E2), then

cos 0=cos 00 sn2 (t + T) + cos Ecn24 (t + T). (14)
To determine cp, the equation (8) gives

=cA + Ca
+Ca2 Cossinl pO
osx sin co (15)
 CHESSIN: On Foucault's Pendulum. 85

The expression of a in function of time is found from (14), where we only need
to put e = 0. Then sin-n = sin 00 sn t (t + T) and cos a = dn i (t + T).
2 2snut+)2
Thus - sin 5<cosa is readily obtained in function of t. Let us put
sin' 0
= 5 a- sin ~Ocos a andJpt0 (t) dt =(P(t)
(p(t) sin' (16)

Then, as easily seen, (P (t) is a periodical function of the period' 2T. We have
furthermore
f (0) if (0) + Uf (n)
sinl-0 - cos 0 +cos0'
1- cos 0= - 2k2 (sn2nqi- sn2y (t + T))
1+ cos0 2dn2t(t+ T),

where we have put -q= k-i and very simple integrations give

dt i t
fts?sn2i
n2 -sn2u(t+ T) - - g 03 ( +
_2_(_ _ )_t
_
*)
dt 1(1\J.1? (at)
J0dnu2 (t+ T) - kI2 ((d yk'u e(yt)

A transformation, which will be easily verified by the reader, gives

e/
21 l02 (yt - ni$) _tn- I (yt)

In these formulas, as is usual,

j'E))= 1fJ k2sn2x dxc (17)

Substituting the above formulas into (15), integrating and putting for the sake
of brevity
N= -sin a+[P + k` (1-)] (18)

F (t) -f (A) 0Y (tt) cosa sincpo p(t), (19)

4k'21zE (yt)
86 CHESSIN: On Foucault's Pendulum.

we shall find
0 = (Po+ caNt+ tn'- E, G(ti)] + 6F(t). (20)
If we put furthermore
(p = tn-1
""' 2 (,t) + &)F(t),

or or [q -F (t)] =
trA(Ftt) 5sys(t (21 )
~~~~~~tg
62~ ~ ~ ~~~~(1
the motion of the axis of the pendulum may be represented in the following way:
1. Equation (14) shows that the axiscof the pendulum oscillates between the
positions 0= 0 and 0 =, never passing through the vertical.
2. The equations (14) and (21) representa closed cone with a plane of symmetry
which would be the plane of oscillations but for the disturbance due to the rotation of
the earth.
3. If we rotate the cone Just defined, about the vertical OZ with thzeconstant
angular velocity caN; the combined motion: of the cone about OZ, and of the axis of
the pendulum on this cone, will represent the motion of the axis of the pendulum
relatively to the earth.
4. The rotation of the cone defined above may take place towards the west or
towards the east, or the cone may befixed relatively to the earth, according to whether
N is less than, greater than, or equal to zero.*
In fact, formula (18) can be written in the following way:
N= - ni sin X + n2 cost sin q43, (22)
nj;k2 + (1-) k2 -
2A
(1-2') cos00, (23)
O0-sin 0cosO[02k,2+(1l)k2+ ? (1- 2;) sin00, (24)
sin.200 2A

and it is easily verified that both n1 and n2 have always positive values. Hence
Nis 0 according as
n2 sin qpo n1 tgA. (25)

It follows from this discussion that the rotation of the cone defined above, at the
same latitude, depends 1) on the amplitude of the oscillations; 2) on the construction

*See Comte de Sparre: Sur le mouvement du pendule conique & la surface de la terre. (Thdse de
Doctorat.)
 CHESSIN: On Foucault's Pendulum. 87

of the pendulum, and 3) on its initial orientation. The dependence on the ampli-
tude and the orientation has been indicated first by Count de Sparre in his doctor
thesis, in which, however, only the motion of a mathematical pendulum was con-
sidered. The influence of the construction of the pendulum on the results, as seen
from the formulas developed in this paper, offers a greater field for experi-
ments which would be very interesting.
Formula (22) shows that, ceteris paribus, the absolute value of N is maximum
or minimum, if the pendulum is started in the plcaneof the meridian; maximum if
initially deviated towards the north; minimum if towards the south.

INlmax. = n1 sin X + n2cos A,

INlmi.. = I sin X -n2 cos xa.

In order to complete the solution it remains to determine the angle 4 in function

of t. This is done very easily, but the limits of this paper do not allow the
reproduction of the calculation here.
To conclude, let us see what the above formulas become in the case of
very small amplitudes of oscillations, neglecting powers of 0" higher than the
second.
Formula (21) becomes

tg = > E2 (tt) = J (21)'

E?2 (alt) TOtgjIt,
and instead of the equation (14) we may write
02 = 02co02ut + E2 sin2yt. (14)'

The elimination of t betwee-n the equations (21)' and (14)' gives the elliptic cone
1 =_ Cos2(P + sin2(p1 (26)
~~~~0~

The intersection of this cone with a horizontal plane at the distance V]/7from
the point 0 and fixed with regard to the cone (26), gives an ellipse with the
semiaxes
a = VIA00,
2A-C./ A
~ __A_ ~
bb=VAe=
~2A
_
2/Iy
- aci sin a.
88 CHESSIN: On Foicault's Pendulum.

A first approximation gives for N the value - 2A2- C sin X. These results are
2A
in perfect accordance with those of Mr. Kamerlingh Onnes.*
A more accurate value of N is obtained by retaining second powers of 00.
Formula (22) then gives

N=-N-- 2A- C sin x + 2a

3 A-AC 01
00,cos X sin po+ TA 0 sin X,
~2AC 2

which differs from the result obtained by Mr. Kamerlingh Onnes, who gives
the value
2A- C c 3 2A C
N-- 2A sin X A00 CosX sinq0po+8~ 2A 0,2,sinX.

t Over de betrekkelijke Beweging. Nieuw Archief voor Wiskunde, Deel V, p. 162.