THE LIBRARY
OF
THE UNIVERSITY
OF CALIFORNIA
LOS ANGELES
LMVSS IFPFN7*
EonDon: C. J. CLAY AND SONS,
CAMBRIDGE UNIVERSITY PRESS WAREHOUSE,
AVE MARIA LANE.
50, WELLINGTON STREET.
Uripjtg: F. A. BROCKHAUS.
ork: THE MACMILLAN COMPANY.
[All Rights reserved.]
SCIENTIFIC PAPERS
BY
JOHN WILLIAM STEUTT,
BARON RAYLEIGH,
D.S<x, F.R.S.,
HONORARY FELLOW OF TRDTTTY COLLEGE, CAMBRIDGE,
OF NATURAL PHILOSOPHY Df THE ROYAL
VOL. III.
1887—1892.
CAMBRIDGE :
AT THE UNIVERSITY PRESS.
1902
108078
(ZTambrtoge:
PRINTED BY J. AND C. F. CLAY,
AT THE UNIVERSITY PRESS.
PREFACE.
4 S in former volumes, the papers here included embrace a wide range of
^*- subjects. In Optics, Arts. 149, 150 deal with the reflexion of light at
a twin plane of a crystal and, besides revealing unexpected peculiarities
respecting polarization., explain some remarkable phenomena observed by
Prof. Stokes. Attention may also be called to Art. 185 in which it is shown
that the light found by Jamin to be reflected from water at the polarizing
angle is to be attributed to a film of grease, and to Art. 157 " On the Limit
to Interference when Light is radiated from moving Molecules."
Several papers treat of capillary questions. In Art. 170 Plateau's
" Superficial Viscosity " is traced to greasy contamination of water surfaees.
The theory of Surface Forces is expounded in Arts. 176, 186, 193, and atten-
tion is called to T. Young's remarkable estimates of molecular magnitudes.
The relative densities of Hydrogen and Oxygen and the composition of
Water are the subjects of Arts. 146, 153, 187.
In Acoustics the most important paper is probably that on Bells
(Art. 164). The modes of vibration and the corresponding partial tones
of a large number of bells are there recorded.
The next volume will bring the Collection down to about the present
time and, it is hoped, may be ready in about a year.
TERLDTO PLACE, WITHAX,
1901.
o3
The works of the Lord are great,
Sought out of all them that have pleasure therein.
CONTENTS.
ART. PAGE
142. On the Maintenance of Vibrations by Forces of Double
Frequency, and on the Propagation of Waves through a
Medium endowed with a Periodic Structure . . . 1
[Philosophical Magazine, xxiv. pp. 145—159, 1887.]
143. On the Existence of Reflection when the Relative Refractive
Index is Unity . 15
[British Association Report, pp. 585, 586, 1887.]
144. On the Stability or Instability of Certain Fluid Motions, II. . 17
[Proceedings of the London Mathematical Society, xix. pp. 67—74, 1887.]
145. Diffraction of Sound 24
[Royal Institution Proceedings, XIL pp. 187 — 198, 1888 ; Nature, xxxvnr.
pp. 208—211, 1888.]
146. On the Relative Densities of Hydrogen and Oxygen. (Pre-
liminary Notice) . . 37
[Proceedings of the Royal Society, xmi. pp. 356—363, 1888.]
147. On Point-, Line-, and Plane-Sources of Sound ... 44
[Proceedings of the London Mathematical Society, xix. pp. 504—507, 1888.]
148. Wave Theory of Light . . 47
Plane Waves of Simple Type 49
Intensity . 51
Resultant of a Large Number of Vibrations of Arbitrary
Phase 52
Propagation of Waves in General ..... 54
Waves Approximately Plane or Spherical. ... 56
Interference Fringes 59
Colours of Thin Plates . .... . . . 63
Newton's Diffusion Rings . . . . . .72
Huygens's Principle. Theory of Shadows ... 74
Fraunhofer's Diffraction Phenomena . . . . . 79
Theory of Circular Aperture 87
Influence of Aberration. Optical Power of Instruments . 100
Theory of Gratings . . . . ,- . . . . 106
VU1 CONTENTS.
ART. PAGE
Theory of Corrugated Waves . . . . . .117
Talbot's Bands ... 123
Diffraction when the Source of Light is not Seen in Focus . 127
Diffraction Symmetrical about an Axis . . . .134
Polarization . 137
Interference of Polarized Light. . . . . .140
Double Refraction . . . . . , . "., . 148
Colours of Crystalline Plates ,. . • , v '• •• 156
Rotatory Polarization. . . . . . .159
Dynamical Theory of Diffraction 163
The Diffraction of Light by Small Particles . . . . 170
Reflexion and Refraction 176
Reflexion on the Elastic Solid Theory . . . . 181
The Velocity of Light . . . . . . .187
[Encyclopedia Britannica, xxiv., 1888.]
149. On the Reflection of Light at a Twin Plane of a Crystal . 190
Equations of a Dialectric Medium, of which the Magnetic
Permeability is Unity throughout ... . . 190
Iso tropic Reflexion 192
Propagation in a Crystal . . . ..' . . .194
Reflexion at a Twin Plane 194
. Incidence in the Plane of Symmetry . . . .195
Plane of Incidence perpendicular to that of Symmetry . 197
Doubly Refracting Power Small 200
Plate bounded by Surfaces parallel to Twin Plane . . 200
[Phil. Mag. xxvi. pp. 241—255, 1888.]
150. On the Remarkable Phenomenon of Crystalline Reflexion
described by Prof. Stokes ...... . . 204
[Phil. Mag. xxvi. pp. 256—265, 1888.]
151. Is the Velocity of Light in an Electrolytic Liquid influenced by
an Electric Current in the Direction of Propagation ? . 213
[Brit. Ass. Report, pp. 341—343, 1888.]
152. On the Bending and Vibration of Thin Elastic Shells, especially
of Cylindrical Form . .. .>,'..,. . . . 217
[Proceedings of the Royal Society, XLV. pp. 105—123, 1888.]
153. On the Composition of Water 233
[Proceedings of the Royal Society, XLV. pp. 425 — 430, 1889.]
154. The History of the Doctrine of Radiant Energy . . . 238
[Philosophical Magazine, xxvn. pp. 265—270, 1889.]
CONTEXTS. IX
ABT. PAG*
155. Note on the Free Vibrations of an Infinitely Long Cylindrical
Shell . . - . . ....... 244
[Proceeding* of tne Royal Society^ XLV. pp. 443—448, 1889.]
156. On the Free Vibrations of an Infinite Plate of Homogeneous
Isotropic Elastic Matter . . . . . . .249
[ProataKngt of tie London Mathematical Society, XX. pp. 225—234, 1889.]
157. On the Limit to Interference when Light is Radiated from
Moving Molecules . ........ 258
[Pkaotoptieal Magazine^ xxvn. pp. 298-304, 1889.]
158. Iridescent Crystals . .- . . * . . . .- . 264
[Prof. Boy. /«£ XH. pp. 447—449, 1889; Jotere, XL pp. 227, 228, 1889.]
159. The Sailing Flight of the Albatross ..... 267
[JTotwe, XL p. 34, 1889.]
160. On the Character of the Complete Radiation at a Given
Temperature . . ........ 268
[PhSofopkifal Magazine, xxrn. pp. 460-469, 1889.]
16L On the Visibility of Faint Interference-Bands . . .277
Magazine, xxvn. pp. 4M— 486, 1889.]
162. On the Uniform Deformation in Two Dimensions of a Cylindrical
Shell of Finite Thickness, with Application to the General
Theory of Deformation of Thin Shells ..... 280
[Proceeding* of the London Mathematical Society, XX. pp. 372—381, 1889.]
163. On Achromatic Interference-Bands ...... 288
Introduction . . . ...... 288
Fresnel's Bands ......... 289
Lloyd's Bands . ....... 292
Limit to Illumination ....... 294
Achromatic Interference-Bands ...... 296
Prism instead of Grating ....... 299
Airy s Theory of the White Centre . . . . 301
Thin Plates . * . . . . . . . 303
Herschel s Bands ..... ... 309
Effect of a Prism upon Newton's Rings .... 311
Analytical Statement. . . . - . . . . 314
Curved Interference-Bands . . . .« " . . 316
[PkOotopUeal Magazine, xxvm. pp. 77—91, 189—906, 1889.]
X CONTENTS.
ART. PAGE
164. On Bells 318
Appendix : On the Bending of a Hyperboloid of Revolution 330
[Philosophical Magazine, xxix. pp. 1 — 17, 1890.]
165. The Clark Standard Cell . . . . ..' . . .333
[The Electrician, p. 285, Jan. 1890.]
166. On the Vibrations of an Atmosphere . . . '." . 335
[Philosophical Magazine, xxix. pp. 173—180, 1890.]
— 167. On the Tension of Recently Formed Liquid Surfaces . .341
[Proceedings of the Royal Society, XLVII. pp. 281—287, 1890.]
168. Measurements of the Amount of Oil Necessary in Order to
Check the Motions of Camphor upon Water . . . 347
[Proceedings of the Royal Society, XLVII. pp. 364 — 367, March, 1890.]
169. Foam ' . . .'"'. . . . 351
[Proceedings of the Royal Institution, xin. pp. 85—97, March, 1890.]
"-470. On the Superficial Viscosity of Water . '. . . . 363
[Proceedings of the Royal Society, XLVIII. pp. 127 — 140, 1890.]
171. On Huygens's Gearing in Illustration of the Induction of
Electric Currents . . :. - . ' '•". . . . . 376
[Philosophical Magazine, xxx. pp. 30—32, 1890.]
172. The Bourdon Gauge , ;fi , . 379
[Nature, XLII. p. 197, 1890.]
173. On Defective Colour Vision . 380
[British Association Report (Leeds}, pp. 728—729, 1890.]
174. Instantaneous Photographs of Water Jets .... 382
[British Association Report (Leeds), p. 752, 1890.]
— 175. On the Tension of Water Surfaces, Clean and Contaminated,
Investigated by the Method of Ripples .... 383
Postscript, Sept. 19 394
[Philosophical Magazine, xxx. pp. 386—400, Nov. 1890.]
— i76. On the Theory of Surface Forces . .... . . 397
[Philosophical Magazine, xxx. pp. 285 — 298, 456 — 475, 1890.]
177. Clerk-Maxwell's Papers . •' , • ,. V . , . 426
[Nature, XLIII. pp. 26, 27, 1890.]
178. On Pin-Hole Photography . . ... . . . . 429
[Philosophical Magazine, xxxi. pp. 87 — 99, 1891.]
COXTEXTSu
179. Some Applications of Photography . . . . . . 441
[/W. Boy. /«*. xm. pp. 261—272, Feb. 1891 ; JTotere, nrr. pp. 249—254, 1891.]
180. On the Sensitiveness of the Bridge Method in its Application
to Periodic Electric Currents ...... 452
[Pneeedimgt of U* Royal Sotiety, TTTT pp. 208— 217, 1891.]
18L On Tan der Waals" Treatment of Laplace's Pressure in the
Virial Equation: Letters to Pro£ Tait ..... 465
[_Va/*rr, xur. pp. 499, 597, 1S91.J
182. On the Virial of a System of Haid Colliding Bodies . . 469
[Salmn, xrv. pp. 80— 83, 1891.]
183. Dynamical Problems in Illustration of the Theory of Gases . 473
Introduction ......... 473
Collision Formula? ........ 473
Permanent State of Free Masses under Bombardment . 474
Another Method of Investigation ..... 479
Progress towards the Stationary State .... 480
Pendulums in place of Free Masses ..... 485
\PkH. Mag. TTTIT pp. 424—443, 1891.]
184. Experiments in Aerodynamics. [Be view of Langley's] . . 491
[JTrfwnr, xrv. pp. 108, 109, 1891.]
185. On Reflexion from Liquid Surfaces in the Neighbourhood of
the Polarizing Angle ........ 496
Postscript (October 11) ....... 511
[PUwpAiad Magadm^ mm. pp. 1-19, Jan. 1892.]
186. On the Theory of Surface Forces. IL Compressible Fluids . 513
[Pfcfettpfeof JfepmiK, xxxnx pp. 209-220, 1892.]
187. On the Relative Densities of Hydrogen and Oxygen. IL . 524
[PHKtBHmy* of *> Boyri Society, t pp. 44S— 163, 1892.]
188u Superheated Steam ......... 538
Heat Engines and Saline Solutions ..... 539
Heat Engines and Saline Solutions ..... 540
. [JTofve, XLV. pp. 375, 378, 438, 512, 1892.]
189. Aberration .......... 542
T. pp. 499-502, 1892.]
190. Remarks on Maxwell's Investigation respecting Bohzmann's
Theorem .......... 554
YTTlff. pp. 356-359, 1892.]
xil CONTENTS.
ART. PAGE
191. On the Physics of Media that are composed of Free and
Perfectly Elastic Molecules in a State of Motion. [Intro-
duction to Waterston's Memoir] ...... 558
[Phil. Trans. 183 A, pp. 1—5, 1892.]
192. Experiments upon Surface-Films ...... 562
The Behaviour of Clean Mercury ..... 562
Drops of Bisulphide of Carbon upon Water . . . 563
Movements of Dust ........ 564
Camphor Movements a Test of Surface-Tension . . 565
Influence of Heat 567
Saponine and Soap ........ 568
Separation of Motes . . . . . .... 569
The Lowering of Tension by the Condensation of Ether
Vapour . . - . . 570
Breath Figures and their Projection .... 570
[Philosophical Magazine, xxxin. pp. 363 — 373, 1892.]
— 193. On the Theory of Surface Forces. III.— Effect of Slight
Contaminations . . . . . . . . .572
[Philosophical Magazine, XXXIII. pp. 468—471, 1892.]
194. On the Question of the Stability of the Flow of Fluids . 575
[Philosophical Magazine, xxxiv. pp. 59—70, 1892.]
195. On the Instability of a Cylinder of Viscous Liquid under
Capillary Force 585
[Philosophical Magazine, xxxiv. pp. 145—154, 1892.]
- 196. On the Instability of Cylindrical Fluid Surfaces . • . . 594
[Philosophical Magazine, xxxiv. pp. 177—180, 1892.]
ILLUSTRATIONS.
Plate I (Figs. 1-4) ^ ^ ,. , . . . ., To face p. 444
* H ( „ 5—7) „ „ 446
„ HI ( „ 8-12) ... 450
142.
OX THE MAINTENANCE OF VIBRATIONS BY FORCES OF
DOUBLE FREQUENCY, AND ON THE PROPAGATION
OF WAVES THROUGH A MEDIUM ENDOWED WITH A
PERIODIC STRUCTURE.
[Philosophical Magazine, xxiv. pp. 145 — 159; 1887.]
THE nature of the question to be first considered may be best explained
by a paragraph from a former paper*, in which the subject was briefly
treated. "There is also another kind of maintained vibration which, from
one point of view, may be regarded as forced, inasmuch as the period is
imposed from without, but which differs from the kind just referred to
(ordinary forced vibrations) in that the imposed periodic variations do not
tend directly to displace the body from its configuration of equilibrium.
Probably the best-known example of this kind of action is that form of
Melde's experiment in which a fine string is maintained in transverse
vibration by connecting one of its extremities with the vibrating prong of a
massive tuning-fork, the direction of motion of the point of attachment being
parallel to the length of the string^. The effect of the motion is to render
the tension of the string periodically variable ; and at first sight there is
nothing to cause the string to depart from its equilibrium condition of
straightness. It is known, however, that under these circumstances the
equilibrium position may become unstable, and that the string may settle
down into a state of permanent and vigorous vibration whose period is the
double of that of the point of attachment +." Other examples of acoustical
interest are mentioned in the paper.
* " On Maintained Vibrations." Phil. Mag. April, 1883, p. 229. [Vol. n. Art. 97.]
t "When the direction of motion is transverse, the case falls under the head of ordinary
forced vibrations."
* " See TyndalTs Soitnd, 3rd ed. eh. in. § 7, where will also be found a general explanation
of the mode of action."
R. III. 1
2 ON THE MAINTENANCE OF [142
My attention was recalled to the subject by Mr Glaisher's Address to the
Astronomical Society*, in which he gives an interesting account of the
treatment of mathematically similar questions in the Lunar Theory by
Mr Hillf and by Prof. Adams J. The analysis of Mr Hill is in many
respects incomparably more complete than that which I had attempted ;
but his devotion to the Lunar Theory leads the author to pass by many
points of great interest which arise when his results are applied to other
physical questions.
By a suitable choice of the unit of time, the equation of motion of the
vibrating body may be put into the form
^+2fc~+ (©0 + 2®, cos 2t)w=0; .................. (1)
where k is a positive quantity, which may usually be treated as small,
representing the dissipative forces. (00 + 20! cos 2£) represents the coefficient
of restitution, which is here regarded as subject to a small imposed periodic
variation of period IT. Thus B0 is positive, and @x is to be treated as
relatively small.
The equation to which Mr Hill's researches relate is in one respect less
general than (1), and in another more general. It omits the dissipative term
proportional to k ; but, on the other hand, as the Lunar Theory demands, it
includes terms proportional to cos 4>t, cos Qt, &c. Thus
.)w = 0; ............ (2)
(3)
where
6 = 2,6,6** ................................................... (4)
n being any integer, and % representing V(- 1). In the present investigation
6_=6n.
It will be convenient to give here a sketch of Mr Hill's method and
results. Remarking that when 0X, 02, &c. vanish, the solution of (3) is
(5)
* Monthly Notices, Feb. 1887.
t " On the Part of the Motion of the Lunar Perigee which is a Function of the Mean Motions
of the Sun and Moon," Acta Mathematica, vm. 1; 1886. Mr Hill's work was first published
in 1877.
t " On the Motion of the Moon's Node, in the case when the orbits of the Sun and Moon are
supposed to have no Eccentricities, and when their Mutual Inclination is supposed to be
indefinitely small." Monthly Notices, Nov. 1877.
1887] VIBRATIONS BY FORCES OF DOUBLE FREQUENCY. 3
where K, K' are arbitrary constants, and c= V(^»), he shows that in the
general case we may assume as a particular solution
iv^bne****, ................................. (6)
the value of c being modified by the operation of Bj, &c., and the original
term &,«** being accompanied by subordinate terms corresponding to the
positive and negative integral values of w.
The multiplication by 6, as given in (4), does not alter the form of (6) ;
and the result of the substitution in the differential equation (3) may be
written
(c+2my>bm-2n&1^nbn = 0, ..................... (7)
which holds for all integral values of m, positive and negative. These
conditions determine the ratios of all the coefficients bn to one of them, e.g.,
6,, which may then be regarded as the arbitrary constant. They also
determine c, the main subject of quest. Mr Hill writes
[»]=(c + 2n)»-e.; ........................... (8)
so that the equations take the form
... + [- 2] &_- e, 6_1-ea&,-e361-e462-...=oJ
...- e, 6_,+[-i]6_1-e16.-e361-eJ63-... = o,
...- e, 6_- e, 6_1+[o]&.-e161-B36i-... = o, >• ....... (9)
...- e3 6_2- ea 6_1-
...- e4 &_*- e3 &_1-
The determinant formed by eliminating the b's from these equations is
denoted by 5) (c) ; so that the equation from which c is to be found is
5>(c) = 0 .................................. (10)
The infinite series of values of c determined by (10) cannot give inde-
pendent solutions of (3), — a differential equation of the second order only.
It is evident, in feet, that the system of equations by which c is determined
is not altered if we replace c by c + 2i», where v is any positive or negative
integer. Neither is any change incurred by the substitution of — c for c.
" It follows that if (10) is satisfied by a root c = c,, it will also have, as roots,
all the quantities contained in the expression
where n is any positive or negative integer or zero. And these are all the
roots the equation admits of; for each of the expressions denoted by [n] is of
two dimensions in c, and may be regarded as introducing into the equation
the two roots 2n + c, and 2n — c,. Consequently the roots are either all real
or all imaginary; and it is impossible that the equation should have any
equal root unless all the roots are integral."
1—2
ON THE MAINTENANCE OF
[142
On these grounds Mr Hill concludes that 1) (c) must be such that
1)(c) = A [COS(TTC) - cos (TTCO)] (11)
identically, where A is some constant independent of c ; whence on putting
c = 0, _ _
in which, if we please, c0 may be replaced by c. The value of A may now be
determined by comparison with the particular case ©j = 0, ©2 = 0, &c., for
which of course c = V©0. Thus if £>'(0) denote the special form then assumed,
i.e. the simple product of the diagonal constituents,
2)'(0) = A [1 —cos (TT\/©O)], (13)
and
l-cos(Trc) sinM^TTc) =S)(0) a,
1 - cos (TT V©o) sin2 (£ TT V©o) 2)' (0) " '
The fraction 3) (0) ^- 2)' (0) is denoted by D (0). It is the determinant
formed from the original one by dividing each row by the constituent in
the diagonal, so as to reduce all the diagonal constituents to unity, and by
making c vanish. Thus
1-COS(7TC) _nrftv n-
l-cos(,rV©o)~
where
+ i ®* ®2 @=» @4
42_@0 42-©0 42-©0 42-(
_©!_ Q! @2 ®3
22-©0"t" 2s -©0 2s -©0 22-(
~02^V0~Oa~^©(;+ X "o^Vo"^^'" ""(1<
(-1 ^-l t-\ f^l
_ U3 ^ PI n £>i
42-©0
The value of D (0) is calculated for the purposes of the Lunar Theory to a
high order of approximation. It will here suffice to give the part which
depends upon the squares of ©1} ©2, &c. Thus
j^_. . _§.'_. . _^+ 1
!-@O + 4-©/9-©0 +
Another determinant, y(0), is employed by Mr Hill, the relation of which
to D (0) is expressed by
(18)
1887] VIBRATIONS BY FORCES OF DOUBLE FREQUENCY. 5
so that the general solution for c may be written
cos(7rc) = l- v(0). ........................... (19)
Mr Hill observes that the reality of c requires that 1 — v(0) should lie
between — 1 and + 1. In the Lunar Theory this condition is satisfied; but
in the application to Acoustics the case of an imaginary c is the one of
greater interest, for the vibrations then tend to increase indefinitely.
COS(TTC) being itself always real, let us suppose that TTC is complex,
so that
c = a + ift,
where a and ft are real. Thus
cos TTC = cos Tra cos ITT ft — sin ira. sin ITT ft ;
and the reality of cos ire requires either (1) that y3 = 0, or (2) that a=n,
n being an integer. In the first case c is real. In the second
cos TTC = ± cos iir/3 = 1 — V(0), ..................... (20)
which gives but one (real) value of ft. If 1 — v (0) be positive,
; ................................. (21)
but if 1 — v (0) be negative,
COS 7TC = — COS ITT ft,
whence
c=±t£ + 2w + l ............................... (22)
The latter is the case with which we have to do when B0, and therefore c, is
nearly equal to unity ; and the conclusion that when c is complex, the real
part is independent of Bj , B2 , &c. is of importance. The complete value of w
may then be written
«<1+2n», .................. (23)
the ratios of bn and also of bn' being determined by (9). After the lapse of a
sufficient time, the second set of terms in e~& become insignificant.
In the application of greatest acoustical interest ©0 (and c) are nearly
equal to unity; so that the free vibrations are performed with a frequency
about the half of that introduced by Q,. In this case the leading equations
in (9) are those which involve the small quantities [0] and [- 1] ; but for the
sake of symmetry, it is advisable to retain also the equation containing [1].
If we now neglect O2, as well as the b's whose suffix is numerically greater
than unity, we find
and [0] [1] [- 1] - 8^ {[1] + [- 1]} = 0 ................... (25)
6 ON THE MAINTENANCE OF [142
For the sake of distinctness it will be well to repeat here that
[0] = c2 - 80, [- 1] = (c - 2)2 - 00, [1] = (c + 2)2 - ©0.
Substituting these values in (25), Mr Hill obtains
(c2 - ©0) {(c2 + 4 - ©0)2 - 16c2} - 2©,2 {c2 + 4 - ©„} = 0,
and neglecting the cube of (c2 - ©0), as well as its product with ©^
and from this again
l)2-©i2] ......................... (26)
It appears, therefore, that c is real or imaginary according as (©0 — I)2 is
greater or less than ®j2. In the problem of the Moon's apse, treated by
Mr Hill,
@0=11588439, @! = - 0-0570440 ;
and in the corresponding problem of the node, investigated by Prof. Adams,
@0 = 117804,44973,149,
@! = 0-01261,68354,6.
In both these cases the value of c is real, though of course not to be
accurately determined by (26).
Mr Hill's results are not immediately applicable to the acoustical problem
embodied in (1), in consequence of the omission of k, representing the dissi-
pation to which all actual vibrations are subject. The inclusion of this term
leads, however, merely to the substitution for (c -I- 2?i)2 — ©0 in (8) of
(c + 2n)2 - 2ik (c + 2n) - ©0 5
so that the whole operation of k is represented if we write (c — ik) in place
of c, and (®0-k2) in place of @0. Accordingly
cos TT (c - ik) = 1 - v'(°)> ........................ (27)
V'(°) differing from y(0) only by the substitution of @0-A;2 for @0.
If 1- v'(°) lies between + 1, (c - ik) is real, so that
c = ik± a + 2n ............................... (28)
In this case both solutions are affected with the factor e~kt, indicating that
whatever the initial circumstances may be, the motion dies away.
It may be otherwise when 1 - v'(°) lies beyond the limits ±1. In the
case of most importance, when ®0 is nearly equal to unity, 1— v'(°) ig
algebraically less than — 1. If
cos iV/3 = - 1 + v'(0), ........................ (29)
we may write c = 1 + i(k ± /3)+ 2w ......................... (30)
1887]
YIBRATIOXS BY FORCES OF DOUBLE FBEQUEXCY.
Here again both motions die down unless ft is numerically greater than L;
in which case one motion dies down, while the other increases without limit.
The critical relation may be written
cos (irk) = -l + V'(0)- -(31)
From (30) we see that, whatever may be the value of t, the vibrations
(considered apart from the rise or subsidence indicated by the exponential
factors) have the same frequency as if 1% as well as 0,, 0,, &c. vanished.
Before leaving the general theory it may be worth while to point out that
Mr Hills method may be applied when the coefficients of dV dr* and dar <fr,
as well as of •». are subject to given periodic variations. We may write
Assuming, as before, w = ~£.mbme?t*~**wtt
we obtain, on substitution, as the coefficient of «•"****"*,
which is to be equated to zero. The equation for c may still be written
where
...[-2,01 [-1, -11 [0, -21 [1, -31 2.-41...J
...[-2,11 [-1,01 [0, -11 [1, -21 2.-31...
3)(e)= ...[-2, 21 [-1,11 [0,01 [1, -11 2, -21...: (36 1
...[-2,31 [-1,21 [°>11 [1.01 2, -11 —
(...[-2,41 [-1,31 [0,21 [1,11 2S01 ...
By similar reasoning to that employed by Mr Hill we may show that
3>(c) = A (cos ve — cos vet)
+ B (sin we — sin vet)...,
where A and B are constants independent of c ; and, further, that
If all the quantities <&r, ¥r, 0r vanish except 4»,, %, 0., X (0) reduces
to the diagonal row simply, say £'(°> Let c,, c. be the roots of
ON THE MAINTENANCE OF [142
0 (40)
so that the equation for c may be written
2) (0), 1 — cos TTC, sin ire,
3)' (0), 1 - cos TTCj , sin TJ-C! ,
1 $)' (0), 1 — cos 7rc2 , sin TTC.J ,
In this equation 2) (0) -r 2)' (0) is the determinant derived from 3) (0) by
dividing each row so as to make the diagonal constituent unity.
If ...^_i, WQ, %... vanish (even though ...<I>_i, 4>0, <$>,... remain finite),
3) (c) is an even function of c, and the coefficient B vanishes in (38). In
this case we have simply
1 - cos TTC _ 3)(0)
l-coswV80~3>'(0)'
exactly as when 3>j, 3>_i, <I>2, <J>_2... vanish.
Reverting to (24), we have as the approximate particular solution, when
there is no dissipation,
e(c-2)tt ecit e(c+»it
-=(c_2)2_@o+e; + (c+2)2_@0 ................ <4
If c be real, the solution may be completed by the addition of a second,
found from (41) by changing the sign of c. Each of these solutions is
affected with an arbitrary constant multiplier. The realized general solution
may be written
Rcosct + S sin ct
~~
from which the last term may usually be omitted, in consequence of the
relative magnitude of its denominator. In this solution c is determined
by (26).
When c2 is imaginary, we take
4#s=ei«-(eo-l)'; ........................... (43)
so that
c2 = 1 + 2ts, c = 1 + is, c-2 = -l+is.
The particular solution may be written
w = e-°t{®1e-it+(l-®0-'2is)eit}; .................. (44)
or, in virtue of (43),
w = e-*'{(l-@0 + <H)1)cos* + 2ssin*}; ............... (45)
or, again,
w = e~st {V(0! + 1 - eo) . cos t + V(®i - 1 + ®o) • sin t] ....... (46)
1887] VIBRATIONS BY FORCES OF DOUBLE FREQUENCY. 9
The general solution is
w = Re-'t{(I-S6+&l)cost + 2ssint}\
+ Se+« {(I -B0 + B^cos t- Zsswt} j '
R, S being arbitrary multipliers.
One or two particular cases may be noticed. If B0 = 1, 2s = B1} and
w = Re'* {cos t + sin t] + S'e* {cost -sint] ............. (48)
Again, suppose that
012 = (B0- I)2, .............................. (49)
so that s vanishes, giving the transition between the real and imaginary
values of c. Of the two terms in (46), one or other preponderates indefinitely
in the two alternatives. Thus, if Bj= 1 — B0, the solution reduces to cos <;
but if Bj = — 1 + B0, it reduces to sin t. The apparent loss of generality
by the merging of the two solutions may be repaired in the usual way by
supposing s infinitely small.
When there are dissipative forces, we are to replace c by (c - ik), and B by
(B0 — &*); but when k is small the latter substitution may be neglected.
Thus, from (26),
c = i + i* + ix/{(6o-i)a-Bis} ................... (so)
Interest here attaches principally to the case where the radical is imaginary ;
otherwise the motion necessarily dies down. If, as before,
4s2 = B12-(B0-l)=) ........................... (51)
c = 1 +ik + is, c-2 = - 1 +ik + is, ............... (52)
and
(c— 2} it dt
or
w = er#+«t {B,^' + (1 - B0 - 2w) e*},
or
w = e-<*-w{(l-Be + B1)cosf + 2ssin*] ............. (53)
This solution corresponds to a motion which dies away.
The second solution (found by changing the sign of s) is
The motion dies away or increases without limit according as s is less or
greater than k.
The only case in which the motion is periodic is when s = k, or
4*i = B1s-(B0-l)a; ........................... (55)
and then
w = (l -Bj-Bi) cost- 2k sin t ................... (56)
10
ON THE MAINTENANCE OF
[142
These results, under a different notation, were given in my former paper*.
If ®o=l, we have by (51), 2s = © ; and from (53), (54),
w = J2«r *+•>« {cos t + sin t] + Se-(k-g» {cos t - sin t] (57)
In the former paper some examples were given drawn from ordinary
mechanics and acoustics. To these may be added the case of a stretched
wire, whose tension is rendered periodically variable by the passage through
it of an intermittent electric current. It is probable that an illustration
might be arranged in which the vibrations are themselves electrical. ©0
would then represent the stiffness of a condenser, ^ resistance, and <E>0 self-
induction. The most practicable way of introducing the periodic term would
be by rendering the self-induction variable with the time (4>j). This could be
effected by the rotation of a coil forming part of the circuit.
The discrimination of the real and imaginary values of c is of so much
importance, that it is desirable to pursue the approximation beyond the point
attained in (26). From (11) we find
1 + cos (ire) .
Tcos(7rV©o)"
from which, or directly, we see that if c = l, corresponding to the transition
case between real and imaginary values,
2>(1) = 0 (59)
If, as we shall now suppose, ©2, ©3... vanish, (59) may be written in the
form
...1, a,, 1, 0, 0, 0...
...0, 1, «,, 1, 0, 0... =0>
...0, 0, 1, alt I, 0...
...0, 0, 0, 1, a,, 1...
where
= ©0-1 ^@o-9 _©0-25
The first approximation, equivalent to (26), is found by considering merely
the central determinant of the second order involving only ax; thus,
Oif-l = 0 (62)
The second approximation is
a? {(a, - 1/V)2- 1} = 0 (63)
* In consequence of an error of sign, the result for a second approximation there stated is
incorrect [rectified in reprint Art. 97].
1887] YTBRATIOXS BY FORCES OF DOUBLE FREQUEXCT. 11
The third is
and so on. The equation (60) is thus equivalent to
a,-— — —... = ±1;. ...(65)
a,- o3- a4-
and the successive approximations are
&«-, .................. (66)
where
are the corresponding convergents to the infinite continued fraction*.
In terms of 0,, 0, , the second approximation to the equation discriminating
the real and imaginary values of c is
One of the most interesting applications of the foregoing analysis is to the
ease of a laminated medium in which the mechanical properties are periodic
functions of one of the coordinates. I was led to the consideration of this
problem in connexion with the theory of the colours of thin plates. It
is known that old, superficially decomposed, glass presents reflected tints
much brighter, and transmitted tints much purer, than any of which a single
transparent film is capable. The laminated structure was proved by Brewster :
and it is easy to see how the effect may be produced by the occurrence of
nearly similar lamina* at nearly equal intervals. Perhaps the simplest case
of the kind that can be suggested is that of a stretched string, periodically
loaded, and propagating transverse vibrations. We may imagine similar
small loads to be disposed at equal intervals. If, then, the wave-length of a
train of progressive waves be approximately equal to the double interval
between the loads, the partial reflexions from the various loads will all concur
in phase, and the result must be a powerful aggregate reflexion, even though
die effect of an individual load may be insignificant.
The general equation of vibration for a stretched string of periodic
density is
x 2-
JT-'S •«*>
* VteidMiM«f*«niM^«rttliaitDmtfeMi&MtioMlMk«ii rf^ttlff Ibk
(Edimb. Prof. raL THL).
12 ON THE MAINTENANCE OF [142
I being the distance in which the density is periodic. We shall suppose that
PI, pz, ... vanish, so that the sines disappear, a supposition which involves no
loss of generality when we restrict ourselves to a simple harmonic variation of
density. If we now assume that w <x eipt, or oc cos pt, we obtain
.)w = 0, ......... (69)
where £ = irxfl, and
*-££, **-££. &e, .................. (70)
and this is of the form of Mr Hill's equation (2).
When c is real, we may employ the approximate solutions (41), (44). The
latter (with £ written for t) gives, when multiplied by cospt or sinpt, the
stationary vibrations of the system. From (41) we get
_ cos [pt + (c - 2) fl cos [pt + cfl m .
(c-2)*-00 0X
in which, if c = 1 nearly, the two terms represent waves progressing with
nearly equal velocities in the two directions. Neither term gains permanently
in relative importance as x is increased or diminished indefinitely.
It is otherwise when the relation of ©0 to 0j is such that c is imaginary.
By (44) the solution for w, assumed to be proportional to eipt, now takes the
form
- 00 - 2w) e*<*+fl}
- ©o + 2is) e*<**+aj ............. (72)
Whatever may be the relative values of R and 8, the first solution
preponderates when # is large and negative, and the second preponderates
when x is large and positive. In either extreme case the motion is composed
of two progressive waves moving in opposite directions, whose amplitudes are
equal in virtue q/"(43).
The meaning of this is that a wave travelling in either direction is
ultimately totally reflected. For example, we may so choose the values of
R and 8 that at the origin of x there is a wave (of given strength) in the
positive direction only, and we may imagine that it here passes into a uniform
medium, and so is propagated on indefinitely without change. But, in order
to maintain this state of things, we have to suppose on the negative side the
coexistence of positive and negative waves, which at sufficient distances from
the origin are of nearly equal and ever-increasing amplitudes. In order
therefore that a small wave may emerge at x = 0, we have to cause intense
waves to be incident upon a face of the medium corresponding to a large
negative x, of which nearly the whole are reflected.
1887] VIBRATIONS BY FORCES OF DOUBLE FREQUENCY. 13
It is important to observe that the ultimate totality of reflexion does not
require a special adjustment between the frequency of the waves and the
linear period of the lamination. The condition that c should be imaginary
is merely that ^ should numerically exceed (1 — €^)- If X be the wave-length
of the vibration corresponding to tf* and to density pt»
..(73)
and thus the limits between real and imaginary values of c are given by
Xs . ft
If ^ exceeds these limits a train of waves is ultimately totally reflected, in
spite of the finite difference between £X and I *.
In conclusion, it may be worth while to point out the application to
such a problem as the stationary vibrations, of a string of variable density
fixed at two points. A distribution of density,
p.+/hCOS -jr— -1-pjCOS-j— + (75)
is symmetrical with respect to the points x = 0 and or = \l. and between
those limits is arbitrary. It is therefore possible for a string of this density
to vibrate with the points in question undisturbed, and the law of displace-
ment will be
, sin ^ + ...'-. ...(76)
I I
When, therefore, the problem is attacked by the method of Mr Hill, the value
of c obtained by the solution of (69) must be equal to 2. By (15) this
requires
D (0) = 0. (77)
* A iMiihii iiijMiiiai ntal rramination of various eases in which a laminated structure leads
to a powerful but highly selected reflexion would be of value. The most frequent examples are
met with in the organic world. It has occurred to me that BecqaereTs reproduction of the
speeUuiu in natural colours upon silver plates may perhaps be explicable in this manner. The
various parts of the film of subchloride of silver with which the metal k coated may be conceived
to be subjected, during exposure, to ttatiomary luminous waves of nearly definite ware-length,
the effect of which might be to impress upon the substance a periodic structure recurring at
.equal to Jkotf the wave-length of the fight; just as a sensitive flame exposed to stationary
i is influenced at the loops but not at the nodes (PkiL M*g. March, 1879, p. 153).
[YoL L p. 406.] In this way the operation of any kind of light would be to produce just such a
of the film as would cause it to reflect copiously that particular kind of fight. I
. from developing this suggestion, in the hope of soon finding an opportunity of
imentauy acquainted with the subject. [1900. I need hardly remind the
of the beautiful coloured photographs which 1L Lippmann has since obtained by this
14 ON THE MAINTENANCE OF VIBRATIONS, ETC. [142
This equation gives a relation between the quantities ©0, ®lt ®2,...; and
this again, by (70), determines p, or the frequency (p/2?r) of vibration.
Since ®0 = 4 nearly, the most important term in (17) is that involving
@j2. The first approximation to (77) gives
whence, by (70),
fl T
To this order of approximation the solution may be obtained with far
greater readiness by the method given in my work on Sound * ; but it is
probable that, if the solution were required in a case where the variation
of density is very considerable, advantage might be taken of Mr Hill's
determinant D (0). There are doubtless other physical problems to which
a similar remark would be applicable.
* Theory of Sound, vol. i. § 140. In comparing the results, it must be borne in mind that
the length of the string in (78) is denoted by %l.
143.
ON THE EXISTENCE OF REFLECTION WHEN THE RELATIVE
REFRACTIVE INDEX IS UNITY.
[British Association Report, pp. 585, 586 ; 1887.]
THE copious undisturbed transmission of light by glass powder when
surrounded by liquid of the same index, as in Christiansen's experiment
[vol. II. p. 433], suggests the question whether the reflection of any particular
ray is really annihilated when the relative refractive index is unity for that
ray. Such would be the case according to Fresnel's formulae, but these are
known to be in some respects imperfect. Mechanical theory would indicate
that when there is dispersion, reflection would cease to be merely a function
of the index or ratio of wave- velocities. We may imagine a stretched string
vibrating transversely under the influence of tension, and in a subordinate
degree of stiffness, to be composed of two parts so related to one another in
respect of mass and stiffness that the wave- velocity is the same in both parts
for a specified wave-length. But, as it is easy to see, this adjustment will
not secure the complete transmission of a train of progressive waves incident
upon the junction, even when the wave-length is precisely that for which
the velocities are the same.
The experiments that I have tried have been upon plate glass immersed
in a mixture of bisulphide of carbon and benzole, of which the first is more
refractive and the second less refractive than the glass ; and it was found that
the reflection of a candle-flame from a carefully cleaned plate remained pretty
strong at moderate angles of incidence, in whatever proportions the liquids
were mixed.
For a closer examination the plate was roughened behind (to destroy the
second reflection), and was mounted in a bottle prism in such a manner that
the incidence could be rendered grazing. When the adjustment of indices
was for the yellow, the appearances observed were as follows : if the incidence
16 ON THE EXISTENCE OF EEFLECTION, ETC. [143
is pretty oblique, the reflection is total for the violet and blue ; scanty, but
not evanescent, for the yellow; more copious again in the red. As the
incidence becomes more and more nearly grazing, the region of total reflection
advances from the blue end closer and closer upon the ray of equal index, and
ultimately there is a very sharp transition between this region and the band
which now looks very dark. On the other side .the reflection revives, but
more gradually, and becomes very copious in the orange and red. On this
side the reflection is not technically total. If the prism be now turned
so that the angle of incidence is moderate, it is found that, in spite of the
equality of index for the most luminous part of the spectrum, there is a
pretty strong reflection of a candle-flame, and apparently without colour.
With the aid of sunlight it was proved that in the reflection at moderate
incidences there was no marked chromatic selection, and in all probability the
blackness of the band in the yellow at grazing incidences is a matter of
contrast only.
Indeed calculation shows that, according to Fresnel's formulse, the reflection
would be nearly insensible at all parts of the spectrum when the index is
adjusted for the yellow. The outstanding reflection is not due to a difference
of wave- velocities, but to some other cause not usually taken into account.
Such a cause might be found in the presence of a film upon the surface
of the glass, of index differing from that of the interior, and not removable by
mere cleaning. The glass plate was accordingly repolished with putty powder,
after which the reflection was very decidedly diminished. But neither by this
nor by any other treatment (e.g. with hydrofluoric acid) has it been found
possible to render the reflection of a candle-flame at moderate incidences even
difficult of observation although the adjustment of indices was as good as
could be.
It would, however, be hardly safe to conclude that no sufficient film was
operative ; and I do not see how the question is to be decided unless an
experiment can be made upon a surface freshly obtained by fracture.
[1899. At the suggestion of Lord Kelvin I have lately repeated these
observations. The residual light reflected at 45° incidence is polarised in
the usual way, i.e. as if it were reflected from an interface between two media
of slightly differing indices.]
144.
ON THE STABILITY ©R INSTABILITY OF CERTAIN FLUID
^MOTIONS, II.
{Proceedings of the London Mathematical Society, xix. pp. 67 — 74: 1887.]
As the question of the stability, or otherwise, of fluid motions is attracting
attention in consequence of Sir W. Thomson's recent work, I think it advisable
to point out an error in the solution which I gave some years ago* of one of
the problems relating to this subject ; and I will take the opportunity to treat
the problem with greater generality.
In the steady laminated motion, the velocity (U) is a function of y only.
In the disturbed motion U + u, v, the small quantities u, v are supposed to be
periodic functions of x, proportional to eikx, and, as dependent upon the time,
to be proportional to e*"*, where n is a constant, real or imaginary. Under
these circumstances the equation determining v (51) is
The vorticity (Z) of the steady motion is ^dU/dy. If throughout any layer
Z be constant, d^U/dy1 vanishes, and wherever n + kU does not also vanish
or v = Ae1* + Be-** ............................... (3)
If there are several layers in each of which Z is constant, the various solutions
of the form (3) are to be fitted together, the arbitrary constants being so
chosen as to satisfy certain boundary conditions. The first of these conditions
is evidently
» Math. Soc. Proe. xi. p. 57 ; 1880. [Vol. i. Art. 6
t [1900. A being the symbol of finite differences.]
B. in.
18 ON THE STABILITY OR INSTABILITY [144
The second may be obtained by integrating (1) across the boundary. Thus
fdv\ A fdU\
(5)
At a fixed wall v = 0.
In the special problem to which attention is here directed, the laminated
motion is supposed to take place between two fixed walls, at
y = 0 and y = b1 + b' + b2; and the vorticity is supposed to be Flg' 1'
constant throughout each of the three layers bounded by ui us
y = 0, y = bi',
There are thus two internal surfaces at 3/ = &i, y = b1 + bf,
where the vorticity changes. The values of U at these surfaces
may be denoted by U1} U%.
In conformity with (4) and with the condition that v = 0
when y = 0, we may take in the first layer
.(6)
V = v2 = v1 + M! sinh k(y-b1); (7)
...(8)
in the second layer
in the third layer
v = va = v2 + M2 sinh k(y~b' — b^. .
The condition that v = 0, when y = 6, + b' + ba, now gives
0 = M2 sinh kb2 + M1 sinh k (ba + b') + sinh k (b2 + b' + &,) (9)
We have still to express the two other conditions (5) at the surfaces of
transition. At the first surface,
v = sinh kb, , A (dv/dy) = kM, ;
at the second surface,
v = M, sinh kV + sinh k (b, + b'), A (dv/dy) = kMa.
If we denote the values of &(dU/dy) at the two surfaces respectively by
A!, A2, our conditions become
By (9) and (10) the values of M1} M2, n are determined.
1887] OF CERTAIN FLUID MOTIONS, II. 19
The equation for n is found by equating to zero the determinant
sinh£&3, smhk(bt + br), sinh k (6, + bf + bj
n + kU,, -A3sinh^', - A, sinh k (6, + 6')
0, n + k £/i , — A! sinh kbt
so that n has the values determined by the quadratic
0, ...(11)
where
...................................................... (12)
+ A, sinh kb, sinh k (h, + b'\ ...... (13)
sinh k(ba + b' + bl) + k 0A sinh kba sinh k (h + V)
sinh kbt sinh k (bt + 6') + AA sinh A:6, sinh Ar6a sinh kb'. . . .(14)
To find the character of the roots; we have to form the expression for
B*—4AC. Having regaiti to
sinh k (6a + 6') sinh k (^ + 6') - sinh k (62 + V + 6,) sinh kb' = sinh kb^ sin
we find
+ A, sinh th sinh k (62 + 6') - A^ sinh kbs sinh Jl- (6, + by?
+ 4A1A,sinhU-61sinhU-6s ..................................... (15)
Hence, if A,, AS have the same sign, that is, if the curve expressing U as a
function of y be of one curvature throughout, 5a — 4^1 C is positive, and the
two values of u are real. Under these circumstances the disturbance is
stable.
We will now suppose that the surfaces at which the vorticity changes are
symmetrically situated, so that
61 = 6J = 6.
In this case we find
(16)
(17)
+ A, A, sinh1 £6 sinh Id)', ...... (18)
2—2
20
ON THE STABILITY OR INSTABILITY
[144
Under this head there are two sub-cases which may be especially noted.
The first is that in which the values of U are the same on
both sides of the median plane, so that the middle layer is a
region of constant velocity without vorticity, and the velocity
curve is that shown in Fig. 2. We may suppose that U = V
in the middle layer, and that U = 0 at the walls, without loss
of generality, since any constant velocity (U0) superposed
upon this system merely alters n by the corresponding quan-
tity — kU0, as is evident from (1). Thus
U, = U2 = V, A2 = Ax = A = - V/b ;
and Bn- - 4, A C = 4 A2 sinh4 kb.
t j,r V sinh kb sinh k (b + b') ± sinh2 kb
xience n + K v — - j— : — : — . . . — =-7- .
b sinh k (26 + 6')
If the middle layer be absent, b' = 0, and
,lv V 2sinh2£6 V
n + kv =-r . , _.. = -r tanh&&, (21)
6 sinh 2kb b
in conformity with (44) of the former paper ; but the more general result (20)
does not agree with (46).
The other case which we shall consider is that in which the velocities U on
the two sides of the median plane are opposite to one another ; so that
U^-U^V, &, = -&1 = -fiV. (22)
Here 5 = 0, and
.(20)
b')
Thus
_n'_ _ _ fr sinh k (26 + V) + 2kf* sinh kb sinh k (b + b')+ p? sinh2 kb sinh W
Here the two values of n are equal and opposite ; and, since A1; A2 are of
opposite signs, the question is open as to whether n is real
or imaginary.
It is at once evident that n is real if ^ be positive, that
is, if Aj and V are of the same sign, as in Fig. 3.
Even when p. is negative, nz is necessarily positive for
great values of k, that is, for small wave-lengths. For we
have ultimately, from (23),
Fig. 3.
We will now inquire for what values of ^ n2 may be nega-
tive when k is very small, that is, when the wave-length is
1887]
OF CERTAIN FLUID MOTIONS, II.
21
very great. Equating the numerator of (23) to zero, and expanding the
hyperbolic sines, we get as a quadratic in p,,
' + 2//6 (b + 6') + 26 + b' = 0,
whence
.(24)
When p lies between these limits (and then only), »2 is negative, and the
disturbance (of great wave-length) increases exponentially with the time.
We may express these results by means of the velocity F, at the wall
where y = 0. We have
The limiting values of F0 are therefore
The velocity curve corresponding to the first limit is shown in Fig. 4 by the
line QPOPty, the point Q being found by drawing a line AQ parallel to OP
to meet the wall in Q. If 6' = 26, QP is parallel to OA, or the velocity is
constant in each of the extreme layers.
At the second b'mit F0 = 0, and the velocity-curve is that shown in Fig. 5.
Fig. 4.
Fig. 5.
It is important to notice that motions represented by velocity-curves
intermediate between these h'mits are unstable in a manner not possible
to motions in which the velocity-curve, as in Fig. 2, is of one curvature
throughout.
According to the first approximation, the motion of Fig. 5 is on the
border-land between stability and instability for disturbances of great wave-
length ; but, if we pursue the calculation, we find that it is really unstable.
Taking, in (23),
1 2
b b"
22 ON THE STABILITY OR INSTABILITY [144
and writing for brevity kb = x, kb' = x', we get
sinh ( 2# + x) - 2xx (2x + x) sinh x sinh (x + x'} }
+ (2x + a/)2 sinh2 x sinh x' }
k*V* x*x* sinh (2a? + #')
from which, on expanding the hyperbolic sines and retaining two terms, we
get, after reduction,
x*_ W
~~ """"
indicating instability.
[January, 1888*. According to (23), we may always, with a prescribed
wave-length, determine two values of //, (or F0), F being regarded as given,
between which n2 will be negative, and the motion unstable. But, if these
values of p. were imaginary, the result would be of no significance in the
present problem. We may, however, write (23) in the form
n2 _ [p, sinh kb sinh kb' + k sinh k (b + 6')}2 - k* sinh2 kb
k*V*~~ If sinh kb' sinh k (26 + b')
from which we see that, whatever be the value of k, it is possible so to
determine p, that the disturbance shall be unstable. The condition is simply
that /* must lie between the limits
, sinh k (b + b') ± sinh kb
sinh kb sinh kb'
,n
................ (26)
, ?7 coth) kbl
- k coth kb + . , [ — , . .
tanhj 2 J
in which the upper alternative corresponds to the superior limit to the
numerical value of p.
When k is very large, the limits are very great and very close. When k
is small, they become
1 2 j 1
~b~v and ~b>
as has already been proved. As k increases from 0 to oo , the numerical value
of the upper limit increases continuously from l/b + 2/b' to oo , and in like
manner that of the inferior limit from 1/6 to oo . The motion therefore cannot
be stable for all values of k, if ^ (being negative) exceed numerically 1/6.
The final condition of complete stability is therefore that algebraically
.
* This paragraph is re-written, and embodies an improvement suggested in a report com-
municated to me by the Secretary.
1887] OF CERTAIN FLUID MOTIONS, II. 23
In the transition case
it is that represented in Fig. 4. If PQ be bent more downwards than is there
shown, as for example in Fig. 5, the steady motion is certainly unstable.
It would be of interest, in some particular case of instability (such as that
of Fig. 5), to calculate for what value of k the instability, measured by in, is
greatest, and to ascertain the degree of this instability.]
Reverting to the general equations (11), (12), (13), (14), (15), let us
suppose that A.2 = 0, amounting to the abolition of the corresponding surface
of discontinuity. We get
B = k(U1+ U^ sinh k (63 + b' + b,) + A, sinh kb, sinh k (62 + V),
so that n = -kU9, ................................. (27)
A, sinh to, sinh k (b, + b*) ,
'
The latter is the general solution for two layers of constant vorticity of
breadths ^ and b' + 62. An equivalent result may be obtained by supposing
in (11), &c., that 6' = 0, or that ^ = 0.
The occurrence of (27) suggests that any value of — kU is admissible as a
value of n, and the meaning of this is apparent from (1). For, at the place
where n + kU=Q, (2) need not be satisfied, or the arbitrary constants in (3)
may change their values. It is evident that, with the prescribed values of n
and k, a solution may be found satisfying the required conditions at the walls
and at the surfaces where dUjdy changes value, as well as equation (4) at the
plane where n + kU = 0. Equation (5) is there satisfied independently of the
value of v. In this motion an additional vorticity is supposed to be communi-
cated at the plane in question, and moves with the fluid at velocity U.
145.
DIFFRACTION OF SOUND.
[Royal Institution Proceedings, XII. pp. 187—198, 1888;
Nature, XXXVIIL pp. 208—211, 1888.]
THE interest of the subject which I propose to bring before you this
evening turns principally upon the connection or analogy between light and
sound. It has been known for a very long time that sound is a vibration ;
and every one here knows that light is a vibration also. The last piece of
knowledge, however, was not arrived at so easily as the first ; and one of the
difficulties which retarded the acceptance of the view that light is a vibration
was that in some respects the analogy between light and sound seemed to be
less perfect than it should be. At the present time many of the students at
our schools and universities can tell glibly all about it ; yet this difficulty is
one not to be despised, for it exercised a determining influence over the great
mind of Newton. Newton, it would seem, definitely rejected the wave
theory of light on the ground that according to such a theory light would
turn round the corners of obstacles, and so abolish shadows, in the way that
sound is generally supposed to do. The fact that this difficulty seemed to
Newton to be insuperable is, from the point of view of the advancement
of science, very encouraging. The difficulty which stopped Newton two
centuries ago is no difficulty now. It is well known that the question
depends upon the relative wave-lengths in the two cases. Light-shadows
are sharp under ordinary circumstances, because the wave-length of light
is so small : sound-shadows are usually of a diffused character, because the
wave-length of sound is so great. The gap between the two is enormous.
I need hardly remind you that the wave-length of C in the middle of the
musical scale is about 4 feet. The wave-length of the light with which we
are usually concerned, the light towards the middle of the spectrum, is about
the forty-thousandth of an inch. The result is that an obstacle which is
immensely large for light may be very small for sound, and will therefore
behave in a different manner.
1888] DIFFRACTIOX OF SOUJfD. 25
That light-shadows are sharp is a familiar fact, but as I can prove it in a
moment I will do so. We have here light from the electric arc thrown on the
screen: and if I hold up my hand thus we have a sharp shadow at any
moderate distance, which shadow can be made sharper still by diminishing
the source of light. Sound-shadows, as I have said, are not often sharp;
but I believe that they are sharper than is usually supposed, the reason being
that when we pass into a sound-shadow — when, for example, we pass into the
shade of a large obstacle, such as a building — it requires some little time to
effect the transition, and the consequence is that we cannot make a very
ready comparison between the intensity of the sound before we enter and its
diminution afterwards. When the comparison is made under more favourable
conditions, the result is often better than would have been expected. It is.
of course, impossible to perform experiments with such obstacles before an
audience, and the shadows which I propose to show you to-night are on
a much smaller scale. I shall take advantage of the sensitiveness of a flame
such as Professor Tyndall has often used here — a flame sensitive to the waves
produced by notes so exceedingly high as to be inaudible to the human ear.
In fact,, all the sounds with which I shall deal to-night will be inaudible to
the audience. I hope that no quibbler will object that they are therefore not
sounds : they are in every respect analogous to the vibrations which produce
the ordinary sensations of hearing.
I will now start the sensitive flame. We must adjust it to a reasonable
degree of sensitiveness. I need scarcely explain the mechanism of these
flames, which yon know are fed from a special gasholder supplying gas at a
high pressure. When the pressure is too high, the flame flares on its own
account (as this one is doing now), independently of external sound. When
the pressure is somewhat diminished, but not too much so — when the flame
"stands on the brink of the precipice," were, I think, TyndalTs words — the
sound pushes it over, and causes it to flare ; whereas, in the absence of such
sound, it would remain erect and unaffected. Now, I believe, the flame
is flaring under the action of a very high note that I am producing here.
That can be tested in a moment by stopping the sound, and seeing whether
the flame recovers or not. It recovers now. What I want to show you,
however, is that the sound-shadows may be very sharp. I will put my hand
between the flame and the source of sound, and you will see the difference.
The flame is at present flaring; if I put my hand here, the flame recovers,
When the adjustment is correct, my hand is a sufficient obstacle to throw a
most conspicuous shadow. The flame is now in the shadow of my hand, and
it recovers its steadiness : I move my hand up, the sound comes to the flame
again, and it flares. When the conditions are at their best, a very small
obstacle is sufficient to make the entire difference, and a sound-shadow may
be thrown across several feet from an obstacle as small as the hand. The
of the divergence from ordinary experience here met with is, that
26 DIFFRACTION OF SOUND. [145
while the hand is a fairly large obstacle in comparison with the wave-length
of the sound I am here using, it would not be a sufficiently large obstacle in
comparison with the wave-lengths with which we have to do in ordinary life
and in music.
Everything then turns upon the question of the wave-length. The wave-
length of the sound that I am using now is about half an inch. That is its
complete length, and it corresponds to a note that would be very high indeed
on the musical scale. The wave-length of middle C being four feet, the
C one octave above that is two feet ; two octaves above, one foot ; three
octaves above, six inches ; four octaves, three inches ; five octaves, one and a
half inch ; six octaves, three-quarters of an inch ; between that and the next
octave, that is to say, between six and seven octaves above middle C, is the
pitch of the note that I was just now using. There is no difficulty in
determining what the wave-length is. The method depends upon the
properties of what are known as stationary sonorous waves as opposed to
progressive waves. If a train of progressive waves are caused to impinge
upon a reflecting wall, there will be sent back or reflected in the reverse
direction a second set of waves, and the co-operation of these two sets of
waves produces one set or system of stationary waves, the distinction being
that, whereas in the one set the places of greatest condensation are continually
changing and passing through every point, in the stationary waves there are
definite points for the places of greatest condensation (nodes), and others
distinct and definite (loops) for the places of greatest motion. The places of
greatest variation of density are the places of no motion : the places of
greatest motion are places of no variation of density. By the operation of a
reflector, such as this board, we obtain a system of stationary waves, in which
the nodes and loops occupy given positions relatively to the board.
You will observe that as I hold the board at different distances behind,
the flame rises and falls — I can hardly hold it still enough. In one position
the flame rises, further off it falls again; and as I move the board back
the flame passes continually from the position of the node — the place of no
motion — to the loop or place of greatest motion and no variation of pressure.
As I move back the aspect of the flame changes ; and all these changes are
due to the reflection of the sound-waves by the reflector which I am holding.
The flame alternately ducks and rises, its behaviour depending upon the
different action of the nodes and loops. The nodes occur at distances from
the reflecting wall, which are even multiples of the quarter of a wave-length ;
the loops are, on the other hand, at distances from the reflector which are odd
multiples, bisecting therefore the intervals between the nodes. I will now
show you that a very slight body is capable of acting as a reflector. This is
a screen of tissue paper, and the effect will be apparent when it is held
behind the flame and the distances are caused to vary. The flame goes up
1888] DIFFRACTION OF SOUND. 27
and down, showing that a considerable proportion of the sonorous intensity
incident upon the paper screen is reflected back upon the flame; otherwise
the exact position of the reflector would be of no moment. I have here,
however, a different sort of reflector. This is a glass plate — I use glass so
that those behind ma}' see through it — and it will slide upon a stand here
arranged for it. When put in this position the flame is very little affected ;
the place is what I call a node — a place where there is great pressure
variation, but no vibratory velocity. If I move the glass back, the flame
becomes vigorously excited ; that position is a loop. Move it back still more
and the flame becomes fairly quiet; but you see that as the plate travels
gradually along, the flame goes through these evolutions as it occupies in
succession the position of a node or the position of a loop. The interest
of this experiment for our present purpose depends upon this — that the
distances through which the glass plate, acting as a reflector, must be
successively moved in order to pass the flame from a loop to the next loop,
or from a node to the consecutive node, is in each case half the wave-length :
so that by measuring the space through which the plate is thus withdrawn
one has at once a measurement of the wave-length, and consequently of the
pitch of the sound, though one cannot hear it.
The question of whether the flame is excited at the nodes or at the
loops, — whether at the places where the pressure varies most or at those
where there is no variation of pressure, but considerable motion of air — is one
of considerable interest from the point of view of the theory of these flames.
The experiment could be made well enough with such a source of sound
as I am now using ; but it is made rather better by using sounds of a lower
pitch and therefore of greater wave-length, the discrimination being then
more easy. Here is a table of the distances which the screen must be from
the flame in order to give the maximum and the minimum effect, the
minimum being practically nothing at all.
TABLE OF MAXIMA AXD MDJOCA.
Max, Min.
1-1
30
4-5
5-9
7-5
8-9
10-3
117
13-0
147
15-9
The distance between successive maxima or successive minima is very
nearly 3 (centims.), and this is accordingly half the length of the wave.
28 DIFFRACTION OF SOUND. [145
But there is a further question behind. Is it at the loops or is it at the
nodes that the flame is most excited? The table shows what the answer
must be, because the nodes occur at distances from the screen which are even
multiples, and the loops at distances which are odd multiples ; and the
numbers in the table can be explained in only one way — that the flame is
excited at the loops corresponding to the odd multiples, and remains quiescent
at the nodes corresponding to the even multiples. This result is especially
remarkable, because the ear, when substituted for the flame, behaves in the
exactly opposite manner, being excited at the nodes and not at the loops.
The experiment may be tried with the aid of a tube, one end of which is
placed in the ear, while the other is held close to the burner. It is then
found the ear is excited the most when the flame is excited least, and
vice versa. The result of the experiment shows, moreover, that the manner
in which the flame is disintegrated under the action of sound is not, as might
be expected, symmetrical in regard to the axis of the flame. If it were
symmetrical, it would be most affected by the symmetrical cause, namely, the
variation of pressure. The fact being that it is most excited at the loop,
where there is the greatest vibratory velocity, shows that the method of
disintegration is unsymmetrical, the velocity being a directed quantity.
In that respect the theory of these flames is different from the theory of
the water-jets investigated by Savart, which resolve themselves into detached
drops under the influence of sonorous vibration. The analogy fails at this
point, and it has been pressed too far by some experimenters on the subject.
Another simple proof of the correctness of the result of our experiment is
that it makes all the difference which way the burner is turned in respect of
the direction in which the sound-waves are impinging upon it. If the
phenomenon were symmetrical, it would make no difference if the flame
were turned round upon its vertical axis. But we find that it does make a
difference. This is the way in which I was using the flame, and you see that
it is flaring strongly. If I now turn the burner round through a right angle,
the flame stops flaring. I have done nothing more than turn the burner
round and the flame with it, showing that the sound-waves may impinge in
one direction with great effect, and in another direction with no effect. The
sensitiveness occurs again when the burner is turned through another right
angle ; after three right angles there is another place of no effect ; and after a
complete revolution of the flame the original sensitiveness recurs. So that if
the flame were stationary, and the sound-waves came, say, from the north or
south, the phenomena would be exhibited ; but if they came from the east or
west, the flame would make no response.
This is of convenience in experimenting, because, by turning the burner
round, I make the flame almost insensitive to a sound, and I am now free to
show the effect of any sound that may be brought to it in the perpendicular
direction. I am going to use a very small reflector — a small piece of
DITFRACTIOX OF SOCXD.
29
looking-glass. Wood would do as well; but looking-glass facilitates the
adjustment, because my assistant, by seeing the reflection, will be able to tell
me when I am holding it in the best position. Now, the sound is being
reflected from the bit of glass, and is causing the flame to flare, though the
same sound, travelling a shorter distance and impinging in another direction,
is incompetent to produce the result (Fig. 1).
I am now going to move the reflector to and fro along the line
perpendicular to that joining the source and the burner, all the while
maintaining the adjustment, so that from the position of the source of sound
the image of the flame is seen in the centre of the mirror. Seen from the
source, it is still as central as before ; but it has lost its effect, and as I move
it to and fro I produce cycles of effect and no effect. What is the cause
of this ? The question depends upon something different from what I have
been speaking of hitherto : and the explanation is, that we are here dealing
with a diffraction phenomenon. The mirror is a small one, and the sound-
waves which it reflects are not big enough to act in the normal manner.
We are really dealing with the same sort of phenomena as arise in optics
when we use small pin-holes for the entrance of our light. It is not very
easy to make the experiment in the present form quite simple, because
the mirror would have to be withdrawn, all the while maintaining a somewhat
complicated adjustment. In order to raise the question of diffraction in its
simplest shape, we must have a direct course for the sound between its origin
and the place of observation, and interpose in the path a screen perforated
with such holes as we desire to try.
30 DIFFRACTION OF SOUND. [145
[1900. Further experiments with the arrangement of Fig. 1 have
recently been made. When the gas pressure is carefully adjusted, the
positions of the mirror corresponding to recovery of the flame may be very
well denned, but they depend upon the orientation of the burner. If for
example the burner is so turned round its axis that the azimuth of maximum
sensibility bisects internally the angle subtended by the source and the
mirror, the positions of the mirror for minimum effect are well defined, and
they are so spaced along the line of motion that the sum of the distances
from the mirror to the source and to the burner increases at each step by one
complete wave-length. But if the burner be again turned upon its axis
through a right angle, the positions of minimum effect are shifted so as to
bisect the intervals between the former ones. In other words the positions
of maximum and minimum are interchanged. These effects are just what
might have been expected, and they clearly depend upon the co-operation of
the direct and the nearly equal reflected sound. In the orientation of the
burner contemplated above where the azimuth of maximum sensibility
passes through the mirror, we might expect the influence of the direct
sound to be eliminated, and then there should be no alternation of effect
as the mirror moves. But this state of things can be attained only im-
perfectly. It is possible so to adjust the orientation of the burner that
the sound of the flaring shall be uniform; but if we use our eyes instead
of our ears, we recognise that the flame still executes periodic evolutions.
The residual variation may depend upon diffraction as above suggested ; but
I think that it may also be connected with a behaviour of the burner
in respect of orientation less simple than that above supposed and applicable
as a first approximation. Unless care be taken, a variation of effect with
position of the mirror would probably be mainly due to imperfect adjust-
ment of orientation of the burner.]
The screen I propose to use is of glass. It is a practically perfect
obstacle for such sounds as we are dealing with ; but it is perforated here
with a hole (20 cm. in diameter), rendered more evident to those at a distance
by means of a circle of paper pasted round it. The edge of the hole
corresponds to the inner circumference of the paper. We shall thus be able
to try the effect of different sized apertures, all the other circumstances
remaining unchanged. The experiment is rather a difficult one before an
audience, because everything turns on getting the exact adjustment of
distances relatively to the wave-length. At present the sound is passing
through this comparatively large hole in the glass screen, and is producing,
as you see, scarcely any effect upon the flame situated opposite to its centre.
But if (Fig. 2) I dimmish the size of the hole by holding this circle of zinc
(perforated with a hole 14 cm. in diameter) in front of it, it is seen that,
although the hole is smaller, we get a far greater effect. That is a funda-
mental phenomenon in diffraction. Now I reopen the larger hole, and the
1888] DIFFRACTION OF SOUND. 31
flame becomes quiet. So that it is evident that in this case the sound
produces a greater effect in passing through a small hole than in passing
Fig. 2.
Source Burner
o * -o
through a larger one. The experiment may be made in another way, by
obstructing the central in place of the marginal part of the aperture in the
glass. When I hold this unperforated disc of zinc (14 cm. in diameter)
centrically in front, we get a greater effect than when the sound is allowed to
pass through both parts of the aperture. The flame is now flaring vigorously
under the action of the sonorous waves passing the marginal part of the
aperture, whereas it will scarcely flare at all under the action of waves
passing through both the marginal and the central hole.
This is a point which I should like to dwell upon a little, for it lies at the
root of the whole matter. The principle upon which it depends is one
that was first formulated by Huygens, one of the leading names in the
development of the undulatory theory of light. In this diagram (Fig. 3)
is represented in section the different parts of the obstacle. C represents the
source of sound, B represents the flame, and APQ is the screen. If we
choose a point P on the screen, so that the whole distance from B to C,
reckoned through P, viz. BPC, exceeds the shortest distance BAC by exactly
half the wave-length of the sound, then the circular area, whose radius is AP,
is the first zone. We take next another point, Q, so that the whole distance
BQC exceeds the previous one by half a wave-length. Thus we get the second
zone represented by PQ. In like manner, by taking different points in
succession such that the last distance taken exceeds the previous one every
time by half a wave-length, we may map out the whole of the obstructing
screen into a series of zones called Huygens' zones. I have here a material
32
DIFFRACTION OF SOUND.
[145
embodiment of that motion, in which the zones are actually cut out of a
piece of zinc. It is easy to prove that the effects of the parts of the wave
Fig. 3.
traversing the alternate zones are opposed, that whatever may be the effect
of the first zone, A P, the exact opposite will be the effect of PQ, and so on.
Thus, if AP and PQ are both allowed to operate, while all beyond Q is cut
off, the waves will neutralise one another, and the effect will be immensely
less than if AP or PQ operated alone. And that is what you saw just now.
When I used the inner aperture only, a comparatively loud sound acted upon
the flame. When I added to that inner aperture the additional aperture PQ,
the sound disappeared, showing that the effect of the latter was equal and
opposite to that of AP, and that the two neutralised each other.
If AG=a, AB = b, AP = x, wave-length = \, the value of x for the
external radius of the nib. zone is
ab
x* = n\ j- ,
a + b'
or, if a = 6,
a? = | n\a.
With the apertures used above, ^ = 49 for n = l; #2 = 100 for n=2;
so that
\a = 100,
the measurements being in centimetres. This gives the suitable distances,
when X is known. In the present case X = T2, a = 83.
Closely connected with this there is another very interesting experiment,
which can easily be tried, and which has also an important optical analogy.
I mean the experiment of the shadow thrown by a circular disc. If a very
small source of light be taken — such a source as would be produced by per-
forating a thin plate in the shutter of the window of a dark room with a pin
and causing the rays of the sun to enter horizontally — and if we interpose in
the path of the light a small circular obstacle and then observe the shadow
1888] DIFFRACTION' OF SOUND. 33
thrown in the rear of that obstacle, a very remarkable peculiarity manifests
itself. It is found that in the centre of the shadow of the obstacle, where the
darkness might be expected to be greatest, there is, on the contrary, no
darkness at all, but a bright spot, a spot as bright as if no obstacle
intervened in the course of the light. The history of this subject is curious.
The feet was first observed by Delisle in the early part of the eighteenth
century, but the observation fell into oblivion. When Fresnel began his
important investigations, his memoir on diffraction was communicated to the
French Academy and was reported on by the great mathematician Poisson.
Poisson was not favourably impressed by Fresnel's theoretical views. Like most
mathematicians of the day, he did not take kindly to the wave theory ; and
in his report on Fresnel's memoir, he made the objection that if the method
were applied, as Fresnel had not then done, to investigate what should
happen in the shadow of a circular obstacle, it brought out this paradoxical
result, that in the centre there would be a bright point. This was regarded
as a reductio ad absurd um of the theory. All the time, as I have mentioned,
the record of Delisle's observations was in existence. The remarks of Poisson
were brought to the notice of Fresnel, the experiment was tried, and the
bright point was rediscovered, to the gratification of Fresnel and the confir-
mation of his theoretical views. I don't propose to attempt the optical
experiment now, but it can easily be tried in one's own laboratory. A long
room or passage must be darkened: a fourpenny bit may be used as the
obstacle, strung up by three hairs attached by sealing-wax. When the
shadow of the obstacle is received on a piece of ground glass, and examined
from behind with a magnifying lens, the bright spot will be seen without
much difficulty. But what I propose to show you is the corresponding
phenomenon in the case of sound. Fresnel's reasoning is applicable, word
for word, to the phenomena we are considering just as much as to that which
he, or rather Poisson, had in view. The disc (Fig. 4), which I shall hang up
now between the source of sound and the flame, is of glass. It is about
Fig. 4.
Due
SNBM
15 inches in diameter. I believe the flame is flaring now from being in the
bright spot. If I make a small motion of the disc I shall move the bright
R. III. 3
34
DIFFRACTION OF SOUND.
[145
spot and the effect will disappear. I am pushing the disc away now, and the
flaring has stopped. The flame is still in the shadow of the disc, but not
at the centre. I bring the disc back, and when the flame comes into the
centre it flares again vigorously. That is the phenomenon which was
discovered by Delisle and confirmed by Arago and Fresnel, but mathemati-
cally it was suggested by Poisson.
Poisson's calculation related only to the very central point in the axis
of the disc. More recently the theory of this experiment has been very
thoroughly examined by a German mathematician, Lommel ; and I have
exhibited here one of the curves given by him embodying the results of his
calculations on the subject (Fig. 5).
The abscissae, measured horizontally, represent distances drawn outwards
from the centre of the shadow 0 ', the ordinates measure the intensity of the
light at the various points. The maximum intensity OA is at the centre.
A little way outwards at B the intensity falls almost, but not quite, to zero.
At G there is a revival of intensity, indicating a bright ring ; and further out
there is a succession of subordinate fluctuations. The curve on the other
side of OA would of course be similar. This curve corresponds to the
Fig. 5.
distances and proportions indicated, a is the distance between the source
of sound and the disc; 6 is the distance between the disc and the flame, the
1888] DIFFRACTION OF SOUXD. 35
place where the intensity is observed- The numbers given are taken
the notes of an experiment which went well If we can get our flame to the
right point of sensitiveness we may succeed in bringing into view not onlr
the central spot, but the revived sound which occurs after we have got away
from the central point and have passed through the ring of silence. There is
the loud central point. If I push the disc a little we enter the ring of
silence B* : a little further, and the flame flares again, being now at C.
Although we have thus imitated the optical experiment, I must not
leave you under the idea that we are working under the same conditions that
prevail in optics. You see the diameter of my disc is 15 inches, and the
length of my sound-wave is about half an inck My disc is therefore about
30 wave-lengths in diameter, whereas the diameter of a disc representing
30 wave-lengths of light would be only about j^ inch. Still the conditions
are sufficiently alike to get corresponding effects, and to obtain this bright
point in the centre of the shadow conspicuously developed.
I will now make an experiment illustrating still further the principle of
Huygens' zones, which I have already roughly sketched. I indicated that
the effect of contiguous zones was equal and opposite, so that the effect of
each of the odd zones is one thing, and of the even zones the opposite thing.
If we can succeed in so preparing a screen as to fit the system of zones.
allowing the one set to pass, and at the same time intercepting the other set.
then we shall get a great effect at the central point, because we shall have
removed those parts which, if they remained, would have neutralised the
remaining parts. Such a system has been cut out of zinc, and is now
hanging before you. When the adjustments are correct there will be
produced, under the action of that circular grating, an effect much greater
than would result if the sound-waves were allowed to pass on without any
obstruction. The only point difficult of explanation is as to what happens
when the system of zones is complete, and extends to infinity, viz. when
there is no obstruction at all. In that case it may be proved that the
aggregate effect of all the zones is, in ordinary cases, half the effect that
would be produced by any one zone alone, whereas if we succeed in stopping
out a number of the alternate zones, we may expect a large multiple of the
effect of one zone. The grating is now in the right position, and you see the
flame flaring strongly, under the action of the sound-waves transmitted
through these alternate zones, the action of the other zones being stopped
by the interposition of the zinc. But the interest of the experiment is
principally in this, that the flame is flaring more than it would do if the
grating were removed altogether. There is now, without the grating, a very
trivial flaringf; but when the grating is in position again — though a great
* With the data gran above the diameter of the silent ring is two-thuds of an inch.
t Under the bat conditions the flame is abeolotely unaffected.
3—2
36 DIFFRACTION OF SOUND. [145
part of the sound is thereby stopped out — the effect is far more powerful
than when no obstruction intervened. The grating acts, in fact, the part of
a lens. It concentrates the sound upon the flame, and so produces the
intense magnification of effect which we have seen.
The exterior radius of the nth zone being x, we have, from the formula
given above :
1 1 n\
so that if a and b be the distances of the source and image from the grating,
the relation required to maintain the focus is as usual,
1 + 1,1
a b f
where f, the focal length, is given by
In the actual grating, eight zones (the first, third, fifth, &c.) are occupied by
metal. The radius of the first zone, or central circle, is 3 inches, so that
a?/n = 9. The focal length is necessarily a function of X. In the present case
X = £ inch nearly, and therefore /= 18 inches. If a and b are the same, each
must be made equal to 36 inches.
146.
ON THE RELATIVE DENSITIES OF HYDROGEN AND
OXYGEN. (PRELIMINARY NOTICE.)
[Proceedings of the Royal Society, XLIII. pp. 356—363, 1888.]
THE appearance of Professor Cooke's important memoir upon the atomic
weights of hydrogen and oxygen* induces me to communicate to the Royal
Society a notice of the results that I have obtained with respect to the
relative densities of these gases. My motive for undertaking this investiga-
tion, planned in 1882f, was the same as that which animated Professor Cooke,
namely, the desire to examine whether the relative atomic weights of the two
bodies really deviated from the simple ratio 1 : 16, demanded by Prout's Law.
For this purpose a knowledge of the densities is not of itself sufficient ; but it
appeared to me that the other factor involved, viz., the relative atomic volumes
of the two gases, could be measured with great accuracy by eudiometric
methods, and I was aware that Mr Scott had in view a redetermination of
this number, since in great part carried out*. If both investigations are
conducted with gases under the normal atmospheric conditions as to tempe-
rature and pressure, any small departures from the laws of Boyle and Charles
will be practically without influence upon the final number representing the
ratio of atomic weights.
In weighing the gas the procedure of Regnault was adopted, the working
globe being compensated by a similar closed globe of the same external
volume, made of the same kind of glass, and of nearly the same weight.
In this way the weighings are rendered independent of the atmospheric
* "The Relative Values of the Atomic Weights of Hydrogen and Oxygen," by J. P. Cooke
and T. W. Richards, Amer. Acad. Proc. Vol. xxm., 1887.
t Address to Section A, British Association Report, 1882 [Vol. n. p. 124].
£ "On the Composition of Water by Volume," by A. Scott, Roy. Soc. Proc., June 16, 1887
(Vol. xun. p. 396).
1081)78
38 ON THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. [146
conditions, and only small weights are required. The weight of the globe
used in the experiments here to be described was about 200 grams, and the
contents were about 1800 c.c.
The balance is by Oertling, and readings with successive releasements of
the beam and pans, but without removal of the globes, usually agreed to
^ mg. Each recorded weighing is the mean of the results of several
releasements.
The balance was situated in a cellar, where temperature was very constant,
but at certain times the air currents, described by Professor Cooke, were very
plainly noticeable. The beam left swinging over night would be found still
in motion when the weighings were commenced on the following morning.
At other times these currents were absent, and the beam would settle down
to almost absolute rest. This difference of behaviour was found to depend
upon the distribution of temperature at various levels in the room. A delicate
thermopile with reflecting cones was arranged so that one cone pointed towards
the ceiling and the other to the floor. When the galvanometer indicated that
the ceiling was the warmer, the balance behaved well, and vice versa. The
reason is of course that air is stable when the temperature increases upwards,
and unstable when heat is communicated below. During the winter months
the ground was usually warmer than the rest of the room, and air currents
developed themselves in the weighing closet. During the summer the air
cooled by contact with the ground remained as a layer below, and the balance
was undisturbed.
The principal difference to be noted between my arrangements and those
of Professor Cooke is that in my case no desiccators were used within the
weighing closet. The general air of the room was prevented from getting
too damp by means of a large blanket, occasionally removed and dried
before a fire*.
In Regnault's experiments the globe was filled with gas to the atmospheric
pressure (determined by an independent barometer), and the temperature was
maintained at zero by a bath of ice. The use of ice is no doubt to be recom-
mended in the case of the heavier gases ; but it involves a cleaning of the
globe, and therefore diminishes somewhat the comparability of the weighings,
vacuous and full, on which everything depends. Hydrogen is so light that,
except perhaps in the mean of a long series, the error of weighing is likely to
be more serious than the uncertainty of temperature. I have therefore con-
tented myself with enclosing the body of the globe during the process of
filling in a wooden box, into which passed the bulbs of two thermometers,
reading to tenths of a degree centigrade. It seems probable that the mean
' I can strongly recommend this method. In twenty-four hours the blanket will frequently
absorb two pounds of moisture.
1888] ON THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. 39
of the readings represents the temperature of the gas to about ^th degree, or
at any rate that the differences of temperature on various occasions and with
various gases will be given to at least this degree of accuracy. Indeed the
results obtained with oxygen exclude a greater uncertainty.
Under these conditions the alternate full and empty weighings can be
effected with the minimum of interference with the surface of the globe.
The stalk and tap were only touched with a glove, and the body of the globe
was scarcely touched at all. To make the symmetry as complete as possible,
the counterpoising globe was provided with a similar case, and was carried
backwards and forwards between the balance room and the laboratory exactly
as was necessary for the working globe.
In my earliest experiments (1885) hydrogen and oxygen were prepared
simultaneously in a U-shaped voltameter containing dilute sulphuric acid.
Since the same quantity of acid can be used indefinitely, I hoped in this way
to eliminate all extraneous impurity, and to obtain hydrogen contaminated
only by small quantities of oxygen, and vice versa. The final purification of
the gases was to be effected by passing them through red-hot tubes, and
subsequent desiccation with phosphoric anhydride. In a few trials I did not
succeed in obtaining good hydrogen, a result which I was inclined to attribute
to the inadequacy of a red heat to effect the combination of the small residue
of oxygen*. Meeting this difficulty, I abandoned the method for a time,
purposing to recur to it after I had obtained experience with the more usual
methods of preparing the gases. In this part of the investigation my expe-
rience runs nearly parallel with that of Professor Cooke. The difficulty of
getting quit of the dissolved air when, as in the ordinary preparation of
hydrogen, the acid is fed in slowly at the time of working, induced me to
design an apparatus whose action can be suspended by breaking an external
electrical contact. It may be regarded as a Smee cell thoroughly enclosed.
Two points of difference may be noted between this apparatus and that of
Professor Cooke. In my manner of working it was necessary that the gene-
rator should stand an internal vacuum. To guard more thoroughly against
the penetration of external air, every cemented joint was completely covered
with vaseline, and the vaseline again with water. Again, the zincs were in
the form of solid sheets, closely surrounding the platinised plate on which the
hydrogen was liberated, and standing in mercury. It was found far better to
work these cells by their own electromotive force, without stimulation by
an external battery. If the plates are close, and the contact wires thick,
the evolution of gas may be made more rapid than is necessary, or indeed
desirable.
* From Professor Cooke's experience it appears not improbable that the impurity may have
been sulphurous acid. Is it certain that in his combustions no hydrogen (towards the close
largely diluted with nitrogen) escapes the action of the cupric oxide?
40 ON THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. [146
Tubes, closed by drowned stopcocks, are provided, in order to allow the
acid to be renewed without breaking joints ; but one charge is sufficient for
a set of experiments (three to five fillings), and during the whole of the time
occupied (10 to 14 days) there is no access of atmospheric air. The removal
of dissolved air (and other volatile impurity) proved, however, not to be so
easy as had been expected, even when assisted by repeated exhaustions with
intermittent evolution of hydrogen; and the results often showed a pro-
gressive improvement in the hydrogen, even after a somewhat prolonged
preliminary treatment. In subsequent experiments greater precautions will
be taken*. Experience showed that good hydrogen could not thus be
obtained from zinc and ordinary " pure " sulphuric acid, or phosphoric acid,
without the aid of purifying agents. The best results so far have been from
sulphuric and hydrochloric acid, when the gas is passed in succession over
liquid potash, through powdered corrosive sublimate, and then through
powdered caustic potash. All the joints of the purifying tubes are connected
by fusion, and a tap separates the damp from the dry side of the apparatus.
The latter includes a large and long tube charged with phosphoric anhydride,
a cotton-wool filter, a blow-off tube sealed with mercury until the filling is
completed, besides the globe itself and the Topler pump. A detailed descrip-
tion is postponed until the experiments are complete. It may be sufficient
to mention that there is but one india-rubber connexion, — that between the
globe and the rest of the apparatus, and that the leakage through this was
usually measured by the Topler before commencing a filling or an evacuation.
The object of giving a considerable capacity to the phosphoric tube was to
provide against the danger of a too rapid passage of gas through the purifying
tubes at the commencement of a filling. Suppose the gas to be blowing off,
all the apparatus except the globe (and the Topler) being at a pressure some-
what above the atmospheric. The tap between the damp and dry sides is
then closed, and that into the globe is opened. The gas which now enters
somewhat rapidly is thoroughly dry, having been in good contact with the
phosphoric anhydride. In this way the pressure on the dry side is reduced
to about 2 inches of mercury, but this residue is sufficient to allow the damp
side of the apparatus to be exhausted to a still lower pressure before the tap
between the two sides of the apparatus is re-opened. When this is done, the
first movement of the gas is retrograde ; and there is no danger at any stage
of imperfect purification. The generator is then re-started until the gas
(after from two to five hours) begins to blow off again.
In closing the globe some precaution is required to secure that the pres-
sure therein shall really be that measured by the barometer. The mercury
seal is at some distance from, and at a lower level than, the rest of the
* Spectrum analysis appears to be incapable of indicating the presence of comparatively
large quantities of nitrogen.
1888] ON THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. 41
apparatus. After removal of the mercury the flow of gas is continued for
about one minute, and then the tap between the dry and damp sides is
closed. From three to five minutes more were usually allowed for the com-
plete establishment of equilibrium before the tap of the globe was turned off.
Experiments on oxygen appeared to show that two minutes was sufficient.
For measuring the atmospheric pressure two standard mercury barometers
were employed.
The evacuations were effected by the Topler to at least 5^5, so that the
residual gas (at any rate after one filling with hydrogen) could be neglected.
I will now give some examples of actual results. Those in the following
tables relate to gas prepared from sulphuric acid, with subsequent purification,
as already described : —
Globe (14), empty.
Left Bight
Oct 27— NOT. 5 <7M+0-394 Gn ii •>.
Nov. 7— Nov. 8 . . . . 22-89
Nov. 9— Nov. 10 . . . . 23-00
Nov. 11— Nov. 12.... 21-72
Globe (14), fulL
Date Left Right ^ BaK^neter ] Tempenftne
:^: in. c.
Nov. 5— 7 . . tfM+0-2400 j Gu 2O-52 29416 147
Hbr. 8— 9.. I 0M+0-2364 jj Gu 19-77 u— 12-3
Nov. 10—11 . . <7H+0-23eO Gu 19-18 22-807 11-2
Nov. 12—14 . . (7M+O2340 <?„ 19^1 3O-135 10-3
The second column shows that globe (14) and certain platinum weights
suspended from the left end of the beam, and the third column that (in
this series) only the counterpoising globe (11) was hung from the right
end. The fourth column gives the mean balance reading in divisions of
the scale, each of which (at the time of the above experiments) represented
O000187 gram. The degree of agreement of these numbers in the first part
of the table gives an idea of the errors due to the balance, and to uncertainties
in the condition of the exteriors of the globes. A minute and unsystematic
42 ON THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. [146
correction depending upon imperfect compensation of volumes (to the extent
of about 2 c.c.) need not here be regarded.
The weight of the hydrogen at each filling is deduced, whenever possible,
by comparison of the "full" reading with the mean of the immediately
preceding and following " empty " readings. The difference, interpreted in
grams, is taken provisionally as the weight of the gas. Thus for the filling of
Nov. 5—
H = 0-154 - 2-25 x 0'000187 = 015358.
The weights thus obtained depend of course upon the temperature and
pressure at the time of filling. Reduced to correspond with a temperature of
12°, and to a barometric height of 30 inches (but without a minute correction
for varying temperature of the mercury) they stand thus —
November 5 015811
8 015807
10 015798
12 015792
Mean 015802
The hydrogen obtained hitherto with similar apparatus and purifying
tubes from hydrochloric acid is not quite so light, the mean of two accordant
series being 015812.
The weighing of oxygen is of course a much easier operation than in the
case of hydrogen. The gas was prepared from chlorate of potash, and from a
mixture of the chlorates of potash and soda. The discrepancies between the
individual weighings were no more than might fairly be attributed to ther-
mometric and manometric errors. The result reduced so as to correspond in
all respects with the numbers for hydrogen is 2'5186*.
But before these numbers can be compared with the object of obtaining
the relative densities, a correction of some importance is required, which
appears to have been overlooked by Professor Cooke, as it was by Regnault.
The weight of the gas is not to be found by merely taking the difference of
the full and empty weighings, unless indeed the weighings are conducted
in vacuo. The external volume of the globe is larger when it is full than
when it is empty, and the weight of the air corresponding to this difference
of volume must be added to the apparent weight of the gas.
By filling the globe with carefully boiled water, it is not difficult to deter-
mine experimentally the expansion per atmosphere. In the case of globe (14)
it appears that under normal atmospheric conditions the quantity to be added
to the apparent weights of the hydrogen and oxygen is 0'00056 gram.
* An examination of the weights revealed no error worth taking into account at present.
1888] OX THE RELATIVE DENSITIES OF HYDROGEN AND OXYGEN. 43
The actually observed alteration of volume (regard being had to the
pressibility of water) agrees very nearly with an d priori estimate, founded
upon the theory of thin spherical elastic shells and the known properties of
glass. The proportional value of the required correction, in my case about
ufa of the weight of the hydrogen, will be for spherical globes proportional
to at. where a is the radius of the globe, and t the thickness of the shell, or
to Vj W, if V be the contents, and W the weight of the glass. This ratio is
nearly the same for Professor Cooke's globe and for mine: but the much
greater departure of his globe from the spherical form may increase the
amount of the correction which ought to be introduced.
In the estimates now to be given, which must be regarded as provisional,
the apparent weight of the hydrogen is taken at 0-15804, so that the real
weight is 0-15860. The weight of the same volume of oxygen under the
same conditions is 2-5186 + O0006 = 2*5192. The ratio of these numbers
is 15 884.
The ratio of densities found by Regnault was 15"964, but the greater part
of the difference may well be accounted for by the omission of the correction
just now considered.
In order to interpret our result as a ratio of atomic weights, we need to
know accurately the ratio of atomic volumes. The number given as most
probable by Mr Scott in May, 1887*, was 1-994, but he informs me that more
recent experiments under improved conditions give 1D965. Combining this
with the ratio of densities, we obtain as the ratio of atomic weights —
2 x 15 884
1-9965
15^12.
It is not improbable that experiments conducted on the same lines, but
with still greater precautions, may raise the final number by one or even
two thousandths of its value.
The ratio obtained by Professor Cooke is 15-953: but the difference
between this number and that above obtained may be more than accounted
for, if I am right in my suggestion that his gas weighings require cor-
rection for the diminished buoyancy of the globe when the internal pressure
is removed.
[1901. Further work upon this subject is recorded in Proc. Roy. Soc.
VoL L. p. 449, 1892.]
* Lor. cit. [190L Dr Scott's final number (Proc. Roy. Soc. YoL un. p. 133, 1%3) was
3-00245.]
147.
ON POINT-, LINE-, AND PLANE-SOURCES OF SOUND.
[Proceedings of the London Mathematical Society,
xix. pp. 504—507, 1888.]
THE velocity -potential at a distance p from a simple source of sound is*
where— a~24>1et'*flt represents the rate at which fluid is being introduced at
the source at time t. In order to apply this to a linear source of unit
intensity, coincident with the axis of y, we have to imagine that the intro-
duction of fluid along the element dy is equal to dy eikat ; so that, if for the
sake of brevity we omit the time factor eikat, we may take as the velocity
potential
If r be the distance of the point at which (j> is to be estimated from the
axis of y,
and *=
if p — rv.
The relation of (3) to Bessel's functions is best studied by the method of
Lipschitzf. Consider the integral I .-.•- --- , where w is a complex variable
J V(l + U'2)
of the form u + iv. If we represent, as usual, simultaneous pairs of values of
u and v by the coordinates of a point, the integral will vanish when taken
round any closed circuit not including the points w=±i. The first circuit
we have to consider is that enclosed by the axes of u and v, and the quadrant
of a circle whose centre is the origin and whose radius is infinite. It is easy
* Theory of Sound, §277. t Crelle, Bd. LVI., 1859.
1888] ON POINT-, LINE-, AND PLAXE-SOUKCES OF SOUND. 45
to see that along this quadrant the integral ultimately vanishes, so that the
result is the same whether we integrate from 0 to oo along the axis of u or
from 0 to too along the axis of v. Thus
1 ~^ ° -^
In like manner, the integral along the axis of u from 0 to x is equal to
that along the course from 0 to t along the axis of v, and then to infinity
along a line through i parallel to u. Thus
d(iv)
F* ert'
Jo V(2
By comparison of (4), (5), or at once by equating the results of integrating
from the point t to tx , and to ac + t, we get
[* e~indv = f er^e-^du = er*-
JiA/(^-l) Jo V(2tu + M*) ~ 7(«
r+ l -
--
2tW 1.8w-""l.2.(8»r)F 1 . 2 . 3 (Sir)*
This is the series in descending powers of r by which is expressed the effect
of a linear source at a great distance.
Equation (4) may be written in the form
or, if we put, as usual,
|J cos(rcos^)cW = Jr.(r), ........................ (7)
^/ *8in(rcos^)cW = Jfi:.(r), ..................... (8)
and separate the real and imaginary parts,
r+dfi TT
J
.(10)
the latter giving Mehler's integral expressive of the Bessel's function of
order zero*.
* Math. Am. T. p. 141.
46
ON POINT-, LINE-, AND PLANE-SOURCES OF SOUND.
[147
By integrating the effect of a linear source, parallel to y, with respect to
a perpendicular coordinate x, we may obtain the
effect of a source uniformly distributed over a Fi8- *•
plane. If the rate of introduction of fluid over
the area dxdy be dxdye*"1, the value of ^ at a
point distant z from the plane, will be found by
integrating (3) with respect to x, connected with r
and z by the relation
see Fig. 1, in which
PQ = r, OP = z.
If" f« e~ikrvdv I f
Thus # — .-J( 4rji ^-^ = --]_
rdr
e~ikri>dv
...(11)
The result of a uniform plane source is of course a train of plane waves
issuing from it symmetrically in both directions. On the positive side
<j> = Ae~ikz, where A is a constant readily determined. For ,^ (z=0)=—ikA ;
and this, representing the half of the rate of introduction of fluid per unit
area, is by supposition equal to J. Thus
^ cos kz + ^ sin kz.
,...(12)
,.(13)
Comparing the two expressions for <£, and having regard to (9) and (10), we
see that
rJ0 (kr) rdr _ cos kz
^(r2 — z*) k
rdr [ f°° e-tdfi TT „ ,, I _ sm ....
F^l/o V(/32 + £V)~2-J ~~^~
If we use the series (6), the identity may be written
This equation is easily verified when kz (and therefore kr) is great. Under
these circumstances the series may be replaced by its first term ; also with
sufficient approximation
Vr 1
since only those elements for which r differs little from z contribute sensibly
to the integral.
148.
WAVE THEORY OF LIGHT.
[Encyclopaedia Brttannica, XXIT., 1888.]
§ 1. A GENERAL statement of the principles of the undulatory theory,
with elementary explanations, has already been given tinder Light [Enc.
Brit. ToL xiv.], and in the article on Ether the arguments which point to
the existence of an all-pervading medium, susceptible in its various parts
of an alternating change of state, have been traced by a master hand : but
the subject is of such great importance, and is so intimately involved in
recent optical investigation and discovery, that a more detailed exposition of
the theory, with application to the leading phenomena, was reserved for a
special article. That the subject is one of difficulty may be at once admitted.
Even in the theory of sound, as conveyed by aerial vibrations, where we are
well acquainted with the nature and properties of the vehicle, the fundamental
conceptions are not very easy to grasp, and their development makes heavy
demands upon our mathematical resources. That the situation is not
improved when the medium is hypothetical will be easily understood. For.
although the evidence is overwhebning in favour of the conclusion that
light is propagated as a vibration, we are almost entirely in the dark as to
what it is that vibrates and the manner of vibration. This ignorance
entails an appearance of vagueness even in those parts of the subject the
treatment of which would not really be modified by the acquisition of a
more precise knowledge, c.g., the theory of the colours of thin plates, and of
the resolving power of optical instruments. But in other parts of the
subject, such as the explanation of the laws of double refraction and of the
intensity of light reflected at the surface of a transparent medium, the
vagueness is not merely one of language ; and if we wish to reach definite
results by the d priori road we must admit a hypothetical element, for
which little justification can be given. The distinction here indicated
should be borne clearly in mind. Many optical phenomena must necessarily
48 WAVE THEORY OF LIGHT. [148
agree with any kind of wave theory that can be proposed ; others may agree
or disagree with a particular form of it. In the latter case we may regard
the special form as disproved, but the undulatory theory in the proper wider
sense remains untouched.
Of such special forms of the wave theory the most famous is that which
assimilates light to the transverse vibrations of an elastic solid. Transverse
they must be in order to give room for the phenomena of polarization. This
theory is a great help to the imagination, and allows of the deduction of
definite results which are at any rate mechanically possible. An isotropic
solid has in general two elastic properties — one relating to the recovery from
an alteration of volume, and the other to the recovery from a state of shear,
in which the strata are caused to slide over one another. It has been shown
by Green that it would be necessary to suppose the luminiferous medium to
be incompressible, and thus the only admissible differences between one
isotropic medium and another are those of rigidity and of density. Between
these we are in the first instance free to choose. The slower propagation of
light in glass than in air may be equally well explained by supposing the
rigidity the same in both cases while the density is greater in glass, or by
supposing that the density is the same in both cases while the rigidity is
greater in air. Indeed there is nothing, so far, to exclude a more complicated
condition of things, in which both the density and rigidity vary in passing
from one medium to another, subject to the one condition only of making
the ratio of velocities of propagation equal to the known refractive index
between the media.
When we come to apply this theory to investigate the intensity of light
reflected from (say) a glass surface, and to the diffraction of light by very
small particles (as in the sky), we find that a reasonable agreement with the
facts can be brought about only upon the supposition that the rigidity is the
same (approximately, at any rate) in various media, and that the density
alone varies. At the same time we have to suppose that the vibration is
perpendicular to the plane of polarization.
Up to this point the accordance may be regarded as fairly satisfactory ;
but, when we extend the investigation to crystalline media in the hope
of explaining the observed laws of double refraction, we find that the
suppositions which would suit best here are inconsistent with the conclusions
we have already arrived at. In the first place, and so long as we hold
strictly to the analogy of an elastic solid, we can only explain double
refraction as depending upon anisotropic rigidity, and this can hardly be
reconciled with the view that the rigidity is the same in different isotropic
media. And if we pass over this difficulty, and inquire what kind of double
refraction a crystalline solid would admit of, we find no such correspondence
with observation as would lead us to think that we are upon the right track.
1888] ELECTROMAGNETIC THEORY. 40
The theory of anisotropic solids, with its twenty-one elastic constants, seems
to be too wide for optical double refraction, which is of a much simpler
character*.
For these and other reasons, especially the awkwardness with which it
lends itself to the explanation of dispersion, the elastic solid theory, valuable
as a piece of purely dynamical reasoning, and probably not without
mathematical analogy to the truth, can in Optics be regarded only as
an illustration.
In recent years a theory has been received with much favour in which
light is regarded as an electromagnetic phenomenon. The dielectric medium
is conceived to be subject to a rapidly periodic " electric displacement." the
variations of which have the magnetic properties of an electric current. On
the basis of purely electrical observations Maxwell calculated the velocity
of propagation of such disturbances, and obtained a value not certainly
distinguishable from the velocity of light. Such an agreement is very
striking : and a further deduction from the theory, that the specific inductive
capacity of a transparent medium is equal to the square of the refractive
index, is supported to some extent by observation. The foundations of the
electrical theory are not as yet quite cleared of more or less arbitrary
hypothesis; but, when it becomes certain that a dielectric medium is
susceptible of vibrations propagated with the velocity of light, there will be
no hesitation in accepting the identity of such vibrations with those to which
optical phenomena are due. In the meantime, and apart altogether from the
question of its probable truth, the electromagnetic theory is very instructive.
in showing us how careful we must be to avoid limiting our ideas too much
as to the nature of the luminous vibrations.
§ 2. Plane Waves of Simple Type.
Whatever may be the character of the medium and of its vibration, the
analytical expression for an infinite train of plane waves is
(1)
in which X represents the wave-length, and V the corresponding velocity of
propagation. The coefficient A is called the amplitude, and its nature
depends upon the medium, and must therefore here be left an open question.
The phase of the wave at a given time and place is represented by a. The
expression retains the same value whatever integral number of wave-lengths
* See Stokes. "Report on Doable Refraction." Brit. Attoe. Report, 1862, p. 453-
H. ni. *
50 WAVE THEORY OF LIGHT. [148
be added to or subtracted from x. It is also periodic with respect to t, and
the period is
T-X./F. .................................... (2)
In experimenting upon sound we are able to determine independently T, X,
and V't but, on account of its smallness, the periodic time of luminous
vibrations eludes altogether our means of observation and is only known
indirectly from X and V by means of (2).
There is nothing arbitrary in the use of a circular function to represent
the waves. As a general rule this is the only kind of wave which can be
propagated without a change of form ; and, even in the exceptional cases
where the velocity is independent of wave-length, no generality is really lost
by this procedure, because in accordance with Fourier's theorem any kind of
periodic wave may be regarded as compounded of a series of such as (1), with
wave-lengths in harmonical progression.
A well-known characteristic of waves of type (1) is that any number of
trains of various amplitudes and phases, but of the same wave-length, are
equivalent to a single train of the same type. Thus
= S.4 cos a . cos — - ( Vt — x) — 2 A sin a . sin — ( Vt - x)
A, A*
l/v
where
P2 = (2A cos a)2 + 2 (A sin a)2, tan <£ = /x ' "'(4' 5)
An important particular case is that of two component trains only.
A cos (Vt -a) + « + A' cos ~(F* _^) + «' = Pcos (F* - *) +</>,
where P* = A* + A'2 + 2AA' co$(a- a') ...................... (6)
The composition of vibrations of the same period is precisely analogous,
as was pointed out by Fresnel, to the composition of forces, or indeed of any
other two-dimensional vector quantities. The magnitude of the force corre-
sponds to the amplitude of the vibration, and the inclination of the force
corresponds to the phase. A group of forces of equal intensity, represented
by lines drawn from the centre to the angular points of a regular polygon,
constitute a system in equilibrium. Consequently, a system of vibrations of
equal amplitude and of phases symmetrically distributed round the period has
a zero resultant.
1888] INTERFERENCE. 51
According to the phase-relation, determined by (a — a'), the amplitude of
the resultant may vary from (A — Af) to {A + A'}. If A' and A are equal,
the minimum resultant is zero, showing that two equal trains of wares may
neutralize one another. This happens when the phases are opposite, or differ
by half a (complete) period, and the effect is usually spoken of as the
interference of light. From a purely dynamical point of view the word is
not very appropriate, the vibrations being simply superposed with as little
interference as can be imagined.
§ 3. Intensity.
The intensity of light of given wave-length must depend upon the
amplitude, but the precise nature of the relation is not at once apparent.
We are not able to appreciate by simple inspection the relative intensities of
two unequal lights : and when we say. for example, that one candle is twice
as bright as another, we mean that two of the latter burning independently
would give us the same light as one of the former. This may be regarded as
the definition : and then experiment may be appealed to to prove that the
intensity of light from a given source varies inversely as the square of the
distance. But our conviction of the truth of the law is perhaps founded quite
as much upon the idea that something not liable to loss is radiated outward?,
and is distributed in succession over the surfaces of spheres concentric with
the source, whose areas are as the squares of the radii. The something can
only be energy : and thus we are led to regard the rate at which energy is
propagated across a given area parallel to the waves as the measure of
intensity: and this is proportional, not to the first power, but to the square
of the amplitude.
Practical photometry is usually founded upon the law of inverse squares
(Enc. Brit. VoL xiv. p. 583) : and it should be remembered that the method
involves essentially the use of a diffusing screen, the illumination of which,
seen in a certain direction, is assumed to be independent of the precise
direction in which the light falls upon it : for the distance of a candle, for
example, cannot be altered without introducing at the same time a change in
the apparent magnitude, and therefore in the incidence of some part at any
rate of the light.
With this objection is connected another which is often of greater
importance, the necessary enfeeblement of the light by the process of
diffusion. And, if to maintain the brilliancy we substitute regular reflectors
for diffusing screens, the method breaks down altogether by the apparent
illumination becoming independent of the distance of the source of light.
The use of a revolving disk with transparent and opaque sectors in order
to control the brightness, as proposed by Fox Talbot*. may often be recom-
* Phil. Xm9. VoL T. p. 331, 18J4.
52 WAVE THEORY OF LIGHT. [148
mended in scientific photometry, when a great loss of light is inadmissible.
The law that, when the frequency of intermittence is sufficient to give a
steady appearance, the brightness is proportional to the angular magnitude
of the open sectors appears to be well established.
§ 4. Resultant of a Large Number of Vibrations of Arbitrary Phase.
We have seen that the resultant of two vibrations of equal amplitude is
wholly dependent upon their phase-relation, and it is of interest to inquire
what we are to expect from the composition of a large number (n) of equal
vibrations of amplitude unity, and of arbitrary phases. The intensity of the
resultant will of course depend upon the precise manner in which the phases
are distributed, and may vary from w2 to zero. But is there a definite intensity
which becomes more and more probable as n is increased without limit ?
The nature of the question here raised is well illustrated by the special
case in which the possible phases are restricted to two opposite phases. We
may then conveniently discard the idea of phase, and regard the amplitudes
as at random positive or negative. If all the signs are the same, the intensity
is w8 ; if, on the other hand, there are as many positive as negative, the result
is zero. But, although the intensity may range from 0 to w2, the smaller
values are much more probable than the greater.
The simplest part of the problem relates to what is called in the theory of
probabilities the " expectation " of intensity, that is, the mean intensity to be
expected after a great number of trials, in each of which the phases are taken
at random. The chance that all the vibrations are positive is 2~n, and thus
the expectation of intensity corresponding to this contingency is 2~n . n-. In
like manner the expectation corresponding to the number of positive vibrations
being (n — 1) is
2-» . n . (n - 2)2,
and so on. The whole expectation of intensity is thus
Now the sum of the (n + 1) terms of this series is simply n, as may be proved
by comparison of coefficients of a? in the equivalent forms
(e* + e~x)n = 2B (1
1.2
The expectation of intensity is therefore n, and this whether n be great or
small.
1888] LARGE XTMBER OF VIBRATIONS. 53
The same conclusion holds good when the phases are unrestricted.
From (4), § 2, if A = 1,
P*-=n + 2Scos(os-ai), ........................ (2)
where under the sign of summation are to be included the cosines of the
| ii (n — 1) differences of phase. When the phases are arbitrary, this sum is as
likely to be positive as negative, and thus the mean value of Pa is n.
The reader must be on his guard here against a fallacy which has misled
some high authorities. We have not proved that when n is large there is any
tendency for a single combination to give the intensity equal to n, but the
quite different proposition that in a large number of trials, in each of which
the phases are rearranged arbitrarily, the mean intensity will tend more and
more to the value n. It is true that even in a single combination there is no
reason why any of the cosines in (2) should be positive rather than negative,
and from this we may infer that when n is increased the sum of the terms
tends to vanish in comparison with the number of terms. But, the number
of terms being of the order n*, we can infer nothing as to the value of the sum
of the series in comparison with n.
Indeed it is not true that the intensity in a single combination
approximates to n, when n is large. It can be proved* that the probability
of a resultant intermediate in amplitude between r and r + dr is
(3)
The probability of an amplitude less than r is thus
-4 (4)
or, which is the same thing, the probability of an amplitude greater than r is
(5)
The accompanying table gives the probabilities of intensities less than the
fractions of n named in the first column. For example, the probability of
intensity less than n is '6321.
-05 -0488
•80
•5506
•10
•0952
i-oo
•6321
•20 -1813
1-50
•7768
•40 -3296
2-00
•8647
•60
•4512
3-00
•9502
It will be seen that, however great n may be, there is a fair chance of
considerable relative fluctuations of intensity in consecutive combinations.
* PkiL Mag. Aug. 1880 [Vol. i. p. 491].
54 WAVE THEORY OF LIGHT. [148
The -mean intensity, expressed by
e-r-/n ri rdr,
nJo
is, as we have already seen, equal to n.
It is with this mean intensity only that we are concerned in ordinary
photometry. A source of light, such as a candle or even a soda flame, may be
regarded as composed of a very large number of luminous centres disposed
throughout a very sensible space; and, even though it be true that the
intensity at a particular point of a screen illuminated by it and at a particular
moment of time is a matter of chance, further processes of averaging must
be gone through before anything is arrived at of which our senses could
ordinarily take cognizance. In the smallest interval of time during which
the eye could be impressed, there would be opportunity for any number of
rearrangements of phase, due either to motions of the particles or to
irregularities in their modes of vibration. And even if we supposed that
each luminous centre was fixed, and emitted perfectly regular vibrations,
the manner of composition and consequent intensity would vary rapidly
from point to point of the screen, and in ordinary cases the mean illumi-
nation over the smallest appreciable area would correspond to a thorough
averaging of the phase-relationships. In this way the idea of the intensity
of a luminous source, independently of any questions of phase, is seen to be
justified, and we may properly say that two candles are twice as bright as one.
§ 5. Propagation of Waves in General.
It has been shown under Optics [Vol. n. p. 387], that a system of rays,
however many reflexions or refractions they may have undergone, are always
normal to a certain surface, or rather system of surfaces. From our present
point of view these surfaces are to be regarded as wave-surfaces, that is,
surfaces of constant phase. It is evident that, so long as the radius of
curvature is very large in comparison with A, each small part of a wave-
surface propagates itself just as an. infinite plane wave coincident with the
tangent plane would do. If we start at time t with a given surface, the
corresponding wave-surface at time t + dt is to be found by prolonging every
normal by the length Vdt, where V denotes the velocity of propagation at
the place in question. If the medium be uniform, so that V is constant, the
new surface is parallel to the old one, and this property is retained however
many short intervals of time be considered in succession. A wave-surface
thus propagates itself normally, and the corresponding parts of successive
surfaces are those which lie upon the same normal. In this sense the normal
may be regarded as a ray, but the idea must not be pushed to streams of
1888] FERMAT'S PRINCIPLE. 55
light limited to pass through small apertures. The manner in which the
phase is determined by the length of the ray, and the conditions under which
energy may be regarded as travelling along a ray, will be better treated under
the head of Huygens's principle, and the theory of shadows (§ 10).
From the law of propagation, according to which the wave-surfaces are
always as far advanced as possible, it follows that the course of a ray is that
for which the time, represented \)\ J'V~lds, is a minimum. This is Fermat's
principle of least time. Since the refractive index (/i) varies as V~l, we may
take fad* as the measure of the retardation between one wave-surface and
another ; and it is the same along whichever ray it may be measured.
The principle that ffjds is a minimum along a ray lends itself readily to
the investigation of optical laws. As an example, we will consider the very
important theory of magnifying power. Let A9, B0 be twTo points upon a
wave-surface before the light enters the object-glass of a telescope, A, B the
corresponding points upon a wave-surface after emergence from the eye-piece,
both surfaces being plane. The value of fads is the same along the ray A* A
as along B9B . and, if from any cause B0 be slightly retarded relatively to Aft,
then B will be retarded to the same amount relatively to A. Suppose now
that the retardation in question is due to a small rotation (0) of the wave-
surface A0B0 about an axis in its own plane perpendicular to AB. The
retardation of B0 relatively to A* is then A0BQ . B ; and in like manner, if <f> be
the corresponding rotation of AB, the retardation is AB . <£. Since these
retardations are the same, we have
or tlie magnifying power is equal to the ratio of the widths of the stream of
light before and, after passing the telescope.
The magnifying power is not necessarily the same in all directions.
Consider the case of a prism arranged as for spectrum work. Passage
through the prism does not alter the vertical width of the stream of light;
hence there is no magnifying power in this direction. What happens in a
horizontal direction depends upon circumstances. A single prism in the
position of minimum deviation does not alter the horizontal width of the
beam. The same is true of a sequence of any number of prisms each in the
position of minimum deviation, or of the combination called by Thollon a
couple, when the deviation is the least that can be obtained by rotating the
couple as a rigid system, although a further diminution might be arrived at
by violating this tie. In all these cases there is neither horizontal nor
vertical magnification, and the instrument behaves as a telescope of power
unity. If, however, a prism be so placed that the angle of emergence differs
from the angle of incidence, the horizontal width of the beam undergoes a
change. If the emergence be nearly grazing, there will be a high magnifying
56 WAVE THEORY OF LIGHT. [148
power in the horizontal direction ; and, whatever may be the character of the
system of prisms, the horizontal magnifying power is represented by the ratio
of widths. Brewster suggested that, by combining two prisms with refracting
edges at right angles, it would be possible to secure equal magnifying power
in the two directions, and thus to imitate the action of an ordinary telescope.
The theory of magnifying power is intimately connected with that of
apparent brightness. By the use of a telescope in regarding a bright body,
such, for example, as the moon, there is a concentration of light upon the
pupil in proportion to the ratio of the area of the object-glass to that of
the pupil*. But the apparent brightness remains unaltered, the apparent
superficial magnitude of the object being changed in precisely the same
proportion, in accordance with the law just established.
These fundamental propositions were proved a long while since by Cotes
and Smith ; and a complete exposition of them, from the point of view of
geometrical optics, is to be found in Smith's treatise f.
§ 6. Waves Approximately Plane or Spherical.
A plane wave of course remains plane after reflexion from a truly plane
surface; but any irregularities in the surface impress themselves upon the
wave. In the simplest case, that of perpendicular incidence, the irregularities
are doubled, any depressed portion of the surface giving rise to a retardation
in the wave-front of twice its own amount. It is assumed that the lateral
dimensions of the depressed or elevated parts are large multiples of the
wave-length ; otherwise the assimilation of the various parts to plane waves
is not legitimate.
In like manner, if a plane wave passes perpendicularly through a parallel
plate of refracting material, a small elevation t at any part of one of the
surfaces introduces a retardation (p — l)t in the corresponding part of the
wave-surface. An error in a glass surface is thus of only one-quarter of the
importance of an equal error in a reflecting surface. Further, if a plate,
otherwise true, be distorted by bending, the errors introduced at the two
surfaces are approximately opposite, and neutralize one another^.
* It is here assumed that the object-glass is large enough to till the whole of the pupil with
light ; also that the glasses are perfectly transparent, and that there is no loss of light by
reflexion. For theoretical purposes the latter requirement may be satisfied by supposing the
transition between one optical medium and another to be gradual in all cases.
t Smith, Compleat System of Optics, Cambridge, 1738. The reader may be referred to a
paper entitled "Notes, chiefly Historical, on some Fundamental Propositions in Optics"
(Phil. Mag. June 1886 [Vol. n. Art. 137]), in which some account is given of Smith's work,
and its relation to modern investigations.
J On this principle Grubb has explained the observation that the effects of bending stress
are nearly as prejudicial in the case of thick object-glasses as in the case of thin ones.
1888] ABERRATION. 57
In practical applications it is of importance to recognize the effects of a
small departure of the waveHSurface from its ideal plane or spherical form.
Let the surface be referred to a system of rectangular coordinates, the axis of
z being normal at the centre of the section of the beam, and the origin being
the point of contact of the tangent plane. If. as happens in many cases, the
surface be one of symmetry round UZ. the equation of the surface may be
represented approximately by
(1)
in which p is the radius of curvature, or focal length, and r = JT + if. If the
surface be truly spherical, A = 1 8/»*, and any deviation of A from this value
indicates ordinary symmetrical spherical aberration.
If, however, the surface be not symmetrical, we may have to encounter
aberration of a lower order of small quantities, and therefore presumably of
higher importance. By taking the axis of x and y coincident with the
directions of principal curvature at O, we may write the equation of the
surface
p, p" being the principal radii of curvature, or focal lengths. The m«:tst
important example of unsymmetrical aberration is in the spectra-cope, where
(if the faces of the prisms may be regarded as at any rate surfaces of
revolution) the wave-surface may by suitable adjustments be rendered
symmetrical with respect to the horizontal plane y = 0. This plane may
then be regarded as primary, p being the primary focal length, at which
distance the spectrum is formed. Under these circumstances £ and £ may
be omitted from (2), which thus takes the form
The constants a and 7 in (3) may be interpreted in terms of the differential
coefficients of the principal radii of curvature. By the usual formula the
radius of curvature at the point jc of the intersection of (3) with the plane
y — 0 is approximately p (1 — Qapjc). Since y = 0 is a principal plane through-
out, this radius of curvature is a principal radius of the surface: so that,
denoting it by p, we have
Again, in the neighbourhood of the origin, the approximate value of the
product of the principal curvatures is
PP
58 WAVE THEORY OF LIGHT. [148
Thus
, / 1 \ _ dp _ dp __ (ioix 2ya
\pp'J ~ p-p p'p p p
whence by (4)
The equation of the normal at the point x, y, z is
. (6)
'
__
p'~ly + tyxy '
and its intersection with the plane £ = p occurs at the point determined
approximately by
7; = p ~ ± - Zpyjy, ............ (7)
terms of the third order being omitted.
According to geometrical optics, the thickness of the image of a luminous
line at the primary focus is determined by the extreme value of £ ; and for
good definition in the spectroscope it is necessary to reduce this thickness as
much as possible. One way of attaining the desired result would be to
narrow the aperture ; but, as we shall see later, to narrow the horizontal
aperture is really to throw away the peculiar advantage of large instruments.
The same objection, however, does not apply to narrowing the vertical
aperture ; and in many spectroscopes a great improvement in definition may
be thus secured. In general, it is necessary that both 7 and a be small.
Since the value of £ does not depend on p', it would seem that in respect of
definition there is no advantage in avoiding astigmatism.
The width of the image when 77 = 0 (corresponding to y = 0) is 3a.pi*;-, and
vanishes when a = 0, i.e., Avhen there is no aberration for rays in the primary
plane. In this case the image reduces to a linear arc. If further 7 = 0, this
arc becomes straight, and then the image at the primary focus is perfect to
this order of approximation. As an example where a = 0, the image of a
luminous point, formed at an equal distance on the further side of a sloped
equi-convex lens, may be mentioned.
At the secondary focus, %=p, and from (6)
^ = -2/7^ ...................... (8)
If 7 = 0, the secondary focal line is formed without aberration, but not
otherwise. Both focal lines are well formed when parallel rays fall upon a
plano-convex lens, sloped at about 30°, the curved side of the lens being
turned towards the parallel rays.
FRESNEL'S EXPERIMENT. 59
[1900. A plane reflecting plate, which reflects approximately plane
waves perpendicularly, may be bent so as to eliminate all the errors of
wave-front represented in (2). For the solution of the equation of bending,
viz.
,
da? dx*dy* dy*
includes arbitrary terms in all powers of x, y below the fourth. Errors of the
same nature in the (unbent) surface of the reflector are eliminated at the
same time.]
§ 7. Interference Fringes.
We have seen (§ 2) that, when two trains of parallel waves of equal
wave-length are superposed, the intensity of the resultant depends upon the
phase-relation of the components : but it is necessarily the same at all points
of the wave-front. It not unfrequently happens that the parallelism of the
component trains is approximate only, and there then arises the phenomenon
known as interference fringes. If the two directions of propagation be
inclined on opposite sides to the axis of x at small angles a, the expressions
for two components of equal amplitudes are
cos ~^— -I Vt — a: cos a— y sin ah , and cos "- \ Vt — xcosa + y sin a- ;
so that the resultant is expressed by
(1)
from which it appears that the vibrations advance parallel to the axis of «r,
unchanged in type, and with a uniform velocity F'coso. Considered as
depending on y, the vibration is a maximum when y sin a is equal to 0. X, 2X,
3X, etc., corresponding to the centres of the bright bands, while for the
intermediate values £X, |X, &c., there is no vibration. This is the interference
of light proceeding from two similar homogeneous and very distant sources.
In the form of experiment adopted by Fresnel the sources 01; 0,* are
situated at a finite distance D from the place of observation (Enc. Brit.
Vol. xiv. p. 606). If A be the point of the screen equidistant from 01? 0..
and P a neighbouring point, then approximately
0>P - 0,P = S{D* + (H + ±by>] - J{& + (u - £&)*} = ub/D,
where 010. = b,AP = u.
* It is scarcely necessary to say that C\ . O* mast not be distinct sources of light ; otherwise
there could be no fixed phase-relation and consequently no regular interference. In Fresnel's
experiment O,, O, are virtual images of one real source 0, obtained by reflexion in two mirrors.
The mirrors may be replaced by a bi-prism. Or, as in Lloyd's arrangement, 01 may be identical
with O, and O» obtained by a grazing reflexion from a single mirror.
60 WAVE THEORY OF LIGHT. [148
Thus, if \ be the wave-length, the places where the phases are accordant
are given by
u = n\D/b, (2)
n being an integer.
If the light were really homogeneous, the successive fringes would be
similar to one another and unlimited in number ; moreover there would be no
place that could be picked out by inspection as the centre of the system. In
practice A, varies, and the only place of complete accordance for all kinds of
light is at A , where u = 0. Theoretically, there is no place of complete
discordance for all kinds of light, and consequently no complete blackness.
In consequence, however, of the fact that the range of sensitiveness of the eye
is limited to less than an "octave," the centre of the first dark band (on
either side) is sensibly black, even when white light is employed ; but it
should be carefully remarked that the existence of even one band is due to
selection, and that the formation of several visible bands is favoured by the
capability of the retina to make chromatic distinctions within the visible
range.
The number of perceptible bands increases pari passu with the approach
of the light to homogeneity. For this purpose there are two methods that
may be used.
We may employ light, such as that from the soda flame, which possesses
ab initio a high degree of homogeneity. If the range of wave-length included
be 5otfoo> a corresponding number of interference fringes maybe made visible.
The above is the number obtained by Fizeau, and Michelson has recently
gone as far as 200,000. The narrowness of the bright line of light seen in
the spectroscope, and the possibility of a large number of Fresnel's bands,
depend upon precisely the same conditions ; the one is in truth as much an
interference phenomenon as the other.
In the second method the original light may be highly composite, and
homogeneity is brought about with the aid of a spectroscope. The analogy
with the first method is closest if we use the spectroscope to give us a line of
homogeneous light in simple substitution for the artificial flame. Or,
following Foucault and Fizeau, we may allow the white light to pass, and
subsequently analyse the mixture transmitted by a narrow slit in the screen
upon which the interference bands are thrown. In the latter case we
observe a channelled spectrum, with maxima of brightness corresponding to
the wave-lengths bu/(nD). In either case the number of bands observable is
limited solely by the resolving power of the spectroscope (§ 13), and proves
nothing with respect to the regularity, or otherwise, of the vibrations of the
original light.
The truth of this remark is strikingly illustrated by the possible formation,
with white light, of a large number of achromatic bands. The unequal
1888] ACHROMATIC BAXDS. 61
widths of the bands tor the various colours, and consequent overlapping and
obliteration, met with in the usual form of the erperiment, depend upon the
constancy of 6 (the- mutual distance of the two sources) while X varies. It is
obvious that, if 6 were proportional to X, the widths of the bands would be
independent of X, and that the various systems would fit together perfectly.
To cam- out the idea in its entirety, it would be necessary to use a diffraction
spectrum as a source, and to duplicate this by Lloyd's method with a single
reflector placed so that 6 = 0 when X = 0. [Phil. Mag. xxvin. p. 77, 1889.]
In practice a sufficiently good result could doubtless be obtained with a
prismatic spectrum (especially if the red and violet were removed by absorbing
agents) under the condition that d(6X) = 0 in the yellow-green. It is
remarkable that, in spite of the achromatic character of the bands, their
possible number is limited still by the resolving power of the instrument
used to form the spectrum.
If a system of Fresnel's bands be examined through a prism, the central
white band undergoes an abnormal displacement, which has been supposed to
be inconsistent with theory. The explanation has been shown by Airy* to
depend upon the peculiar manner in which the white band is in general
formed.
"Any one of the kinds of homogeneous light composing the incident heterogeneous
light will produce a series of bright and dark bars, unlimited in number as far as the
mucture of light from the two pencils extends, and undistinguishable in quality. The
consideration, therefore, of homogeneous light will never enable us to determine which is
the point that the eye immediately turns to as the centre of the fringes. What then is
the physical circumstance that determines the centre of the fringes ?
u The answer is very easy. For different colours the bars have different breadths. If
then the bars of all colours coincide at one part of the mixture of light, they will not
coincide at any other part; but at equal distances on both sides from that place of
coincidence they will be equally far from a state of coincidence. If then \ve can find where
the bars of all colours coincide, that point is the centre of the fringes.
" It appears then that the centre of the fringes is not necessarily the point where the
two pencils of light have described equal paths, but is determined by considerations
of a perfectly different land. The distinction is important in this and in other
The effect in question depends upon the dispersive power of the prism.
If v be the b'near shifting due to the prism of the originally central band,
v must be regarded as a function of X. Measured from the original centre,
the position of the n* bar is now
w + nXD/6.
* " Remarks on Mr Potter's Experiment on Interference." Pkil. Mag. Vol. n. p. 161, 1833.
62 WAVE THEORY OF LIGHT. [148
The coincidence of the various bright bands occurs when this quantity is as
independent as possible of \, that is, when n is the nearest integer to
b dv
n = — =;
or, as Airy expresses it in terms of the width of a band (h), n = — dv/dh.
The apparent displacement of the white band is thus not v simply, but
-*a .......... : .......................... w
The signs of dv and dh being opposite, the abnormal displacement is in
addition to the normal effect of the prism. But, since dv/dh, or dv/d\, is not
constant, the achromatism of the white band is less perfect than when no
prism is used.
If a grating were substituted for the prism, v would vary as h, and
(4) would vanish, so that in all orders of spectra the white band would bo
seen undisplaced.
The theoretical error, dependent upon the dispersive power, involved in
the method of determining the refractive index of a plate by means of the
displacement of a system of interference fringes (Enc. Brit. Vol. xiv. p. 607)
has been discussed by Stokes*. In the absence of dispersion the retardation
R due to the plate would be independent of X, and therefore completely
compensated at the point determined by u = DR/b ; but when there is
dispersion it is accompanied by a fictitious displacement of the fringes on the
principle explained by Airy.
More recently the matter has engaged the attention of Goran f, who thus
formulates the general principle: — "Dans un systeme de /ranges $ interferences
prodaites a I' aide d'une lumiere heterogene ay ant un spectre continu, il existe
toujours une /range achromatique qui joue le role de /range centrale et qui se
trouve au point de champ ou les radiations les plus intenses presentent une
difference de phase maximum ou minimum."
In Fresnel's experiment, if the retardation of phase due to an interposed
plate, or to any other cause, be F (\), the whole relative retardation of the
two pencils at the point u is
(5)
and the situation of the central, or achromatic, band is determined, not by
<f> = 0, but by d<f>jd\ = 0, or
u = \*DF'(\)/b ............................... (6)
* Brit. Assoc. Rep., 1850. f Jour, de Physique, i. p. 293, 1882.
1888] AIRTS THE«»RY. 63
In the theoretical statement we have supposed the source of light to be
limited to a mathematical point, or to be extended only in the vertical
direction (parallel to the bands). Such a vertical extension, while it increases
illumination, has no prejudicial effect upon distinctness, the various systems
due to different points of the luminous line being sensibly superposed. On
the other hand, the horizontal dimension of the source must be confined
within narrow limits, the condition obviously being that the displacement of
the centre of the system incurred by using in succession the two edges only
of the slit should be small in comparison with the width of an interference
band.
Before quitting this subject it is proper to remark that Fresnel's bands
are more influenced by diffraction than their discoverer supposed. On this
account the fringes are often unequally broad and undergo fluctuations of
brightness. A more precise calculation has been given by H. F. Weber* and
by H. Strove^ but the matter is too complicated to be further considered
here. The observations of Strove appear to agree well with the comet-ted
theory.
§ 8. Colours of Thin Plates.
When plane waves of homogeneous light (X) fall upon a parallel plate of
index /*, the resultant reflected wave is made up of an infinite number of
components, of which the most important are the first, reflected at the upper
surface of the plate, and the second, transmitted at the upper surface,
reflected at the under surface, and then transmitted at the upper surface.
It is readily proved (Enc. Brit Vol. xiv. p. 608) that so far as it depends
[Kg. 0-1
upon the distances to be travelled in the plate and in air the retardation (5)
of the second wave relatively to the first is given by
8 = 2^cosa', (1)
* Wied. Amm. nn. p. 407. t Wted. Am*. XT. p. 49.
64 WAVE THEORY OF LIGHT. [148
where t denotes the thickness of the plate, and a' the angle of refraction
corresponding to the first entrance. [1900. ABF=2a, BCD = 2a',
= 2pBC - 2BC sin a' sin a
= 2pBC(l - sin2 a') = 2/*£ cos a'.]
If we represent all the vibrations by complex quantities, from which finally
the imaginary parts are to be rejected, the retardation 8 may be expressed by
the introduction of the factor e~iKS, where i=\/(— 1), and « = 27T/X.
At each reflexion or refraction the amplitude of the incident wave must
be supposed to be altered by a certain factor. When the light proceeds from
the surrounding medium to the plate, the factor for reflexion will be supposed
to be 6, and for refraction c ; the corresponding quantities when the progress
is from the plate to the surrounding medium will be denoted by e, f.
Denoting the incident vibration by unity, we have then for the first com-
ponent of the reflected wave b, for the second cefe~'*s, for the third ce?f€~-{*s,
and so on. Adding these together, and summing the geometric series, we
find
6+rS <2>
In like manner for the wave transmitted through the plate we get
The quantities b, c, e,f are not independent. The simplest way to find
the relations between them is to trace the consequences of supposing 8 = 0 in
(2) and (3). For it is evident d priori that with a plate of vanishing
thickness there must be a vanishing reflexion and a total transmission.
Accordingly,
b + e = 0, c/=l-e2, ........................... (4)
the first of which embodies Arago's law of the equality of reflexions, as well
as the famous " loss of half an undulation." Using these we find for the
reflected vibration,
and for the transmitted vibration
The intensities of the reflected and transmitted lights are the squares of
the moduli of these expressions. Thus
Intensity of reflected light
_ (1 - cos *:S)2 + sin2 /c8 4e« sin2 (j*&) m
(1 - e2 cos *S)2 + e4 sin2 rcS ~ 1 - 2e2 cos rc8 + e4 ' ......... "'
1888] NEWTON'S RINGS. 65
Intensity of transmitted light
-(8)
the sum of the two expressions being unity.
According to (7) not only does the reflected light vanish completely when
5 = 0, but also whenever |c5 = mr, • being an integer, that is, whenever
8 = nX. When the first and third medium are the same, as we have here
supposed, the central spot in the system of Newton's rings is black, even
though the original light contain a mixture of all wave-lengths. The general
explanation of the colours of Newton's rings is given under "Light™ [E*f. Brit.
YoL XIT.]T to which reference must be made. If the light reflected from a
plate of any thickness be examined with a spectroscope of sufficient resolving
power (| 13), the spectrum will be traversed by dark bands, of which the
centres correspond to those wave-lengths which the plate is incompetent to
reflect. It is obvious that there is no limit to the fineness of the bands
which may be thus impressed upon a spectrum, whatever may be the
character of the original mixed light.
[1900. As ordinarily observed, Xewton's rings depend upon the variable
thickness of the thin plate, which is seen in focus. This disposition implies
that the rays which proceeding from a given part of the plate and filling the
aperture of the eye are ultimately brought to a point upon the retina, are
incident at various obliquities. The confusion is least when the incidence is
approximately perpendicular, and it is usually of no importance when the
whole retardation is small, as when coloured bands are formed from white
light. But when we proceed to high interference the difficulty arising from
variable obliquity increases,, and it becomes necessary to pay great attention
to the perpendicularity of the incidence, and perhaps to contract the aperture
of the eye. A stage is soon reached at which it is better to abandon this
procedure altogether and to focus the eye, not upon the plate, but for an
infinite distance, so as to combine at one point of the retina rays which are
incident in a give* direction. If the surfaces of the plate are absolutely
parallel, an ideal ring system is then formed, the centre of the system
corresponding to perpendicular incidence, and each ring to a definite degree
of obliquity. Accurately parallel surfaces may be obtained very simply from
a layer of water resting upon mercury (Nature, XJ.VTLL p. 212, 1893). In this
method no slit, or limitation of the beam, otherwise than in the pupil of the
eye, is anywhere required.
The illumination depends upon the intensity of the monochromatic source
and upon the reflecting power of the surfaces. If R denote the intensity of
reflected light, as given in (7),
66 WAVE THEORY OF LIGHT. [148
If e = 1 absolutely,
l/R = R = l
for all values of 8. If e = 1 very nearly, R = 1 nearly for all values of 8 for
which sin(£/cS) is not very small. The field will be of the full brightness
corresponding to the source, but will be traversed by narrow black lines.
This condition of things may be approximated to in the case of the layer
of water over mercury by making the reflexion very oblique. The experiment
in this form succeeds, but the high obliquity is inconvenient. In the
researches of MM. Fabry and Perot the transmitted light is employed with
an incidence approximately perpendicular. If a transparent plate could be
composed of material for which e = 1 nearly, the transmitted light ( 1 — R)
would nearly vanish except when sin (^tc&) is close to zero. The field would
be dark in general, but be traversed by narrow bright lines. Unfortunately
there is no transparent material giving nearly complete reflexion at perpen-
dicular incidence, but MM. Fabry and Perot have obtained very interesting
results by the use of lightly silvered glass surfaces. The silvered surfaces
may include a plate of air, of which the thickness can then be regulated, or
they may be the external surfaces of a plate of glass, which needs to be very
accurately formed. This arrangement constitutes a spectroscope, inasmuch as
it allows the structure of a complex spectrum line to be directly observed.
If for example we look at a soda flame, we see in general two distinct systems
of narrow bright circles corresponding to the two D-lines. With particular
values of the thickness of the plate of air the two systems may coincide so as
to be seen as a single system, but a slight alteration of thickness will cause a
separation. One peculiarity of the light from a soda flame will at once strike
the observer more conspicuously than with any other form of spectroscope.
If the flame contains but little soda, the lines of the two systems are very
unequal in brightness, but the difference greatly diminishes when the supply
of soda is increased, as would be necessary from the first in other methods of
observation. In using this apparatus the eye of the observer must be focused
for infinity, and the adjustment of the reflecting surfaces to parallelism must
be very exact. A small movement of the eye in any direction should not
entail an expansion or contraction of the rings.
In Michelson's apparatus the colours reflected from a thin plate are
obtained without actual approximation of the reflecting surfaces. By means
of it Michelson has made a very thorough and successful comparison of the
standards of length and the wave-lengths of the radiation obtained by electric
discharge from cadmium vapour in a vacuum tube.]
The relations between the factors b, c, e, f have been proved, independently
of the theory of thin plates, in a general manner by Stokes*, who called to his
* " On the Perfect Blackness of the Central Spot in Newton's Eings, and on the Verification
of Fresnel's Formulae for the Intensities of Reflected and Refracted Rays." Camb. and Dub.
Hath, Jour. Vol. iv. p. 1, 1849; reprint Vol. n. p. 89.
1888] PRINCIPLE OF REVERSIBILITY. 67
aid the general mechanical principle of reversibility. If the motions constitut-
ing the reflected and refracted rays to which an incident ray gives rise be
supposed to be reversed, they will reconstitute a reversed incident ray. This
gives one relation : and another is obtained from the consideration that there
is no ray in the second medium, such as would be generated by the operation
alone of either the reversed reflected or refracted rays. Space
does not allow of the reproduction of the argument at length, but
a few words may perhaps give the reader an idea of how the
conclusions are arrived at. The incident ray {I A) being 1, the
reflected (A R) and refracted (AF) rays are denoted by 6 and c.
When 6 is reversed, it gives rise to a reflected ray 6s along AI,
and a refracted ray be along A G (say). When c is reversed, it
gives rise to cf along AI, and ce along AG. Hence
fr» + c/= 1, which agree with (4).
It is here assumed that there is no change of phase in the
act of reflexion or refraction, except such as can be represented by a change
of sign. Professor Stokes has, however, pushed the application of his method
to the case where changes of phase are admitted, and arrives at the conclusion
that " the sum of the accelerations of phase at the two reflexions is equal to
the sum of the accelerations at the two refractions, and the accelerations of
the two refractions are equal to each other." The accelerations are supposed
to be so measured as to give like signs to c and /" and unlike to 6 and e.
The same relations as before obtain between the factors 6, c, e,f, expressing
the ratios of amplitudes*.
When the third medium differs from the first, the theory of thin plates is
more complicated, and need not here be discussed. One particular case,
however, mav be mentioned- When a thin transparent film is backed by a
perfect reflector, no colours should be visible, all the light being ultimately
reflected, whatever the wave-length may be. The experiment may be tried
with a thin layer of gelatin on a polished silver plate. In other cases where
* It would appear, however, that these laws cannot be properly applied to the calculation of
reflexion from a thin plate. This is sufficiently proved by the fact that the resultant expression
for the intensity founded upon them does not vanish with the thickness. The truth is that the
method of deducing the aggregate reflexion from the consideration of the successive partial
reflexions and refractions is applicable only when the disturbance in the interior of the plate is
fully represented by die transverse waves considered in the argument, whereas the occurrence of
a change of phase is probably connected with the existence of additional superficial waves (§ 27).
The existence of these superficial waves may be ignored when the reflected and refracted waves
are to be considered only at distances from the surface exceeding a few wave-lengths, but in the
application to thin plates this limitation is violated. If indeed the method of calculating the
the expressions (2) and (3), merely understanding by b, c, e, /, factors which may be complex ;
and the same formal relations (4) would still hold good. These do not agree with those found by
Stokes by the method of reversion ; and the discrepancy indicates that, when there are
of phase, the action of a thin plate cannot be calculated in the usual war.
5—2
68 WAVE THEORY OF LIGHT. [148
a different result is observed, the inference is that either the metal does not
reflect perfectly, or else that the material of which the film is composed is not
sufficiently transparent.
Theory and observation alike show that the transmitted colours of a thin
plate, e.g., a soap film or a layer of air, are very inferior to those reflected.
Specimens of ancient glass, which have undergone superficial decomposition,
on the other hand, sometimes show transmitted colours of remarkable
brilliancy. The probable explanation, suggested by Brewster, is that we
have here to deal not merely with one, but with a series of thin plates of
not very different thicknesses. It is evident that with such a series the
transmitted colours would be much purer, and the reflected much brighter,
than usual. If the thicknesses are strictly equal, certain wave-lengths must
still be absolutely missing in the reflected light ; while on the other hand a
constancy of the interval between the plates will in general lead to a special
preponderance of light of some other wave-length for which all the component
parts as they ultimately emerge are in agreement as to phase*.
All that can be expected from a physical theory is the determination of
the composition of the light reflected from or transmitted by a thin plate in
terms of the composition of the incident light. The further question of
the chromatic character of the mixtures thus obtained belongs rather to
physiological optics, and cannot be answered without a complete knowledge
of the chromatic relations of the spectral colours themselves. Experiments
upon this subject have been made by various observers, and especially by
Maxwell f, who has exhibited his results on a colour diagram as used by
Newton. A calculation of the colours of thin plates, based upon Maxwell's
data, and accompanied by a drawing showing the curve representative of the
entire series up to the fifth order, has recently been published J ; and to this
the reader who desires further information must be referred, with the remark
that the true colours are not seen in the usual manner of operating with a
plate of air enclosed between glass surfaces, on account of the contamination
with white light reflected at the other surfaces of the glasses. This objection
is avoided when a soap film is employed, to the manifest advantage of the
darker colours, such as the red of the first order. The colours of Newton's
scale are met with also in the light transmitted by a somewhat thin plate
of doubly-refracting material, such as mica, the plane of analysis being
perpendicular to that of primitive polarization.
* The analytical investigations and formulae given by Stokes for a pile of plates (Proc. Boy.
Soe. Vol. xi. p. 545, 1860) may be applied to this question, provided that we understand the
quantities r, f, £, f , Ac., to be complex, so as to express the luminous displacement in phase as
well as in amplitude, instead of real quantities relating merely to intensities.
t Maxwell, "Theory of Compound Colours," Phil. Trtuu., 1860.
* EtKn, Trtau., 1887 [Vol. n. p. 498].
: — ;
The aaiae aeries rfcdlmnis«K»al»»oth^
the eentore- ®£ tine- ifflnDminatted ana when fight J
through a smaJIE imMmui apertrwe- im an otherwise
Tie eolooars oi? Mr&iofa. we KEITO fteem
peipendkabr meifence,. *> that IL&K-
greatly departed, from when the- thin
atnjol the- mjeideiuee- i* sacSn nftaifr aT
cofflseqiTKaaice ol tifoe- ptDiwerfiml! fifisj»eirs»3iiD,.
firom one etD&Hmr to araMher. Umier ttBDese- (OBwnnifflfitemiittes ttfl»& aeiriksi <aitf' ffldbums:
entirely alters its eftaaraeter,. and the hands fcMmmqiKmdimg to a
e vem kae ttftieir ciofiscaiUBOBB,. fcgGioioniiniig: senahijr Mock ami
naaaj aftrannsidJifiMB*. Hfr gemeraJl espiknitBimani «otf ttiniis.
s sagg?5i5Ced fcy Xewtnoin^ Iraitt it «&se* moid -syyfew m>
&X3& mi acerMr^miEe \irMn ttfine-
tibaC pfiame- waiTes of whitta
* raprani a pflatte rf am; wftmdn o* bamnded agaak .Dm ttlbfr
Iff » he- the-, imfifes <of «&» £&•&$, «" ttfins- aon^Cg- nrff'
aami ttBae- ro^tamfiBttBOfflv eapreaeeii fty ttBftt-
ftseea"— jft-fttana"Hna
and the uriaiiliiiiiHi in jpnaw is 2tf cwff a"/ X X ftitjniisr as tiosiiiali iOf
The fesfi thing to he nolieed as ttfeate,. wiceffl
i af he«Mnes as aaaall as ice pbaa^, and that «t5HiHefignBan«iIhr irfee
to a gwen thaekmess QJ vianr nmiaefei Gese- tftajE an
Hawse the giase- aadaees meed mott le $®> d»o»r a^
A aexnad jfeatme is *&*» Sim.iiiu^iHI hnEffiianz^' <aif' frftufr Hwfti
to (7) the intenaitty dT the idhaetted ffijght wteni alt a TmaMnnmmmu (I-SHIL |«c£ = I))
is 4rfjf|l -H-^f- At peipendiealnr ^JM'^A^rg' € is- afcioum i,. ami H£K- iinuieciscnj-
is somewhat, anaffl: hat^, as cneaf ap|avadbes aenoi. <r af ^inQjasiis? mmmiiij «| fto n.
and the Hnlli*ntj^ is anarh nmereased.
Bat the peealiarilQr which
innnBnoe of a T-iiiiT^yn* in X
af X of ittsdtf' inmiiaaet the zrefiapifittixofln off ptoase,. flrntt,, arnmnft wa\re* -otf dkntter
-^±-~.-^—.-. •.:•-: L_ :•-: :•-:-•,.: _: : ' .- -.fv.: _..-, •-. L_ - : •.-.- :•-:.":'-.-.:._
fejr tine- greater loMiii^mitnt anii ©i}cua£i'|i]Baniii «fliinmiiiniiiimiiiu)ffli iim the
of cos a". We wffl Einiwstoigrattfr tBue leiainufinsniiiiiff tmna&r wMj^n tdire
3f zte- df a ^unadnkom df ' X
ThBiil.. HO. X*& VdL m. p. Jflfl^
70 WAVE THEORY OF LIGHT. [148
In order that X-1 cos a' may be stationary, we must have
X sin a' da + cos a'd\ = 0,
where (a being constant)
cos a da = sin a dp.
Thus
. X da
cot2 a = — -T- , (9)
giving a' when the relation between a and X is known.
According to Cauchy's formula, which represents the facts very well
throughout most of the visible spectrum,
a = A + B\-2, (10)
so that
If we take, as for Chance's " extra-dense flint," B = '984 x 10~10, and as for
the soda lines, ^ = T65, X = 5'89 x 10~8, we get
a = 79° 30'.
At this angle of refraction, and with this kind of glass, the retardation of
phase is accordingly nearly independent of wave-length, and therefore the
bands formed, as the thickness varies, are approximately achromatic. Perfect
achromatism would be possible only under a law of dispersion
fj? = A' - B'tf.
If the source of light be distant and very small, the black bands are
wonderfully fine and numerous. The experiment is best made (after
Newton) with a right-angled prism, whose hypothenusal surface may be
brought into approximate contact with a plate of black glass. The bands
should be observed with a convex lens, of about 8 inches focus. If the eye
be at twice this distance from the prism, and the lens be held midway
between, the advantages are combined of a large field and of maximum
distinctness.
If Newton's rings are examined through a prism, some very remarkable
phenomena are exhibited, described in his twenty-fourth observation*:
"When the two object-glasses are laid upon one another, so as to make
the rings of the colours appear, though with my naked eye I could not
discern above eight or nine of those rings, yet by viewing them through a
prism I could see a far greater multitude, insomuch that I could number
more than forty And I believe that the experiment may be improved
to the discovery of far greater numbers But it was but one side of these
rings, namely, that towards which the refraction was made, which by the
* Newton's Optics. See also Place, Pogg. Ann. cxiv. p. 504, 1861.
1888] ACHROMATIC BANDS. 71
refraction was rendered distinct, and the other side became more confused
than when viewed with the naked eye
" I have sometimes so laid one object-glass upon the other that to the
naked eye they have all over seemed uniformly white, without the least
appearance of any of the coloured rings ; and yet by viewing them through
a prism great multitudes of those rings have discovered themselves."
Newton was evidently much struck with these " so odd circumstances " ;
and he explains the occurrence of the rings at unusual thicknesses as due to
the dispersing power of the prism. The blue system being more refracted
than the red, it is possible under certain conditions that the wth blue ring
may be so much displaced relatively to the corresponding red ring as at one
part of the circumference to compensate for the different diameters. A white
stripe may thus be formed in a situation where without the prism the
mixture of colours would be complete, so far as could be judged by the eye.
The simplest case that can be considered is when the " thin plate " is
bounded by plane surfaces inclined to one another at a small angle. By
drawing back the prism (whose edge is parallel to the intersection of the
above-mentioned planes) it will always be possible so to adjust the effective
dispersing power as to bring the 7ith bars to coincidence for any two assigned
colours, and therefore approximately for the entire spectrum. The formation
of the achromatic band, or rather central black band, depends indeed upon
the same principles as the fictitious shifting of the centre of a system of
Fresnel's bands when viewed through a prism.
But neither Newton nor, as would appear, any of his successors has
explained why the bands should be more numerous than usual, and under
certain conditions sensibly achromatic for a large number of alternations. It
is evident that, in the particular case of the wedge-shaped plate above
specified, such a result would not occur. The width of the bands for any
colour would be proportional to X, as well after the displacement by the
prism as before ; and the succession of colours formed in white light and the
number of perceptible bands would be much as usual.
The peculiarity to be explained appears to depend upon the curvature
of the surfaces bounding the plate. For simplicity suppose that the lower
surface is plane (y = 0), and that the approximate equation of the upper
surface is y = a + ba?} a being thus the least distance between the plates.
The black of the ?ith order for wave-length \ occurs when
%n\ = a + bxt; (12)
and thus the width (&e) at this place of the band is given by
$\ = 2bxSx, (13)
or
72 WAVE THEORY OF LIGHT. [148
If the glasses be in contact, as is usually supposed in the theory of
Newton's rings, a = 0, and So; oc X*, or the width of the band of the wth order
varies as the square root of the wave-length, instead of as the first power.
Even in this case the overlapping and subsequent obliteration of the bands
is greatly retarded by the use of the prism, but the full development of the
phenomenon requires that a should be finite. Let us inquire what is the
condition in order that the width of the band of the nth order may be
stationary, as A, varies. By (14) it is necessary that the variation of
X2/(i?A- - a) should vanish. Hence a = \n\, so that the interval between
the surfaces at the place where the nth band is formed should be half due
to curvature and half to imperfect contact at the place of closest approach.
If this condition be satisfied, the achromatism of the nth band, effected by the
prism, carries with it the achromatism of a large number of neighbouring
bands, and thus gives rise to the remarkable effects described by Newton.
[1901. For further developments see Phil. Mag. Vol. XXVIIL p. 200, 1889.]
§ 9. Newton s Diffusion Rings.
In the fourth part of the second book of his Optics Newton investigates
another series of rings, usually (though not very appropriately) known as the
colours of thick plates. The fundamental experiment is as follows. At the
centre of curvature of a concave looking-glass, quicksilvered behind, is placed
an opaque card, perforated by a small hole through which sunlight is
admitted. The main body of the light returns through the aperture ; but
a series of concentric rings are seen upon the card, the formation of which
was proved by Newton to require the co-operation of the two surfaces of the
mirror. Thus the diameters of the rings depend upon the thickness of the
glass, and none are formed when the glass is replaced by a metallic speculum.
The brilliancy of the rings depends upon imperfect polish of the anterior
surface of the glass, and may be augmented by a coat of diluted milk, a
device used by the Due de Chaulnes. The rings may also be well observed
without a screen in the manner recommended by Stokes. For this purpose
all that is required is to place a small flame at the centre of curvature of the
prepared glass, so as to coincide with its image. The rings are then seen
surrounding the flame and occupying a definite position in space.
The explanation of the rings, suggested by Young and developed by
Herschel, refers them to interference between one portion of light scattered
or diffracted by a particle of dust and then regularly refracted and reflected,
and another portion first regularly refracted and reflected and then diffracted
at emergence by the same particle. It has been shown by Stokes* that no
* Camb. Trans. Vol. ix. p. 147, 1851.
1888] DIFFUSION RINGS. 73
regular interference is to be expected between portions of light diffracted by
different particles of dust.
In the memoir of Stokes will be found a very complete discussion of the
whole subject, and to this the reader must be referred who desires a fuller
knowledge. Our limits will not allow us to do more than touch upon one or
two points. The condition of fixity of the rings when observed in air, and of
distinctness when a screen is used, is that the systems due to all parts of the
diffusing surface should coincide : and it is fulfilled only when, as in Newton's
experiments, the source and screen are in the plane passing through the
centre of curvature of the
As the simplest for actual calculation, we will consider a little further the
case where the glass is plane and parallel, of thickness t and index /*, and is
supplemented by a lens at whose focus the source of light is placed. This
lens acts both as collimator and as object-glass, so
that the combination of lens and plane mirror Fi£- -•
replaces the concave mirror of Newton's experiment.
The retardation is calculated in the same way as
for thin plates. In Fig. 2 the diffracting particle is
situated at B, and we have to find the relative
retardation of the two rays which emerge finally
at inclination 0, the one diffracted at emergence
following the path ABDBIE, and the other dif-
fracted at entrance and following the path ABFGH.
The retardation of the former from B to / is 2ftt + BI, and of the latter
from B to the equivalent place G is 2pBF. Now FB = t sec ff, ff being the
angle of refraction ; BI = 2* tan ff sin 0 : so that the relative retardation E is
given by
B = 2fd (1 + p-1 tan ff sin 6 - sec ff} = 2/rt (1 - cos ff}.
If 0, ff be small, we may take
R = 2tPf*, (1)
as sufficiently approximate.
The condition of distinctness is here satisfied, since R is the same for
every ray emergent parallel to a given one. The rays of one parallel svstern
are collected by the lens to a focus at a definite point in the neighbourhood
of the original source.
The formula (1) was discussed by Herschel, and shown to agree with
Newton's measures. The law of formation of the rings follows immediatelv
from the expression for the retardation, the radius of the ring of n01 order
being proportional to n and to the square root of the wave-length.
74 WAVE THEORY OF LIGHT. [148
§ 10. Huygens's Principle. Theory of Shadows.
The objection most frequently brought against the undulatory theory
in its infancy was the difficulty of explaining in accordance with it the
existence of shadows. Thanks to Fresnel and his followers, this department
of Optics is now precisely the one in which the theory has secured its greatest
triumphs.
The principle employed in these investigations is due to Huygens, and
may be thus formulated. If round the origin of waves an ideal closed
surface be drawn, the whole action of the waves in the region beyond may be
regarded as due to the motion continually propagated across the various
elements of this surface. The wave motion due to any element of the
surface is called a secondary wave, and in estimating the total effect regard
must be paid to the phases as well as the amplitudes of the components. It
is usually convenient to choose as the surface of resolution a wave-front, i.e., a
surface at which the primary vibrations are in one phase.
Any obscurity that may hang over Huygens's principle is due mainly to
the indefmiteness of thought and expression which we must be content to
put up with if we wish to avoid pledging ourselves as to the character of
the vibrations. In the application to sound, where we know what we are
dealing with, the matter is simple enough in principle, although mathematical
difficulties would often stand in the way of the calculations we might wish to
make. The ideal surface of resolution may be there regarded as a flexible
lamina ; and we know that, if by forces locally applied every element of the
lamina be made to move normally to itself exactly as the air at that place
does, the external aerial motion is fully determined. By the principle of
superposition the whole effect may be found by integration of the partial
effects due to each element of the surface, the other elements remaining
at rest.
We will now consider in detail the important case in which uniform plane
waves are resolved at a surface coincident with a wave-front (OQ). We
imagine the wave-front divided into elementary rings or zones, called
Huygens's zones, by spheres described round P (the point at which the
aggregate effect is to be estimated), the first sphere,
touching the plane at 0, with a radius equal to PO, Fl8- 3-
and the succeeding spheres with radii increasing
at each step by ^X. There are thus marked out
a series of circles, whose radii x are given by
a? + r* = (r + -£nX)2, or x* = n\r nearly ; so that the
rings are at first of nearly equal area. Now the
effect upon P of each element of the plane is
proportional to its area; but it depends also upon
1888] LAW OF SECONDARY WAVE. 75
the distance from P, and possibly upon the inclination of the secondary ray
to the direction of vibration and to the wave-front. These questions will be
further considered in connexion with the dynamical theory; but under all
ordinary circumstances the result is independent of the precise answer that
may be given. All that it is necessary to assume is that the effects of the
successive zones gradually diminish, whether from the increasing obliquity of
the secondary ray or because (on account of the limitation of the region of
integration) the zones become at last more and more incomplete. The
component vibrations at P due to the successive zones are thus nearly equal
in amplitude and opposite in phase (the phase of each corresponding to that
of the infinitesimal circle midway between the boundaries), and the series
which we have to sum is one in which the terms are alternately opposite in
sign and, while at first nearly constant in numerical magnitude, gradually
diminish to zero. In such a series each term may be regarded as very nearly
indeed destroyed by the halves of its immediate neighbours, and thus the
sum of the whole series is represented by half the first term, which stands
over uncompensated. The question is thus reduced to that of finding the
effect of the first zone, or central circle, of which the area is TrXr.
We have seen that the problem before us is independent of the law of
the secondary wave as regards obliquity; but the result of the integration
necessarily involves the law of the intensity and phase of a secondary wave
as a function of r, the distance from the origin. And we may in fact, as was
done by A. Smith*, determine the law of the secondary wave, by comparing
the result of the integration with that obtained by supposing the primary
wave to pass on to P without resolution.
Now as to the phase of the secondary wave, it might appear natural to
suppose that it starts from any point Q with the phase of the primary wave,
so that on arrival at P it is retarded by the amount corresponding to QP.
But a little consideration will prove that in that case the series of secondary
waves could not reconstitute the primary wave. For the aggregate effect of
the secondary waves is the half of that of the first Huygens zone, and it is
the central element only of that zone for which the distance to be travelled is
equal to r. Let us conceive the zone in question to be divided into infini-
tesimal rings of equal area. The effects due to each of these rings are equal
in amplitude and of phase ranging uniformly over half a complete period.
The phase of the resultant is midway between those of the extreme elements,
that is to say, a quarter of a period behind that due to the element at the
centre of the circle. It is accordingly necessary to suppose that the secondary
waves start with a phase one-quarter of a period in advance of that of the
primary wave at the surface of resolution.
* Camb. Math. Journ. Vol. in. p. 46, 1843.
76 WAVE THEORY OF LIGHT. [148
Further, it is evident that account must be taken of the variation of phase
in estimating the magnitude of the effect at P of the first zone. The middle
element alone contributes without deduction ; the effect of every other must
be found by introduction of a resolving factor, equal to cos 0, if 6 represent
the difference of phase between this element and the resultant. Accordingly,
the amplitude of the resultant will be less than if all its components had the
same phase, in the ratio
-H*
cos Odd : IT,
r-
J-
or 2 : TT. Now 2 area/7r =* 2Xr ; so that, in order to reconcile the amplitude of
the primary wave (taken as 'unity) with the half effect of the first zone, the
amplitude, at distance r, of the secondary wave emitted from the element of
area dS must be taken to be
By this expression, in conjunction with the quarter-period acceleration of
phase, the law of the secondary wave is determined.
That the amplitude of the secondary wave should vary as r~l was to be
expected from considerations respecting energy; but the occurrence of the
factor X"1, and the acceleration of phase, have sometimes been regarded as
mysterious. It may be well therefore to remember that precisely these laws
apply to a secondary wave of sound, which can be investigated upon the
strictest mechanical principles.
The recomposition of the secondary waves may also be treated analytically.
If the primary wave at 0 be cos kat, the effect of the secondary wave pro-
ceeding from the element dS at Q is
J Of 7 O
— - cos k (at - p + ^X) = — — - sin k (at — p).
If dS=2-nxdx, we have for the whole effect
2-7T f°° sin k (at — p) xdx
X Jo p
or, since xdx = pdp, & = 27r/X,
— k I sin k (at — p) dp = — cos k (at — p)\ .
In order to obtain the effect of the primary wave, as retarded by traversing
the distance r, viz. cos k (at — r), it is necessary to suppose that the integrated
term vanishes at the upper limit. And it is important to notice that without
some further understanding the integral is really ambiguous. According to
the assumed law of the secondary wave, the result must actually depend upon
the precise radius of the outer boundary of the region of integration, supposed
1888] THEORY OF SHADOWS. 77
to be exactly circular. This case is, however, at most very special and excep-
tional. We may usually suppose that a large number of the outer rings are
incomplete, so that the integrated term at the upper limit mav properlv be
taken to vanish. If a formal proof be desired, it may be obtained bv
introducing into the integral a factor such as e~*f, in which h is ultimatelv
made to diminish without limit.
When the primary wave is plane, the area of the first Huygens zone is
ir\r, and, since the secondary waves vary as r-1, the intensity is independent
of r, as of course it should be. If, however, the primary wave be spherical,
and of radius a at the wave-front of resolution, then we know that at a
distance r further on the amplitude of the primary wave will be diminished
in the ratio a : (r + a). This may be regarded as a consequence of the altered
area of the first Huygens zone. For, if a- be its radius, we have
so that
Xar
.
nearlv.
o + r
Since the distance to be travelled by the secondary waves is still r, we see
how the effect of the first zone, and therefore of the whole series, is pro-
portional to a;(a + r). In like manner may be treated other cases, such as
that of a primary wave-front of unequal principal curvatures.
The general explanation of the formation of shadows may also be con-
veniently based upon Huygens's zones. If the point under consideration be
so far away from the geometrical shadow that a large number of the earlier
zones are complete, then the illumination, determined sensibly by the first
zone, is the same as if there were no obstruction at all. If. on the other
hand, the point be well immersed in the geometrical shadow, the earlier
zones are altogether missing, and, instead of a series of terms beginning with
finite numerical magnitude and gradually diminishing to zero, we have now
to deal with one of which the terms diminish to zero at both ends. The sum
of such a series is very approximately zero, each term being neutralized by
the halves of its immediate neighbours, which are of the opposite sign. The
question of light or darkness then depends upon whether the series begins or
ends abruptly. With few exceptions, abruptness can occur only in the presence
of the first term, viz. when the secondary wave of least retardation is unob-
structed, or when a ray passes through the point under consideration.
According to the undulatory theory the light cannot be regarded strictly
as travelling along a ray: but the existence of an unobstructed ray implies
that the system of Huygens's zones can be commenced, and, if a large number
of these zones are fully developed and do not terminate abruptly, the illu-
mination is unaffected by the neighbourhood of obstacles. Intermediate
78 WAVE THEORY OF LIGHT. [148
cases in which a few zones only are formed belong especially to the province
of diffraction.
An interesting exception to the general rule that full brightness requires
the existence of the first zone occurs when the obstacle assumes the form of a
small circular disk parallel to the plane of the incident waves. In the earlier
half of the 18th century* Delisle found that the centre of the circular shadow
was occupied by a bright point of light, but the observation passed into
oblivion until Poisson brought forward as an objection to Fresnel's theory
that it required at the centre of a circular shadow a point as bright as if no
obstacle were intervening. If we conceive the primary wave to be broken up
at the plane of the disk, a system of Huygens's zones can be constructed which
begin from the circumference ; and the first zone external to the disk plays
the part ordinarily taken by the centre of the entire system. The whole
effect is the half of that of the first existing zone, and this is sensibly the
same as if there were no obstruction.
When light passes through a small circular or annular aperture, the
illumination at any point along the axis depends upon the precise relation
between the aperture and the distance from it at which the point is taken.
If, as in the last paragraph, we imagine a system of zones to be drawn
commencing from the inner circular boundary of the aperture, the question
turns upon the manner in which the series terminates at the outer boundary.
If the aperture be such as to fit exactly an integral number of zones, the
aggregate effect may be regarded as the half of those due to the first and
last zones. If the number of zones be even, the action of the first and last
zones are antagonistic, and there is complete darkness at the point. If on the
other hand the number of zones be odd, the effects conspire ; and the illumi-
nation (proportional to the square of the amplitude) is four times as great as
if there were no obstruction at all.
The process of augmenting the resultant illumination at a particular point
by stopping some of the secondary rays may be carried much furtherf. By
the aid of photography it is easy to prepare a plate, transparent where the
zones of odd order fall, and opaque where those of even order fall. Such a
plate has the power of a condensing lens, and gives an illumination out of all
proportion to what could be obtained without it. An even greater effect
(fourfold) would be attained if it were possible to provide that the stoppage
of the light from the alternate zones were replaced by a phase-reversal
without loss of amplitude.
In such experiments the narrowness of the zones renders necessary a
pretty close approximation to the geometrical conditions. Thus in the case
of the circular disk, equidistant (r) from the source of light and from the
* Verdet, Lemons d'Optique Physique, i. § 66.
t Soret, Fogg. Ann. CLVI. p. 99, 1875.
1888] DiFFBAcmox. 79
screen upon which the shadow is observed, the width of the first exterior zone
is given bv
4(2*)'
2x being the diameter of the disk. If 2r=1000cm., 2x=lcm.,X=6xlO-*cm.,
then dx = "0015 cm. Hence, in order that this zone may be perfectly formed,
there should be no error in the circumference of the order of "001 cm.* The
experiment succeeds in a dark room of the length above mentioned, with a
threepenny bit (supported by three threads) as obstacle, the origin of light
being a small needle-hole in a plate of tin, through which the sun's rays shine
horizontally after reflexion from an external mirror. In the absence of a
heliostat it is more convenient to obtain a point of light with the aid of a
lens of short focus.
The amplitude of the light at any point in the axis, when plane waves are
incident perpendicularly upon an annular aperture, is. as above,
cos k (at — r,) — cos k (at — rs) = 2 sin kat . sin k (r, — rs),
TJ, r, being the distances of the outer and inner boundaries from the ]>:>int
in question. It is scarcely necessary to remark that in all such cases the
calculation applies in the first instance to homogeneous light, and that.
in accordance with Fourier's theorem, each homogeneous component of a
mixture may be treated separately. When the original light is white, the
presence of some components and the absence of others will usually give rise
to coloured effects, variable with the precise circumstances of the case.
Although what we have to say upon the subject is better postponed until
we consider the dynamical theory, it is proper to point out at once that there
is an element of assumption in the application of Huygens's principle to the
calculation of the effects produced by opaque screens of limited extent.
Properly applied, the principle could not fail : but, as may readily be proved
in the case of sonorous waves, it is not in strictness sufficient to assume the
expression for a secondary wave suitable when the primary wave is un-
disturbed, with mere limitation of the integration to the transparent parts
of the screen. But, except perhaps in the case of very fine gratings, it is
probable that the error thus caused is insignificant: for the incorrect
estimation of the secondary waves will be limited to distances of a few
wave-lengths only from the boundary of opaque and transparent parts.
§11. Fraunhofer's Diffraction Phenomena.
A very general problem in diffraction is the investigation of the dis-
tribution of light over a screen upon which impinge divergent or convergent
spherical waves after passage through various diffracting apertures. When
* It is easy to see that the radius of the bright spot is of the same order of magnitude.
80
WAVE THEORY OF LIGHT.
[148
the waves are convergent and the recipient screen is placed so as to contain
the centre of convergency — the image of the original radiant point, the
calculation assumes a less complicated form. This class of phenomena was
investigated by Fraunhofer (upon principles laid down
by Fresnel), and are sometimes called after his name.
We may conveniently commence with them on account
of their simplicity and great importance in respect to
the theory of optical instruments.
If / be the radius of the spherical wave at the place
of resolution, where the vibration is represented by
cos kat, then at any point M (Fig. 4) in the recipient
screen the vibration due to an element dS of the wave-
front is (§ 9)
— — sin k (at — p),
p being the distance between M and the element dS.
Taking coordinates in the plane of the screen with the centre of the wave
as origin, let us represent M by £, 77, and P (where dS is situated) by x, y, z.
Then
p* = (x - £)» + (y
so that
In the application with which we are concerned, f , 77 are very small quantities ;
and we may take
At the same time dS may be identified with dxdy, and in the denominator p
may be treated as constant and equal to f. Thus the expression for the
vibration at M becomes
- 1, ((si
V**
sin k -
dxdy;
,
V** I /
and for the intensity, represented by the square of the amplitude,
(1)
^, [//
cos t
...(2)
, ,
This expression for the intensity becomes rigorously applicable when f is
indefinitely great, so that ordinary optical aberration disappears. The
incident waves are thus plane, and are limited to a plane aperture coincident
with a wave-front. The integrals are then properly functions of the direction
in which the light is to be estimated.
In experiment under ordinary circumstances it makes no difference
whether the collecting lens is in front of or behind the diffracting aperture.
1888] THEOREMS OF BRIDGE. 81
It is usually most convenient to employ a telescope focused upon the radiant
point, and to place the diffracting apertures immediately in front of the
object-glass. What is seen through the eye-piece in any case is the same as
would be depicted upon a screen in the focal plane.
Before proceeding to special cases it may be well to call attention to some
general properties of the solution expressed by (2)*.
If, when the aperture is given, the wave-length (proportional to tr*}
varies, the composition of the integrals is unaltered, provided f and 17 are
taken [directly] proportional to X. A diminution of X thus leads to a simple
proportional shrinkage of the diffraction pattern, attended by an augmentation
of brilliancy in proportion to X~*.
If the wave-length remains unchanged, similar effects are produced by an
increase in the scale of the aperture. The linear dimension of the diffraction
pattern is inversely as that of the aperture, and the brightness at corre-
sponding points is as the square of the area of aperture.
If the aperture and wave-length increase in the same proportion, the size
and shape of the diffraction pattern undergo no change.
We will now apply the integrals (2) to the case of a rectangular aperture
of width a parallel to x and of width b parallel to y. The limits of in-
tegration for x may thus be taken to be — ^a and +^o, and for y to be
- 46, +|6. We readily find (with substitution for t of
~' ' '
as representing the distribution of light in the image of a mathematical point
when the aperture is rectangular,, as is often the case in spectroscopes.
The second and third factors of (3) being each of the form sin'tiV, we
have to examine the character of this function. It vanishes when M = »MB% in
being any whole number other than zero. When u = 0. it takes the value
unity. The maxima occur when
M = tanHT .................................... (4)
and then
sin5 w/it8 = «*?». .............................. (5)
To calculate the roots of (5) we may assume
where y is a positive quantity which is small when v is large. Substituting
this, we find coty= U — y, whence
3 15 315 "
Bridge, PfttL Mm9. NOT. 1858.
K. m.
82 WAVE THEORY OF LIGHT. [148
This equation is to be solved by successive approximation. It will readily be
found that
(6)
In the first quadrant there is no root after zero, since tan u > u, and in the
second quadrant there is none because the signs of u and tan u are opposite.
The first root after zero is thus in the third quadrant, corresponding to m = 1.
Even in this case the series converges sufficiently to give the value of the root
with considerable accuracy, while for higher values of m it is all that could be
desired. The actual values of u/ir (calculated in another manner by Schwerd)
are 1-4303, 2'4590, 3-4709, 4'4747, 5-4818, 6'4844, &c.
Since the maxima occur when u = (in -f J) TT nearly, the successive values
are not very different from
444
_ . __ _ X-Tf,
97T2' 257T2' 497T2'
The application of these results to (3) shows that the field is brightest at
the centre £ = 0, t] = 0, viz. at the geometrical image. It is traversed by dark
lines whose equations are
| = mf\/a, 77 = mf\/b.
Within the rectangle formed by pairs of consecutive dark lines, and not far
from its centre, the brightness rises to a maximum ; but these subsequent
maxima are in all cases much inferior to the brightness at the centre of the
entire pattern (f = 0, 77 = 0).
By the principle of energy the illumination over the entire focal plane
must be equal to that over the diffracting area ; and thus, in accordance with
the suppositions by which (3) was obtained, its value when integrated from
£= — oo to f = + oo , and from 77 = — oo to 17 = + oo should be equal to ab.
This integration, employed originally by Kelland* to determine the absolute
intensity of a secondary wave, may be at once effected by means of the
known formula
+°°2M , [+°°smu ,
— du = I - - aii
r+
/'-
It will be observed that, while the total intensity is proportional to ab, the
intensity at the focal point is proportional to a2fe2. If the aperture be
increased, not only is the total brightness over the focal plane increased with
it, but there is also a concentration of the diffraction pattern. The form of
(3) shows immediately that, if a and b be altered, the coordinates of any
characteristic point in the pattern vary as a"1 and b~\
The contraction of the diffraction pattern with increase of aperture is of
fundamental importance with reference to the resolving power of optical
* Ed. Trans, xv. 315.
1888] RECTANGULAR APERTURE. 83
instruments. According to common optics, where images are absolute, the
diffraction pattern is supposed to be infinitely small, and two radiant points,
however near together, form separated images. This is tantamount to an
assumption that X is infinitely smalL The actual finiteness of X imposes a
limit upon the separating or resolving power of an optical instrument.
This indefiniteness of images is sometimes said to be due to diffraction by
the edge of the aperture, and proposals have even been made for curing it by
causing the transition between the interrupted and transmitted parts of the
primary wave to be less abrupt. Such a view of the matter is altogether
misleading. What requires explanation is not the imperfection of actual
images so much as the possibility of their being as good as we find them.
At the focal point (£ = 0, y = 0) all the secondary waves agree in phase.
and the intensity is easily expressed, whatever be the form of the aperture.
From the general formula (2), if A be the area of aperture,
I* = A*j\*f*. ................................. (7)
The formation of a sharp image of the radiant point requires that the
illumination become insignificant when £ 17 attain small values, and this
insignificance can only arise as a consequence of discrepancies of phase among
the secondary waves from various parts of the aperture. So long as there is
no sensible discrepancy of phase, there can be no sensible diminution of
brightness as compared with that to be found at the focal point itself. We
may go further, and lay it down that there can be no considerable loss of
brightness until the difference of phase of the waves proceeding from the
nearest and furthest parts of the aperture amounts to JX.
When the difference of phase amounts to X, we may expect the resultant
illumination to be very much reduced. In the particular case of a rectangular
aperture the course of things can be readily followed, especially if we conceive
f to be infinite. In the direction (suppose horizontal) for which ij = Q,
£!/= sin 0, the phases of the secondary waves range over a complete period
when sin0 = Xo, and, since all parts of the horizontal aperture are equally
effective, there is in this direction a complete compensation and consequent
absence of illumination. When sin 6 = f X/a, the phases range one and a half
periods, and there is revival of illumination. We may compare the brightness
with that in the direction 0 = 0. The phase of the resultant amplitude is the
same as that due to the central secondary wave, and the discrepancies of
phase among the components reduce the amplitude in the proportion
or — 2 : 3r; so that the brightness in this direction is 4/Shr* of the maximum
at 0 = 0. In like manner we may find the illumination in any other direction,
and it is obvious that it vanishes when sin 0 is any multiple of X/a.
84
WAVE THEORY OF LIGHT.
[148
The reason of the augmentation of resolving power with aperture will
now be evident. The larger the aperture the smaller are the angles through
which it is necessary to deviate from the principal direction in order to bring
in specified discrepancies of phase — the more concentrated is the image.
In many cases the subject of examination is a luminous line of uniform
intensity, the various points of which are to be treated as independent
sources of light. If the image of the line be £ = 0, the intensity at any
point £, 77 of the diffraction pattern may be represented by
.(8)
the same law as obtains for a luminous point when horizontal directions are
alone considered. The definition of a fine vertical line, and consequently the
resolving power for contiguous vertical lines, is thus independent of the
vertical aperture of the instrument, a law of great importance in the theory
of the spectroscope.
The distribution of illumination in the image of a luminous line is shown
by the curve ABC (Fig. 5), representing the value of the function sin2 u/u*
from u = 0 to u = 2?r. The part corresponding to negative values of u is
similar, OA being a line of symmetry.
Let us now consider the distribution of brightness in the image of a
double line whose components are of equal strength, and at such an angular
interval that the central line in the image of one coincides with the first zero
of brightness in the image of the other. In Fig. 5 the curve of brightness for
one component is ABC, and for the other OA'G' ; and the curve representing
half the combined brightnesses is E'BE. The brightness (corresponding to B)
midway between the two central points AA' is '8106 of the brightness at the
central points themselves. We may consider this to be about the limit of
closeness at which there could be any decided appearance of resolution,
though doubtless an observer accus-
tomed to his instrument would re-
cognize the duplicity with certainty.
The obliquity, corresponding to u = TT,
is such that the phases of the se-
condary waves range over a complete
period, i.e. such that the projec-
tion of the horizontal aperture upon
this direction is one wave-length.
We conclude that a double line cannot
be fairly resolved unless its components
subtend an angle exceeding that sub-
tended by tlie wave-length of light at a
distance equal to the horizontal aper-
1888] DOUBLE LINE. 85
t we. This rule is convenient on account of its simplicity ; and it is
sufficiently accurate in view of the necessary uncertainty as to what exactly
is meant by resolution.
On the experimental confirmation of the theory of the resolving power of
rectangular apertures, see Optics, Enc. Brit. Vol. XVII. p. 807, [Vol. n. p. 411].
If the angular interval between the components of a double line be half as
great again as that supposed in the figure, the brightness midway between is
•1802 as against T0450 at the central lines of each image. Such a falling off
in the middle must be more than sufficient for resolution. If the angle
subtended by the components of a double line be twice that subtended by the
wave-length at a distance equal to the horizontal aperture, the central bands
are just clear of one another, and there is a line of absolute blackness in the
middle of the combined images.
Since the limitation of the width of the central band in the image of a
luminous line depends upon discrepancies of phase among the secondary
waves, and since the discrepancy is greatest for the waves which come from
the edges of the aperture, the question arises how far the operation of the
central parts of the aperture is advantageous. If we imagine the aperture
reduced to two equal narrow slits bordering its edges, compensation will
evidently be complete when the projection on an oblique direction is equal to
£X, instead of A, as for the complete aperture. By this procedure the width
of the central band in the diffraction pattern is halved, and so far an ad-
vantage is attained. But, as will be evident, the bright bands bordering the
central band are now not inferior to it in brightness ; in fact, a band similar
to the central band is reproduced an indefinite number of times, so long as
there is no sensible discrepancy of phase in the secondary waves proceeding
from the various parts of the same slit. Under these circumstances the
narrowing of the band is paid for at a ruinous price, and the arrangement
must be condemned altogether.
A more moderate suppression of the central parts is, however, sometimes
advantageous. Theory and experiment alike prove that a double line, of
which the components are equally strong, is better resolved when, for
example, one-sixth of the horizontal aperture is blocked off by a central
screen ; or the rays quite at the centre may be allowed to pass, while others
a little further removed are blocked off. Stops, each occupying one-eighth of
the width, and with centres situated at the points of trisection, answer well
the required purpose.
It has already been suggested that the principle of energy requires that
the general expression for I* in (2) when integrated over the whole of the
plane £, ij should be equal to A, where A is the area of the aperture. A
86 WAVE THEORY OF LIGHT. [148
general analytical verification has been given by Stokes*. The expression
for P may be written in the form
the integrations with respect to of, y as well as those with respect to x, y
being over the area of the aperture ; and for the present purpose this is to be
integrated again with respect to £, rj over the whole of the focal plane.
In changing the order of integration so as to take first that with respect
to £, rj, it is proper, in order to avoid ambiguity, to introduce under the
integral sign the factor e*"*^11, the + or — being chosen so as to make the
elements of the integral vanish at infinity. After the operations have been
performed, a and ft are to be supposed to vanish.
Thus JSFd£d<r) = Limit of
C+x t ,, ,. T,N ,„ 2acosH
Now J _^ fF* cos (A£
and thus
T i k (X' ~ X} 1 I f0- 7
Let —^ — - = an, dx ~*-jr du.
The limits for u are ultimately — oc and + x , and we have
. f 2adar/ 2/ f+x du 2f
it 7 - jrr-> - v^ = i I ^ . = i • "* =
2 ^ ^ */-, I+V A;
In like manner the integration for y' may be performed ; and we find
4 ...................... (10)f
We saw that /02 (the intensity at the focal point) was equal to
If A' be the area over which the intensity must be I0- in order to give the
actual total intensity in accordance with
ff+c
L//«=/L
* Ed. Trans, xx. p. 317, 1853.
t It is easy to show that this conclusion is not disturbed by the introduction at every point of
an arbitrary retardation p, a function of .1-, y. The terms (p' - p) are then to be added under the
cosine in (9) ; but they are ultimately without effect, since the only elements which contribute
are those for which in the limit x' —x, y' — y, and therefore p'=p.
1888] TOTAL INTENSITY PROPORTIONAL TO APERTURE. 87
the relation between A and A' is A A' = X2/2. Since A' is in some sense the
area of the diffraction pattern, it may be considered to be a rough criterion of
the definition, and we infer that the definition of a point depends principally
upon the arm of the aperture, and only in a very secondary degree upon the
shape when the area is maintained constant.
§ 12. Theory of Circular Aperture.
We will now consider the important case where the form of the aperture
is circular. Writing for brevity
Wf=p, Wf=q, .............................. (1)
we have for the general expression (§11) of the intensity
Kf*r- = & + C\ .............................. (2)
where S=(fsm(px + qy)dxdy, C = ffcos(px + qy)dxdy ....... (3,4)
When, as in the application to rectangular or circular apertures, the form is
symmetrical with respect to the axes both of x and y, $ = 0, and C reduces to
G = JJcos px cos qy dxdy ............................ (5)
In the case of the circular aperture the distribution of light is of course
symmetrical with respect to the focal point p = 0, q = 0 ; and G is a function
of p and q only through V(/>2 + ^2)- It is thus sufficient to determine the
intensity along the axis of p. Putting q = 0, we get
(7=11 cos pxdxdy = 2 I cos px \I(R2 — a?) dx,
R being the radius of the aperture. This integral is the Bessel's function of
order unity, defined by
z /**
Jv(z}=-\ cos (z cos <f>) sin- <t>d<b ................... (6)
7T Jo
Thus, if x = R cos <£,
C-,*2^-; ........................... (7)
pR
and the illumination at distance r from the focal point is
.
~\2/2'
The ascending series for Ji(z), used by Aiiy* in his original investigation of
the diffraction of a circular object-glass, and readily obtained from (6), is
2 2*4 2s . 4s . 6 2» 4»6« .8 ................
On the Diffraction of an Object-Glass with Circular Aperture," Camb. Trans. 1834.
88 WAVE THEORY OF LIGHT. [148
When z is great, we may employ the semi-convergent series
3.5.7.9.1.3.5
8.16.24.82
3 1 3.5.7.1.3
8.16.24
_
8.16.24.32.40 "'"
A table of the values of 2^-1 Jl (z) has been given by Lomrnel *, to whom is
due the first systematic application of Bessel's functions to the diffraction
integrals.
The illumination vanishes in correspondence with the roots of the
equation J1(z) = Q. If these be called z1} z2, z3, ... the radii of the dark
rings in the diffraction pattern are
A*
being thus inversely proportional to R.
The integrations may also be effected by means of polar coordinates,
taking first the integration with respect to 0 so as to obtain the result for an
infinitely thin annular aperture. Thus, if
x = p cos <f>, y = p sin <£,
(7=11 cospxdxdy = I I cos (pp cos 0) pdpdO.
Now by definition
os(^os^)^ = l-g + 2^-237^62+ ....... (11)
The value of G for an annular aperture of radius r and width dr is thus
(12)
For the complete circle,
„ 27r
= -
.,
pR
as before.
In these expressions we are to replace p by &£//, or rather, since the
diffraction pattern is symmetrical, by kr/f, where r is the distance of any
point in the focal plane from the centre of the system.
* Schlomilch, xv. p. 166, 1870.
1888] CIRCULAR APERTURE.
The roots of «/„ (z) after the first may be found from
•050661 -053041 -262051
and those of Ji(z) from
*_. -151982 -015399 '245835
7T~ " '
.(13)
.(14)
formula; derived by Stokes* from the descending series f. The following
table gives the actual values: —
i
- for J0 (z) =0 - for J, (-*) =0
i
- for J0 (z)=Q - for Jx (z)=0
1
•7655
1-2197
6
5-7522 6-2439
2
1-7571
2-2330
7
6-7519 7-2448
3
2-7546
3-2383
8
7-7516 8-2454
4
3-7534
4-2411
9
8-7514 9-2459
o
4-7527 5-2428
10 i 9*7513 10-2463
In both cases the image of a mathematical point is thus a symmetrical
ling system. The greatest brightness is at the centre, where
For a certain distance outwards this remains sensibly unimpaired, and then
gradually diminishes to zero, as the secondary waves become discrepant in
phase. The subsequent revivals of brightness forming the bright rings are
necessarily of inferior brilliancy as compared with the central disk.
The first dark ring in the diffraction pattern of the complete circular
aperture occurs when
rjf= 1-2197 XX/2E ............................ (15)
We may compare this with the corresponding result for a rectangular
aperture of width a,
Bf-V«;
and it appears that in consequence of the preponderance of the central parts,
the compensation in the case of the circle does not set in at so small an
obliquity as when the circle is replaced by a rectangular aperture, whose
side is equal to the diameter of the circle.
* Camb. Trans, n. 1850.
t The descending series for J9 (z) appears to have been first given by Sir W. Hamilton in a
memoir on "Fluctuating Functions," Hoy. Irish Tram. 1840.
90
WAVE THEORY OF LIGHT.
[148
Again, if \ve compare the complete circle with a narrow annular aperture
of the same radius, we see that in the latter case the first dark ring occurs at
a much smaller obliquity, viz.
r/f= -7655 x \/2R.
It has been found by Herschel and others that the definition of a telescope
is often improved by stopping off a part of the central area of the object-glass;
but the advantage to be obtained in this way is in no case great, and anything
like a reduction of the aperture to a narrow annulus is attended by a develop-
ment of the external luminous rings sufficient to outweigh any improvement
due to the diminished diameter of the central area*.
The maximum brightnesses and the places at which they occur are easily
determined with the aid of certain properties of the Bessel's functions. It is
known f that
Jo' (*) = -•/!(*); ........................... (16)
, 18
The maxima of 0 occur when
dz\ z
or by (17) when J3(z) = 0. When z has one of the values thus determined,
The accompanying table is given by Lommel^, in which the first column
gives the roots of J.2 (z) — 0, and the second and third columns the
corresponding values of the functions specified. It appears that the
maximum brightness in the first ring is only about ^ of the brightness
at the centre.
z
Zz-^(z)
4 *-«/»»{*)
•oooooo
+ 1-000000
i-oooooo
5-135630
- -132279
•017498
8-417236
+ -064482
•004158
11-619857
- -040008
•001601
14-795938
+ -027919
•000779
17-959820
- -020905
•000437
* Airy, loc. cit. " Thus the magnitude of the central spot is diminished, and the brightness
of the rings increased, by covering the central parts of the object-glass."
t Todhunter's Laplace's Functions, ch. xxxi. J Loc. cit.
1888] TOTAL ILLUMINATION. 91
We will now investigate the total illumination distributed over the area
of the circle of radius r. We have
where
(20)
Thus
27r j r-rdr = ^^jl'-zdz = TnR2 . 2 | z~lJ{- (z) dz.
Now by (17), (18)
, («)-/.(«)-//(*);
so that
and
If r, or z, be infinite, J»(z), J*(z) vanish, and the whole illumination is
expressed by irB?, in accordance with the general principle. In any case the
proportion of the whole illumination to be found outside the circle of radius r
is given by
For the dark rings Jj (z) = 0 : so that the fraction of illumination outside
any dark ring is simply J?(z). Thus for the first, second, third, and fourth
dark rings we get respectively '161, O90, '062, "047, showing that more
than ^ths of the whole light is concentrated within the area of the second
dark ring*.
When z is great, the descending series (10) gives
<«•>
so that the places of maxima and minima occur at equal intervals.
The mean brightness varies as z~* (or as r~J), and the integral found by
multiplying it by zdz and integrating between 0 and x converges.
It may be instructive to contrast this with the case of an infinitely
narrow annular aperture, where the brightness is proportional to J9~(z).
When z is great,
/z
Jf(z)zdz is not
convergent.
* Phil. May. March 1881. [VoL i. Art. 73.]
92 WAVE THEORY OF LIGHT. [148
The efficiency of a telescope is of course intimately connected with the
size of the disk by which it represents a mathematical point. The resolving
power upon double stars of telescopes of various apertures has been investi-
gated by Dawes and others (Enc. Brit. Vol. xvn. p. 807) [Vol. I. p. 411],
with results that agree fairly well with theory.
If we integrate the expression (8) for 72 with respect to 77, we shall
obtain a result applicable to a linear luminous source of which the various
parts are supposed to act independently.
From (19), (20)
-«f4r
"'J" J 0 J Z.I]
since if = r2 — £-'.
If we write
?=&r!2£/V; ............... • ................. (23)
we get
This integral has been investigated by H. Struve*, who, calling to his aid
various properties of Bessel's functions, shows that
(25)
of which the right-hand member is readily expanded in powers of £. By
means of (24) we may verify that
f +°°d£ f +a°/2efy = 7T.K2.
J -* J_.
Contrary to what would naturally be expected, the subject is more easily
treated without using the results of the integration with respect to x and y,
by taking first of all, as in the investigation of Stokes (§ 11), the integration
with respect to 77. Thus
™ = Limit of
\2/2|
k
~f {£ (x> —*) + '? (y — y)\ dx dy dx dy' dij\ (26)
and
T*T (27>
..
^Tted. ^nn. xvii. 1008, 1882.
1888] LUMINOUS LINE.
We have now to consider
2pdydy'
'
In the integration with respect to y1 every element vanishes in the limit
(£ = 0), unless y = y. If the range of integration for y includes the
value y, then
otherwise it vanishes.
The limit of (28) may thus be denoted by V*F, where Y is the common
part of the ranges of integration for y' and y corresponding to any values
of x and x. Hence
r*I-dij = X-1/-1 [[F cos ^ (x -x)dx dx
= X-1/-1 IT F cos *|? cos ^ dx dxf, ......... (29)
if, as for the present purpose, the aperture is symmetrical with respect to the
axis of y.
In the application to the circle we may write
r™, = .a-/- rr rectos &*,&,
J -x JO JO J J
where F is the smaller of the two quantities 2<J(I& — x-), 2 >J(R> — x2},
i.e., corresponds to the larger of the two abscissse x', x. If we take
Y = 2 n/(R* — a?), and limit the integration to those values of x' which are
less than x, we should obtain exactly the half of the required result. Thus
f +"l*dr, = 16V-1/-1 / fVcfi1 ~ a?) cos ^ cos ^ dx M
J - oo .'0./0 / /
- sm
J
/•+<» ATto Jfc riw
df r-dr,= ~.?£. cos2 ft sin (258) d/3, ............ (30)
J -oo 7T f Jo
Hence, writing as before f = ^TrR^/Xf, we get
riw
o
in which we may replace rf£/£ by rf£/£ in agreement with the result obtained
by Struve.
The integral in (30) may be written in another form. We have
sin /3 dp :
94
and thus
WAVE THEORY OF LIGHT.
[148
('"{ si
Jo
sn cos
cos 2 sin £ sin
W = i/Ml-
= f*"sina(?8in/8)sin/8d/9 (31)
Jo
The integral is thus expressible by means of the function K^* and we have
J - 00
The ascending series for Kl(z) is
and this is always convergent. The descending semi-convergent series is
JKr1(^=-
7T
/fa\
~ V ITT" j '
(!2-4)(32-4)
cosu-iTTMl- 1>2 (8^~ +-
-4 (l2_4)(32-4)(52-4)
... , ...(34)
the series within braces being the same as those which occur in the
expression of the function Ji(z}.
When £ (or £) is very great,
-0 7 — /( o ,.
so that the intensity of the image of a luminous line is ultimately inversely
as the square of the distance from the central axis, or geometrical image.
r
Intensity
On the axis itself ...
o-oo
1
First minimum 3'55
First maximum 4'65
&
Second minimum ...
Second maximum ...
6-80
8-00
if*
Third minimum
9-60
jb
Third maximum
Fourth minimum ...
11-00
13-20
sis
TooOo
* Theory of Sound, § 302.
DOUBLE LINE.
95
1888]
As is evident from its composition, the intensity remains finite for all
values of f ; it is, however, subject to fluctuations presenting maxima and
minima, which have been calculated by Ch. Andre*, using apparently the
method of quadratures.
The results are also exhibited by M. Andre in the form of a curve, of
which Fig. 6 is a copy.
It will be seen that the distribution of brightness does not differ greatly
from that due to a rectangular aperture
whose width (perpendicular to the
luminous line) is equal to the diameter
of the circular aperture. It will be
instructive to examine the image of a
double line, wrhose components present
an interval corresponding to f =TT, and
to compare the result with that already
found for a rectangular aperture (§ 11).
We may consider the brightness at
distance f proportional to
1 <&£-
Fig. 6.
ry-t- (35)
9 10
In the compound image the illumination at the geometrical focus of one
of the luminous lines is represented by
and the illumination midway between the geometrical images of the
two lines is
= -0164,
We find by actual calculation from the series,
L (0) = -3333, so that
and
The corresponding number for the rectangular aperture was -811 : so that, as
might have been expected, the resolving power of the circular aperture is
distinctly less than that of the rectangular aperture of equal width. Hence
a telescope will not resolve a double line unless the angular interval between
them decidedly exceeds that subtended by the wave-length of light at a
distance equal to the diameter of the object-glass. Experiment shows that
resolution begins when the angular interval is about a tenth part greater
than that mentioned.
* Ann. d. VEcole Normale, v. p. 310, 1876.
96
WAVE THEORY OF LIGHT.
[148
If we integrate (30) with respect to £ between the limits — oo and + oo ,
we obtain irB?, as has already been remarked. This represents the whole
illumination over the focal plane due to a radiant point whose image is at 0,
or, reciprocally, the illumination at 0 (the same as at any other point) due
to an infinitely extended luminous area. If we take the integration from %
(supposed positive) to oo we get the illumination at 0 due to a uniform
luminous area extending over this region, that is to say, the illumination at a
point situated at distance £ outside the border of the geometrical image of a
large uniform area. If the point is supposed to be inside the geometrical
image and at a distance f from its edge, we are to take the integration
from - oo to £. Thus, if we choose the scale of intensities so that the full
intensity is unity, then the intensity at a distance corresponding to + £
(outside the geometrical image) may be represented by I (+ £), and that at a
distance — £ by I (— £), where
and
(20 = i -
r!(2£) (36)
This is the result obtained by Struve, who gives the following series
for
The ascending series, obtained at once by integration from (33), is
I)'-'
(37)
When % is great, we have approximately from the descending series
27T3/2 £6/2
Thus " at great distances from the edge of the geometrical image the
intensity is inversely proportional to the distance, and to the radius of the
object-glass."
The following table, abbreviated from that given by Struve, will serve
to calculate the enlargement of an image due to diffraction in any case that
may arise.
f I(f)
f
*(fl
f
I(f)
0-0 -5000
2-5
•0765
7-0
•0293
0-5
•3678
3-0
•0630
9-0
•0222
1-0
•2521
4-0
•0528
11-0
•0186
1-5
•1642
5-0
•0410
15-0
•0135
2-0
•1073
6-0
•0328
1888] STRUVE'S THEOREMS. 97
It may perhaps have struck the reader that there is some want of rigour
in our treatment of (30) when we integrate it over the whole focal plane
of £, ij, inasmuch as in the proof of the formulae £ and i) are supposed to
be small. The inconsistency becomes very apparent when we observe that
according to the formulae there is no limit to the relative retardation of
secondary waves coining from various parts of the aperture, whereas in reality
this retardation could never exceed the longest line capable of being drawn
within the aperture. It will be worth while to consider this point a little
further, although our limits forbid an extended treatment.
The formula becomes rigorous if we regard it as giving the illumination
on the surface of a sphere of very large radius f, in a direction such that
£ =/sin 0cos<f>, 17 =/sin 0 sin </> :
it may then be written
I- = \~* f~"2 ffff cos k [(x' — x) sin 0 cos <£ + (i/ — y) sin 0 sin <f>] dx dy dx' dy.
The whole intensity over the infinite hemisphere is given by
j» -_ ^2 I T* * A) 7/3 7/4v ^^lft\
Jo Jo
According to the plan formerly adopted, we postpone the integration with
respect to x, y, x', y', and take first that with respect to 0 and <f>. Thus for a
single pair of elements of area dxdy, dx' dy we have to consider
//cos k {(x — x) sin 0 cos <£ + (/ — y) sin 0 sin </>} sin 0 dd d$,
or, if we write
x' — x = r cos a, y — y = r sin a,
|| *cos (kr sin 0 cos <£) sin 0 d0 d<f>.
Jo Jo
Now it may be proved (e.g., by expansion in powers of kr) that
2-n- 7 - : (39)
Jo Jo ***
and thus
X = — 1 1 1 1 —f— dx dy dx dy1, (40)
r being the distance between the two elements of area dxdy, dx' dy.
In the case of a circular area of radius R, we have*
'sin AT , 2-n-R?
M si
-
' Theory of Sound, § 302.
98 WAVE THEORY OF LIGHT. [148
and thus
= "" 2 j1 " ~ kR }
When kR = cc,
J = TrR2, as before.
It appears therefore that according to the assumed law of the secondary wave
the total illumination is proportional to the area of aperture, only under the
restriction that the linear dimensions of the aperture are very large in
comparison with the wave-length.
A word as to the significance of (39) may not be out of place. We know
that
^ = cos k (sin 0 cos <j> . x + sin 0 sin <f>.y + cos 0 . z} (42)
satisfies Laplace's extended equation (V2 + k2) >/r = 0, being of the form
coska;', where x' is drawn in an oblique direction; and it follows that
JJVjr sin 0d0 d<f> satisfies the same equation. Now this, if the integration be
taken over the hemisphere 0 = 0 to 0 - |?r, must become a function of r, or
Hence, putting x = r, y = Q, 2=0, we get
riwfai
=l I
But the only function of r which satisfies Laplace's equation continuously
through the origin is A sin kr/(kr) ; and that A = 2?r is proved at once by
putting r = 0. The truth of the formula may also be established inde-
pendently of the differential equation by equating the values of
/•Jir /-2ir
-Jrsm0d0d<b,
Jo Jo
when x = r, y = 0, z = 0, and when x = 0, y - 0, z = r. Thus
ri'T2"1 rjirr2ir
cos (kr sin 0 cos 0) sin 0d0d<f> = cos (kr cos ff) sin 6dddd> =
J 0 J 0 J 0 J 0
The formula itself may also be written
(43)
The results of the preceding theory of circular apertures admit of an
interesting application to coronas, such as are often seen encircling the sun
and moon. They are due to the interposition of small spherules of water,
which act the part of diffracting obstacles. In order to the formation of a
well-defined corona it is essential that the particles be exclusively, or pre-
ponderatingly, of one size.
1888] CORONAS OR GLORIES. 99
If the origin of light be treated as infinitely small, and be seen in focus,
whether with the naked eye or with the aid of a telescope, the whole of the
light in the absence of obstacles would be concentrated in the immediate
neighbourhood of the focus. At other parts of the field the effect is the same,
by Babinet's principle, whether the imaginary screen in front of the object-
glass is generally transparent but studded with a number of opaque circular
disks, or is generally opaque but perforated with corresponding apertures.
Consider now the light diffracted in a direction many times more oblique
than any with which we should be concerned, were the whole aperture
uninterrupted, and take first the effect of a single small aperture. The light
in the proposed direction is that determined by the size of the small aperture
in accordance with the laws already investigated, and its phase depends upon
the position of the aperture. If we take a direction such that the light (of
given wave-length) from a single aperture vanishes, the evanescence continues
even when the whole series of apertures is brought into contemplation.
Hence, whatever else may happen, there must be a system of dark rings
formed, the same as from a single small aperture. In directions other than
these it is a more delicate question how the partial effects should be com-
pounded. If we make the extreme suppositions of an infinitely small source
and absolutely homogeneous light, there is no escape from the conclusion that
the light in a definite direction is arbitrary, that is, dependent upon the
chance distribution of apertures. If, however, as in practice, the light be
heterogeneous, the source of finite area, the obstacles in motion, and the
discrimination of different directions imperfect, we are concerned merely with
the mean brightness found by varying the arbitrary phase-relations, and this
is obtained by simply multiplying the brightness due to a single aperture by
the number of apertures (n)*. The diffraction pattern is therefore that due
to a single aperture, merely brightened n times.
In his experiments upon this subject Fraunhofer employed plates of glass
dusted over with lycopodium, or studded with small metallic disks of uniform
size ; and he found that the diameters of the rings were proportional to the
length of the waves and inversely as the diameter of the disks.
In another respect the observations of Fraunhofer appear at first sight to
be in disaccord with theory; for his measures of the diameters of the red
rings, visible when white light was employed, correspond with the law
applicable to dark rings, and not to the different law applicable to the
luminous maxima. Verdet has, however, pointed out that the observation
in this form is essentially different from that in which homogeneous red
light is employed, and that the position of the red rings would correspond
to the absence of blue-green light rather than to the greatest abundance of
* See §4.
100 WAVE THEORY OF LIGHT. [148
red light. Verdet's own observations, conducted with great care, fully confirm
this view, and exhibit a complete agreement with theory.
By measurements of coronas it is possible to infer the size of the particles
to which they are due, an application of considerable interest in the case of
natural coronas — the general rule being the larger the corona the smaller the
water spherules. Young employed this method not only to determine the
diameters of cloud particles (e.g. T<jVo inch), but also those of fibrous material,
for which the theory is analogous. His instrument was called the eriometer*.
§ 13. Influence of Aberration. Optical Power of Instruments.
Our investigations and estimates of resolving power have thus far
proceeded upon the supposition that there are no optical imperfections,
whether of the nature of a regular aberration or dependent upon irregu-
larities of material and workmanship. In practice there will always be a
certain aberration, or error of phase, which we may also regard as the
deviation of the actual wave-surface from its intended position. In general,
we may say that aberration is unimportant, when it nowhere (or at any rate
over a relatively small area only) exceeds a small fraction of the wave-
length (\). Thus in estimating the intensity at a focal point, where, in the
absence of aberration, all the secondary waves would have exactly the same
phase, we see that an aberration nowhere exceeding %\ can have but little
effect.
The only case in which the influence of small aberration upon the entire
image has been calculated^ is that of a rectangular aperture, traversed by
a cylindrical wave with aberration equal to ex3. The aberration is here
unsymmetrical, the wave being in advance of its proper place in one half of
the aperture, but behind in the other half. No terms in x or x2 need be
considered. The first would correspond to a general turning of the beam ;
and the second would imply imperfect focusing of the central parts. The
effect of aberration may be considered in two ways. We may suppose the
aperture (a) constant, and inquire into the operation of an increasing aberra-
tion ; or we may take a given value of c (i.e. a given wave-surface) and
examine the effect of a varying aperture. The results in the second case
show that an increase of aperture up to that corresponding to an extreme
aberration of half a period has no ill effect upon the central band (§ 11), but
it increases unduly the intensity of one of the neighbouiing lateral bands ;
and the practical conclusion is that the best results will be obtained from an
aperture giving an extreme aberration of from a quarter to half a period, and
that with an increased aperture aberration is not so much a direct cause of
* "Chromatics," in Vol. in. of Supp. to Enc. Brit. 1817.
t "Investigations in Optics," Phil. Mag. Nov. 1879. [Vol. i. p. 428.]
1888] SYMMETRICAL ABERRATION. 101
deterioration as an obstacle to the attainment of that improved definition
which should accompany the increase of aperture.
If, on the other hand, we suppose the aperture given, we find that
aberration begins to be distinctly mischievous when it amounts to about a
quarter period, i>. when the wave-surface deviates at each end by a quarter
wave-length from the true plane.
For the focal point itself the calculations are much simpler. We will
consider the case of a circular object-glass with a symmetrical aberration
proportional to hp*. The vibration will be represented by
in which the radius of the aperture is supposed to be unity. The intensity is
thus expressed by
, ............ (1)
the scale being such that the intensity is unity when there is no aberration
<*=0).
Bv integration by parts it can be shown that
so that
.islw.w- -1
Hence, when A = \rt
= 132945/ V2, 2 sin
Similarly, when A = $r,
J.s = -8003;
and when A=»,
/.»=-3947.
These numbers represent the influence of aberration upon the intensity at
the central point, upon the understanding that the focusing is that adapted
102 WAVE THEORY OF LIGHT. [148
to a small aperture, for which h might be neglected. If a readjustment of
focus be permitted, the numbers will be sensibly raised. The general con-
clusion is that an aberration between the centre and circumference of a
quarter period has but little effect upon the intensity at the central point
of the image.
As an application of this result, let us investigate what amount of
temperature disturbance in the tube of a telescope may be expected to
impair definition. According to Biot and Arago, the index fi for air at £ C.
and at atmospheric pressure is given by
= '°0029
^ " ~ 1 + -0037 t '
If we take 0° C. as standard temperature,
Thus, on the supposition that the irregularity of temperature t extends
through a length I, and produces an acceleration of a quarter of a wave-length,
|\=l-l^xlO-8;
or, if we take X = 5'3 x 10~5,
1*4*12,
the unit of length being the centimetre.
We may infer that, in the case of a telescope tube 12 cm. long, a stratum
of air heated 1° C. lying along the top of the tube, and occupying a moderate
fraction of the whole volume, would produce a not insensible effect. If the
change of temperature progressed uniformly from one side to the other, the
result would be a lateral displacement of the image without loss of definition ;
but in general both effects would be observable. In longer tubes a similar
disturbance would be caused by a proportionally less difference of temperature.
We will now consider the application of the principle to the formation of
images, unassisted by reflexion or refraction*. The function of a lens in
forming an image is to compensate by its variable thickness the differences
of phase which would otherwise exist between secondary waves arriving at
the focal point from various parts of the aperture (Optics, Enc. Brit. Vol. xvn.
p. 802 [Vol. II. p. 398]). If we suppose the diameter of the lens to be given
(2.R), and its focal length f gradually to increase, the original differences of
phase at the image of an infinitely distant luminous point diminish without
limit. When f attains a certain value, say flt the extreme error of phase to
be compensated falls to ^\. But, as we have seen, such an error of phase
causes no sensible deterioration in the definition; so that from this point
onwards the lens is useless, as only improving an image already sensibly as
* Phil. Mag. March 1881. [Vol. i. p. 513.]
1888] IMAGES BY SIMPLE 1PEKTCRES. 103
perfect as the aperture admits oil Throughout the operation of increasing
the focal length, the resolving power of the instrument, which depends only
upon the aperture, remains unchanged; and we thus arrive at the rather
startling conclusion that a telescope of any degree of resolving power might
be constructed without an object-glass, if only there were no limit to the
admissible focal length. This last proviso, however, as we shall see, takes
away almost all practical importance from the proposition.
To get an idea of the magnitudes of the quantities involved, let us take
the case of an aperture of £ inch, about that of the pupil of the eye. The
distance/",, which the actual focal length must exceed, is given by
so that
/ = 2^X. .................................... (4)
Thus, if X - 40^0, R = &, we find /t = 800 inches [inch = 2--S4 cm.].
The image of the sun thrown upon a screen at a distance exceeding 66 feet.
through a hole 1 inch in diameter, is therefore at least as well defined as that
seen direct.
As the minimum focal length increases with the square of the aperture, a
quite impracticable distance would be required to rival the resolving power of
a modern telescope. Even for an aperture of 4 inches, ft would have to be
5 miles.
A similar argument may be applied to find at what point an achromatic
lens becomes sensibly superior to a single one. The question is whether.
when the adjustment of focus is correct for the central rays of the spectrum.
the error of phase for the most extreme rays (which it is necessary to consider)
amounts to a quarter of a wave-length. If not, the substitution of an achro-
matic lens will be of no advantage. Calculation shows that, if the aperture
be 1 inch, an achromatic lens has no sensible advantage if the focal length be
greater than about 11 inches. If we suppose the focal length to be 66 feet, a
single lens is practically perfect up to an aperture of IT inch.
Some estimates of the admissible aberration in a spherical lens have
already been given under Optics, Enc. BriL YoL XVTL p. 807 [VoL II. p. 413].
In a similar manner we may estimate the least visible displacement of the
eye-piece of a telescope focused upon a distant object, a question of interest
in connexion with range-finders. It appears* that a displacement S/"frorn the
true focus will not sensibly impair definition, provided
(5)
being the diameter of aperture. The linear accuracy required is thus a
* PUL May. zx. p. 3*4, 1885. [VoL n. p. 430.]
104 WAVE THEORY OF LIGHT. [148
function of the ratio of aperture to focal length. The formula agrees well with
experiment.
The principle gives an instantaneous solution of the question of the
ultimate optical efficiency in the method of " mirror-reading," as commonly
practised in various physical observations. A rotation by which one edge of
the mirror advances ^X (while the other edge retreats to a like amount)
introduces a phase-discrepancy of a whole period where before the rotation
there was complete agreement. A rotation of this amount should therefore
be easily visible, but the limits of resolving power are being approached ; and
the conclusion is independent of the focal length of the mirror, and of the
employment of a telescope, provided of course that the reflected image is seen
in focus, and that the full width of the mirror is utilized.
A comparison with the method of a material pointer, attached to the parts
whose rotation is under observation, and viewed through a microscope, is of
interest. The limiting efficiency of the microscope is attained when the
angular aperture amounts to 180° (Microscope, Enc. Brit. Vol. xvi. p. 267 ;
Optics, Enc. Brit. Vol. xvn. p. 807 [Vol. n. p. 412]); and it is evident that a
lateral displacement of the point under observation through i\ entails (at the
old image) a phase-discrepancy of a whole period, one extreme ray being
accelerated and the other retarded by half that amount. We may infer that
the limits of efficiency in the two methods are the same when the length of
the pointer is equal to the width of the mirror.
An important practical question is the amount of error admissible in
optical surfaces. In the case of a mirror, reflecting at nearly perpendicular
incidence, there should be no deviation from truth (over any appreciable
area) of more than |\. For glass, fi — 1 = | nearly ; and hence the admissible
error in a refracting surface of that material is four times as great.
Fig. 7.
In the case of oblique reflexion at an angle <£, the error of retardation due
to an elevation BD (Fig. 7) is
QQ' -QS = BD sec 0(1- cos SQQ) = BD sec <f> (1 + cos 20) = 2BD cos 0 ;
from which it follows that an error of given magnitude in the figure of a
surface is less important in oblique than in perpendicular reflexion. It must,
1888] OPTICAL POWER OF PRISMS. 105
however, be borne in uiind that errors can sometimes be compensated by
altering adjustments. If a surface intended to be flat is affected with a sb'ght
general curvature, a remedy may be found in an alteration of focus, and the
remedy is the less complete as the reflexion is more obh'que.
The formula expressing the optical power of prismatic spectroscopes is
given with examples under Optics, Enc. Brit. VoL XVIL p. 807 [Vol. n. p. 412],
and mav readily be investigated upon the principles of the wave theory. Let
A,B, (Fig. 8) be a plane wave-surface of the light before it falls upon the
prisms, AB the corresponding wave-surface for a particular part of the
Fig. 8.
spectrum after the light has passed the prisms, or after it has passed the eye-
piece of the observing telescope. The path of a ray from the wave-surface
AtB# to A or B is determined by the condition that the optical distance,
fpds, is a minimum (Optics, Enc. Brit. Vol. xvn. p. 798) ; and, as AB is by
supposition a wave-surface, this optical distance is the same for both points.
Thus
(6)
We have now to consider the behaviour of light belonging to a neighbouring
part of the spectrum. The path of a ray from the wave-surface A9B9 to the
point A is changed : but in virtue of the minimum property the change may
be neglected in calculating the optical distance, as it influences the result by
quantities of the second order only in the changes of refrangibility. Accord-
ingly, the optical distance from AtB9 to A is represented by f(p + 8/x) ds, the
integration being along the original path A,... A; and similarly the optical
distance between A^Bt and B is represented by /(/* + Sfi)ds, the integration
being along Bt ...B. In virtue of (6) the difference of the optical distances to
A and B is
/S/ids (along B.... B) -ftp ds (along A,.. .A) ............. (7)
The new wave-surface is formed in such a position that the optical distance is
constant ; and therefore the dispersion, or the angle through which the wave-
surface is turned by the change of refrangibility, is found simply by dividing
(7) by the distance AB. If, as in common flint-glass spectroscopes, there
is only one dispersing substance, J'8/t ds = fy* . s, where s is simply the thickness
traversed by the ray. If L and f, be the thicknesses traversed by the extreme
rays, and a denote the width of the emergent beam, the dispersion B is
given by
-*,)/«>
106 WAVE THEORY OF LIGHT. [148
or, if £j be negligible,
(8)
The condition of resolution of a double line whose components subtend an
angle 9 is that 0 must exceed \ja. Hence, in order that a double line may
be resolved whose components have indices /* and /* + fyi, it is necessary
that t should exceed the value given by the following equation : —
(9)
For applications of these results, see Spectroscope (Enc. Brit. Vol. xxn. p. 373).
§ 14. Theory of Gratings.
The general explanation of the mode of action of gratings has been given
under Light (Enc. Brit. Vol. XIV. p. 607). If the grating be composed of
alternate transparent and opaque parts, the question may be treated by
means of the general integrals (§ 11) by merely limiting the integration
to the transparent parts of the aperture. For an investigation upon these
lines the reader is referred to Airy's Tracts and to Verdet's Lemons. If,
however, we assume the theory of a simple rectangular aperture (§ 11), the
results of the ruling can be inferred by elementary methods, which are
perhaps more instructive.
Apart from the ruling, we know that the image of a mathematical line
will be a series of narrow bands, of which the central one is by far the
brightest. At the middle of this band there is complete agreement of phase
among the secondary waves. The dark lines which separate the bands are
the places at which the phases of the secondary waves range over an integral
number of periods. If now we suppose the aperture AB to be covered by a
great number of opaque strips or bars of width d, separated by transparent
intervals of width a, the condition of things in the directions just spoken of is
not materially changed. At the central point there is still complete agree-
ment of phase ; but the amplitude is diminished in the ratio of a : a + d.
In another direction, making a small angle with the last, such that the
projection of AB upon it amounts to a few wave-lengths, it is easy to see
that the mode of interference is the same as if there were no ruling. For
example, when the direction is such that the projection of AB upon it
amounts to one wave-length, the elementary components neutralize one
another, because their phases are distributed symmetrically, though dis-
continuously, round the entire period. The only effect of the ruling is to
diminish the amplitude in the ratio a : a + d ', and, except for the difference
in illumination, the appearance of a line of light is the same as if the aperture
were perfectly free.
BRIGHTNESS OF SPECTRA. 107
The lateral (spectral ) images occur in such directions that the projection
of the element (a + d\ of the grating upon them is an exact multiple of A.
The effect of each of the « elements of the grating is then the same ; and.
unless this vanishes on account of a particular adjustment of the ratio a : rf,
the resultant amplitude becomes comparatively very great. These directions,
in which the retardation between A and B is exactly nwiX, may be called the
principal directions. On either side of any one of them the illumination is
distributed according to the same law as for the central image (m = 0),
vanishing,, for example, when the retardation amounts to (IMM ± 1) X. In
ojnsidering the relative brightnesses of the different spectra, it is therefore
sufficient to attend merely to the principal directions, provided that the
whole deviation be not so great that its cosine differs considerably from unity.
We have now to consider the amplitude due to a single element, which we
may conveniently regard as composed of a transparent part a bounded by two
opaque parts of width £d. The phase of the resultant effect is by synimetry
that of the component which comes from the middle of a. The feet that the
other components have phases differing from this by amounts ranging between
± amw (a + d) causes the resultant amplitude to be less than for the central
image (where there is complete phase agreement! If Bm denote the
brightness of the w1* lateral image, and B, that of the central image, we have
Samr]* 'a + d\* . , am*
- -- , \=(- j sin' — , ........ (1)
a+rfj \amrj a + d
If B denote the brightness of the central image when the whole of the space
occupied by the grating is transparent, we have
B9:B = a*:
and thus
The sine of an angle can never be greater than unity: and consequent Iv
under the most favourable circumstances only 1 WITT* of the original light can
be obtained in the a*1* spectrum. We conclude that, with a grating composed
of transparent and opaque parts, the utmost light obtainable in any one
spectrum is in the first, and there amounts to IT*, or about J^. and that for
this purpose a and d must be equaL When d = a. the general formula
becomes
........................... (3)
showing that, when m is even, Bm vanishes, and that, when m is odd,
The third spectrum has thus only | of the brilliancy of the first.
108 WAVE THEORY OF LIGHT. [148
Another particular case of interest is obtained by supposing a small
relatively to (a + d). Unless the spectrum be of very high order, we have
simply
Bm:B={a/(a + d)\> ; ........................... (4)
so that the brightnesses of all the spectra are the same.
The light stopped by the opaque parts of the grating, together with that
distributed in the central image and lateral spectra, ought to make up the
brightness that would be found in the central image, were all the apertures
transparent. Thus, if a = d, we should have
which is true by a known theorem. In the general case
a a\ m= . „ / rmra
' ~
a formula which may be verified by Fourier's theorem.
According to a general principle formulated by Babinet, the brightness of
a lateral spectrum is not affected by an interchange of the transparent and
opaque parts of the grating. The vibrations corresponding to the two parts
are precisely antagonistic, since if both were operative the resultant would be
zero. So far as the application to gratings is concerned, the same conclusion
may be derived from (2).
From the value of Bm : B0 we see that no lateral spectrum can surpass the
central image in brightness; but this result depends upon the hypothesis
that the ruling acts by opacity, which is generally very far from being the
case in practice. In an engraved glass grating there is no opaque material
present by which light could be absorbed, and the effect depends upon a
difference of retardation in passing the alternate parts. It is possible to
prepare gratings which give a lateral spectrum brighter than the central
image, and the explanation is easy. For if the alternate parts were equal
and alike transparent, but so constituted as to give a relative retardation of
£\, it is evident that the central image would be entirely extinguished, while
the first spectrum would be four times as bright as if the alternate parts were
opaque. If it were possible to introduce at every part of the aperture of the
grating an arbitrary retardation, all the light might be concentrated in any
desired spectrum. By supposing the retardation to vary uniformly and
continuously we fall upon the case of an ordinary prism ; but there is then no
diffraction spectrum in the usual sense. To obtain such it would be necessary
that the retardation should gradually alter by a wave-length in passing over
any element of the grating, and then fall back to its previous value, thus
springing suddenly over a wave-length. It is not likely that such a result
1888]
OBLIQUE INCIDENCE.
109
Fig. 9.
Fig. 10.
will ever be fully attained in practice ; but the case is worth stating, in order
to show that there is no theoretical limit to the con-
centration of light of assigned wave-length in one spectrum,
and as illustrating the frequently observed unsymmetrical
character of the spectra on the two sides of the central
image*.
We have hitherto supposed that the light is incident
perpendicularly upon the grating : but the theory is easily
extended. If the incident rays make an angle 0 with the
normal (Fig. 9), and the diffracted rays make an angle <£
(upon the same side), the relative retardation from each
element of width (a + d) to the next is (a + d) (sin 0 + sin <f>) ; and this is the
quantity which is to be equated to mX. Thus
sin 6 + sin 0 = 2 sin £ (0 + <f>) . cos £ (6 -</>) = m\[(a + d). (5 )
The " deviation " is (0 + <f>), and is therefore a minimum when 6 = <f>, i.e.
when the grating is so situated that the angles of incidence and diffraction
are equal.
In the case of a reflexion grating the same method applies. If 0 and <f>
denote the angles with the normal made by the incident and diffracted rays,
the formula (5) still holds, and, if the deviation
be reckoned from the direction of the regularly
reflected rays, it is expressed as before by (0 + <f>\
and is a minimum when 0 = $, that is, when the
diffracted raj's return upon the course of the
incident rays.
In either case (as also with a prism) the posi-
tion of minimum deviation leaves the width of
the beam unaltered, i.e. neither magnifies nor
diminishes the angular width of the object under view.
From (5) we see that, when the light falls perpendicularly upon a grating
(0 = 0), there is no spectrum formed (the image corresponding to in = 0 not
being counted as a spectrum), if the grating interval a- or (a + d) is less
than X. Under these circumstances, if the material of the grating be
completely transparent, the whole of the light must appear in the direct
image, and the ruling is not perceptible. From the absence of spectra
Fraunhofer argued that there must be a microscopic limit represented by X ;
and the inference is plausible, to say the least f. Fraunhofer should, however,
have fixed the microscopic limit at |X, as appears from (5), when we suppose
0 = \vy </> = £TT.
* Phil. Mag. n.vu. 193, 1874. [Vol. i. p. 215.]
t •' Notes on some Fundamental Propositions in Optics," Phil. Mag. June 1886. [Vol. n.
p. 513.]
110 WAVE THEORY OF LIGHT. [148
We will now consider the important subject of the resolving power of
gratings, as dependent upon the number of lines FJ(r
(n) and the order of the spectrum observed (TO).
Let BP (Fig. 11) be the direction of the principal
maximum (middle of central band) for the wave-
length X in the mth spectrum. Then the relative
retardation of the extreme rays (corresponding to
the edges A, B of the grating) is mn\. If BQ be
the direction for the first minimum (the darkness
between the central and first lateral band), the
relative retardation of the extreme rays is (mn + 1)X. Suppose now that
X H- SX is the wave-length for which BQ gives the principal maximum, then
(mn + l)\ = mn(\ + SX) ;
whence
8X/X=l/?wn (6)
According to our former standard, this gives the smallest difference of
wave-lengths in a double line which can be just resolved ; and we conclude
that the resolving power of a grating depends only upon the total number of
lines, and upon the order of the spectrum, without regard to any other con-
siderations. It is here of course assumed that the n lines are really utilized.
In the case of the D-lines the value of SX/X is about 1/1000 ; so that to
resolve this double line in the first spectrum requires 1000 lines, in the
second spectrum 500, and so on.
It is especially to be noticed that the resolving power does not depend
directly upon the closeness of the ruling. Let us take the case of a grating
1 inch broad, and containing 1000 lines, and consider the effect of interpolating
an additional 1000 lines, so as to bisect the former intervals. There will be
destruction by interference of the first, third, and odd spectra generally;
while the advantage gained in the spectra of even order is not in dispersion,
nor in resolving power, but simply in brilliancy, which is increased four times.
If we now suppose half the grating cut away, so as to leave 1000 lines in half
an inch, the dispersion will not be altered, while the brightness and resolving
power are halved.
There is clearly no theoretical limit to the resolving power of gratings,
even in spectra of given order. But it is possible that, as suggested by
Rowland*, the structure of natural spectra may be too coarse to give
opportunity for resolving powers much higher than those now in use. How-
ever this may be, it would always be possible, with the aid of a grating of
given resolving power, to construct artificially from white light mixtures of
* Compare also Lippich, Pogg. Ann. cxxxix. p. 465, 1870 ; Eayleigh, Nature, Oct. 2, 1873,
[Vol. i. p. 183.]
1888] RESOLVING POWER AND DISPERSION. Ill
slightly different wave-lengths whose resolution or otherwise would discriminate
between powers inferior and superior to the given one*.
If we define as the 'dispersion " in a particular part of the spectrum the
ratio of the angular interval d0 to the corresponding increment of wave-length
rfX, we may express it by a very simple formula. For the alteration of
wave-length entails, at the two limits of a diffracted wave-front, a relative
retardation equal to mnd\. Hence, if a be the width of the diffracted beam,
and dd the angle through which the wave-front is turned,
or dispersion = inn/a ............................ (7)
The resolving power and the width of the emergent beam fix the optical
character of the instrument. The latter element must eventually be decreased
until less than the diameter of the pupil of the eye. Hence a wide beam
demands treatment with further apparatus (usually a telescope) of high
magnifying power.
In the above discussion it has been supposed that the ruling is accurate.
and we have seen that by increase of m a high resolving power is attainable
with a moderate number of lines. But this procedure (apart from the question
of illumination) is open to the objection that it makes excessive demands
upon accuracy. According to the principle already laid down, it can make
but little difference in the principal direction corresponding to the first
spectrum, provided each line lie within a quarter of an interval (« + d) from
its theoretical position. But, to obtain an equally good result in the wth
spectrum, the error must be less than 1/m of the above amount t.
There are certain errors of a systematic character which demand special
consideration. The spacing is usually effected by means of a screw, to each
revolution of which corresponds a large number (e.g. one hundred) of lines.
In this way it may happen that, although there is almost perfect periodicity
with each revolution of the screw after (say) 100 lines, yet the 100 lines
themselves are not equally spaced. The " ghosts " thus arising were first
described by Quincke*, and have been elaborately investigated by Peirce§,
both theoretically and experimentally. The general nature of the effects to be
expected in such a case may be made clear by means of an illustration already
* The power of a grating to construct light of nearly definite wave-length is well illustrated
by Young's comparison with the production of a musical note by reflexion of a sudden sound
from a row of palings. The objection raised by Herschel (Light, § 703) to this comparison
depends on a misconception.
f It must not be supposed that errors of this order of magnitude are unobjectionable in all
cases. The position of the middle of the bright band representative of a mathematical line can
be fixed with a spider-line micrometer within a small fraction of the width of the band, just as
the accuracy of astronomical observations far transcends the separating power of the instrument.
* Pogg. Ann. CILVI. p. 1, 1872. § Am. Jour. Math. n. p. 330, 1879.
112
WAVE THEORY OF LIGHT.
[148
employed for another purpose. Suppose two similar and accurately ruled
transparent gratings to be superposed in such a manner that the lines are
parallel. If the one set of lines exactly bisect the intervals between the
others, the grating interval is practically halved, and the previously existing
spectra of odd order vanish. But a very slight relative displacement will
cause the apparition of the odd spectra. In this case there is approximate
periodicity in the half interval, but complete periodicity only after the whole
interval. The advantage of approximate bisection lies in the superior
brilliancy of the surviving spectra; but in any case the compound grating
may be considered to be perfect in the longer interval, and the definition is
as good as if the bisection were accurate.
The effect of a gradual increase in the interval (Fig. 12) as we pass across
the grating has been investigated by Cornu*, who thus explains an anomaly
observed by Mascart. The latter found that certain gratings exercised a
converging power upon the spectra formed upon one side, and a corresponding
j. 13.-7/.
I. 14.— Xs. Fig. 15.— .r?/2
Fig. 12.— x2.
diverging power upon the spectra on the other side. Let us suppose that the
light is incident perpendicularly, and that the grating interval increases from
the centre towards that edge which lies nearest to the spectrum under
observation, and decreases towards the hinder edge. It is evident that the
waves from both halves of the grating are accelerated in an increasing degree,
as we pass from the centre outwards, as compared with the phase they would
possess were the central value of the grating interval maintained throughout.
Fig. 16.— xy. Fig. 17.— x*y.
Fig.
The irregularity of spacing has thus the effect of a convex lens, which
accelerates the marginal relatively to the central rays. On the other side the
effect is reversed. This kind of irregularity may clearly be present in a
degree surpassing the usual limits, without loss of definition, when the
telescope is focused so as to secure the best effect.
It may be worth while to examine further the other variations from
correct ruling which correspond to the various terms expressing the deviation
of the wave-surface from a perfect plane. If sc and y be coordinates in the
C. R. LXXX. p. 645, 1875.
1888] ABERRATIONS IN GRATINGS. 113
plane of the wave-surface, the axis of y being parallel to the lines of the
grating, and the origin corresponding to the centre of the beam, we have
as an approximate equation to the wave-surface (§ 6)
...-, (8)
and, as we have just seen, the term in of corresponds to a linear error in the
spacing. In like manner, the term in y3 corresponds to a general curvature
of the lines (Fig. 13), and does not influence the definition at the (primary)
focus, although it may introduce astigmatism*. If we suppose that everything
is symmetrical on the two sides of the primary plane y = 0, the coefficients
B, j3, 8 vanish. In spite of any inequality between p and p', the definition
will be good to this order of approximation, provided a and 7 vanish. The
former measures the thickness of the primary focal line, and the latter
measures its curvature. The error of ruling giving rise to a is one in which
the intervals increase or decrease in both directions from the centre outwards
(Fig. 14), and it may often be compensated by a slight rotation in azimuth of
"the object-glass of the observing telescope. The term in 7 corresponds to
a variation of curvature in crossing the grating (Fig. 15).
When the plane zx is not a plane of symmetry, we have to consider the
terms in xy, x*y, and y*. The first of these corresponds to a deviation from
parallelism, causing the interval to alter gradually as we pass along the lines
(Fig. 16). The error thus arising may be compensated by a rotation of the
object-glass about one of the diameters y = + x. The term in x*y corresponds
to a deviation from parallelism in the same direction on both sides of the
central line (Fig. 17); and that in y* would be caused by a curvature such
that there is a point of inflexion at the middle of each line (Fig. 18).
All the errors, except that depending on or, and especially those depending
on 7 and 8, can be diminished, without loss of resolving power, by contracting
the vertical aperture. A linear error in the spacing, and a general curvature
of the lines, are eliminated in the ordinary use of a grating.
The explanation of the difference of focus upon the two sides as due to
unequal spacing was verified by Cornu upon gratings purposely constructed
with an increasing interval. He has also shown how to rule a plane surface
with lines so disposed that the grating shall of itself give well-focused
spectra.
* "In the same way we may conclude that in flat gratings any departure from a straight line
has the effect of causing the dost in the slit and the spectrum to have different foci— a fact
sometimes observed" (Rowland, "On Concave Gratings for Optical Purposes," Phil. Mag.
September 1883).
R. IIL 8
114 WAVE THEORY OF LIGHT. [148
A similar idea appears to have guided Rowland to his brilliant invention
of concave gratings, by which spectra can be photo-
graphed without any further optical appliance. In
these instruments the lines are ruled upon a spherical
surface of speculum metal, and mark the intersections
of the surface by a system of parallel and equidistant
planes, of which the middle member passes through
the centre of the sphere. If we consider for the
present only the primary plane of symmetry, the figure
is reduced to two dimensions. Let AP (Fig. 19)
represent the surface of the grating, 0 being the centre
of the circle. Then, if Q be any radiant point and Qf its image (primary
focus) in the spherical mirror AP, we have
1 1 _ 2 cos <ft
#! u a
where v1 = AQ', u = AQ, a = OA, <b — angle of incidence QAO, equal to the
angle of reflexion Q'AO*. If Q be on the circle described upon OA as
diameter, so that u = a cos <£, then Q' lies also upon the same circle ; and in
this case it follows from the symmetry that the unsymmetrical aberration
(depending upon a) vanishes.
This disposition is adopted in Rowland's instrument; only, in addition to
the central image formed at the angle <j>' = </>, there are a series of spectra
with various values of <j>', but all disposed upon the same circle. Rowland's
investigation is contained in the paper already referred to ; but the following
account of the theory is in the form adopted by Glazebrook-f*.
In order to find the difference of optical distances between the courses
QAQ', QPQ', we have to express QP - QA, PQ' - AQ'. To find the former,
we have, if OA Q = </>, A OP = a,
QP2 = w2 + 4a2 sin2 ^«o — 4>au sin %w sin (^«o — <£)
= (u + a sin </> sin o>)2 — a2 sin2 </> sin2 &> + 4a sin2£&> (a — u cos <£).
Now as far as &>4
4 sin2 |«o = sin2 to + \ sin4 o>,
and thus to the same order
QP2 = (u+ a sin <f> sin to)2 — a cos <j> (u — a cos </>) sin2 a> + |a (a — u cos <£) sin4 a>.
* This formula may be obtained as in Optics, Enc. Brit. Vol. xvn. p. 800, equation (3)
[Vol. u. p. 390], and may indeed be derived from that equation by writing tj>'=<p, fj.= - 1.
t Phil. Mag. June 1883, Nov. 1883.
ROWLAND'S CONCAVE GRATINGS. 115
But if we now suppose that Q lies on the circle u = a cos 0, the middle term
vanishes, and we get, correct as far as o>4,
so that
QP— u = a sin <£ sin <a + |a sin<£ tan <£ sin4a>, ............ (9)
in which it is to be noticed that the adjustment necessary to secure the
disappearance of sin* o> is sufficient also to destroy the term in sin* o».
A similar expression can be found for Q'P - Q[A ; and thus, if Q[A = v,
Q[ A 0 = <£', where v = a cos <f>', we get
QP + PQ7 - QA - AQf = a sin a> (sin </> - sin <//)
+ \a sin4 a> (sin $ tan <f> + sin </>' tan $'). . . .(10)
If <f>' = <f>, the term of the first order vanishes, and the reduction of the
difference of path via P and via A to a term of the fourth order proves not
only that Q and Q7 are conjugate foci, but also that the foci are exempt from
the most important term in the aberration. In the present application <£>' is
not necessarily equal to <f> ; but if P correspond to a line upon the grating,
the difference of retardations for consecutive positions of P, so far as expressed
by the term of the first order, will be equal to + m\ (m integral), and
therefore without influence, provided
tr (sin 0 — sin 0') = + m\, ........................ (11)
where a denotes the constant interval between the planes containing the
lines. This is the ordinary formula for a reflecting plane grating, and it
shows that the spectra are formed in the usual directions. They are here
focused (so far as the rays in the primary plane are concerned) upon the
circle OtyA, and the outstanding aberration is of the fourth order.
In order that a large part of the field of view may be in focus at once,
it is desirable that the locus of the focused spectrum should be nearly
perpendicular to the line of vision. For this purpose Rowland places the
eye-piece at 0, so that 0 = 0, and then by (11) the value of 0' in the »ith
spectrum is
o-sin<f>'=± mX ............................... (12)
If o> now relate to the edge of the grating, on which there are altogether
n lines, no- = 2a sin a>, and the value of the last term in (10) becomes
J$n<r sins at sin <f>' tan £',
or ^mn\ sin* to tan <f>' ............................ (13)
This expresses the retardation of the extreme relatively to the central
ray, and is to be reckoned positive, whatever may be the signs of «•> and <£'.
a— 2
116 WAVE THEORY OF LIGHT. [148
If the semi-angular aperture («) be T^, and tan <£' = 1, mn might be as great
as four millions before the error of phase would reach i\. If it were desired
to use an angular aperture so large that the aberration according to (13)
would be injurious, Rowland points out that on his machine there would
be no difficulty in applying a remedy by making tr slightly variable towards
the edges. Or, retaining a- constant, we might attain compensation by so
polishing the surface as to bring the circumference slightly forward in
comparison with the position it would occupy upon a true sphere.
It may be remarked that these calculations apply to the rays in the
primary plane only. The image is greatly affected with astigmatism ; but
this is of little consequence, if 7 in (8) be small enough. Curvature of the
primary focal line having a very injurious effect upon definition, it may be
inferred from the excellent performance of these gratings that 7 is in fact
small. Its value does not appear to have been calculated. The other
coefficients in (8) vanish in virtue of the symmetry.
The mechanical arrangements for maintaining the focus are of great
simplicity. The grating at A and the eye-piece at 0 are rigidly attached to
a bar AO, whose ends rest on carriages, moving on rails OQ, AQ at right
angles to each other. A tie between C and Q can be used if thought
desirable.
The absence of chromatic aberration gives a great advantage in the
comparison of overlapping spectra, which Rowland has turned to excellent
account in his determinations of the relative wave-lengths of lines in the
solar spectrum*.
For absolute determinations of wave-lengths plane gratings are used. It
is found f that the angular measurements present less difficulty than the
comparison of the grating interval with the standard metre. There is also
some uncertainty as to the actual temperature of the grating when in use.
In order to minimize the heating action of the light, it might be submitted
to a preliminary prismatic analysis before it reaches the slit of the spectro-
meter, after the manner of Von Helmholtz (Optics, Enc. Brit. Vol. xvn.
p. 802 [Vol. ii. p. 397]).
Bell found further that it is necessary to submit the gratings to
calibration, and not to rest satisfied with a knowledge of the number of lines
and of the total width. It not unfrequently happens that near the
beginning of the ruling the interval is anomalous. If the width of this
region be small, it has scarcely any effect upon the angular measurements,
and should be left out of account in estimating the effective interval.
* Phil. Mag. March 1887. t Bell, Phil. Mag. March 1887.
1888] GRATINGS. 117
§ 15. Theory of Corrugated Waves.
The theory of gratings is usually given in a form applicable only to the
case where the alternate parts are transparent and opaque. Even then it is
very improbable that the process of simply including the transparent parts
and excluding the opaque parts in the integra-
tions of § 11 gives an accurate result. The
condition of things in actual gratings is much
more complicated, and all that can with confi-
dence be asserted is the approximate periodicity
in the interval a. The problem thus presents
itself — to determine the course of events on
the further side of the plane 2 = 0 when the
amplitude and phase over that plane are periodic
functions of x; and the first step in the solution
would naturally be to determine the effect corresponding to the infinitesimal
strip ydx over which the amplitude and phase are constant. In Fig. 20 QQ'
represents the strip in question, of which the effect is to be estimated at P,
viz. (0, 0, z) ;
QR = y, RP = r, QP = p.
If we assume the law of secondary wave determined in § 10 so as to suit the
resolution of an infinite uniform primary wave, we have, as the effect of QQ',
The development of this expression for the operation of a linear source would
take us too far*. We must content ourselves with the limiting form
assumed when kr is great, as it would almost always be in optics. Under
these circumstances the denominator may be simplified by writing
so that (1) becomes
Now
and thus we obtain
[sin*(a*-r)-cos*(a*-r);=-~ sin *(a« - r - |X), ...(2)
which gives the effect of a linear source at a great distance. The occurrence
of the factor r~* is a consequence of the cylindrical expansion of the waves.
* Theory of Sound, § £41.
118 WAVE THEORY OF LIGHT. [148
The whole effect is retarded one-eighth of a period in comparison with that of
the central element, instead of one-quarter of a period as in the case of a
uniform wave extending over the whole plane.
The effect of the uniform plane wave can be recovered by integrating (2)
with respect to x from — oo to -f oo , on the supposition that kr is great.
We have
dx rdr_ <\/r . d (r - z)
yV ~ >Jr . x ~ V(r + z) . VO - z) '
and in this, since the only elements which contribute sensibly to the integral
are those for which (r — z) is small, we may write
V2'
The integral can then be evaluated by the same formula as before, and we
get finally cos k (at — z), the same as if the primary wave were supposed to
advance without resolution. The recomposition of the primary wave by
integration with rectangular coordinates is thus verified, but only under the
limitation, not really required by the nature of the case, that the point at
which the effect is to be estimated is distant by a very great number of
wave-lengths from the plane of resolution.
We will now suppose that the amplitude and phase of the primary wave
at the plane of resolution z = 0 are no longer constants, but periodic
functions of x. Instead of cos kat simply, we should have to take in general
A cos (px +f) cos kat + B cos (px + g) sin kat ;
but it will be sufficient for our purpose to consider the first term only, in
which we may further put for simplicity A = l, /= 0. The effect of the
linear element at ae, 0, upon a point at f, z, will be, according to (2),
dx
where r is the distance, expressed by r'2 = z" + (x - £)2.
Thus, if we write x- £ + a, the whole effect is
~ j -oo 2V("r) tsin (kat +P% -fr-kr+ pot)
+ sin (kat -pg-^-rr-kr- pa)}, . . .(3)
where r2 = z2 + a2.
In the two terms of this integral the elements are in general of rapidly
fluctuating sign ; and the only important part of the range of integration in
(for example) the first term is in the neighbourhood of the place where
pa. — kr is stationary in value, or where
pda-kdr = 0 ............................... (4)
AMPLITUDE AND PHASE VARIABLE ALONG WAVE-FRONT. 119
In general ada — rdr = 0, so that if the values of a and r corresponding
to (4) be called a., r,, we have
p
Now, in the neighbourhood of these values, if a = a, -f alt
in which by (5) the term of the first order vanishes. Using this in (3), we
get for the first term
~ f
-
- cos (kat + P%-ITT- kr9 + pa9) sin hif],
where for brevity h is written for
The integration is effected by means of the formula
I co8hu-du=l smhu'du=.
and we find
The other term in (3) gives in like manner
so that the complete value is
it cos;
.(6)
When p = 0, we fall back on the uniform plane wave travelling with
velocity a. In general the velocity is not a, but
p*) ............................... (7)
The wave represented by (6) is one in which the amplitude at various
points of a wave-front is proportional to cospg, or cos/xr: and, beyond the
reversals of phase herein implied, the phase is constant, so that the wave-
surfaces are given by z = constant. The wave thus described moves forward
at the velocity given by (7), and with type unchanged.
The above investigation may be regarded as applicable to gratings which
give spectra of the first order only. Although k vary, there is no separation
120 WAVE THEORY OF LIGHT. [148
of colours. Such a separation requires either a limitation in the width of the
grating (here supposed to be infinite), or the use of a focusing lens.
It is important to remark that p has been assumed to be less than k, or <r
greater than X ; otherwise no part of the range of integration in (3) is exempt
from rapid fluctuation of sign, and the result must be considered to be zero.
The principle that irregularities in a wave-front of periods less than \ cannot
be propagated is of great consequence. Further light will be thrown upon it
by a different investigation to be given presently.
The possibility of the wave represented by (6) is perhaps sufficiently
established by the preceding method, but the occurrence of the factor
k/^(k2 — p2) shows that the law of the secondary wave (determined originally
from a consideration of uniform plane waves) was not rightly assumed.
The correct law applicable in any case may be investigated as follows.
Let us assume that the expression for the wave of given periodic time is
(8)
and let us inquire what the value of F ' (x, y) must be in order that the
application of Huygens's principle may give a correct result. From (8)
d <r*f\
and -j-l -- = •
dp\ p J pz
We propose now to find the limiting value of dty/dz when z is very small.
The value of the integral will depend upon those elements only for which x
and y are very small, so that we replace F (x, y) in the limit by ^(0, 0).
Also, in the limit,
so that
Limit ^ = - 2-7T eikat F(0, 0).
The proper value of eikatF(x, y) is therefore that of — dty/dz at the same
point (x, y, 0) divided by 2?r, and we have in general
In the case of the uniform plane wave,
- ik eik(at-z> ;
1888] DIFFERENTIAL EQUATION. 121
SO that
agreeing with what we have already found for the secondary wave in this
case.
But, if -^ = cos px . i
and
The occurrence of the anomalous factor in (6) is thus explained.
It must be admitted that the present process of investigation is rather
artificial : and the cause lies in the attempt to dispense with the differential
equation satisfied by ^, viz.,
on which in the case of sound the whole theory is based. It is in feet easy to
verify that any value of -^r included under (8), where
^-(f
satisfies the equation
When there is no question of resolution by Huygens's principle, the distinction
between £ , 17 and x, y may be dropped.
Starting from the differential equation, we may recover previous results
very simply. If -^r be proportional to cospx cos qy, we have
0 ......................... (11)
If 4s -/>* - 5s = /*',/* being real, the solution of (11) is
where A and B are independent of z. Restoring the factors involving t, x, y,
we may write
+ = cos px cos qy{A***+'»+B **«-**}, ............... (12)
of which the first term may be dropped when we contemplate waves travelling
in the positive direction only. The corresponding realized solution is of the
type
[tat- </(**-?-&.*}. ............ (13)
122 WAVE THEORY OF LIGHT. [148
When &2 > (p2 + q2}, the wave travels without change of type and with velocity
= ......................... (")
We have now to consider what occurs when k*<(pz + q*). If we write
k? — p* — q* = — p?, we have in place of (12)
>/r = cospx cos qy {Aeikat+*z + Beikat-*z} ; ............... (15)
and for the realized solution corresponding to (13)
•^r = cos px cos qy e~>tz cos kat ...................... (16)
We conclude that under these circumstances the motion rapidly diminishes
as z increases, and that no wave in the usual sense can be propagated at all.
It follows that corrugations of a reflecting surface (no matter how deep)
will not disturb the regularity of a perpendicularly reflected wave, provided
the wave-length of the corrugation do not exceed that of the vibration. And,
whatever the former wave-length may be in relation to the latter, regular
reflexion will occur when the incidence is sufficiently oblique.
The first form of solution may be applied to give an explanation of the
appearances observed when a plane wave traverses a parallel coarse grating
and then impinges upon a screen held at varying distances behind*. As the
general expression of the wave periodic with respect to x in distance a- we
may take
A0 cos (kat — kz) + Al cos (px +/i) cos (kat — ^z)
+ Bl cos (px + g^) sin (kat — fj^z) + A2 cos ( '2px + /2) cos (kat — yu.2 z) + . . . ,
where
p = 2-TT/o-, k = 2?r/X, and tf = &- p*, f^ = kz- 4p2,
the series being continued as long as p, is real. We shall here, however, limit
ourselves to the first three terms, and in them suppose Al and B1 to be small
relatively to A0. The intensity may then be represented by
A<? + 2A0A1 cos (px 4/) cos (kz — ^z)
+ <2A0Bl cos (px + g) sin (kz — ^z). . . .(17)
The stripes thrown upon the screen in various positions are thus periodic
functions of z, and the period is
if X be supposed small in comparison with <r. It may be noticed that, if the
position of the screen be altered by the half of this amount, the effect is
equivalent to a shifting parallel to x through the distance \<r. Hence, if the
grating consists of alternate transparent and opaque parts of width \a, the
stripes seen upon the screen are reversed when the latter is drawn back
* Phil. Mag. March 1881, " On Copying Diffraction Gratings and on some Phenomena
connected therewith." [Vol. i. p. 504.]
1888] REVERSALS OF DARK AND BRIGHT BANDS. 123
through the distance cr*/A. In this case we may suppose B^ to vanish, and
(17) then shows that the field is uniform when the screen occupies positions
midway between those which give the most distinct patterns. These results
are of interest in connexion with the photographic reproduction of gratings.
§ 16. Talbot's Bands.
These very remarkable bands are seen under certain conditions when a
tolerably pure spectrum is regarded with the naked eye, or with a telescope,
half the aperture being covered by a thin plate, e.g., of glass or mica. The view
of the matter taken by the discoverer* was that any ray which suffered in
traversing the plate a retardation of an odd number of half wave-lengths
would be extinguished, and that thus the spectrum would be seen interrupted
by a number of dark bars. But this explanation cannot be accepted as it
stands, being open to the same objection as Arago's theory of stellar scintilla-
tion f. It is as far as possible from being true that a body emitting homo-
geneous light would disappear on merely covering half the aperture of vision
with a half-wave plate. Such a conclusion would be in the face of the
principle of energy, which teaches plainly that the retardation in question
leaves the aggregate brightness unaltered. The actual formation of the
bands comes about in a very curious way, as is shown by a circumstance
first observed by Brewster. When the retarding plate is held on the side
towards the red of the spectrum, the bands are not seen. Even in the contrary
case, the thickness of the plate must not exceed a certain limit, however pure
the spectrum may be. A satisfactory explanation of these bands was first given
by AiryJ, but we shall here foitow the investigation of Stokes §, limiting
ourselves, however, to the case where the retarded and unretarded beams are
contiguous and of equal width. The aperture of the unretarded beam may
thus be taken to be limited by x = — h, x = 0, y = - /, y = + I ; and that of the
beam retarded by R to be given by x = Q, x = h, y = -l, y= + /. For the
former (1) § 11 gives
on integration and reduction.
* Phil. Mag. x. p. 364; 1837.
t On account of inequalities in the atmosphere giving a variable refraction, the light from a
star would be irregularly distributed over a screen. The experiment is easily made on a laboratory
scale, with a small source of light, the rays from which, in their course towards a rather distant
screen, are disturbed by the neighbourhood of a heated body. At a moment when the eye, or
object-glass of a telescope, occupies a dark position, the star vanishes. A fraction of a second
later the aperture occupies a bright place, and the star reappears. According to this view the
chromatic effects depend entirely upon atmospheric dispersion.
J Phil. Trans. 1840, p. 225 ; 1841, p. 1. § Ibid. 1848, p. 227.
124 WAVE THEORY OF LIGHT. [148
For the retarded stream the only difference is that we must subtract R
from at, and that the limits of x are 0 and +h. We thus get for the
disturbance at £, vj due to. this stream
2lh f . knl 2/ . k£h
If we put for shortness r for the quantity under the last circular function in
(1), the expressions (1), (2) may be put under the forms it sin T, #sin(T — a)
respectively ; and, if / be the intensity, 7 will be measured by the sum of the
squares of the coefficients of sin T and cos r in the expression
u sin T + v sin (T — a),
so that
/= u? + v2 + 2uv cos a,
which becomes on putting for u, v, and a. their values, and putting
~ . ir _ -rr
7=«wsm v 2+2cos— .............
If the subject of examination be a luminous line parallel to t), we shall obtain
what we require by integrating (4) with respect to rj from — oo to + oo . The
constant multiplier is of no especial interest, so that we may take as applicable
to the image of a line
, 2 . 7r
I
J
If R = £\, I vanishes at % = 0 ; but the whole illumination, represented by
I df;, is independent of the value of R. If R = 0, 7 = ^sin2 — |-, in
-oo ? A/
agreement with §11, where a has the meaning here attached to 2 A.
The expression (5) gives the illumination at f due to that part of the
complete image whose geometrical focus is at £=0, the retardation for this
component being R. Since we have now to integrate for the whole illumin-
ation at a particular point 0 due to all the components which have their foci
in its neighbourhood, we may conveniently regard 0 as origin. £ is then the
coordinate relatively to 0 of any focal point 0' for which the retardation is R;
and the required result is obtained by simply integrating (5) with respect to f
from — oo to + oo . To each value of £ corresponds a different value of X, and
(in consequence of the dispersing power of the plate) of R. The variation
of A, may, however, be neglected in the integration, except in 27rR/\, where a
small variation of X entails a comparatively large alteration of phase. If
we write
(6)
1888] TALBOT'S BANDS. 125
we must regard p as a function of £, and we may take with sufficient approxi-
mation under an ordinar circumstances
(7)
where p' denotes the value of p at 0, and ts is a constant which is positive
when the retarding plate is held at the side on which the blue of the spectrum
is seen. The possibility of dark bands depends upon vr being positive. Only
in this case can
retain the constant value — 1 throughout the integration, and then only when
w-2»*/X/f ................................. (8)
and
COB /»' = -! .................................. (9)
The first of these equations is the condition for the formation of dark bands,
and the second marks their situation, which is the same as that determined
by the imperfect theory.
The integration can be effected without much difficulty. For the first
term in (5) the evaluation is effected at once by a known formula. In the
second term if we observe that
cos [/>' + (w - '2irhi\f) £} = cos {p - g1 £ • = cos p cos 0, £ + sin p sin gl %,
we see that the second part vanishes when integrated, and that the remaining
integral is of the form
w=l sin'Ajf cos^^,
J -x q~
where
h^irh/Xf, g^v-ZirhlXf. ..................... (10)
By differentiation with respect to g1 it may be proved that
w = 0 from^1 = -x to gl = — 2hl,
w = $ir(2h1+gl) from ^ = -2^ to ^ = 0,
w = ^7r(2A1 — <7,) from ^ = 0 to gl = 2hl,
w = Q from g^ =• 2A, to gl = x .
The integrated intensity, F, or
is thus
/'= 2-n-hly ................................. (11)
when gl numerically exceeds 2A, ; and; when gl lies between ± 2A,,
/ = ^{2A1 + (2A1-^)cosp'} ...................... (12)
It appears therefore that there are no bands at all unless «r lies between 0
and +4V. and that within these limits the best bands are formed at the
126 WAVE THEORY OF LIGHT. [148
middle of the range when •or = 2^. The formation of bands thus requires
that the retarding plate be held upon the side already specified, so that «r be
positive ; and that the thickness of the plate (to which -or is proportional) do
not exceed a certain limit, which we may call 27V At the best thickness T0
the bands are black, and not otherwise.
The linear width of the band (e) is the increment of f which alters p by
2?r, so that
e = 27r/-5r (13)
With the best thickness
*r = 27rhl\f, (14)
so that in this case
e = \fjh (15)
The bands are thus of the same width as those due to two infinitely narrow
apertures coincident with the central lines of the retarded and unretarded
streams, the subject of examination being itself a fine luminous line.
If it be desired to see a given number of bands in the whole or in any part
of the spectrum, the thickness of the retarding plate is thereby determined,
independently of all other considerations. But in order that the bands may
be really visible, and still more in order that they may be black, another
condition must be satisfied. It is necessary that the aperture of the pupil be
accommodated to the angular extent of the spectrum, or reciprocally. Black
bands will be too fine to be well seen unless the aperture (2&) of the pupil be
somewhat contracted. One-twentieth to one-fiftieth of an inch is suitable.
The aperture and the number of bands being both fixed, the condition of
blackness determines the angular magnitude of a band and of the spectrum.
The use of a grating is very convenient, for not only are there several spectra
in view at the same time, but the dispersion can be varied continuously by
sloping the grating. The slits may be cut out of tin-plate, and half covered
by mica or " microscopic glass," held in position by a little cement.
If a telescope be employed there is a distinction to be observed, according
as the half-covered aperture is between the eye and the ocular, or in front of
the object-glass. In the former case the function of the telescope is simply to
increase the dispersion, and the formation of the bands is of course independent
of the particular manner in which the dispersion arises. If, however, the half-
covered aperture be in front of the object-glass, the phenomenon is magnified
as a whole, and the desirable relation between the (unmagnified) dispersion
and the aperture is the same as without the telescope. There appears to be
no further advantage in the use of a telescope than the increased facility of
accommodation, and for this of course a very low power suffices.
The original investigation of Stokes, here briefly sketched, extends also to
the case where the streams are of unequal widths h, k, and are separated by an
1888] EXPERIMENTAL CONDITIONS. 127
interval 2g. In the case of unequal widths the bands cannot be black ; but if
h = k, the finiteness of 2g does not preclude the formation of black bands.
The theory of Talbot's bands with a half-covered circular aperture has
been treated by H. Struve*.
§ 17. Diffraction when the Source of Light is not Seen in Focus.
The phenomena to be considered under this head are of less importance
than those investigated by Fraunhofer, and will be treated in less detail ; but,
in view of their historical interest and of the ease with which many of the
experiments may be tried, some account of their theory could not be excluded
from such a work as the present. One or two examples have already attracted
our attention when considering Huygens's zones, viz., the shadow of a circular
disk, and of a screen circularly perforated ; but the most famous problem of
this class — first solved by Fresnel — relates to the shadow of a screen bounded
by a straight edge.
In theoretical investigations these problems are usually treated as of two
dimensions only, everything being referred to the plane passing through the
luminous point and perpendicular to the diffracting edges, supposed to be
straight and parallel. In strictness this idea is appropriate only when the
source is a luminous line, emitting cylindrical waves, such as might be
obtained from a luminous point with the aid of a cylindrical lens. When,
in order to apply Huygens's principle, the wave is supposed to be broken up,
the phase is the same at every element of the surface of resolution which lies
upon a line perpendicular to the plane of reference, and thus the effect of the
whole line, or rather infinitesimal strip, is related in a
constant manner (§15) to that of the element which
lies in the plane of reference, and may be considered to
be represented thereby. The same method of represen-
tation is applicable to spherical waves, issuing from a
point, if the radius of curvature be large ; for, although
there is variation of phase along the length of the
infinitesimal strip, the whole effect depends practically
upon that of the central parts where the phase is sensibly
constant f.
In Fig. 21 APQ is the arc of the circle representa-
tive of the wave-front of resolution, the centre being at 0, and the radius 0 A
* St Petersburg Trans. mi. No. 1, 1883.
+ In experiment a line of light is sometimes substituted for a point in order to increase the
illumination. The various parts of the line are here independent sources, and should be treated
accordingly. To assume a cylindrical form of primary wave would be justifiable only when there
is synchronism among the secondary waves issuing from the various centres.
128 WAVE THEORY OF LIGHT. [148
being equal to a. B is the point at which the effect is required, distant a + b
from 0, so that AB = b, AP = s, PQ = ds.
Taking as the standard phase that of the secondary wave from A, we may
represent the effect of PQ by
where 8 = BP — AP is the retardation at B of the wave from P relatively to
that from A.
Now
S = (a + b)s*/2ab, (1)
so that, if we write
27rS^7r(a + 6)s2 = 7r
\ " ab\ 2 '
the effect at B is
( ab\ H f 2-rrt f . M [
i^-7 - ST}- scos - Icos *TTV 2. dv + sin --- l8
(2(a + 6)j ( r J T J
the limits of integration depending upon the disposition of the diffracting
edges. When a, b, \ are regarded as constant, the first factor may be
omitted, — as indeed should be done for consistency's sake, inasmuch as
other factors of the same nature have been omitted already.
The intensity 72, the quantity with which we are principally concerned,
may thus be expressed
I*={fcos%TTv'*.dv}2+ {Jsm^7rv*.dv}2 ................... (4)
These integrals, taken from v = 0, are known as Fresnel's integrals ; we will
denote them by C and S, so that
rv rv
\ COS^TTV* .dv, S=l siu^TTV^.dv ................ (5)
Jo Jo
When the upper limit is infinity, so that the limits correspond to the
inclusion of half the primary wave, C and S are both equal to £, by a known
formula; and on account of the rapid fluctuation of sign the parts of the
range beyond very moderate values of v contribute but little to the result.
Ascending series for C and S were given by Knockenhauer, and are
readily investigated. Integrating by parts, we find
*1 dv = e''-*
and, by continuing this process,
**+
S 35 357
1888] DIFFRACTION INTEGRALS. 129
By separation of real and imaginary parts,
C = ilfcos£7rwa + JVsiniTTW5, S = Msinfarir> - N cos |irt^ ...(6)
^--+-
~1.3 1.3. 5. 7 + 1. 375T.~9.il ~
These series are convergent for all values of v, but are practically useful only
when v is small.
Expressions suitable for discussion when v is large were obtained by
Gilbert*. Taking
u, ................................. (9)
we may write c+iS = -j-. ......................... (10)
V(27r)7o V«
Again, by a known formula,
i__jr-£-*
V«~VWo V* '
Substituting this in (10), and inverting the order of integration, we get
0+«_j _f *r*~i.. j_ r**!^i. ...(i2)
tr^Jg^xJo W2./o V* i- a;
Thus, if we take
1 f-e-«Var.<fcr 1 f- *-**<&
~W2 Jo 1+** ' = W2j0 V*.(1+*T -(
(7 = £-£costt + .ErsinM, S=^-Gsini<-.H'cosM. ...(14)
The constant parts in (14), viz. ^, may be determined by direct integration of
(12), or from the observation that by their constitution G and H vanish when
ti = oo , coupled with the fact that C and S then assume the value £.
Comparing the expressions for C, S in terms of M, JV, and in terms of
G, H, we find that
G = $ (cosw + sin u) - M, H = $ (cos u - sin u) + N, ...... (15)
formulae which may be utilized for the calculation of G, H when u (or v) is
small. For example, when u = 0, M = 0, N = 0, and consequently G = H = %.
Descending series of the semi-convergent class, available for numerical
calculation when u is moderately large, can be obtained from (12) by writing
x = uy, and expanding the denominator in powers of y. The integration of
the several terms may then be effected by the formula
* Mem. couronne* de PAcad. de Bruxellet, xxu. 1. See also Verdet. Lermu. § 86.
B. III. 9
130 WAVE THEORY OF LIGHT. [148
and we get in terms of v
1 1.3.5 1.3.5.7.9
G=-- + -.:.,
1 1.8 1.8.5.7
The corresponding values of C and 8 were originally derived by Cauchy,
without the use of Gilbert's integrals, by direct integration by parts.
From the series for G and H just obtained it is easy to verify that
-, -l... ....(18)
dv dv
We now proceed to consider more particularly the distribution of light
upon a screen PBQ near the shadow of a straight edge A. At a point P
within the geometrical shadow of the obstacle, the half of the wave to the
right of C (Fig. 22), the nearest point on the wave-front, is wholly intercepted,
and on the left the integration is to be taken from s = CA to s = x . If V be
the value of v corresponding to CA, viz.,
........................ <>»>
we may write
sin \Trtf . dv\\ ............ (20)
+ ( I
or, according to our previous notation,
/2 = (i-CV)2 + (i-SF)2 = (?2+tf2 ................ (21)
Now in the integrals represented by G and H every element diminishes as V
increases from zero. Hence, as GA increases, viz., as
the point P is more and more deeply immersed in
the shadow, the illumination continuously decreases,
and that without limit. It has long been known
from observation that there are no bands on the
interior side of the shadow of the edge.
The law of diminution when V is moderately
large is easily expressed with the aid of the series
(16), (17) for G, H. We have ultimately # = 0,
H=(TT FT1, so that
/2=1/7T2F2,
or the illumination is inversely as the square of the distance from the shadow
of the edge.
For a point Q outside the shadow the integration extends over more than
half the primary wave. The intensity may be expressed by
(22)
STRAIGHT EDGE. 131
and the maxima and minima occur when
whence
sini7rF* + cos£7rF2 = <r (23)
When F= 0, viz., at the edge of the shadow, 72 = \ ; when V= <x , P = 2, on
the scale adopted. The latter is the intensity due to the uninterrupted
wave. The quadrupling of the intensity in passing outwards from the edge
of the shadow is, however, accompanied by fluctuations giving rise to bright
and dark bands. The position of these bands determined by (23) may be
very simply expressed when V is large, for then sensibly G = 0, and
|7rF2 = f7r+/wr (24)
7i being an integer. In terms of 8, we have from (2)
S = (f + £n)X. (25)
The first maximum in fact occurs when S = |X — '0046 X, and the first
minimum when B = £ X — '001 6 X*, the corrections being readily obtainable
from a table of G by substitution of the approximate value of V.
The position of Q corresponding to a given value of V, that is, to a band
of given order, is by (19)
.(26)
2a
By means of this expression we may trace the locus of a band of given order
as b varies. With sufficient approximation we may regard BQ and b as
rectangular coordinates of Q. Denoting them by x, y, so that AB is axis of
y and a perpendicular through A the axis of x, and rationalizing (26), we have
2CW? -V-\y-- FsaXy = 0,
which represents a hyperbola with vertices at 0 and A.
From (24), (26) we see that the width of the bands is of the order
V{fcX(a + 6)/o|. From this we may infer the limitation upon the width of
the source of light, in order that the bands may be properly formed. If to be
the apparent magnitude of the source seen from A, tab should be much
smaller than the above quantity, or
If a be very great in relation to 6, the condition becomes
o><V(X/6), (28)
so that if b is to be moderately great (1 metre), the apparent magnitude of
the sun must be greatly reduced before it can be used as a source.
* Verdet, Leymt, § 90.
9—2
132
WAVE THEORY OF LIGHT.
[148
The values of V for the maxima and minima of intensity, and the
magnitudes of the latter, were calculated by Fresnel. An extract from his
results is given in the accompanying table.
v
I2
First maximum
1-2172
2-7413
First minimum
1-8726 1-5570
Second maximum ...
2-3449 2-3990
Second minimum ...
2-7392
1-6867
Third maximum
3-0820
2-3022
Third minimum
3-3913
1-7440
A very thorough investigation of this and other related questions,
accompanied by fully worked-out tables of the functions concerned, will bo
found in a recent paper by Lommel*.
When the functions G and S have once been calculated, the discussion of
various diffraction problems is much facilitated by the idea, due to Cornuf,
of exhibiting as a curve the relationship between C and S, considered as the
rectangular coordinates (x, y) of a point. Such a curve is shown in Fig. 23,
where, according to the definition (5) of C, S,
— I cos \irif- .dv, y = I sin ^TTVZ . dv
Jo Jo
The origin of coordinates 0 corresponds to v = 0 ; and the asymptotic points
J, J', round which the curve revolves in an ever-closing spiral, correspond to
v=± oo.
The intrinsic equation, expressing the relation between the arc a (measured
from 0) and the inclination <£ of the tangent at any point to the axis of x,
assumes a very simple form. For
so that
Accordingly,
.(30)
.(81)
.(32)
* "Die Beugungserscheinungen geradlinig begrenzter Schirme," Abh. bayer, Akad. der Wilts.
n. Cl. xv. Bd. in. Abth., 1886.
t Journal de Physique, in. p. 1, 1874. A similar suggestion has recently been made
independently by Fitzgerald.
1888] CORXrS SPIRAL. 133
and for the curvature,
£ - ™- -c33)
Cornu remarks that this equation suffices to determine the general
character of the curve. For the osculating circle at any point includes the
whole of the curve which lies beyond: and the successive convolutions
envelop one another without intersection.
The utility of the curve depends upon the fact that the elements of arc
represent, in amplitude and phase, the component vibrations due to the
coircsponding portions of the primary wave-front. For by (30) da = dw, and
by (2) rfr is proportional to ds. Moreover by (2) and (31 1 the retardation of
phase of the elementary vibration from PQ (Fig. 21) is 2»S A, or <£. Hence,
in accordance with the rale for compounding vector quantities, the resultant
vibration at Br due to any finite part of the primary wave, is represented in
amplitude and phase by the chord joining the extremities of the corresponding
In applying the curve in special cases of diffraction to exhibit the effect
at any point P (Fig. 22), the centre of the curve O is to be considered to
correspond to that point C of the primary wave-front which lies nearest to P.
The operative part, or parts, of the curve are of course those which represent
the unobstructed portions of the primary wave.
Let us reconsider, following Gornu, the diffraction of a screen unlimited
on one side, and on the other terminated by a straight edge. On the
illuminated side, at a distance from the shadow, the vibration is represented
by Jf. The coordinates of J, J' being (i,±X(-i-& & ™ 2; and the
phase is | period in arrear of that of the element at 0. As the point under
contemplation is supposed to approach the shadow, the vibration is represented
by the chord drawn from J to a point on the other half of the curve, which
134 WAVE THEORY OF LIGHT. [148
travels inwards from J' towards 0. The amplitude is thus subject to
fluctuations, which increase as the shadow is approached. At the point 0 the
intensity is one-quarter of that of the entire wave, and after this point is
passed, that is, when we have entered the geometrical shadow, the intensity
falls off gradually to zero, without fluctuations. The whole progress of the
phenomenon is thus exhibited to the eye in a very instructive manner.
We will next suppose that the light is transmitted by a slit, and inquire
what is the effect of varying the width of the slit upon the illumination at
the projection of its centre. Under these circumstances the arc to be
considered is bisected at 0, and its length is proportional to the width of the
slit. It is easy to see that the length of the chord (which passes in all cases
through 0) increases to a maximum near the place where the phase-
retardation is f of a period,, then diminishes to a minimum when the
retardation is about f of a period, and so on.
If the slit is of constant width and we require the illumination at various
points on the screen behind it, we must regard the arc of the curve as of
constant length. The intensity is then, as always, represented by the square
of the length of the chord. If the slit be narrow, so that the arc is short, the
intensity is constant over a wide range, and does not fall off to an important
extent until the discrepancy of the extreme phases reaches about a quarter
of a period.
We have hitherto supposed that the shadow of a diffracting obstacle is
received upon a diffusing screen, or, which comes to nearly the same thing,
is observed with an eye-piece. If the eye, provided if necessary with a
perforated plate in order to reduce the aperture, be situated inside the
shadow at a place where the illumination is still sensible, and be focused
upon the diffracting edge, the light which it receives will appear to come
from the neighbourhood of the edge, and will present the effect of a silver
lining. This is doubtless the explanation of a " pretty optical phenomenon,
seen in Switzerland, Avhen the sun rises from behind distant trees standing on
the summit of a mountain*."
§ 18. Diffraction Symmetrical about an Axis.
The general problem of the diffraction pattern due to a source of light
concentrated in a point, when the system is symmetrical about an axis, has
been ably investigated by Lommelf. We must content ourselves here with
a very slight sketch of some of his results.
* Necker, Phil. Mag. Nov. 1832; Fox Talbot, Phil. Mag. June 1833. "When the sun is
about to emerge every branch and leaf is lighted up with a silvery lustre of indescribable
beauty The birds, as Mr Necker very truly describes, appear like flying brilliant sparks."
Talbot ascribes the appearance to diffraction ; and he recommends the use of a telescope.
t Abh. der bayer. Akad. der Wiss. n. Cl. xv. Bd. n. Abth.
1888] LOMMEL'S INVESTIGATION. 135
Spherical waves, centred upon the axis, of radius a fall upon the diffracting
screen ; and the illumination is required on a second screen, like the first
perpendicular to the axis, at a distance (a 4- b) from the source. We have
first to express the distance (d) between an element dS of the wave-front and
a point M in the plane of the second screen. Let £ denote the distance of M
from the axis of symmetry ; then, if we take an axis of x to pass through M,
the coordinates of M are (f, 0, 0). On the same system the coordinates of
dS are
a sin 6 cos <£, a sin B sin <f>, a(l - cos 0)4- 6;
and the distance is given by
d>- = fc + £* - 2of sin B cos $ + 4o (a 4- 6) sin2 \0.
In this expression f and 6 are to be treated as small quantities. Writing p
for o sin 6, we get approximately
The vibration at the wave-front of resolution being denoted by a"1 cos ^trt
the integral expressive of the resultant of the secondaiy waves is (§ 17)
n2ir^-£W.. ...(2)
Substituting pdpdfy for dS, and for d its value from (1), we obtain as the
expression for the intensity at the point f,
where
(4)
(5)
and the following abbreviations have been introduced
£t±»-W, **_t ................... (6)
X 2ab X6
The range of integration is for £ from 0 to 2ir. The limits for p depend
upon the particular problem in hand; but for the sake of detiniteness we will
suppose that in the analytical definitions of C and S the limits are 0 and r, so
as to apply immediately to the problem of a circular aperture of radius r.
If we introduce the notation of Bessel's functions, we have
=2irj r
...(7, 8)
• Used now in an altered
136 WAVE THEORY OF LIGHT. [148
By integration by parts of these expressions Lommel develops series
suitable for calculation. Setting
he finds in the first place
U1 + ^U2], 8 = ^*^--17,, (10,11)
\y \y ) ( \y w
where
-..., ............ (12)
j,(z}-jt(z}+j6(z)- ................ (13)
The series are convenient when y is less than z.
The second set of expressions are
where
These series are suitable when z\y is small.
When the primary wave is complete, r = oo , and we have at once from
the second set of expressions
~ 2?r . Z2 2-^ £2
so that
as we know it should be.
In the application to the problem of the shadow of a circular disk the
limits of integration are from r to oo . If these integrals be denoted by C',
S', we have
(20)
(21)
1888] SHADOW OF CIRCULAR DISC. 137
and
C^ + S'^^F.'+F,'), ........................ (22)
When the point where the illumination is required is situated upon the axis,
£ I, z are zero. Hence F. = 1, F, = 0, and
** (a+bf
the same as if the primary wave had come on unbroken. This is Poisson's
theorem, already found (§ 10) by a much simpler method, in which attention
is limited from the first to points upon the axis. The distribution of light at
other points upon the screen is to be found from (23) by means of the series
(16), (17) for F, and F,. Lommel gives curves for the intensity when y = ir,
2ir, Sir, . . . QTT. The bright central spot is accompanied by rings of varying
intensity.
The limit of the geometrical shadow [£/(a + 6) = r/a] corresponds to y = z.
In this case
V.=J.(*)-J,(*)+JM- ... = ± {./;(*) + 008*}, ......... (24)
F! = Jt (z) -J3(z) + «/,(*)- ... = £sins. ..................... (25)
The numbers computed for special values of y and z apply to a whole class
of problems. Since
both y and z remain unchanged, even when X is constant, if we suppose
fcoca, roc£«Va (26)
We may fall back upon Fraunhofer's phenomena by supposing a = b = x , or
more generally 6 = — a, so that y = 0.
Under these circumstances
z
But it is unnecessary to add anything further under this head
§ 19. Polarization.
A ray of ordinary light is symmetrical with respect to the direction of
propagation. If, for example, this direction be vertical, there is nothing
that can be said concerning the north and south sides of the ray that is
not equally true concerning the east and west sides. In polarized light this
symmetry is lost. Huygens showed that when a ray of such light falls upon
138 WAVE THEORY OF LIGHT. [148
a crystal of Iceland spar, which is made to revolve about the ray as an axis,
the phenomena vary in a manner not to be represented as a mere revolution
with the spar. In Newton's language, the ray itself has sides, or is polarized.
Mains discovered that ordinary light may be polarized by reflexion as well
as by double refraction ; and Brewster proved that the effect is nearly complete
when the tangent of the angle of incidence is equal to the refractive index, or
(which comes to the same) when the reflected and refracted rays are perpen-
dicular to one another. The light thus obtained is said to be polarized in the
plane of reflexion.
Reciprocally, the character of a polarized ray may be revealed by submitting
it to the test of reflexion at the appropriate angle. As the normal to the
reflecting surface revolves (in a cone) about the ray, there are two azimuths
of the plane of incidence, distant 180°, at which the reflexion is a maximum,
and two others, distant 90° from the former, at which the reflexion (nearly)
vanishes. In the latter case the plane of incidence is perpendicular to that
in which the light must be supposed to have been reflected in order to acquire
its polarization.
The full statement of the law of double refraction is somewhat complicated,
and scarcely to be made intelligible except in terms of the wave theory ; but,
in order merely to show the relation of double refraction in a uniaxal crystal,
such as Iceland spar, to polarized light, we may take the case of a prism so
cut that the refracting edge is parallel to the optic axis. By traversing such a
prism, in a plane perpendicular to the edge, a ray of ordinary light is divided
into two, of equal intensity, each of which is refracted according to the ordinary
law of Snell. Whatever may be the angle and setting of the prism, the
phenomenon may be represented by supposing half the light to be refracted
with one index (1'65), and the other half with the different index (1'48). The
rays thus arising are polarized, — the one more refracted in the plane of
refraction, and the other in the perpendicular plane. If these rays are now
allowed to fall upon a second similar prism, held so that its edge is parallel to
that of the first prism, there is no further duplication. The ray first refracted
Avith index 1'65 is refracted again in like manner, and similarly the ray first
refracted with index T48 is again so refracted. But the case is altered if the
second prism be caused to rotate about the incident ray. If the rotation be
through an angle of 90°, each ray is indeed refracted singly ; but the indices
are exchanged. The ray that suffered most refraction at the first prism now
suffers least at the second, and vice versa. At intermediate rotations the
double refraction reasserts itself, each ray being divided into two, refracted
with the above-mentioned indices, and of intensity dependent upon the
amount of rotation, but always such that no light is lost (or gained) on the
whole by the separation.
LAW OF MALTS. 139
The law governing the intensity was formulated by Mains, and has been
verified by the measures of Arago and other workers. If 0 be the angle of
rtrlalinrn from the position in which one of the rays is at a maximum, while
the other vanishes, the intensities are proportional to cos*0 and safO. On
the same scale, if we neglect the loss by reflexion and absorption, the intensity
of the incident light is represented by unity.
A similar law applies to the intensity with which a polarized ray is reflected
from a glass surface at the Brewsterian angle. If 0 be reckoned from the
azimuth of marimnm reflexion, the intensity at other angles may be repre-
sented by cosr0. vanishing when 0 = 90*.
The phenomena here briefly sketched force upon us the view that the
vibrations of tight are transverse to the direction of propagation. In ordinary
light the vibrations are as much in one transverse direction as in another ; and
when such tight falls upon a doubly refracting, or reflecting, medium, the
vibrations are resolved into two definite directions^ constituting two rays
polarized in perpendicular planes, and differently influenced by the medium.
In this case the two rays are necessarily of equal intensity.
Consider, for example, the application of this idea to the reflexion of a ray
of ordinary tight at the Brewsterian, or polarizing, angle. The incident light
may be resolved into two, of equal intensity, and polarized respectively in and
perpendicular to the plane of incidence. Now we know that a ray polarized
in the plane perpendicular to that of incidence will not be reflected, will in
fact be entirely transmitted ; and the necessary consequence is that all the
light reflected at this angle will be polarized in the plane of incidence. The
operation of the plate is thus purely selective, the polarized component, which
is missing in the reflected tight, being represented in undue proportion in the
transmitted light
If the incident light be polarized, suppose at an angle 0 with the plane of
incidence, the incident vibration may be resolved into cos 0 in thv «.<K- plane
and sin 0 in the other. The latter polarized component is not reflected. The
reflected light is thus in all cases polarized in the plane of reflexion ; and it?
intensity, proportional to the square of the vibration, is represented by Acosrl.
if A be the intensity in which light is reflected when polarized in the plane of
reflexion. The law of Mains is thus a necessary consequence of the principle
of resolution.
The idea of transverse vibrations was admitted with reluctance, even by
Young and Fresnel themselves. A perfect fluid, such as the ethereal medium
was then supposed to be, is essentially incapable of transverse vibrations. But
there seems to be no reason d priori for preferring one kind of vibration
to another; and the phenomena of polarization prove conclusively that, if
luminous vibrations are analogous to those of a material medium, it is to
140 WAVE THEORY OF LIGHT. [148
solids, and not to fluids, that we must look. An isotropic solid is capable
of propagating two distinct kinds of waves, — the first dependent upon rigidity.
or the force by which shear is resisted, and the second analogous to waves of
sound and dependent upon compressibility. In the former the vibrations are
transverse to the direction of propagation, that is, they may take place in any
direction parallel to the wave-front, and they are thus suitable representatives
of the vibrations of light. In this theory the luminiferous ether is distinctly
assimilated to an elastic solid, and the velocity of light depends upon the
rigidity and density assigned to the medium.
The possibility of longitudinal waves, in which the displacement is
perpendicular to the wave-front, is an objection to the elastic solid theory
of light, for there is nothing known in optics corresponding thereto. If,
however, we suppose with Green that the medium is incompressible, the
velocity of longitudinal waves becomes infinite, and the objection is in
great degree obviated. Such a supposition is hardly a departure from the
original idea, inasmuch as, so far as we know, there is nothing to prevent
a solid material possessing these properties, and an approximation is actually
presented by such bodies as jelly, in which the velocity of longitudinal
vibrations is a large multiple of that of transverse vibrations.
§ 20. Interference of Polarized Light.
The conditions of interference of polarized light are most easily deduced
from the phenomena of the colours of crystalline plates, if we once admit
Young's view that the origin of the colours is to be sought in the interference
of the differently refracted rays. Independently of any hypothesis of this
kind, the subject was directly investigated by Fresnel and Arago*, who
summarized their conclusions thus: —
(1) Under the same conditions in which two rays of ordinary light appear
to destroy one another, two rays polarized in contrary (viz., perpendicular)
directions are without mutual influence.
(2) Two rays of light polarized in the same direction act upon one another
like ordinary rays ; so that, with these two kinds of light, the phenomena of
interference are identical
(3) Two rays originally polarized in opposite directions may afterwards be
brought to the same plane of polarization, without thereby acquiring the power
to influence one another.
(4) Two rays polarized in opposite directions, and afterwards brought to
similar polarizations, react in the same manner as natural rays, if they are
derived from a beam originally polarized in one direction.
* Fresnd'8 War**, VoL i. p. 521.
1888] FLASK P€HLABIZH> UKWY. 141
The fact that oppositely polarized rays cannot be made to interfere may
of itself be regarded as a proof that the vibrations are transverse: and the
principle, once admitted, gives an intelligible account of all the varied
phenomena in this field of optics. The only points on which any difficulty
arises are as to the nature of ordinary unpolarized light, and the rales
according to which intensity is to be calculated. It will be proper to
consider these questions somewhat fully.
In ordinary (plane) polarized light the vibrations are supposed to be m.
one direction only. If x and y be rectangular coordinates in the plane of the
wave, we may take, as representing a regular vibration of plane-polarized light,
where <f> = 2-jrt/r, and a, a. denote constants. It must be remembered, however,
that in optics a regular vibration of this kind never presents itself In the
simplest case of approximately monochromatic light, the amplitude and phase
must be regarded (| 4) as liable to incessant variation, and all that we are
able to appreciate is the mean intensity, represented by J^o2). If a number
of these irregular streams of light are combined, the intensity of the mixture
cannot be calculated from a mere knowledge of the separate intensities, unless
we have assurance that the streams are independent, that is, without mutual
phase-relations of a durable character. For instance, two thoroughly similar
streams combine into one of four-fold intensity, if the phases are the same ;
while, if the phases are opposed, the intensity tails to zero. It is only when
the streams are independent, so that the phase-relation is arbitrary and
variable from moment to moment, that the apparent resultant intensity is
necessarily the double of the separate intensities.
If any number of independent vibrations of type (1) be superposed, nhe
resultant is
[2^! cos «J cos
and the momentary intensity is
[2% eos « + ^ sn
or
The phase-relations being unknown, this quantity is quite indeterminate.
But, since each cosine varies from moment to moment, and on the whole is
as much positive as negative, the mewi intensity is
that is to aay, is to be found by simple addition of the separate intensities.
Let us now dispense with the restriction to one direction of vibration, and
consider in the first place the character of a regular vibration, of given
frequency. The general expression will be
142 WAVE THEORY OF LIGHT. [148
where a, a, b, /9 are constants. If @ = a, the vibrations are executed entirely
in the plane as/y = a/6, or the light is plane-polarized. Or if ft = TT — a, the
light is again plane-polarized, the plane of vibration being x\y = — a/6. In
other cases the vibrations are not confined to one plane, so that the light is
not plane-polarized, but, in conformity with the path denoted by (2), it is said
to be elliptically -polarized. If one of the constituents of elliptically-polarized
light be suitably accelerated or retarded relatively to the other, it may be
converted into plane-polarized light, and so identified by the usual tests. Or,
conversely, plane-polarized light may be converted into elliptically-polarized
by a similar operation. The relative acceleration in question is readily effected
by a plate of doubly refracting crystal cut parallel to the axis.
If ft = a + ^TT, whether in the first instance or after the action of a
crystalline plate,
# = acos(<£ — a), y= ± 6 sin (<£ - a) (3)
The maxima and minima values of the one coordinate here occur synchronously
with the evanescence of the other, and the coordinate axes are the principal
axes of the elliptic path.
An important particular case arises when further 6 = a. The path is then
a circle, and the light is said to be circularly-polarized. According to the
sign adopted in the second equation (3), the circle is described in the one
direction or in the other.
Circularly-polarized light can be resolved into plane-polarized components
in any two rectangular directions, which are such that the intensities are equal
and the phases different by a quarter period. If a crystalline plate be of such
thickness that it retards one component by a quarter of a wave-length (or
indeed by any odd multiple thereof) relatively to the other, it will convert
plane-polarized light into circularly-polarized, and conversely, — in the latter
case without regard to the azimuth in which it is held.
The property of circularly-polarized light whereby it is capable of resolution
into oppositely plane-polarized components of equal intensities is possessed
also by natural unpolarized light; but the discrimination may be effected
experimentally with the aid of the quarter- wave plate. By this agency the
circularly-polarized ray is converted into plane-polarized, while the natural
light remains apparently unaltered. The difficulty which remains is rather to
explain the physical character of natural light. To this we shall presently
return ; but in the meantime it is obvious that the constitution of natural
light is essentially irregular, for we have seen that absolutely regular, i.e.,
absolutely homogeneous, light is necessarily (elliptically) polarized.
In discussing the vibration represented by (2), we have considered the
amplitudes and phases to be constant ; but in nature this is no more attain-
able than in the case of plane-polarized light. In order that the elliptic
1888] CHARACTER OF NATURAL LIGHT. 143
polarization may be of a definite character, it is only necessary that die ratio
of amplitudes and the difference of phases should be absolute constants, and
this of course is consistent with the same degree of irregularity as was
admitted for plane vibrations.
The intensity of elliptically-polarized light is the sum of the intensities of
its rectangular components. This we may consider to be an experimental
feet, as well as a consequence of the theory of transverse vibrations. In what-
ever form such a theory may be adopted, the energy propagated will certainly
conform to this law. When the constants in (2) are regarded as subject to
variation, the apparent intensity is represented by
We are now in a position to examine the constitution which must be
ascribed to natural light. The conditions to be satisfied are that when
resolved in any plane the mean intensity of the vibrations shall be inde-
pendent of the orientation of the plane, and, further, that this property
shall be unaffected by any previous relative retardation of the rectangular
components into which it may have been resolved. The original vibration
being represented by
or, as we may write it, since we are concerned only with phase differences,
*=««**, y=6cos(*-S), .................. (5)
let us suppose that die second component is subjected to a retardation e.
Thus
x = acosf, y = 6cos(£-S-e), ............... (6)
in which a, b, & will be regarded as subject to rapid variation, while e remains
constant. If the vibration represented by (6) be now resolved in a direction
x, making an angle •» with f. we have
^ cos •» + 6 cos (^ — 5 — ^) sin *»
«* + 6sin» cos (£ + e)] cos £ + 6 sin « sin (8 + e) sin (f> :
and the intensity is
o*cosItt + 6isin-« + 2a6cus» sin « cos (8 + «). ............ (7)
Of this expression we take the mean, w and e remaining constant. Thus the
may be written
+ J^(^)sm*« + 2Jr[aAcos(S + €)]cos.sin«. ...(8)
In order now that the stream may satisfy the conditions laid down as
necessary for natural light, (8) must be independent of • and e; so that
M (ab oos S)= M (ab sin 8) = 0. ..................... (10)»
Yenfct, Le^ufOpti^ Pky^me, Vol. «. p. 85.
144 WAVE THEORY OF LIGHT. [148
In these equations a2 and 62 represent simply the intensities, or squares of
amplitudes, of the x and y vibrations ; and the other two quantities admit
also of a simple interpretation. The value of y may be written
y = bcos8 cos <£ + b sin 8 sin</>; (11)
from which we see that b cos B is the coefficient of that part of the y vibration
which has the same phase as the x vibration. Thus ab cos 8 may be
interpreted as the product of the coefficients of the parts of the x and y
vibrations which have the same phase. Next suppose the phase of y
accelerated by writing ^TT + <f> in place of 0. We should thus have
y— — b cos 8 sin <£ + 6 sin 8 cos <f>,
and ab sin 8 represents the product of the coefficients of the parts which are
now in the same phase, or (which is the same) the product of the coefficients
of the x vibration and of that part of the y vibration which was 90° behind
in phase. In general, if
x = h cos <f> + h' sin tf>, y = kcos <£ + k' sin </>, (12)
the first product is hk -f h'k' and the second is hk' — h'k.
Let us next examine how the quantities which we have been considering
are affected by a transformation of coordinates in accordance with the
formulae
x' = as cos &) + y sin &>, y' = — x sin w + y cos o> (13)
We find
x = cos <f> {a cos ft> + b sin o> cos 8} + sin <f> . b sin 8 sin &>, (14)
y' == cos 0 { — a sin G> + b cos &> cos 8} + sin <£ . 6 sin 8 cos &> ; . . .(15)
whence
amp.2 of x' = a2 cos2 a> + 62 sin2 a> + 2ab cos 8 sin ai cos a>, . . .(16)
amp.2 of y' = a2 sin2 &> + 62 cos2 G> — 2a& cos 8 sin o> cos a>. . . .(17)
In like manner
First product = (b- — a?) sin &> cos a> + a& cos £ (cos2 G> - sin2 &>), (18)
Second product = ab sin 8 (19)
The second product, representing the circulating part of the motion, is thus
unaltered by the transformation.
Let us pass on to the consideration of the mean quantities which occur in
(9), (10), writing for brevity
B, M(abcos8)=C, M
1888] UXPOLARIZED LIGHT. 145
From (16), (17), (18), (19), if A', E\ C', D' denote the corresponding
quantities after transformation,
A'=A coss« + .Bsin1a» + 2(7cos« sinw, (20)
B' = A sin* « + B cos1 « - 2C cos » sinw, (21)
(7/ = C(cosa«-8in»w) + (5-^l)co8« sin», (22)
D' = D. (23)
These formulae prove that, if the conditions (9), (10), shown to be necessary
in order that the light may behave as natural light, be satisfied for one set
of axes, they are equally satisfied with any other. It is thus a matter of
indifference with respect to what axes the retardation e is supposed to be
introduced, and the conditions (9), (10) are sufficient, as well as necessary, to
characterize natural light.
Reverting to (8), we see that, whether the light be natural or not, its
character, so far as experimental tests can show, is determined by the values
of A, B, C, D. The effect of a change of axes is given by (20), &c., and it is
evident that the new axes may always be so chosen that C' = 0. For this
purpose it is only necessary to take to such that
tan 2o> = '2Cj(A - B).
If we choose these new axes as fundamental axes, the values of the constants
for any others inclined to them at angle a> will be of the form
A = AI cos* o> + Bl sin* co 1
B = At sin1 » + Bl cos*« f (24)
C = (BJ -AJ cos o> sinw J
If A! and BI are here equal, then C = 0, A = B for all values of a>. In
this case, the light cannot be distinguished from natural light by mere
resolution ; but if D be finite, the difference may be made apparent with the
aid of a retarding plate.
If A! and Bl are unequal, they represent the maximum and minimum
values of A and B. The intensity is then a function of the plane of resolution,
and the light may be recognized as partially polarized by the usual tests.
If either Aj or 5, vanishes, the light is plane-polarized*.
When several independent streams of light are combined, the values, not
only of A and B, but also of C and D, for the mixture, are to be found by
simple addition. It must here be distinctly understood that there are no
permanent phase-relations between one component and another. Suppose,
for example, that there are two streams of light, each of which satisfies the
relations A =B, (7= 0, but makes the value of D finite. If the two values
of D are equal and opposite, and the streams are independent, the mixture
* In this case D, necessarily vanishes.
K. III. 10
146 WAVE THEORY OF LIGHT. [148
constitutes natural light. A particular case arises when each component is
circularly-polarized (D = ± A = ± B), one in the right-handed and the other
in the left-handed direction. The intensities being equal, the mixture is
equivalent to natural light, but only under the restriction that the streams
are without phase-relation. If, on the contrary, the second stream be similar
to the first, affected merely with a constant retardation, the resultant is not
natural, but completely (plane) polarized light.
We will now prove that the most general mixture of light may be
regarded as compounded of one stream of light elliptically-polarized in a
definite manner, and of an independent stream of natural light. The
theorem is due to Stokes*, but the method that we shall follow is that of
Verdetf.
In the first place, it is necessary to observe that the values of the
fundamental quantities A, B, C, D are not free from restriction. It will be
shown that in no case can C2 + D2 exceed AB.
In equations (2), expressing the vibration at any moment, let alt blt a1} fti,
be the values of a, b, ct, B during an interval of time proportional to m^ , and
in like manner let the suffixes 2, 3, ... correspond to times proportional to
m-i, m3, .... Then
AB = mfafbf + ra22a22622 + . . . -f m^ (afb? + afb?) +... .
Again, by (12),
C = m^bi (cos «! cos & + sin o^ sin &) + ...
cos §!
D = WjeZj&j sin 8l + w^a^ sin
where, as before,
«i = ft-«i, S2 =
Thus,
From these equations we see that AB—C*-D2 reduces itself to a sum of
terms of the form
each of which is essentially positive.
The only case in which the sum can vanish is when
S1 = g2 = S3=...,
and further b1:a1 = b2:a2 = b3:a3= ... .
Under these conditions the light is reduced to be of a definite elliptic
* "On the Composition and Kesolution of Streams of Light from Different Sources," Camb.
Phil. Trans. 1852. f Loc. cit. p. 94.
1888] ANALYSIS OF GEXERAL CASE. 147
character, although the amplitude and phase of the system as a whole may
be subject to rapid variation. The elliptic constants are given by
(25)
In general AB exceeds (C* + D*)i but it will always be possible to find
a positive quantity H. which when subtracted from A and B (themselves
necessarily positive) shall reduce the product to equality with C* + DS, in
accordance with
..................... (26)
The original light may thus be resolved into two groups. For the first group
the constants are H, H, 0, 0 : and for the second A - H. B - H. C. D. Each
of these is of a simple character; for the first represents natural light, and
the second light eltiptically-polarized. It is thus proved that in general
a stream of light may be regarded as composed of one stream of natural light
and of another elliptieally-polarized. The intensity of the natural light is
2ff , where from (26)
(27*
The elliptic constants of the second component are given by
«?>*=£»- H)(A -H), tanS=D C, ............ (28)
and
M(a*)=A-H. .............................. (29)
If D=0, and therefore by (28) B = 0, the second component is plane-polarized.
This is regarded as a particular case of elliptic polarization. Again, if A = B.
(7 = 0, the polarization is circular.
The laws of interference of polarized light, discovered by Fresnel and
Arago, are exactly what the theory of transverse vibrations would lead us to
expect, when once we have cleared up the idea of unpolarized light. Ordinary
sources, such as the sun, emit nnpolarized light. If this be resolved in two
opposite directions, the polarized components are not only each irregular, but
there is no permanent phase-relation between them. Xo tight derived from
one can therefore ever interfere regularly with tight derived from the other.
K. however, we commence with plane-polarized light, we have only one
series of irregularities to deal with. When resolved in two rectangular
directions, the components cannot then interfere, but only on account of the
perpendicularity. If brought back by resolution to the same plane of polari-
zation, interference becomes possible, because the same series of irregularities
are to be found in both components.
10—2
148 WAVE THEORY OF LIGHT. [148
§ 21. Double Refraction.
The construction by which Huygens explained the ordinary and extra-
ordinary refraction of Iceland spar has already been given (Light, Enc. Brit.
Vol. XIV. p. 610). The wave-surface is in two sheets, composed of a sphere
and of an ellipsoid of revolution, in contact with one another at the
extremities of the polar axis. In biaxal crystals the wave-surface is of
a more complicated character, including that of Huygens as a particular case.
It is not unimportant to remark that the essential problem of double
refraction is to determine the two velocities with which plane waves are
propagated, when the direction of the normal to the wave-front is assigned.
When this problem has been solved, the determination of the wave-surface is
a mere matter of geometry, not absolutely necessary for the explanation of
the leading phenomena, but convenient as affording a concise summary of
the principal laws. In all cases the wave-surface is to be regarded as the
envelope at any subsequent time of all the plane wave-fronts which at
a given instant may be supposed to be passing through a particular point.
In singly refracting media, where the velocity of a wave is the same in all
directions, the wave-normal coincides with the ray. In doubly refracting
crystals this law no longer holds good. The principles by which the
conception of a ray is justified (§ 10), when applied to this case, show that
the centre of the zone system is not in general to be found at the foot of the
perpendicular upon the primary wave-front. The surface whose contact with
the primary wave-front determines the element from which the secondary
disturbance arrives with least retardation is now not a sphere, but whatever
wave-surface is appropriate to the medium. The direction of the ray,
corresponding to any tangent plane of the wave-surface, is thus not the
normal, but the radius vector drawn from the centre to the point of contact.
The velocity of propagation (reckoned always perpendicularly to the
wave-front) may be conceived to depend upon the direction of the wave-front,
or wave-normal, and upon what we may call (at any rate figuratively) the
direction of vibration. If the velocity depended exclusively upon the wave-
normal, there could be no double, though there might be extraordinary,
refraction, i.e., refraction deviating from the law of Snell ; but of this nothing
is known in nature. The fact that there are in general two velocities for one
wave-front proves that the velocity depends upon the direction of vibration.
According to the Huygenian law, confirmed to a high degree of accuracy
by the observations of Brewster and Swan*, a ray polarized in a principal
plane (i.e., a plane passing through the axis) of a uniaxal crystal suffers
ordinary refraction only, that is, propagates itself with the same velocity in
* Edin. Trans. Vol. xvi. p. 375.
1888] FRESXEL'S HYPOTHESES. 149
all directions. The interpretation which Fresnel put upon this is that the
vibrations (understood now in a literal sense) are perpendicular to the plane
of polarization, and that the velocity is constant because the direction of
vibration is in all cases similarly related (perpendicular) to the axis. The
development of this idea in the fertile brain of Fresnel led him to the
remarkable discovery of the law of refraction in biaxal crystals.
The hypotheses upon which Fresnel based his attempt at a mechanical
theory are thus summarized by Verdet : —
(1) The vibrations of polarized light are perpendicular to the plane of
polarization ;
(2) The elastic forces called into play during the propagation of a system
of plane waves (of rectilinear transverse vibrations) differ from the elastic
forces developed by the parallel displacement of a single molecule only by
a constant factor, independent of the particular direction of the plane of the
wave;
(3) When a plane wave propagates itself in any homogeneous medium,
the components parallel to the wave-front of the elastic forces called into
play by the vibrations of the wave are alone operative;
(4) The velocity of a plane wave which propagates itself with type
unchanged in any homogeneous medium is proportional to the square root of
the effective component of the elastic force developed by the vibrations.
Fresnel himself was perfectly aware that his theory was deficient in
rigour, and indeed there is little to be said in defence of his second hypothesis.
Nevertheless, the great historical interest of this theory, and the support that
experiment gives to Fresnel's conclusion as to the actual form of the wave-
surfece in biaxal crystals, render some account of his work in this field
imperative.
The potential energy of displacement of a single molecule from its position
of equilibrium is ultimately a quadratic function of the three components
reckoned parallel to any set of rectangular axes. These axes may be so
chosen as to reduce the quadratic function to a sum of squares, so that the
energy may be expressed,
K=ia»f» + **V + **£', (1)
where £, 17, £ are the three component displacements. The corresponding
forces of restitution, obtained at once by differentiation, are
X = a*£, F=6->, Z=<*£ (2)
The force of restitution is thus in general inclined to the direction of
displacement. The relation between the two directions X, Y, Z and £, rj, £ is
the same as that between the normal to a tangent plane and the radius vector
p to the point of contact in the ellipsoid
150 WAVE THEORY OF LIGHT. [148
If a?, b2, c2 are unequal, the directions of the coordinate axes are the only ones
in which a displacement calls into operation a parallel force of restitution.
If two of the quantities a?, b'2, c2 are equal, the ellipsoid (3) is of revolution,
and every direction in the plane of the equal axes possesses the property in
question. This is the case of a uniaxal crystal. If the three quantities
a2, 62, c2 are all equal, the medium is isotropic.
If we resolve the force of restitution in the direction of displacement, we
obtain a quantity dependent upon this direction in a manner readily ex-
pressible by means of the ellipsoid of elasticity (3). For, when the total
displacement is given, this quantity is proportional to
that is to say, to the inverse square of the radius vector p in (3).
We have now to inquire in what directions, limited to a particular plane,
a displacement may be so made that the projection of the force of restitution
upon the plane may be parallel to the displacement. The answer follows at
once from the property of the ellipsoid of elasticity. For, if in any section of
the ellipsoid we have a radius vector such that the plane containing it and
the normal to the corresponding tangent plane is perpendicular to the plane
of the section, the tangent line to the section must be perpendicular to the
radius vector, that is, the radius vector must be a principal axis of the section.
There are therefore two, and in general only two, directions in any plane
satisfying the proposed condition, and these are perpendicular to one another.
If, however, the plane be one of those of circular section, every line of
displacement is such that the component of the force, resolved parallel to the
plane, coincides with it.
According to the principles laid down by Fresnel, we have now complete
data for the solution of the problem of double refraction. If the direction of
the wave-front be given, there are (in general) only two directions of vibration
such that a single wave is propagated. If the actual displacements do not
conform to this condition, they will be resolved into two of the required
character, and the components will in general be propagated with different
velocities. The two directions are the principal axes of the section of (3)
made by the wave-front, and the velocities of propagation are inversely
proportional to the lengths of these axes.
The law connecting the lengths of the axes with the direction (7, m. n) of
the plane is a question of geometry*; and indeed the whole investigation of
ithe wave-surface may be elegantly carried through geometrically with the
aid of certain theorems of MacCullagh respecting apsidal surfaces (Salmon,
* See Salmon's Analytical Geometry of Three Dimensions, Dublin 1882, § 102.
1888] LAW OF VELOCITY. 151
ch. xiv.). For this, however, we have not space, and must content ourselves
with a sketch of the analytical method of treatment.
If r be the velocity of propagation in direction I, TO, w, the wave-surface
is the envelope of planes
lx+ my + nz = v ............................... (4)
where r is a function of /, m, n, whose form is to be determined. If (X, /*, v)
be the corresponding direction of vibration, then
fX+mf4+nv = 0 ............................... (5)
According to the principles laid down by Fresnel, we see at once that the
force of restitution (a*X, #*/*, c*v), corresponding to a displacement unity, is
equivalent to a force r* along (X, /*, v), together with some force (P) along
(/, m, «). Resolving parallel to the coordinate axes, we get
IP mP nP
Multiplying these by I, m, n respectively, and taking account of (5). we see
that
is the relation sought for between r and (/, m, n). In this equation 6, c are
the velocities when the direction of propagation is along x, the former being
applicable when the vibration is parallel to y, and the latter when it is
parallel to x.
The directions of vibration are determined by (5) and by the consideration
that (I, TO, n), (X, /*, v), and (a*X, 61/*, cfv) lie in a plane, or (as we may put it )
are all perpendicular to one direction (f, g, h). Thus
lf+ mg+ nh=0 }
"* = ° ......................... 8
The determinant expressing the result of the elimination of /: g : h may be
put into the form
which with (5) suffices to determine (X, /*, v) as a function of (I, m, n).
The feet that the system of equations (5), (8) is symmetrical as between
(X, /*, v) and (f,g, h) proves that the two directions of vibration corresponding
to a given (/, TO, ») are perpendicular to one another.
The direct investigation of the wave-surface from (4) and (7) was first
effected by Ampere, but his analytical process was very laborious. Fresnel had
152 WAVE THEORY OF LIGHT. [148
indeed been forced to content himself with an indirect method of verification.
But in the following investigation of A. Smith* the eliminations are effected
with comparatively little trouble.
In addition to (4) and (7), we know that
To find the equation to the envelope, we have to differentiate these equations,
making I, m, n, v vary. Eliminating the differentials by the method of
multipliers, we obtain the following: —
and
p m2
I n '
The equations (11), (12), (13) multiplied by Z, m, n respectively, and added,
y = A (15)
The same equations, squared and added, give
x? + f + 2* = A* + B/v.
If we put r2 for aP + y^ + z2, and for J. the value just found, we obtain
B = v(r*-vz) (16)
If these values of A and B be substituted in (11),
, r2-a2
or
If we substitute this value of I, and the corresponding values of m, n in (4),
we get
(18)
=
t* . a« ' r2 _ tf r2 - c2 r2 r2 r2 '
whence
as the equation of the wave-surface.
By (6) equation (11) may be written
from which and the corresponding equations we see that the direction (x, y, z)
lies in the same plane as (I, m, n) and (X, p, v). Hence in any tangent plane
Camb. Trans, vi. 1835.
1888]
FRESNELS SURFACE.
153
of the wave-surface the direction of vibration is that of the line joining the
foot of the perpendicular and the point of contact (x, y, z).
The equation (18) leads to another geometrical definition of Fresnel's
wave-surface. If through the centre of the ellipsoid reciprocal to the
ellipsoid of elasticity (3), viz.,
.(19)
a plane be drawn, and on the normal to this plane two lengths be marked off
proportional to the axes of the elliptic section determined by the plane, the
locus of the points thus obtained, the apsidal surface of (19), is the wave-
surface (18).
Fully developed in integral powers of the coordinates, (18) takes the form
(«» + y* + z*) (a*x* + &Y + c2^) - a2 (ft2 + c2)*2
-62(c2 + a2)y2-c2(a2 + 6s)^+a262c2=0 ....... (20)
The section of (20) by the coordinate plane y = 0 is
a?c3) = 0, .................. (21)
Fig. 24.
representing a circle and an ellipse (Fig. 24). That the sections by each of
the principal planes would be a circle and an ellipse might have been foreseen
independently of a general solution of the envelope problem. The forms of
the sections prescribed in (21) and the two similar
equations are sufficient to determine the character of
the wave-surface, if we assume that it is of the fourth
degree, and involves only the even powers of the
coordinates. It was somewhat in this way that the
equation was first obtained by Fresnel.
If two of the principal velocities, e.g., a and 6, are
equal, (20) becomes
(a* + y* + * - a*) (a'a? + asy2 + c2^ - a2c2) = 0, . . .(22)
so that the wave-surface degenerates into the Huygen-
ian sphere and ellipsoid of revolution appropriate to a
uniaxal crystal. The two sheets touch one another at the points x = 0, y = 0.
z= ±a. If c> a, as in Iceland spar, the ellipsoid is external to the sphere.
On the other hand, if c < a, as in quartz, the ellipsoid is internal.
We have seen that when the wave-front is parallel to the circular sections
of (3), the two wave-velocities coincide. Thus in (7), if a2, lr, c2 be in descend-
ing order of magnitude, we have m = 0, r = b ; so that
_c* a'-c2'
.(23)
154 WAVE THEORY OF LIGHT. [148
In general, if 0, 6' be the angles which the normal to the actual wave-
front makes with the optic axes, it may be proved that the difference of the
squares of the two roots of (7) is given by
v* - v* = (a2 - c2) sin 6 sin & (24)
In a uniaxal crystal the optic axes coincide with the axis of symmetry, and
there is no distinction between & and 6.
Since waves in a biaxal crystal propagated along either optic axis have but
one velocity, it follows that tangent planes to the wave-surface, perpendicular
to these directions, touch both sheets of the surface. It may be proved
further that each plane touches the surface not merely at two, but at an
infinite number of points which lie upon a circle.
The directions of the optic axes, and the angle included between them,
are found frequently to vary with the colour of the light. Such a variation
is to be expected, in view of dispersion, which renders a2, b'2, c2 functions of the
wave-length.
A knowledge of the form of the wave-surface determines in all cases the
law of refraction according to the construction of Huygens. We will suppose
for simplicity that the first medium is air, and that the surface of separation
between the media is plane. The incident wave-front at any moment of
time cuts the surface of separation in a straight line. On this line take any
point, and with it as centre construct the wave-surface in the second medium
corresponding to a certain interval of time. At the end of this interval the
trace of the incident wave-front upon the surface will have advanced to a new
position, parallel to the former. Planes drawn through this line so as to touch
the wave-surface give the positions of the refracted wave-fronts. None other-
could satisfy the two conditions — (1) that the refracted wave-front should
move within the crystal with the normal velocity suitable to its direction,
and (2) that the traces of the incident and refracted waves upon the surface
of separation should move together. The normal to a refracted wave lies
necessarily in the plane of incidence, but the refracted ray, coinciding with
the radius vector of the wave-surface, in general deviates from it. In most
cases it is sufficient to attend to the wave-normal.
As in total reflexion by simply refracting media, it may happen that no
tangent planes can be drawn to satisfy the prescribed conditions, or that but
one such can be drawn.
When the crystal is uniaxal, one wave is refracted according to the
ordinary law of Snell. The accuracy of both the sphere and the ellipsoid
of the Huygenian construction has been fully verified by modern obser-
vations*.
* Stokes, Proc. Roy. Soc. Vol. xx. p. 443, 1872 ; Glazebrook, Phil. Trans. 1880, p. 421 ;
Hastings, Amer. Jour. Jan. 1888.
1888] CONICAL REFRACTION. 155
The simplest case of uniaxal refraction is when the axis of the crystal is
perpendicular to the plane of incidence, with respect to which every thing
then becomes symmetrical. The section of the wave-surface with which we
have to deal reduces to two concentric circles : so that both waves are refracted
according to the ordinary law, though of course with different indices.
In biaxal crystals one wave follows the ordinary law of refraction, if the
plane of incidence coincide with a principal plane of the crystal This
consequence of his theory was verified by Fresnel himself, and subsequently
by Rudberg and others. But the most remarkable phenomena of biaxal
refraction are undoubtedly those discovered by Hamilton and Lloyd, generally
known as conical refraction.
In general there are two refracted rays, corresponding to two distinct
waves. But the refracted waves coalesce when they are perpendicular to
either optic axis, and (as we have seen) this wave touches the wave-surface
along a circle. Thus corresponding to one wave direction there are an
infinite number of rays, lying upon a cone. The division of a single incident
ray into a cone of refracted rays is called internal conical refraction. If the
second face of the crystal is parallel to the first, each refracted ray resumes
on emergence its original direction, so that the emergent bundle forms a
hollow cylinder.
External conical refraction depends upon the singular points in the
principal plane of zx, where the two sheets of the surface cross one another
(Fig. 24). At such a point (P) an infinite number of tangent planes may
be drawn to the surface, and each of the perpendiculars from 0 represents a
wave direction, corresponding to the single ray OP. On emergence these
waves will be differently refracted: and thus corresponding to a single
internal ray there are an infinite number of external rays, lying upon a cone.
It has already been admitted that the dynamical foundations of Fresnel's
theory are unsound: and it must be added that the rigorous theory of
crystalline solids investigated by Cauchy and Green does not readily lend
itself to the explanation of Fresnel's laws of double refraction. On this
subject the reader should consult Pro£ Stokes's Report. Sir W. Thomson
has recently shown* that an originally isotropic medium, pressed unequally
in different directions, may be so constituted as to vibrate in accordance with
Fresnel's laws.
It may perhaps be worth while to remark that the equations, analogous
to (2) § 24, which lead to these laws are
<25>
h "On Caochy's and Green's Doctrine of Extraneous Force to explain dynamically Fresnel'a
Kinematics of Doubk Refraction," PhiL Xag. Feb. 1888.
156 WAVE THEORY OF LIGHT. [148
where a, b, c are the principal wave- velocities. If we here assume
e/09 = pfPt
and substitute in (25), the condition of transversality leads at once to the
desired results. But the equations (25) are not applicable to the vibrations
of a crystalline solid.
In the electromagnetic theory double refraction is attributed to aeolotropic
inductive capacity, and appears to offer no particular difficulty.
If the present position of the theory of double refraction is still somewhat
unsatisfactory, it must be remembered that the uncertainty does not affect
the general principle. Almost any form of wave-theory involving transverse
vibrations will explain the leading phenomenon, viz., the bifurcation of the
ray. It is safe to predict that when ordinary refraction is well understood
there will be little further trouble over double refraction.
The wave-velocity is not the only property of light rendered unsyni-
metrical by crystalline structure. In many cases the two polarized rays
are subject to a different rate of absorption. Tourmalines and other crystals
may be prepared in plates of such thickness that one ray is sensibly stopped
and the other sensibly transmitted, and will then serve as polarizing (or
analysing) apparatus. Although for practical purposes Nicol's prisms (Light,
Enc. Brit. Vol. xiv. p. 612) are usually to be preferred, the phenomenon of
double absorption is of great theoretical interest. The explanation is doubtless
closely connected with that of double refraction.
§ 22. Colours of Crystalline Plates.
When polarized light is transmitted through a moderately thin plate of
doubly refracting crystal, and is then analysed, e.g., with a Nicol, brilliant
colours are often exhibited, analogous in their character to the tints of
Newton's scale. With his usual acuteness, Young at once attributed these
colours to interference between the ordinary and extraordinary waves, and
showed that the thickness of crystal required to develop a given tint, inversely
proportional to the doubly refracting power, was in agreement with this view.
But the complete explanation, demanding a fuller knowledge of the laws of
interference of polarized light, was reserved for Fresnel and Arago. The
subject is one which admits of great development*; but the interest turns
principally upon the beauty of the effects, and upon the facility with which
many of them may be obtained in experiment. We must limit ourselves to a
brief treatment of one or two of the simpler cases.
* See Verdet's Lemons, Vol. n.
1888] COLOURS OF SELEXITE. 157
The incident vibration being plane-polarized, we will suppose that its
plane makes an angle a with the principal plane of the crystal. On entering
the crystal it is accordingly resolved into the two components represented by-
cos a cos tf>, sin a cos <£, where <f> = Zirt/r.
In traversing the crystal both waves are retarded, but we are concerned
only with the difference of the retardations. Denoting the difference by p, we
may take as the expressions of the waves on emergence
cos a cos <f>. sin a cos(^ — p).
It may be remarked that, in the absence of dispersion, p would be inversely
proportional to X ; but in feet there are many cases where it deviates greatly
from this law.
Now let the plane of analysation be inclined at the angle /8 to that of
primitive polarization (Fig. 25). Then for the sum of the two resolved
components we have
cos a cos (a — /8) cos <f> + sin a sin (a - /8) cos (<f> — p),
of which the intensity is
{cos a cos (a — /8) + sin a sin (a — £) cos pf + sin* 2 sin1 (a — £)sin*/8
= cos*/8-sin22sin2(a-/3)sins£p. ...(1)
If in (1) we write ft -f %ir in place of ft, we get
sins)3 + sin2asin2(a-y8)sin1^; (2)
and we notice that the sum of (1) and (2) is unity under all circumstances.
The effect of rotating the analyser through 90' is thus always to transform
the tint into its complementary. The two complemen-
tary tints may be seen at the same time if we employ
a double-image prism. In the absence of an analyser
we may regard the two images as superposed, and there
is no colour.
These expressions may be applied at once to the
explanation of the colours of thin plates of mica or
selenite. In this case the retardation p is proportional
to the thickness, and approximately independent of the
precise direction of the light, supposed to be nearly perpendicular to the plate,
viz., nearly parallel to a principal axis of the crystal.
The most important cases are when ft = 0, ft = far. In the latter the field
would be dark were the plate removed : and the actual intensity is
sin*2a sin1^ ..(3)
The composition of the light is thus independent of the azimuth of the
plate (a); but the intensity varies greatly, vanishing four times during the
158 WAVE THEORY OF LIGHT. [148
complete revolution. The greatest brightness occurs when the principal
plane bisects the angle between the planes of polarization and analysis.
If /S = 0, the light is complementary to that represented by (3).
If two plates be superposed, the retardations are 'added if the azimuths
correspond ; but they are subtracted if one plate be rotated relatively to the
other through 90°. It is thus possible to obtain colour by the superposition
of two nearly similar plates, although they may be too thick to answer the
purpose separately.
If dispersion be neglected, the law of the colours in (3) is the same as
that of the reflected tints of Newton's scale. The thicknesses of the plates of
mica (acting by double refraction) and of air required to give the same colour
are as 400 : 1. When a plate is too thick to show colour, its action may be
analysed with the aid of a spectroscope.
Still thicker plates may be caused to exhibit colour, if the direction of the
light within them makes but a small angle with an optic axis. Let us suppose
that a plate of Iceland spar, or other uniaxal crystal (except quartz), cut
perpendicularly to the axis, is interposed between the polarizing and analysing
apparatus, and that the latter is so turned that the field is originally dark.
The ray which passes perpendicularly is not doubly refracted, so that the
centre of the field remains dark. At small angles to the optic axis the
relative retardation is evidently proportional to the square of the inclination,
so that the colours are disposed in concentric rings. But the intensity is not
the same at the various parts of the circumference. In the plane of polari-
zation and in the perpendicular plane there is no double refraction, or rather
one of the refracted rays vanishes. Along the corresponding lines in the field
of view there is no revival of light, and the ring system is seen to be traversed
by a black cross.
In many crystals the influence of dispersion is sufficient to sensibly
modify the proportionality of p to X. In one variety of uniaxal apophyllite
Herschel found the rings nearly achromatic, indicating that p was almost
independent of \. Under these circumstances a much larger number of rings
than usual became visible.
In biaxal crystals, cut so that the surfaces are equally inclined to the
optic axes, the rings take the form of lemniscates.
A medium originally isotropic may acquire the doubly refracting property
under the influence of strain ; and, if the strain be homogeneous, the conditions
are optically identical with those found in a natural crystal. The principal
axes of the wave-surface coincide with those of strain. If the strain be sym-
metrical, the medium is optically uniaxal. In general, if P, Q, R be the
principal stresses, the difference of velocities for waves propagated parallel
to JK is evidently proportional to (P — Q), and so on.
1888] DOUBLE REFRACTION DUE TO STRAIN. 159
More often it happens that the strain is not homogeneous. Even then
the small parts may be compared to crystals, but the optical constants vary
from point to point. The comparatively feeble doubly refracting power thus
developed in glass may best be made evident by the production of the colours
of polarized light. Thus, in an experiment due to Brewster, a somewhat
stout slab of glass, polished on the edges, is interposed between crossed
Xicols. When the slab is bent in a plane perpendicular to that of vision, a
revival of light takes place along the edges, where the elongation and
contraction is greatest. If the width (in the direction of vision) be sufficient,
the effect may be increased until the various colours of Xewton's scale are
seen. These colours vary from point to point of the thickness in the plane of
bending, the " neutral axis " remaining dark. The optic axis, being every-
where coincident with the direction of elongation (or contraction), is parallel
to the length of the slab. To this direction the plane of polarization should
be inclined at about 45°.
The condition of internal strain is not necessarily due to forces applied
from without. Thus, if glass originally free from strain be unequally heated.
the accompanying expansions give rise to internal strains which manifest
themselves in polarized light. If the heating be moderate, so as not to
approach the softening point, the state of ease is recovered upon cooling, and
the double refraction disappears. But if the local temperature be raised
further, the hot parts may relieve themselves of the temporary strain, and
then upon cooling they and other parts may be left in a condition of
permanent strain. Sudden cooling of glass heated to the softening point
leads to a similar result. The outer parts harden while the interior is still at
a higher temperature, so that, when the whole is cooled down, the outside,
being as it were too large for the inside, is in a condition of radial tension
and circumferential compression. An examination in polarized light shows
that the strains thus occasioned are often very severe. If any small part be
relieved by fracture from the constraint exercised upon it by the remainder,
the doubly refracting property almost or wholly disappears. In this respect
unannealed glass differs essentially from a crystal, all parts of which are
similar and independent. It may be remarked that it is difficult to find large
pieces of glass so free from internal strain as to show no revival of light when
examined between crossed Nicols.
§ 23. Rotatory Polarization.
In general a polarized ray travelling along the axis of a uniaxal crystal
undergoes no change : but it was observed by Arago that, if quartz be used
in this experiment, the plane of polarization is found to be rotated through
an angle proportional to the thickness of crystal traversed. The subject was
further studied by Biot, who ascertained that the rotation due to a given
160 WAVE THEORY OF LIGHT. [148
thickness is inversely as the square of the wave-length of the light, thus
varying very rapidly with the colour. In some specimens of quartz (called
in consequence right-handed) the rotation is to the right, while in others it
is to the left. Equal thicknesses of right- and left-handed quartz may thus
compensate one another.
Fresnel has shown that the rotation of the plane may be interpreted as
indicating a different velocity of propagation of the two circularly-polarized
components into which plane-polarized light may always be resolved. In
ordinary media the right- and left-handed circularly-polarized rays travel at
the same speed, and at any stage of their progress recompound a ray
rectilinearly-polarized in a fixed direction. But it is otherwise if the
velocities of propagation of the circular components be even slightly different.
The first circularly-polarized wave may be expressed by
^ = rcos(nt — hz), ^ = r sin (nt — k^z) ; (1)
and the second (of equal amplitude) by
£2 = r cos (nt — k^z), ?;2 = - r sin (nt - kzz) (2)
The resultant of (1) and (2) is
£=!j + |-2= 2r cos ^ (&2 - &i) z . cos {nt — 1(^ + ka)z],
f) = iji+ % = 2r sin ^ (k2 — k^z. cos {nt — ^(k1 + k2) z} ;
so that
V£=tani(fca-&i)*, (3)
which shows that for any fixed value of z the light is plane-polarized. The
direction of this plane, however, varies with z. Thus, if ?;/£ = tan 6, so that 0
gives the angular position of the plane in reference to £ we have
0=i(fc-*i)*, (4)
indicating a rotation proportional to z. The quantities kl} k, are inversely as
the wave-lengths of the two circular components for the same periodic time.
When the relative retardation amounts to an entire period, (k, — k1)z = 2ir,
and then, by (4), 0 = TT. The revolution of the plane through two right
angles restores the original state of polarization. In quartz the rotation is
very rapid, amounting in the case of yellow light to about 24<° for each
millimetre traversed.
It is interesting to observe with what a high degree of accuracy the
comparison of the velocities of the two waves can be effected. If the plane
of polarization be determined to one minute of angle, a relative retardation of
X/10800 is made manifest. If I be the thickness traversed, v and v + Sv the
two velocities, the relative retardation is l&v/v. To take an example, suppose
that / = 20 inches, X = ^fav inch ; so that if Sv/v exceed 10~8, the fact might
be detected, [inch = 2'54 cm.]
1888] MAGNETIC ROTATIOX. 161
In quartz the rotation of the plane depends upon the crystalline structure,
but there are many liquids, e.g., oil of turpentine and common syrup, which
exhibit a like effect. In such cases the rotation is of course independent of
the direction of the light; it must be due to some peculiarity in the
constitution of the molecules.
A remarkable connexion has been observed between the rotatory property
and the crystalline form. Thus Herschel found that in many specimens the
right-handed and left-handed varieties of quartz could be distinguished by
the disposition of certain subordinate faces. The crystals of opposite kinds
are symmetrical in a certain sense, but are yet not siiperposable. The
difference is like that between otherwise similar right- and left-handed
screws. The researches of Pasteur upon the rotatory properties of tartaric
acid have opened up a new and most interesting field of chemistry. At that
time two isomeric varieties were known, — ordinary tartaric acid, which
rotates to the right, and racemic acid, which is optically inactive, properties
of the acids shared also by the salts. Pasteur found that the crystals of
tartaric acid and of the tartrates possessed a right-handed structure, and
endeavoured to discover corresponding bodies with a left-handed structure.
After many trials crystallizations of the double racemate of soda and ammonia
were obtained, including crystals of opposite kinds. A selection of the
right-handed specimens yielded ordinary dextro-tartaric acid, while a similar
selection of the left-handed crystals gave a new variety — laevo-tartaric acid,
rotating the plane of polarization to the left in the same degree as ordinary
tartaric acid rotates it to the right. A mixture in equal proportions of the
two kinds of tartaric acid, which differ scarcely at all in their chemical
properties*, reconstitutes racemic acid.
The possibility of inducing the rotatory property in bodies otherwise free
from it was one of the finest of Faraday's discoveries. He found that, if
heavy glass, bisulphide of carbon, &c., are placed in a magnetic field, a ray of
polarized light, propagated along the lines of magnetic force, suffers rotation.
The laws of the phenomenon were carefully studied by Verdet, whose
conclusions may be summed up by saying that in a given medium the
rotation of the plane for a ray proceeding in any direction is proportional to
the difference of magnetic potential at the initial and final points. In
bisulphide of carbon, at 18° and for a difference of potential equal to unity
C.G.S., the rotation of the plane of polarization of a ray of soda light is
•04202 minute of angle "f*.
A very important distinction should be noted between the magnetic
rotation and that natural to quartz, syrup, &c. In the latter the rotation is
* It would seem that the two varieties could be chemically distinguished only by their
relations with bodies themselves right-handed or left-handed,
t Phil. Tram. 1885, p. 343. [VoL n. p. 377.]
L III. 11
162 WAVE THEORY OF LIGHT. [148
always right-handed or always left-handed with respect to the direction of
the ray. Hence when the ray is reversed the absolute direction of rotation
is reversed also. A ray which traverses a plate of quartz in one direction,
and then after reflexion traverses the same thickness again in the opposite
direction, recovers its original plane of polarization. It is quite otherwise
with the rotation under magnetic force. In this case the rotation is in the
same absolute direction even though the ray be reversed. Hence, if a ray be
reflected backwards and forwards any number of times along a line of
magnetic force, the rotations due to the several passages are all accumulated.
The non-reversibility of light in a magnetized medium proves the case to be
of a very exceptional character, and (as was argued by Thomson) indicates
that the magnetized medium is itself in rotatory motion independently of the
propagation of light through it*.
The importance of polarimetric determinations has led to the contrivance
of various forms of apparatus adapted to the special requirements of the case.
If the light be bright enough, fairly accurate measurements may be made by
merely rotating a Nicol until the field appears dark. Probably the best form
of analyser, when white light is used and the plane is the same for all the
coloured components, is the Jelletf, formed by the combination of two
portions of Iceland spar. By this instrument the field of view is duplicated,
and the setting is effected by turning it until the two portions of the field,
much reduced in brightness, appear equally dark. A similar result is attained
in the Laurent, which, however, is only applicable to homogeneous light.
In this apparatus, advantage is taken of the action of a half-wave plate. In
passing such a plate the plane of polarization is as it were reflected by the
principal section, that is, rotated until it makes the same angle with the
principal section as at first, but upon the further side. The plate covers
only half of the field of view, and the eye is focused upon the dividing edge.
The planes of polarization of the two halves of the field are different, unless
the original plane be parallel (or perpendicular) to the principal section. In
the Laurent analyser the half-wave plate is rigidly combined with a Nicol in
such a position that the principal section of the latter makes a small but
finite angle with that of the plate. The consequence is that the two halves
of the field of view cannot be blackened simultaneously, but are rendered
equally dark when the instrument is so turned that the principal section of
the plate is parallel to the plane of original polarization, which is also that
of the uncovered half of the field. A slight rotation in either direction
darkens one half of the field and brightens the other half.
In another form of " half-shade " polarimeter, invented by Poynting, the
half-wave plate of the Laurent is dispensed with, a small rotation of one half
* Maxwell's Electricity and Magnetism, Vol. n. chap. xxi.
t A description is given in Glazebrook's Physical Optics, London 1883.
1888] POLARLMETRY. 163
of the field with respect to the other half being obtained by quartz (cut
perpendicularly to the axis) or by syrup. In the simplest construction the
syrup is contained in a small cell with parallel glass sides, and the division
into two parts is effected by the insertion of a small piece of plate glass
about ^ inch thick, a straight edge of which forms the dividing line. If the
syrup be strong, the difference of thickness of ^ inch gives a relative
rotation of about 2°. In this arrangement the sugar cell is a fixture, and
only the Xicol rotates. The reading of the divided circle corresponds to the
mean of the planes for the two halves of the field, and this of course differs
from the original position of the plane before entering the sugar. This
circumstance is usually of no importance, the object being to determine the
rotation of the plane of polarization when some of the conditions are altered.
A discussion of the accuracy obtainable in polarirnetry will be found in
a recent paper by Lippich*.
In Soleil's apparatus, designed for practical use in the estimation of the
strength of sugar solutions, the rotation due to the sugar is compensated bv
a wedge of quartz. Two wedges, one of right-handed and the other of left-
handed quartz, may be fitted together, so that a movement of the combination
in either direction increases the thickness of one variety traversed and
diminishes that of the other. The linear movement required to compensate
the introduction of a tube of syrup measures the quantity of sugar present.
§ 24. Dynamical Theory of Diffraction.
The explanation of diffraction phenomena given by Fresnel and his
followers is independent of special views as to the nature of the ether, at
least in its main features ; but in the absence of a more complete foundation
it is impossible to treat rigorously the mode of action of a solid obstacle such
as a screen. The full solution of problems of this kind is scarcely to be
expected. Even in the much simpler case of sound, where we know what we
have to deal with, the mathematical difficulties are formidable ; and we are not
able to solve even such an apparently elementary question as the transmission
of sound past a rigid infinitely thin plane screen, bounded by a straight edget,
or perforated with a circular aperture. But, without entering upon matters
of this kind, we may inquire in what manner a primary wave may be resolved
into elementary secondary waves, and in particular as to the law of intensity
and polarization in a secondary wave as dependent upon its direction of
propagation, and upon the character as regards polarization of the primary
wave. This question is treated by Stokes in his " Dynamical Theory of
Diffraction "J on the basis of the elastic solid theory.
* Win. Ber. LXXXT. 9th Feb. 1882. See also Phil. Trant. 1885, p. 360. [Vol. n. p. 378.]
t [1901. We owe to Sommerfeld some advance in this direction.]
J Camb. Phil. Tratu. Vol. n. p. 1 ; Stokes' Collected Papers, VoL n. p. 243.
11—2
164 WAVE THEORY OF LIGHT.
Let x, y, z be the coordinates of any particle of the medium in its natural
state, and £ 17, £ the displacements of the same particle at the end of time t,
measured in the directions of the three axes respectively. Then the first of
the equations of motion may be put under the form
where a2 and 62 denote the two arbitrary constants. Put for shortness
d£ dij d£ $ /i \
-rr^ + -T= — h -y- = o, W
dx dy dz
and represent by V2| the quantity multiplied by 62. According to this
notation, the three equations of motion are
..-(2)
It is to be observed that 8 denotes the dilatation of volume of the element
situated at (x, y, z). In the limiting case in which the medium is regarded
as absolutely incompressible 8 vanishes ; but, in order that equations (2) may
preserve their generality, we must suppose a at the same time to become
infinite, and replace a2 8 by a new function of the coordinates.
These equations simplify very much in their application to plane waves.
If the ray be parallel to OX, and the direction of vibration parallel to OZ, we
have £ = 0, 77 = 0, while £ is a function of x and t only. Equation (1) and the
first pair of equations (2) are thus satisfied identically. The third equation
3-*$ <3>
of which the solution is
?=/(6«-a), (4)
where /is an arbitrary function.
The question as to the law of the secondary waves is thus answered by
Stokes. " Let £ = 0, 77 = 0, f =f (bt — x) be the displacements corresponding
to the incident light ; let Ox be any point in the plane P (of the wave-front),
dS an element of that plane adjacent to Ol ; and consider the disturbance due
to that portion only of the incident disturbance which passes continually
across dS. Let 0 be any point in the medium situated at a distance from
the point Ol which is large in comparison with the length of a wave ; let
OlO = r, and let this line make an angle 6 with the direction of propagation
1888] STOKES* LAW OF SECONDARY WAVE. 165
of the incident light, or the axis of at, and <£ with the direction of vibration,
or axis of z. Then the displacement at 0 will take place in a direction
perpendicular to 0,0, and lying in the plane ZO^O', and, if f be the
displacement at 0. reckoned positive in the direction nearest to that in which
the incident vibrations are reckoned positive,
In particular, if
/(fe-*) = csinY<fc-*). ..................... (5)
we shall have
s^)^n^cos(6l-r)L ............ (6)"
It is then verified that, after integration with respect to dS, (6) gives the
same disturbance as if the primarv wave had been supposed to pass on
unbroken.
The occurrence of sin <f> as a factor in (6) shows that the relative
intensities of the primarv light and of that diffracted in the direct ion $
depend upon the condition of the former as regards polarization. If the
direction of primary vibration be perpendicular to the plane of diffraction
(containing both primary and secondary rays), sin if> = l : but, if the primary
vibration be in the plane of diffraction, sin $ = cos 0. This result was
employed by Stokes as a criterion of the direction of vibration: and his
experiments, conducted with gratings, led him to the conclusion that the
vibrations of polarized tight are executed in a direction perpendicular to the
plane of polarization.
The factor (1 +cos0) shows in what manner the secondary disturbance
depends upon the direction in which it is propagated with respect to the
front of the primary wave.
1C as suffices for all practical purposes, we limit the application of the
formulae to points in advance of the plane at which the wave is supposed to
be broken up, we may use simpler methods of resolution than that above
considered. It appears indeed that the purely mathematical question has no
definite answer. In illustration of this the analogous problem for sound may
be referred to. Imagine a flexible lamina to be introduced so as to coincide
with the plane at which resolution is to be effected. The introduction of the
lamina (supposed to be devoid of inertia) will make no difference to the
propagation of plane parallel sonorous waves through the position which it
occupies. At every point the motion of the lamina will be the same as
would have occurred in its absence, the pressure of the waves impinging from
behind being just what is required to generate the waves in front. Now it is
166 WAVE THEORY OF LIGHT. [148
evident that the aerial motion in front of the lamina is determined by what
happens at the lamina without regard to the cause of the motion there
existing. Whether the necessary forces are due to aerial pressures acting
on the rear, or to forces directly impressed from without, is a matter
of indifference. The conception of the lamina leads immediately to two
schemes, according to which a primary wave may be supposed to be broken
up. In the first of these the element dS, the effect of which is to be
estimated, is supposed to execute its actual motion, while every other
element of the plane lamina is maintained at rest. The resulting aerial
motion in front is readily calculated*; it is symmetrical with respect to the
origin, i.e., independent of 6. When the secondary disturbance thus obtained
is integrated with respect to dS over the entire plane of the lamina, the
result is necessarily the same as would have been obtained had the primary
wave been supposed to pass on without resolution, for this is precisely the
motion generated when every element of the lamina vibrates with a common
motion, equal to that attributed to dS. The only assiimption here involved
is the evidently legitimate one that, when two systems of variously distri-
buted motion at the lamina are superposed, the corresponding motions in
front are superposed also.
The method of resolution just described is the simplest, but it is only one
of an indefinite number that might be proposed, and which are all equally
legitimate, so long as the question is regarded as a merely mathematical one,
without reference to the physical properties of actual screens. If, instead of
supposing the motion at dS to be that of the primary wave, and to be zero
elsewhere, we suppose the force operative over the element dS of the lamina
to be that corresponding to the primary wave, and to vanish elsewhere, we
obtain a secondary wave following quite a different lawf. In this case the
motion in different directions varies as cos 6, vanishing at right angles to
the direction of propagation of the primary wave. Here again, on integration
over the entire lamina, the aggregate effect of the secondary waves is
necessarily the same as that of the primary.
In order to apply these ideas to the investigation of the secondary wave
of light, we require the solution of a problem, first treated by Stokes J, viz.,
the determination of the motion in an infinitely extended elastic solid due
to a locally applied periodic force. If we suppose that the force impressed
upon the element of mass Ddxdydz is
DZdxdydz,
being everywhere parallel to the axis of Z, the only change required in our
equations (1), (2) is the addition of the term Z to the second member of the
third equation (2). In the forced vibration, now under consideration, Z, and
* Theory of Sound, § 278. f Loc. cit. equation (10).
J Loc. cit. §§ 27—30.
1888] FORCE LOCALLY APPLIED. 167
the quantities £ 17, £ B expressing the resulting motion, are to be supposed
proportional to eimt, where t = V( — 1), and n = 2w/T, T being the periodic
time. Under these circumstances the double differentiation with respect to t
of any quantity is equivalent to multiplication by the factor — n*. and thus
our equations take the form
It will now be convenient to introduce the quantities w,, vrt, vt, which
express the rotations of the elements of the medium round axes parallel to
those of coordinates, in accordance with the equations
v =d% _ *7 cr_^_^ _d£_d% g
dy dx' ^l ~~ dz dy' m* da dz"
In terms of these we obtain from (7), by differentiation and subtraction.
= dZ dy \ (9)
The first of equations (9) gives
w, = 0. (10)
For v1 we have
where r is the distance between the element djrdydz and the point where vrl
is estimated, and
k=n b = '2ir\, .............................. (12)
X being the wave-length.
We will now introduce the supposition that the force Z acts only within
a small space of volume T, situated at (x, y, z), and for simplicity suppose
that it is- at the origin of coordinates that the rotations are to be estimated.
Integrating by parts in (11), we get
'
in which the integrated terms at the limits vanish, Z being finite only within
the region T. Thus
This sedation may be verified in the same manner as Poisson's theorem, in which i=0.
168 WAVE THEORY OF LIGHT. [148
Since the dimensions of T are supposed to be very small in comparison with \,
the factor— (- _ \ is sensibly constant; so that, if Z stand for the mean
dy\ r J
value of Z over the volume T, we may write
TZ y d fe-*r\
•&-.= -- — . -. -y- I I
47T&2 r dr\ r J
In like manner we find
TZ x
From (10), (13), (14) we see that, as might have been expected, the rotation
at any point is about an axis perpendicular both to the direction of the force
and to the line joining the point to the source of disturbance. If the
resultant rotation be -cr, we have
TZ V(#2 + ?/) d_ fe-ikr\ _ TZ sin <f> d^ /e~ik
~~~~
_
~
<£ denoting the angle between r and z. In differentiating e~ikrjr with respect
to r, we may neglect the term divided by r2 as altogether insensible, kr being
an exceedingly great quantity at any moderate distance from the origin of
disturbance. Thus
ik.TZsind) e~ikr
4^ '—' ..................... (L
which completely determines the rotation at any point. For a disturbing-
force of given integral magnitude it is seen to be everywhere about an axis
perpendicular to r and to the direction of the force, and in magnitude
dependent only upon the angle (</>) between these two directions and upon
the distance (r).
The intensity of light is, however, more usually expressed in terms of the
actual displacement in the plane of the wave. This displacement, which we
may denote by £', is in the plane containing z and r, and perpendicular to
the latter. Its connexion with w is expressed by -or = d^'/dr ; so that
where the factor emt is restored.
Retaining only the real part of (16), we find, as the result of a local
application of force equal to
DTZcosnt, .............................. (17)
the disturbance expressed by
, = cos(nt-kr)
47T&2 ' r '
1888] LAW OF SECONDARY WAVE. 169
The occurrence of sin <j> shows that there is no disturbance radiated in
the direction of the force, a feature which might have been anticipated from
considerations of symmetry.
We will now apply (18) to the investigation of a law of secondary
disturbance, when a primary wave
f = sin(irf-fcr) ........................... (19)
is supposed to be broken up in passing the plane x = 0. The first step is to
calculate the force which represents the reaction between the parts of the
medium separated by x = 0. The force operative upon the positive half is
parallel to OZ, and of amount per unit of area equal to
PkD cos nt ;
and to this force acting over the whole of the plane the actual motion on the
positive side may be conceived to be due. The secondary disturbance
corresponding to the element dS of the plane may be supposed to be that
caused by a force of the above magnitude acting over dS and vanishing
elsewhere; and it only remains to examine what the result of such a force
would be.
Now it is evident that the force in question, supposed to act upon the
positive half only of the medium, produces just double of the eft'ect that
would be caused by the same force if the medium were undivided, and on
the latter supposition (being also localized at a point) it comes under the
head already considered. According to (18), the effect of the force acting at
dS parallel to OZ, and of amount equal to
Z&kDdScosnt,
will be a disturbance
regard being had to (12). This therefore expresses the secondary dis-
turbance at a distance r and in a direction making an angle <f> with OZ (the
direction of primary vibration) due to the element dS of the wave-front.
The proportionality of the secondary disturbance to sin <£ is common to
the present law and to that given by Stokes, but here there is no dependence
upon the angle 0 between the primary and secondary rays. The occurrence
of the factor (Xr)"1, and the necessity of supposing the phase of the
secondary wave accelerated by a quarter of an undulation, were first
established by Archibald Smith, as the result of a comparison between the
primary wave, supposed to pass on without resolution, and the integrated
effect of all the secondary waves (§ 10). The occurrence of factors such as
sin ^, or | (1 + cos 0), in the expression of the secondary wave has no
influence upon the result of the integration, the effects of all the elements
170 WAVE THEORY OF LIGHT. [148
for which the factors differ appreciably from unity being destroyed by mutual
interference.
The choice between various methods of resolution, all mathematically
admissible, would be guided by physical considerations respecting the mode
of action of obstacles. Thus, to refer again to the acoustical analogue in
which plane waves are incident upon a perforated rigid screen, the circum-
stances of the case are best represented by the first method of resolution,
leading to symmetrical secondary waves, in which the normal motion is
supposed to be zero over the unperforated parts. Indeed, if the aperture is
veiy small, this method gives the correct result, save as to a constant factor.
In like manner our present law (20) would apply to the kind of obstruction
that would be caused by an actual physical division of the elastic medium,
extending over the whole of the area supposed to be occupied by the
intercepting screen, but of course not extending to the parts supposed to be
perforated. In the present state of our ignorance this law seems to be at
least as plausible as any other.
§ 25. The Diffraction of Light by Small Particles.
The theory of the diffraction, dispersion, or scattering of light by small
particles, as it has variously been called, is of importance, not only from its
bearings upon fundamental optical hypotheses, but on account of its appli-
cation to explain the origin and natiire of the light from the sky. The view,
suggested by Newton and advocated in more recent times by such authorities
as Herschel* and Clausiusf, that the light of the sky is a blue of the first
order reflected from aqueous particles, was connected with the then prevalent
notion that the suspended moisture of clouds and mists was in the form of
vesicles or bubbles. Experiments such as those of BriickeJ pointed to a
different conclusion. When a weak alcoholic solution of mastic is agitated
with water, the precipitated gum scatters a blue light, obviously similar in
character to that from the sky. Not only would it be unreasonable to
attribute a vesicular structure to the mastic, but (as Briicke remarked) the
dispersed light is much richer in quality than the blue of the first order.
Another point of great importance is well brought out in the experiments of
Tyndall§ upon clouds precipitated by the chemical action of light. Whenever
the particles are sufficiently fine, the light emitted laterally is blue in colour,
and, in a direction perpendicular to the incident beam, is completely polarized.
About the colour there can be no primd facie difficulty ; for, as soon as
the question is raised, it is seen that the standard of linear dimension, with
* Article "Light," Enc. Metrop. 1830, § 1143.
t Pogg. Ann. Vols. LXXII. LXXVI. LXXXVIII. ; Crelle, Vols. xxxiv. xxxvi.
J Pogg. Ann. Vol. LXXXIII. § Phil. Mag. [4], Vol. cxxxvn. p. 388.
1888] PARTICLES DENSER THAX EXTTBOXMZST. in
reference to which the particles are called small, is the ware-length of light,
and that a given set of particles would (on any conceivable view as to then-
mode of action) produce a continually increasing disturbance aa we pass along
the spectrum towards the more refrangible end-
On the other hand, that the direction of complete polarization should be
independent of the refracting power of the matter eompjsing the cloud has
been considered mysterious. Of course, on the theory of thin plates, this
direction would be determined by Brewsters law: but, if the particles of
foreign matter are small in all their dimensions, the circumstances are
materially different from those under which Brewster's law is applicable.
The investigation of this question upon the elastic solid theory will
depend upon how we suppose the solid to vary from one optical medium to
another. The slower propagation of light in glass or water than in air or
vacuum may be attributed to a greater density, or to a less rigidity, in the
former case : or we may adopt the more complicated supposition that both
these quantities vary, subject only to the condition which restricts the ratio
of velocities to equality with the known refractive index. It will presently
appear that the original hypothesis of Fresnel. that the rigidity remains the
same in both media, is the only one that can be reconciled with the facts :
and we will therefore investigate upon this basis the nature of the secondary
waves dispersed by small particles.
Conceive a beam of plane-polarized light to move among a number of
particles, all small compared with any of the wave-lengths. According to onr
hypothesis, the foreign matter may be supposed to lead the ether, so as to
increase its inertia without altering its resistance to distortion. If the
particles were away, the wave would pass on unbroken and no light would be
emitted laterally. Even with the particles retarding the motion of the
ether, the same will be true if, to counterbalance the increased inertia.
suitable forces are caused to act on the ether at all points where the inertia
is altered. These forces have the same period and direction as the un-
disturbed luminous vibrations themselves. The light actually emitted
laterally is thus the same as would be caused by forces exactly the opposite
of these acting on the medium otherwise free from disturbance, and it only
remains to see what the effect of such force would be.
On account of the smallness of the particles, the forces acting throughout
the volume of any individual particle are all of the same intensity and
direction, and may be considered as a whole. The determination of the
motion in the ether, due to the action of a periodic force at a given point, is
a problem with which we have recently been occupied (| 24>. But, before
applying the solution to a mathematical investigation of the present question,
it may be well to consider the matter for a few moments from a more general
point of view.
172 WAVE THEORY OF LIGHT. [148
In the first place, there is necessarily a complete symmetry round the
direction of the force. The disturbance, consisting of transverse vibrations,
is propagated outwards in all directions from the centre ; and, in consequence
of the symmetry, the direction of vibration in any ray lies in the plane
containing the ray and the axis of symmetry ; that is to say, the direction of
vibration in the scattered or diffracted ray makes with the direction of
vibration in the incident or primary ray the least possible angle. The
symmetry also requires that the intensity of the scattered light should vanish
for the ray which would be propagated along the axis ; for there is nothing
to distinguish one direction transverse to the ray from another. The
application of this is obvious. Suppose, for distinctness of statement, that
the primary ray is vertical, and that the plane of vibration is that of the
meridian. The intensity of the light scattered by a small particle is constant,
and a maximum, for rays which lie in the vertical plane running east and
west, while there is no scattered ray along the north and south line. If the
primary ray is unpolarized, the light scattered north and south is entirely
due to that component which vibrates east and west, and is therefore
perfectly polarized, the direction of its vibration being also east and west.
Similarly any other ray scattered horizontally is perfectly polarized, and the
vibration is performed in the horizontal plane. In other directions the
polarization becomes less and less complete as we approach the vertical.
The observed facts as to polarization are thus readily explained, and the
general law connecting the intensity of the scattered light with the wave-
length follows almost as easily from considerations of dimensions.
The object is to compare the intensities of the incident and scattered
light, for these will clearly be proportional. The number (i) expressing the
ratio of the two amplitudes is a function of the following quantities : — (T)
the volume of the disturbing particle ; (r) the distance of the point under
consideration from it ; (X) the wave-length ; (b) the velocity of propagation of
light; (D) and (D') the original and altered densities: of which the first
three depend only upon space, the fourth on space and time, while the fifth
and sixth introduce the consideration of mass. Other elements of the
problem there are none, except mere numbers and angles, which do not
depend upon the fundamental measurements of space, time, and mass. Since
the ratio (i), whose expression we seek, is of no dimensions in mass, it follows
at once that D and D' occur only under the form D : D', which is a simple
number and may therefore be disregarded. It remains to find how i varies
with T, r, \, b.
Now, of these quantities, b is the only one depending on time ; and
therefore, as i is of no dimensions in time, b cannot occur in its expression.
Moreover, since the same amount of energy is propagated across all
spheres concentric with the particle, we recognize that i varies as r. It is
1888] LAW OF INVERSE FOURTH POWERS. 173
equally evident that i varies as T, and therefore that it must be proportional
to T/\-r, T being of three dimensions in space. In passing from one part of
the spectrum to another, X is the only quantity which varies, and we have
the important law : —
When light is scattered by particles which are very small compared with
any of the wave-lengths, the ratio of the amplitudes of the vibrations of the
scattered and incident lights varies inversely as the square of the wave-
length, and the ratio of intensities as the inverse fourth power.
The light scattered from small particles is of a much richer blue than the
blue of the first order as reflected from a very thin plate. From the general
theory (§ 8), or by the method of dimensions, it is easy to prove that in the
latter case the intensity varies as AT2, instead of X"4.
The principle of energy makes it clear that the light emitted laterally is
not a new creation, but only diverted from the main stream. If / represent
the intensity of the primary light after traversing a thickness x of the turbid
medium, we have
where h is a constant independent of X. On integration,
log (///.) = -AX-*, ........................... (1)
if 70 correspond to x = 0, — a law altogether similar to that of absorption, and
showing how the light tends to become yellow and finally red as the thickness
of the medium increases*.
Captain Abney has found that the above law agrees remarkably well with
his observations on the transmission of light through water in which particles
of mastic are suspended "f.
We may now investigate the mathematical expression for the disturbance
propagated in any direction from a small particle upon which a beam of light
strikes. Let the particle be at the origin of coordinates, and let the
expression for the primary vibration be
C=sin(n*-^) ............................... (2)
The acceleration of the element at the origin is — na sin nt ; so that the force
which would have to be applied to the parts where the density is D' (instead
of D), in order that the waves might pass on undisturbed, is per unit of
volume
- (D' - D) n? sin nt.
To obtain the total force which must be supposed to act, the factor T
(representing the volume of the particle) must be introduced. The opposite
* "On the Light from the Sky, its Polarization and Colour," PAH. Mag. Feb. 1871.
t Proc. Boy. Soc. May 1886.
174 WAVE THEORY OF LIGHT. [148
of this, conceived to act at 0, would give the same disturbance as is actually
caused by the presence of the particle. Thus by (18) (§ 24) the secondary
disturbance is expressed by
, _ D' - D n*T sin <ft sin (nt- kr)
~~D 4wF"~ "r~~
D' - D -rrT sin
The preceding investigation is based upon the assumption that in passing
from one medium to another the rigidity of the ether does not change. If
we forego this assumption, the question is necessarily more complicated ; but,
on the supposition that the changes of rigidity (Atf) and of density (AD)
are relatively small, the results are fairly simple. If the primary wave be
represented by
£=6-***, .................................... (4)
the component rotations in the secondary wave are
where
The expression for the resultant rotation in the general case would be rather
complicated, and is not needed for our purpose. It is easily seen to be about
an axis perpendicular to the scattered ray (x, y, z), inasmuch as
Let us consider the more special case of a ray scattered normally to the
incident ray, so that a- = 0. We have
If AJV, AD be both finite, we learn from (7) that there is no direction
perpendicular to the primary (polarized) ray in which the secondary light
vanishes. Now experiment tells us plainly that there is such a direction, and
therefore we are driven to the conclusion that either A^ or AD must vanish.
In strictness the force must be supposed to act upon the medium in its actual condition,
whereas in (18) the medium is supposed to be absolutely uniform. It is not difficult to prove
that (3) remains unaltered, when this circumstance is taken into account ; and it is evident in
any case that a correction would depend upon the square of (D' - D).
1888] DIFFRACTION BY A SMALL CYLINDER. 175
The consequences of supposing AiY to be zero have already been traced.
They agree very well with experiment, and require us to suppose that the
vibrations are perpendicular to the plane of polarization. So for as (7) is
concerned, the alternative supposition that AD vanishes would answer equally
well, if we suppose the vibrations to be executed in the plane of polarization :
but let us now revert to (5X which gives
P&Nxy
According to these equations there would be, in all, six directions from 0
along which there is no scattered light, — two along
the axis of y normal to the original ray, and four **• 26-
(y = 0, z = ± x) at angles of 45" with that ray. So :
long as the particles are small no such vanishing of
light in oblique directions is observed, and we are
thus led to the conclusion that the hypothesis of a
finite AA* and of vibrations in the plane of polari-
zation cannot be reconciled with the facts. Xo
form of the elastic solid theory is admissible except A
that in which the vibrations are supposed to be
perpendicular to the plane of polarization, and the
difference between one medium and another to be a difference of density
only*.
Before leaving this subject it may be instructive to show the application
of a method, similar to that used for small particles, to the case of an
obstructing cylinder? whose axis is parallel to the fronts of the primary waves.
We will suppose (1) that the variation of optical properties depends upon a
difference of density (I? — D). and is small in amount : (2) that the diameter
of the cylinder is very small in comparison with the wave-length «»f light.
Let the axis of the cylinder be the axis of z (Fig. 26). and (as before) let
the incident light be parallel to x. The original vibration is thus, in the
principal cases, parallel to either z or y. We will take first the former case,
where the disturbance due to the cylinder must evidently be symmetrical
round OZ and parallel to it. The element of the disturbance at A . due to
PQ (dz), will be proportional to dz in amplitude, and will be retarded in phase
by an amount corresponding to the distance r. In calculating the effect of
the whole bar we have to consider the integral
t/r sin (nt — kr)
* See a paper, "On the Scattering of Light bj Small Particles," P*»7. Mmg. June 1871.
(ToL i. p. 104-]
176 WAVE THEORY OF LIGHT. [148
The integral on the left may be treated as in § 15, and we find
r~l sin (tit - kr) dz = ^(\/R) sin (nt -kR~l TT),
showing that the total effect is retarded £X behind that due to the central
element at 0. We have seen (3) that, if a be the sectional area, the effect of
the element PQ is
D' - D 7T<r
where <f> is the angle OP A. In strictness this should be reckoned perpendicular
to PA, and therefore, considered as a contribution to the resultant at A,
should be multiplied by sin <f>. But the factor sin2 <f>, being sensibly equal to
unity for the only parts which are really operative, may be omitted without
influencing the result. In this way we find, for the disturbance at A,
(9)
corresponding to the incident wave sin (nt — kx).
When the original vibration is parallel to y, the disturbance due to the
cylinder will no longer be symmetrical about OZ. If a be the angle between
OX and the scattered ray, which is of course always perpendicular to OZ, it is
only necessary to introduce the factor cos a in order to make the previous
expression (9) applicable.
The investigation shows that the light diffracted by an ideal wire-grating
would, according to the principles of Fresnel, follow the law of polarization
enunciated by Stokes. On the other hand, this law would be departed from,
were we to suppose that there is any difference of rigidity between the
cylinder and the surrounding medium.
§ 26. Reflexion and Refraction.
So far as the directions of the rays are concerned, the laws of reflexion and
refraction were satisfactorily explained by Huygens on the principles of the
wave-theory. The question of the amount of light reflected, as dependent
upon the characters of the media and upon the angle of incidence, is a much
more difficult one, and cannot be dealt with d priori without special hypotheses
as to the nature of the luminous vibrations, and as to the cause of the difference
between various media. By a train of reasoning, not strictly dynamical, but
of great ingenuity, Fresnel was led to certain formulae, since known by his
name, expressing the ratio of the reflected to the incident vibration in terms
of one constant (/i). If 6 be the angle of incidence and 6l the angle of
refraction, Fresnel's expression for light polarized in the plane of incidence is
sin (0-0,)
sin (0 + 0,)' ................................. *!'
1888] FRESNEL'S FORMULA FOR REFLEXION. 177
where the relation between the angles 6, ft, and //, (the relative refractive
index) is, as usual,
sin 0 = /& sin ft (2)
In like manner, for light polarized perpendicularly to the plane of incidence,
Fresnel found
tan (0- ft)
tan(0 + ft)'
In the particular case of perpendicular incidence, both formulae coincide with
one previously given by Young, viz.,
0*-1)/0* + 1) (4)
Since these formula? agree fairly well with observation, and are at any rate
the simplest that can at all represent the facts, it may be advisable to consider
their significance a little in detail. As 6 increases from 0 to £ TT, the sine-
formula increases from Young's value to unity. We may see this most easily
with the aid of a slight transformation : —
sin (6 - ft) _ 1 - tan ft/tan 0 _ fi - cos 0/cos ft
sin (0 + ft) ~~ 1 + tan ft/ tan 0 ~~ /* -f cos 0/cos ft '
Now, writing cos 6 /cos ft in the form
/( l-sin*0 }
V (1 - /*~8 sin2 0} '
we recognize that, as 6 increases from 0 to JTT, cos 0/ cos ft diminishes
continuously from 1 to 0, and therefore (1) increases from (/*- !)/(/* + !)
to unity.
It is quite otherwise with the tangent-formula. Commencing at Young's
value, it diminishes, as 6 increases, until it attains zero, when 0 + 01 = ^7r, or
sin 0j = cos 0 ; or by (2) tan 0 = ft. This is the polarizing angle denned by
Brewster. It presents itself here as the angle of incidence for which there
is no reflexion of the polarized light under consideration. As the angle of
incidence passes through the polarizing angle, the reflected vibration changes
sign, and increases in numerical value until it attains unity at a grazing
incidence (0 = ^TT).
We have hitherto supposed that the second medium (into which the light
enters at the refracting surface) is the denser. In the contrary case, total
reflexion sets in as soon as sin 0 = /i~1, at which point 01 becomes imaginary.
We shall be able to follow this better in connexion with a mechanical theory.
If light falls upon the first surface of a parallel plate at the polarizing
angle, the refracted ray also meets the second surface of the plate at the
appropriate polarizing angle. For if fi be the index of the second medium
relatively to the first, the tangent of the angle of incidence, which is also the
K. III. 12
178 WAVE THEORY OF LIGHT. [148
cotangent of the angle of refraction, is equal to p. At the second surface
(the third medium being the same as the first) the angles of incidence and
refraction are interchanged, and therefore the condition for the^ polarizing
angle is satisfied, since the index for the second refraction is pr\
The principal formulae apply to light polarized in, and perpendicular to,
the plane of incidence. If the plane of polarization make an angle a with
that of incidence, the original vibration may be resolved into two —cos a
polarized in the plane of incidence, and sin a polarized in the perpendicular
plane. These components are reflected according to the laws already con-
sidered, and reconstitute plane-polarized light, of intensity
tan* (0- ft) _
If the incident light be polarized in a plane making 45° with the plane of
incidence, or be circularly-polarized (§ 20), or be unpolarized, (5) applies to
the reflected light, with substitution of £ for cos2 a and sin2 a. If ft denote in
the general case the angle between the plane of incidence and that in which
the reflected light is polarized,
a result the approximate truth of which has been verified by Fresnel and
Brewster.
The formulae for the intensities of the refracted light follow immediately
from the corresponding formulae relative to the reflected light in virtue of the
principle of energy. The simplest way to regard the matter is to suppose the
refracted light to emerge from the second medium into a third medium
similar to the first without undergoing loss from a second reflexion, a
supposition which would be realized if the transition between the two media
were very gradual instead of abrupt. The intensities of the different lights
may then be measured in the same way ; and the supposition that no loss of
energy is incurred when the incident light gives rise to the reflected and
refracted lights requires that the sum of the squares of the vibrations
representing the latter shall be equal to the square of the vibration
representing the former, viz., unity. We thus obtain, in the two cases
corresponding to (1) and (3),
_ sin2 (9 - ft) _ sin 20 sin 2ft ^
sm2(0 + ft)~ sin2(0 + ft) ' '
1 _ tan2 (O ~ ft) _ sin 20 sin 2ft , ,
tan2 (8 + ft) ~ sin2 (8 + 0,) cos2 (8 - ft) '
A plate of glass, or a pile of parallel plates, is often convenient as a
polarizer, when it is not necessary that the polarization be quite complete.
1888] PILE OF PLATES. 179
At the precise angle of incidence (tan"1 /A) there would be, according to
Fresnel's formulae, only one kind of polarized light reflected, even when the
incident light is unpolarized. The polarization of the transmitted light, on
the other hand, is imperfect ; but it improves as the number of plates is
increased.
If we suppose that there is no regular interference, the intensity (r) of the
light reflected from a plate is readily calculated by a geometric series when
the intensity (p) of the light reflected from a single surface is known. The
light reflected from the first surface is p. That transmitted by the first
surface, reflected at the second, and then transmitted at the first, is p (1 — pf.
The next component, reflected three times and transmitted twice, is p*(l — pf,
and so on. Hence
.}=2?~ ............... (9)
The intensity of the light reflected from a pile of plates has been
investigated by Provostaye and Desains*. If <p(m) be the reflexion from
TO plates, we may find as above for the reflexion from (TO + 1) plates,
1 - r<f>(m)
By means of this expression we may obtain in succession the values of <f> ( 2 ),
c., in terms of 0(1), viz., r. The general value is
as may easily be verified by substitution.
The corresponding expression for the light transmitted by a pile of
m plates is
The investigation has been extended by Stokes so as to cover the case in
which the plates exercise an absorbing influence f.
The verification of Fresnel's formula? by direct photometric measurement
is a matter of some difficulty. The proportion of perpendicularly incident
light transmitted by a glass plate has been investigated by RoodJ ; but the
deficiency may have been partly due to absorption. If we attempt to deal
directly with the reflected light, the experimental difficulties are much
increased; but the evidence is in favour of the approximate correctness of
* Ann. d. Chim. ixx. p. 159, 1850.
t Proc. Roy. Soc. xi. p. 545, 1862.
£ Am. Jour. VoL L. July 1870.
12—2
180 WAVE THEORY OF LIGHT. [148
Fresnel's formulae when light is reflected nearly perpendicularly from a
recently polished glass surface. When the surface is old, even though
carefully cleaned, there may be a considerable falling off of reflecting
power*.
We have seen that according to Fresnel's tangent-formula there would be
absolutely no reflexion of light polarized perpendicularly to the plane of
incidence, when the angle of incidence is tan"1 n, or, which comes to the
same thing, common light reflected at this angle could be perfectly ex-
tinguished with a Nicol's prism.
It was first observed by Airy that in the case of the diamond and other
highly refracting media this law is only approximately in accordance with the
facts. It is readily proved by experiment that, whatever be the angle of
incidence, sunlight reflected from a plate of black glass is incapable of being
quenched by a Nicol, and is therefore imperfectly plane-polarized. [1901. If
however the glass has recently been repolished with putty powder, the
reflexion is much reduced.]
This subject has been studied by Jamin. The character of the reflected
vibration can be represented, as regards both amplitude and phase, by the
situation in a plane of a point P relatively to the origin of coordinates 0.
The length of the line OP represents the amplitude, while the inclination of
OP to the axis of x represents the phase. According to Fresnel's formula
appropriate to light polarized perpendicularly to the plane of incidence, P is
situated throughout on the axis of x, passing through 0 when the angle of
incidence is tan"1 /A. Jamin found, however, that in general P does not pass
through 0, but above or below it. When P is on the axis of y, the amplitude
is a minimum, and the phase is midway between the extreme phases. For
one class of bodies the phase is in arrear of that corresponding to perpendicular
incidence, and for another class of bodies in advance. In a few intermediate
cases P passes sensibly through 0 ', and then the change of phase is sudden,
and the minimum amplitude is zero.
In the case of metals the polarization produced by reflexion is still more
incomplete. Light polarized perpendicularly to the plane of incidence is
reflected at all angles, the amount, however, decreasing as the angle of
incidence increases from 0° to about 75°, and then again increasing up to a
grazing incidence. The most marked effect is the relative retardation of one
polarized component with respect to the other. At an angle of about 75° this
retardation amounts to a quarter-period.
The intensity of reflexion from metals is often very high. From silver,
even at perpendicular incidence, as much as 95 per cent, of the incident light
* "On the Intensity of Light reflected from Certain Surfaces at nearly Perpendicular
Incidence," Proc. Roy. Soc. 1886. [Vol. u. p. 522.]
1888] METALLIC REFLEXION. 181
is reflected. There is reason for regarding the high reflecting power of
metals as connected with the intense absorption which they exercise. Many
aniline dyes reflect in abnormal proportion from their surfaces those rays of
the spectrum to which they are most opaque. The peculiar absorption
spectrum of permanganate of potash is reproduced [with reversal] in the
light reflected from a surface of a crystal*.
§ 27. Reflexion on the Elastic Solid Theory.
On the theory which assimilates the aether to an elastic solid, the investi-
gation of reflexion and refraction presents no very serious difficulties, but the
results do not harmonize very well with optical observation. It is, however,
of some importance to understand that reflexion and refraction can be explained,
at least in their principal features, on a perfectly definite and intelligible
theory, which, if not strictly applicable to the aether, has at any rate a distinct
mechanical significance. The refracting surface and the wave-fronts may for
this purpose be supposed to be plane.
When the vibrations are perpendicular to the plane of incidence (z = 0),
the solution of the problem is very simple. We suppose that the refracting
surface is x = 0, the rigidity and density in the first medium being N, D, and
in the second N1} A- The displacements in the two media are in general
denoted by £, 17, £; £1, ^i, £\; but in the present case £, 77, £,, ^ all vanish.
Moreover f, £i are independent of z. The equations to be satisfied in the
interior of the media are accordingly (§ 24)
At the boundary the conditions to be satisfied are the continuity of displace-
ment and of stress ; so that, when x = 0,
?-?" *£-*£ ......................... <3>
The incident waves may be represented by
£
where
(4)
* Stokes, "On the Metallic Reflection exhibited by Certain Non-Metallic Substances,'
Phil. Mag. Dec. 1853.
182 WAVE THEORY OF LIGHT. [148
and ax + by = const, gives the equation of the wave-fronts. The reflected and
refracted waves may be represented by
£ _ £'ei (-ax+by+et) ^ (5)
The coefficient of t is necessarily the same in all three waves on account of
the periodicity, and the coefficient of y must be the same since the traces of
all the waves upon the plane of separation must move together. With regard
to the coefficient of #, it appears by substitution in the differential equations
that its sign is changed in passing from the incident to the reflected wave ;
in fact
where V, Vl are the velocities of propagation in the two media given by
Now b/\/(a? + 62) is the sine of the angle included between the axis of # and
the normal to the plane of waves — in optical language, the sine of the angle
of incidence, and 6/\/(«i2 + &2) is in like manner the sine of the angle of
refraction. If these angles be denoted (as before) by 6, 0U (7) asserts that
sin 6 : sin 8l is equal to the constant ratio V : V1} the well-known law of sines.
The laws of reflexion and refraction follow simply from the fact that the
velocity of propagation normal to the wave-fronts is constant in each medium,
that is to say, independent of the direction of the wave-front, taken in con-
nexion with the equal velocities of the traces of all the waves on the plane of
separation (F/sin 6 = Fj/sin #]).
The boundary conditions (3) now give
whence
r = ^a + JVW (10)
a formula giving the reflected wave in terms of the incident wave (supposed
to be unity). This completes the symbolical solution. If a, (and 0,) be real,
we see that, if the incident wave be
f = cos (ax + by + ct),
or in terms of F, X, and 0,
2-rr
X
the reflected wave is
- Nt cot 6l ZTT ,
The formula for intensity of the reflected wave is here obtained on the
supposition that the waves are of harmonic type; but, since it does not
1888] ALTERNATIVE SUPPOSITIONS. 183
involve X and there is no change of phase, it may be extended by Fourier's
theorem to waves of any type whatever. It may be remarked that when
the first and second media are interchanged the coefficient in (12) simply
changes sign, retaining its numerical value.
The amplitude of the reflected wave, given in general by (12), assumes
special forms when we introduce more particular suppositions as to the
nature of the difference between media of diverse refracting power. Accord-
ing to Fresnel and Green the rigidity does not vary, or N = Nt. In this case
N cot 6 - J^cot 0l _ cot 6 - cot 0! _ sin (0X - 0)
Ncot0 + Nlcot0l ~ cot 0 + cot 0l ~ suT(0^+0) '
If, on the other hand, the density is the same in various media,
N, : N = V* : V* = sin* 0, : sin' 0,
and then
Ncot0- N, cot 0t = tan (ft - 0)
Ncot0 + N! cot 0, ~ tan (0l + 0) '
If we assume the complete accuracy of Fresnel's expressions, either alternative
agrees with observation; only, if N = Nlt light must be supposed to vibrate
normally to the plane of polarization; while, if D = D,, the vibrations are
parallel to that plane.
An intermediate supposition, according to which the refraction is regarded
as due partly to a difference of density and partly to a difference of rigidity,
could scarcely be reconciled with observation, unless one variation were very
subordinate to the other. But the most satisfactory argument against the
joint variation is that derived from the theory of the diffraction of light by
small particles (§ 25).
We will now, limiting ourselves for simplicity to Fresnel's supposition
(Nl = N), inquire into the character of the solution when total reflexion
sets in. The symbolical expressions for the reflected and refracted waves are
..(14)
and so long as a, is real they may be interpreted to indicate
..................... (15)
fl -f- tti
(16)
corresponding to the incident wave
(17)
184 WAVE THEORY OF LIGHT. [148
In this case there is a refracted wave of the ordinary kind, conveying away
a part of the original energy. When, however, the second medium is the
rarer (Fx> V), and the angle of incidence exceeds the so-called critical angle
(sin-1 (F/FO), there can be no refracted wave of the ordinary kind. In what-
ever direction it may be supposed to lie, its trace must necessarily outrun the
trace of the incident wave upon the separating surface. The quantity alt as
denned by our equations, is then imaginary, so that (13) and (14) no longer
express the real parts of the symbolical expressions (5) and (6).
If -iai be written in place of alf the symbolical equations are
g + idi cn-ax+bv+et) £ = 2tt gH-ia^x+by+ct) -}
' a — ia^ a — iaj
from which, by discarding the imaginary parts, we obtain
), ........................... (18)
where
tane = a1'/a .................................. (20)
Since x is supposed to be negative in the second medium, we see that the
disturbance is there confined to a small distance (a few wave-lengths) from
the surface, and no energy is propagated into the interior. The whole of the
energy of the incident waves is to be found in the reflected waves, or the
reflexion is total. There is, however, a change of phase of 2e, given by (20),
or in terms of F, F1} and 6,
................ (21)
The principal application of the formulae being to reflexions when the
second medium is air, it will be convenient to denote by p the index of the
first medium relatively to the second, so that fi = F3/ F. Thus
tan e = V{tana 6 - sec2 0//i2}. ..' .................... (22)
The above interpretation of his formula sin (6l - 0)/sin (6l + 6), in the case
where 6l becomes imaginary, is due to the sagacity of Fresnel. His argument
was perhaps not set forth with full rigour, but of its substantial validity there
can be no question. By a similar process Fresnel deduced from his tangent-
formula for the change of phase (2e') accompanying total reflexion when the
vibrations are executed in the plane of incidence,
tan e' = p ^{p? tan2 6 - sec2 6} ...................... (23)
The phase-differences represented by 2e and 2e' cannot be investigated
experimentally, but the difference (2e'-2e) is rendered evident when the
incident light is polarized obliquely so as to contribute components in both
the principal planes. If in the act of reflexion one component is retarded
1888] TOTAL REFLEXION. 185
more or less than the other, the resultant light is no longer plane but
elliptically polarized.
From (22) and (23) we have
tan (e - e) = cos 6 V[l - ^ cosec* 0},
whence
The most interesting case occurs when the difference of phase amounts to
a quarter of a period, corresponding to light circularly polarized. If, however,
we put cos (2e' — 2e) = 0, we find
Vsin2 0= I + /S± V{(1 + M2)2- V},
from which it appears that, in order that sin 0 may be real, /*2 must exceed
3 + \/8. So large a value of p.- not being available, the conversion of plane-
polarized into circularly-polarized light by one reflexion is impracticable.
The desired object may, however, be attained by two successive reflexions.
The angle of incidence may be so accommodated to the index that the altera-
tion of phase amounts to | period, in which case a second reflexion under
the same conditions will give rise to light circularly polarized. Putting
(26 — 26') = ^, we get
2/i2sin^=(H-v/|)[(l + /*2)sin26>+l], ............... (25)
an equation by which 0 is determined when p is given. It appears that,
when fj, = 1*51, 6 = 48° 37' or 54° 37'. These results were verified by Fresnel
by means of the rhomb shown in Fig. 27.
The problem of reflexion upon the elastic solid theory, when the vibrations
are executed in the plane of incidence, is more complicated,
on account of the tendency to form waves of dilatation. Fi8- 27-
In order to get rid of these, to which no optical phenomena
correspond, it is necessary to follow Green in supposing that
the velocity of such waves is infinite, or that the media are
incompressible*. Even then we have to introduce in the
neighbourhood of the interface waves variously called
longitudinal, pressural, or surface waves; otherwise it is
impossible to satisfy the conditions of continuity of strain
and stress. These waves, analogous in this respect to those
occurring in the second medium when total reflexion is in
progress (19), extend to a depth of a few wave-lengths only, and they are so
* The supposition that the velocity is zero, favoured by some writers, is inadmissible. Even
dilatational waves involve a shearing of the medium, and must therefore be propagated at a finite
rate, unless the resistance to compression were negative. But in that case the equilibrium would
be unstable. [1901. Lord Kelvin has since (Phil. Mag. xxvi. p. 415, 1888) shown that, if the
medium be held fast at the boundary, negative resistance to compression need not involve
instability.]
186 WAVE THEORY OF LIGHT. [148
constituted that there is neither dilatation nor rotation. On account of them
the final formulae are less simple than those of Fresnel. If we suppose the
densities to be the same in the two media, there is no correspondence what-
ever between theory and observation. In this case, as we have seen, vibrations
perpendicular to the plane of incidence are reflected according to Fresnel's
tangent-formula ; and thus vibrations in the plane of incidence should follow
the sine-formula. The actual result of theory is, however, quite different.
In the case where the relative index does not differ greatly from unity,
polarizing angles of 22£° and 67^° are indicated, a result totally at variance
with observation. As in the case of diffraction by small particles, an elastic
solid theory, in which the densities in various media are supposed to be equal,
is inadmissible. If, on the other hand, following Green, we regard the
rigidities as equal, we get results in better agreement with observation.
To a first approximation indeed (when the refraction is small) Green's
formula coincides with Fresnel's tangent-formula; so that light vibrating
in the plane of incidence is reflected according to this law, and light vibrating
in the perpendicular plane according to the sine-formula. The vibrations are
accordingly perpendicular to the plane of polarization.
The deviations from the tangent-formula, indicated by theory when the
refraction is not very small, are of the same general character as those
observed by Jamin, but of much larger amount. The minimum reflexion
at the surface of glass (fj, = f) would be -fa*, nearly the half of that which
takes place at perpendicular incidence, and very much in excess of the truth.
This theory cannot therefore be considered satisfactory as it stands, and
various suggestions have been made for its improvement. The only varia-
tions from Green's suppositions admissible in strict harmony with an elastic
solid theory is to suppose that the transition from one medium to the other
is gradual instead of abrupt, that is, that the transitional layer is of thickness
comparable with the wave-length. This modification would be of more
service to a theory which gave Fresnel's tangent-formula as the result of
a sudden transition than to one in which the deviations from that formula
are already too great.
It seems doubtful whether there is much to be gained by further discussion
upon this subject, in view of the failure of the elastic solid theory to deal with
double refraction. The deviations from Fresnel's formulae for reflexion are
comparatively small; and the whole problem of reflexion is so much concerned
with the condition of things at the interface of two media, about which we
know little, that valuable guidance can hardly be expected from this quarter.
It is desirable to bear constantly in mind that reflexion depends entirely upon
an approach to discontinuity in the properties of the medium. If the thick-
ness of the transitional layer amounted to a few wave-lengths, there would be
no sensible reflexion at all.
* Green's Papers, by Ferrers, p. 333.
1888] GREEN'S THEORY. 187
Another point may here be mentioned. Our theories of reflexion take no
account of the fact that one at least of the media is dispersive. The example
of a stretched string, executing transverse vibrations, and composed of two
parts, one of which in virtue of stiffness possesses in some degree the
dispersive property, shows that the boundary conditions upon which reflexion
depends are thereby modified. We may thus expect a finite reflexion at the
interface of two media, if the dispersive powers are different, even though the
indices be absolutely the same for the waves under consideration, in which
case there is no refraction. But a knowledge of the dispersive properties of
the media is not sufficient to determine the reflexion without recourse to
hypothesis*.
§ 28. The Velocity of Light.
According to the principles of the wave-theory, the dispersion of refraction
can only be explained as due to a variation of velocity with wave-length or
period. In aerial vibrations, and in those propagated through an elastic
solid, there is no such variation ; and so the existence of dispersion was at
one time considered to be a serious objection to the wave-theory. Dispersion
in vacuo would indeed present some difficulty, or at least force upon us views
which at present seem unlikely as to the constitution of free aether. The
weight of the evidence is, however, against the existence of dispersion in vacuo.
" Were there a difference of one hour in the times of the blue and red rays
reaching us from Algol, this star would show a well-marked coloration in its
phases of increase or decrease. No trace of coloration having been noticed,
the difference of times cannot exceed a fraction of an hour. It is not at all
probable that the parallax of this star amounts to one-tenth of a second, so
that its distance, probably, exceeds two million radii of the earth's orbit, and
the time which is required for its light to reach us probably exceeds thirty
years, or a quarter of a million hours. It is therefore difficult to see how
there can be a difference as great as four parts in a million between the
* The reader who desires to pursue this subject may consult Green, " On the Laws of
Reflexion and Refraction of Light at the Common Surface of Two Non-Crystallized Media,"
Camb. Tram. 1838 (Green's Works, London 1871, pp. 242, 283); Lorenz, "Deber die Reflexion
des Lichts an der Granzflache zweier isotropen, durchsicbtigen Mittel," Pogg. Ann. cxi. p. 460
(1860), and " Bestimmnng der Schwingungsrichtung des Licht aethers darch die Reflexion and
Brechong des Lichtes," ibid. cxiv. p. 238 (1861) ; Strutt (Rayleigh), " On the Reflexion of Light
from Transparent Matter," Phil. Mag. [4] xin. (1871); Von der Miihll, "Ueber die Reflexion und
Brechung des Lichtes an der Grenze unkrystallinischen Medien," Math. Ann. v. 470 (1872), and
" Ueber Greens Theorie der Reflexion nnd Brechung des Lichtes," Math. Ann. xxvn. 506 (1886) ;
Thomson, Baltimore Lectures; Glazebrook, "Report on Optical Theories," Brit. A*». Rep. 1886;
Rayleigh, " On Reflexion of Vibrations at the Confines of Two Media between which the Transition
is gradual," Proc. Math. Soc. xi. ; and Walker, "An Account of Cauchy's Theory of Reflexion and
Refraction of Light," Phil. Mag. xxm. p. 151 (1887). References to recent German writers,
Ketteler, Lommel, Voigt, &c., will be found in Glazebrook's Report.
188 WAVE THEORY OF LIGHT. [148
velocities of light coming from near the two ends of the bright part of
the spectrum*."
For the velocity of light in vacuo, as determined in kilometres per second
by terrestrial methods (Light, Enc. Brit. Vol. xiv. p. 585), Newcomb gives the
following tabular statement : — ;
Michelson, at Naval Academy, in 1879 299,910
Michelson, at Cleveland, 1882 299,853
Newcomb, at Washington, 1882, using only results supposed to be
nearly free from constant errors 299,860
Newcomb, including all determinations 299,810
To these may be added, for reference —
Foucault, at Paris, in 1862 298,000
Cornu, at Paris, in 1874 298,500
Cornu, at Paris, in 1878 300,400
This last result, as discussed by Listing 299,990
Young and Forbes, 1880-1881 301,382
Newcomb concludes, as the most probable result —
Velocity of light in vacuo = 299,860 + 30 kilometres [per second].
.It should be mentioned that Young and Forbes inferred from their observa-
tions a difference of velocities of blue and red light amounting to about
2 per cent., but that neither Michelson nor Newcomb, using Foucault's
method, could detect any trace of such a difference.
When we come to consider the propagation of light through ponderable
media, there seems to be little reason for expecting to find the velocity
independent of wave-length. The interaction of matter and ffither may well
give rise to such a degree of complication that the differential equation
expressing the vibrations shall contain more than one constant. The law
of constant velocity is a special property of certain very simple media. Even
in the case of a stretched string, vibrating transversely, the velocity becomes
a function of wave-length as soon as we admit the existence of finite stiffness.
As regards the law of dispersion, a formula, derived by Cauchy from
theoretical considerations, was at one time generally accepted. According
to this,
and there is no doubt that even the first two terms give a good representation
of the truth in media not very dispersive, and over the more luminous portion
of the spectrum. A formula of this kind treats dispersion as due to the
smallness of wave-lengths, giving a definite limit to refraction (A) when
the wave-length is very large. Recent investigations by Langley on the law
of dispersion for rock-salt in the ultra-red region of the spectrum are not
* Newcomb, Astron. Papers, Vol. n. parts in. and iv. , Washington 1885.
1888] DISPERSION OF REFRACTION. 189
very favourable to this idea. The phenomena of abnormal dispersion indicate
a close connexion between refraction and absorption, and Helmholtz has
formulated a general theory of dispersion based upon the hypothesis that it
may be connected with an absorbing influence operative upon invisible
portions of the spectrum. Upon this subject, which is as yet little under-
stood, the reader may consult Glazebrook's " Report on Optical Theories*."
[1901. Since this article was written, great advances have been made by the
German physicists, of whom Rubens may specially be named.]
The limits of this article do not permit the consideration of the more
speculative parts of our subject. We will conclude by calling attention to
two recent experimental researches by Michelson, the results of which cannot
fail to give valuable guidance to optical theorists. The first of these t was a
repetition under improved conditions of a remarkable experiment of Fizeau,
by which it is proved that when light is propagated through water, itself in
rapid movement in the direction of the ray, the velocity is indeed influenced,
but not to the full extent of the velocity of the water (»). Within the limits
of experimental error, the velocity agrees with a formula suggested by Fizeau
on the basis of certain views of Fresnel, viz.,
(2)
F0 being the velocity when the medium is stationary. In the case of water,
(/i? — I)//*1 = '437. Conformably with (2), a similar experiment upon air,
moving at a velocity of 25 metres per second, gave no distinct effect.
From the result of the experiments upon water we should be tempted to
infer that at the surface of the earth, moving through space, the aether still
retains what must be coarsely called relative motion. Nevertheless, the
second research above alluded to* appears to negative this conclusion, and
to prove that, at any rate within the walls of a building, the aether must
be regarded as fully partaking in the motion of material bodies.
* Brit. Assoc. Rep. 1886. In this matter, as in most others, the advantage lies with the
electro-magnetic theory. See J. W. Gibbs, Amer. Journ. xxin. 1882.
f " Influence of Motion of the Medium on the Velocity of Light," by A. Michelson and
E. W. Morley, Amer. Journ. mi. May, 1886.
J "On the Relative Motion of the Earth and the Luminiferous ^ther," by Michelson and
Morley, Phil. Mag. Dec. 1887.
149.
ON THE REFLEXION OF LIGHT AT A TWIN PLANE OF
A CRYSTAL.
[Phil Mag. Vol. xxvi. pp. 241—255, 1888.]
MY object in the present paper is to calculate cl priori the reflexion of
light at the surface between twin crystals, and to obtain formulae analogous
to those discovered by Fresnel for the case where both media are isotropic.
It is evident that success can only be attained upon the basis of a theory
capable of explaining at once Fresnel's laws of double refraction in crystals
and those just referred to, governing the intensity of reflexion when light
passes from one isotropic medium to another. So far as I am aware the
electric theory of Maxwell is the only one satisfying these conditions*; and
I have accordingly employed the equations of this theory. It will be
remembered that the electric theory of double refraction was worked out
by Maxwell himself, and that the application to the problem of reflexion
was successfully effected by von Helmholtz and Lorentzf. The present
investigation starts, however, independently from the fundamental equations,
as given in Maxwell's Electricity and Magnetism.
Equations of a Dialectric Medium, of which the Magnetic Permeability
is Unity throughout.
In Maxwell's notation the various components are represented as
follows:—
Electric Displacement /, g, h;
Current u} v, w \
Magnetic Force (or Induction) a, b, c;
Electromagnetic Momentum F, G, H;
Electromotive Force % . . . P, Q, R;
* See Prof. Willard Gibbs's excellent "Comparison of the Elastic and the Electrical Theories
of Light with respect to the Law of Double Refraction and the Dispersion of Colours" (Am.
Journ. Sci. June, 1888), which reaches me while revising the present investigation for the press.
t References to the works of previous writers will be found in Glazebrook's " Report on
Optical Theories," Brit. Assoc. Rep. 1886.
1888] REFLEXION AT A TWIN PLANE. 191
and the equations connecting them may be written
"£ -*• -S <»
**y , "** _ /\ (9\
dy dz~
dc db da dc _db da
4wrtt — -j~ —f . *rjri? — ~~; 5~~ . vW — - ~s~" ^ ~j . ...... ( o)
d V dz dz rf* tf_r rfw
rfj? d(? rfF dH dG dF
0=5 -T-, 6 = ^ -j— , c=^ 1-, (4)
ul/ OL-Z dz CUC dJC CIV
dF_d® _^_^? p__^_^ /-x
dt dx' ^ dt dy' dt dz"
In (1) it is assumed that the medium is a perfect insulator. Equations (4)
and (5) may be replaced by
dt = dz~dy' dt=dx~dz' dt=dy~fa' ^6)
from which 4> disappears. Thus
du _ d dc d db
~dt =dydt~dzdi
d?P &P_dL(dQ dR)
dy* + dz* dx\dy+ dz}
where as usual
In (7) and the similar equations in g and A there is involved no assumption as
to the homogeneity or isotropy of the dielectric medium. If, however, these
conditions are fulfilled,
dP dQ dR = Q
P, Q, R being proportional to f, g, h: and the equations then assume a
specially simple form.
The boundary conditions which must be satisfied at the transition from
one homogeneous medium to another are obtained without difficulty from
the differential equations. We will suppose that the surface of transition* is
the plane x = 0. The first condition follows immediately from (2). It is
that /" must be continuous across the surface x = 0. Equation (7) shows
that dQ/dy + dRjdz must be continuous. From the similar equation in g,
viz.: —
47r^ = S#~diA=VSQ~avidi + rfy + ^j' (8)
192 ON THE REFLEXION OF LIGHT [149
we see not only that dc/dt, or c, must be continuous, but also that Q must be
continuous. In like manner from the corresponding equation in h it follows
that R and b must be continuous. The continuity of Q and R secures that
of dQ/dy + dR/dz ; so that it is sufficient to provide for the continuity of
/ Q, R, b, c ............................... (A)*
Isotropic Reflexion.
If both media are isotropic, the problem of reflexion of plane waves is
readily solved. When the electric displacements are perpendicular to the
plane of incidence (xy}, f and g vanish, while h and the other remaining
functions are independent of z. The only boundary conditions requiring
attention are that R and 6 should be continuous, or by (6) that R and dRjdx
should be continuous. This leads, as is well known, to Fresnel's sine-formula
as the expression for the reflected wave.
When the electric displacements are in the plane of incidence, h = 0, and
(as before) all the remaining functions are independent of z. As an
introduction to the more difficult investigation before us, it may be well to
give a sketch of the solution for this case. In the upper medium we have as
the relation between force and displacement,
P, Q, R = 4>7rV*(f, g, h), ........................ (9)
and in the lower,
P, Q, R = 4arVl'(f, g, h), ........................ (10)
V, V1 being the two wave-velocities, whose ratio gives the refractive index.
Since h = 0, R = 0 ; and since R = 0, dP/dz = 0, it follows by (6) that 6 = 0.
The only conditions (A) requiring further consideration are thus the
continuity of/, Q or V*g, and c.
As the expression for the incident wave we take
f=qei(px+qy+8t)^ g = — peHp*+W+*U } .................. (H)
the ratio of the coefficients being determined by the consideration that the
directions f, g, h and p, q, r are perpendicular f. In like manner for the
reflected wave we have
and for the refracted wave
f=q0iei(PiX+qy+st)} g = -
* Of these conditions the first is really superfluous. If we differentiate (7) &c. with
respect to x, y, z respectively and add, we see that the truth of (2) is involved. In some
cases it would shorten the analytical expressions if we took P, Q, E as fundamental variables,
in place of/, g, h.
t In the present case r=0.
1888] AT A TWIN PLAXE OF A CRYSTAL. 193
The coefficient of y is the same for all the waves, since their traces on the
plane x = 0 must move together. The multipliers ff, 0l determine the
amplitudes of the reflected and refracted waves, and may be regarded as the
quantities whose expression is sought. The velocity of propagation in the
first medium is s/*/(p* + q*), so that
(14)
We have now to consider the boundary conditions. The continuity of/,
when x = 0, requires that
1 + 0' = 6,: ................................. (15)
and the continuity of V*g requires that
ff) = v?Plel ............................ (16)
These two equations suffice for the determination of ff, 0l ; and we may infer
that the third boundary condition is superfluous. It is easily proved to be
so; for in the upper medium,
dc _ dP dQ _ ^ f df
dt dy dx \dy
when x = 0. In the lower medium, when x = 0,
so that by (14) the continuity of dcjdt leads to the same condition as the
continuity of/
The usual formula for the reflected wave is readily obtained from (15),
(16). If <f>, fa be the angles of incidence and refraction,
so that
1 — ff _ sin* £, cot <£, _ sin Zfa
1 + ff ~ sin* <f> cot <f> ~ sin 2<f> '
Accordingly,
ff = sin 2<ft - sin 2<fc = tan (<j> - <fc) ,17^
sin 2<f> + sin 2<^ tan (^ + &) '
The insertion of this value of ff in (12) gives the expression for the reflected
wave corresponding to the incident wave (11). The ratio of amplitudes in
the two cases, being proportional to V(/3 +5r*)» ^ represented by ff, and (17)
is the well-known tangent-formula of Fresnel.
194
ON THE REFLEXION OF LIGHT [149
Propagation in a Crystal.
In a homogeneous crystalline medium, the relation of force to strain may
be expressed
P, Q, JR = 47r«/, bfg, cfh) ..................... (18)
where a,, 61; c, are the principal wave- velocities. We here suppose that the
axes of coordinates are chosen so as to be parallel to the principal axes of
the crystal. The introduction of these relations into (7), &c., gives
where
n = afdfjdx + b^dg/dy + tfdh/dz ................... (20)
The principal problem of double refraction is the investigation of the
form of the wave-surface. By means of (19) we can readily determine the
law of velocity (F) for various directions of wave-front (I, m, n). For this
purpose we assume
/,<7,* = (X, A*, *)«*", ........................... (21)
where
- Vt, ........................... (22)
and k = 2-rr -•- wave-length. In accordance with (2) we must have
l\ + mfi + nv = Q,.... ..................... .....(23)
signifying that the electric displacement is in the plane of the wave-front.
If we now write
n = U0eik<a,
and substitute the values of/, g, h from (21) in (19) we find
X(F2-a12) = ^-inoJ, &c.,
so that by (23)
which is Fresnel's law of velocities, leading to the wave-surface discovered
by him.
Reflexion at a Twin Plane.
We are now prepared for the consideration of our special problem, viz.,
the reflexion of plane waves at a twin surface of a crystal. We suppose
that the plane of separation is x = 0, and we assume that there is a plane
perpendicular to this (z — 0), with respect to which each twin is symmetrical.
The only difference between the two media is that which corresponds to a
rotation through 180° about the axis of x, perpendicular to the twin plane.
1888] AT A TWIN PLAXE OF A CRYSTAL 195
In consequence of the symmetry the axis of z is a principal axis in both
media : but the axes of x and y are not principal axes. For the relation
between force and strain in the first medium we may take
P ~*ir(Af+ Bg\ Q = 4/r(fl/+ Cg\ R = torDh ....... (25)
In the second medium we may in the first instance assume similar
expressions with accented letters; but the peculiar relation between the
two media demands that A' = A, C1 = C, U = D, B' = - B. Thus for the
second twin medium,
P = ±ir(Af- Bg), Q = 47r( - B/+ Cg\ R = lirDh, ...... (26)
the only difference being the change in the sign of B. If B vanish, all
optical distinction between the twins disappears, and there can be no
reflexion. The magnitude of B depends upon the intensity of the double
refraction in the twins, and also upon the angles between the principal axes
and the twin plane. If one of these angles were to vanish, B would
disappear, in spite of a powerful double refraction.
For a general solution of the problem of reflexion from a twin plane, we
should have to suppose the plane of incidence to be inclined at an arbitrary
angle to the plane of symmetry (JT, y) : but we may limit ourselves without
much loss of interest to the two principal cases, when the plane of incidence
(1) coincides with the plane of symmetry, (2) is perpendicular to it.
Incidence in the Plane of Symmetry.
Under the first head there are two problems which may be considered
separately. The simplest is that which arises when the vibrations are
perpendicular to the plane of incidence, that is, are parallel to 2. It is not
difficult to see that in this case the difference between the twins never comes
into operation, and that accordingly the reflexion vanishes : but it may be
well to apply the general method.
Since /, g, and therefore -by (25), (26)] P and Q, vanish throughout,
while h and R are independent of z, the two first of equations (7) are
satisfied identically, and the third becomes
, d*h d*R <PR
*7rdf= <& + df'
or by (25)
(»>
This equation applies to both media, since there is no change in the value of
D. Thus, so far as the equations to be satisfied in the interior are concerned,
the incident wave may be supposed to continue its course without alteration.
13—2
196 ON THE REFLEXION OF LIGHT [149
It is equally evident that the general boundary conditions are also
satisfied. For/ Q, c vanish throughout, and by (6) the continuity of R and
6 merely requires the continuity of h and dhjdx. Since all the conditions
are satisfied by supposing the incident wave to pass on without alteration, it
is clear that there can be no reflected wave.
We have next to consider the case when the vibrations are executed in
the plane of incidence, so that h vanishes, while (as before) all the remaining
functions are independent of z. On account of the symmetry there can be
but one reflected and but one refracted wave, and in each A must vanish.
We may, therefore, take the following expressions as applicable to the various
waves : —
Incident wave:
f=qei<px+qy+*t)> g — —p ei(px+gy+*t) (28)
satisfying
*&+•*> \ ............ (29)
Refracted wave:
f=q6l e{ < *>!*+«/+««> , g = —pl0iei <P.*+«/+«<> ............. (30)
The coefficient of the time (s) is necessarily the same throughout on account
of the periodicity ; and the coefficient of y is the same, since the traces of all
three waves upon the plane of separation x = 0 must move together. The
relations between p,q,s; p', q,s; pltq,s are to be obtained by substitution in
the differential equations. Of these the equation in h is satisfied identically,
since R = 0. The other equations for the upper medium are by (7), (8), (25),
These must be satisfied by the incident and reflected waves. On substitution
we find that both equations lead to the same conditions, viz. : —
st = Aqi-2Bpq-\-Cp\ ........................ (31)
a quadratic equation of which the two roots give p and p' in terms of q and s.
In the second medium we get in like manner for the refracted wave
s* = Aq* + IBp^q + Cp^, ........................ (32)
the sign of B being changed. Equating the two values of s2, we find
or C (p - p^ = 2Bq .................................. (33)
We have now to consider the boundary conditions (A). The functions R and
b vanish throughout ; but it remains to provide for the continuity of f, Q,
and c, when # = 0. The first of these conditions gives at once
ei .................................. (34)
1888] AT A TWIN PLAXE OF A CRYSTAL. 197
Again, the continuity of Q, equal to Bf+ Cg in the first medium, and to
— Bf + Cg in the second, gives
Bq-Cp + 0'(Bq-Cp') = -8l(Bq + Cp1) ............. (35)
The continuity of c leads, when regard is paid to (31), (32), merely to the
repetition of the condition (34).
If we eliminate 0l between (34), (35), we find
d' {2Bq - Cp' - Cp,} =C(p-Pl)- 2Bq = 0 by (33).
Hence 6' vanishes. Neither in this case, nor when the vibrations are
perpendicular to the plane of incidence, is there any reflexion of light
incident in the plane of symmetry. And this conclusion may of course be
extended to natural light, and to light plane or elliptically polarized in any
way whatever.
Plane of Incidence perpendicular to that of Symmetry.
We have now to consider the case when the plane of incidence is the
plane y = 0, perpendicular to that of symmetry. Here/, g, h are all finite,
but they (as well as P, Q, R, &c.) are independent of the coordinate y. The
problem is more complicated than when the plane of incidence coincides
with that of symmetry, because an incident wave is here attended by two
reflected waves, and two refracted waves.
The equation of the incident wave in the upper medium may be
expressed
or, since by (2) \p + vr = 0,
£gih = (r,n,-p)ei(*>x+rz+* ...................... (36)
The differential equations to be satisfied in the upper medium assume the
form
If we substitute for/, g, h from (36), the first and third equations give
# = r(Ar + Bp)+pD, ....................... (37)
and the second equation gives
(38)
198 ON THE REFLEXION OF LIGHT [149
These two equations determine p and p, when r, s are given. Since the
elimination of fj, leads to a quadratic in jt>2, it is evident that there are four
admissible values ±plt ±p2, corresponding to waves of given periodicity,
whose trace on the plane of separation moves with a given velocity. Of these
two (say with the + sign) are waves approaching the surface, and two are
waves receding from it. If we limit ourselves to a single incident wave
(+p^ we shall have still to take into account two reflected waves corre-
sponding to — pi, — p2. The equations show that the value of p, is the same
whether p be positive or negative ; we shall suppose that /*, corresponds to
±pi, IH to ±p2.
In applying the equations to the second medium we have to change the
sign of B ; and it is evident that they are satisfied by the same values of p
as before, and that the preceding values of //. are to be taken negatively.
Hence in the second medium — /Zj corresponds to + pl} — ^2 to ±p.2. For the
purposes of our present problem, where there is no incident wave in the
second medium, we are concerned only with +p± and + p%.
The complete specification of the system of waves corresponding to a
single incident wave (p^ in the first medium is thus: —
Incident wave :
f,g,h = (r>pl,-pl)W*0+r*w.t (39)
Two reflected waves:
f,g,h = (r, fjL1} pi) & ei <-*>,*+«+«<>
+ (r. fr, Pi) 0" ei(-v**+rz+st ; (40)
Two refracted* waves :
f, g,h = (r, - /*, , -Pl) 6, e«P,*+«+*»
+ (r,-^,-p,)d,ei^+rz+^ (41)
The next step is the introduction of the boundary conditions (A). The
continuity of /requires that
@1 + (r+6" = fi1+0i (42)
The continuity of R, or Dh, or h, gives with equal facility
pi®i-pi&-P*0"=Pl01+Ps0a (43)
Again, the continuity of Q, equal to Bf+ Cg in the first medium and to
— Bf+ Cg in the second, gives
(Br + CK) e, + (Br + C^) & + (Br + C&) 6"
= -(Br + Cftl)0l-(Br+Cp,)0a , (44)
The continuity of b, or db/dt, or by (6) d R/das - dP/dz, is found, when regard
is paid to (37), to be already secured by (42) ; and we have only further to
* It should be noticed that one of the refracted waves is not refracted in the literal sense,
being parallel to the incident wave.
1888] AT A TWIN PLANE OF A CRYSTAL. 199
consider the continuity of dc/dt, or by (6) of dQjdx, since P is here
independent of y. Thus
(45)
The coefficients which occur in (44), (45) may be expressed more briefly
in terms of the velocities of the various waves. For
........................... (46)
and thus by (38),
^V* ............. (47)
Setting now
<r, .................. (47')
the four equations of condition take the form
.(48)
[ V Vfffff = 0l 1
If we equate the values of 0,, 02 obtained from the first and second pairs of
equations (48), we find
1v Br«r = ,
and from these again
-.
(«r - a) (-era - 1) (tsr - a) (-our - 1 >
by which the two reflected waves are determined.
These reflected waves correspond to the incident wave (Gl5 pl} /*,), and it-
is the wave ff which is reflected according to the ordinary law. If there be
a second incident wave (0a, pit /*,), the corresponding reflected waves are to
be found from (50), (51) by interchanging & ', 0" , and by writing for or, a the
reciprocals of these ratios. If both incident waves coexist,
- ...... (52)
(53)
It will be observed that although the fronts of the two incident waves
0,, Bs are not parallel, they are the waves that would be generated by the
double refraction of a single wave incident from an isotropic medium upon
a face of the crystal parallel to the twin plane.
200 ON THE REFLEXION OF LIGHT [149
Doubly Refracting Power Small.
Thus far our equations are general. But the interpretation will be very
much facilitated if we introduce a supposition, which does not deviate far
from the reality of nature, viz. that the doubly refracting energy is com-
paratively small. There is no new limitation upon the direction of the
principal axes relatively to those of coordinates, but we assume that A, G, D
are nearly equal, and that B is small. We may imagine the two twin
crystals to be bounded by faces parallel to the twin face, and to be embedded
in an isotropic medium of nearly similar optical power. Under these
circumstances pi , p9, ', V\, V% are nearly equal, so that approximately -or = 1,
a = /j>2/f*i > and we may write (52), (53) in the form
(55)
It should be remarked that the intensities of the waves represented by 0,, &c.
are not simply proportional to &*, &c. Referring to (39), (40), we see that
the intensity of ©„ 0' is measured by (r» + p2 + /4i*)(©i8, 0'*); and that of
6S, 0" by (r*
Plate bounded by Surfaces parallel to Twin Plane.
Let us now regard the waves ©j, H2 as due to the passage into the
crystal of waves from an isotropic medium, under such conditions (of gradual
transition, if necessary) that there is no loss by reflexion. The interface is
supposed to be parallel to the twin reflecting plane, and the optical power to
be so nearly equal to that of the crystal that the refraction is negligible.
Then, if the vibration parallel to y (perpendicular to the plane of incidence)
be M, and that in plane of incidence be N, we have
M = fr®, + ^®2, -2V=V(/>2 + »l8){@i + ®2] ....... (56,57)
In like manner, if the vibrations of the emergent reflected wave perpendicular
and parallel to the plane of incidence be M', N',
M' = fJL10' + ^0", N' = </(p* + i*) [ff + B"} ....... (58,59)
If we are prepared to push to an extreme our supposition as to the smallness
of the doubly refracting power, (D, 0 in these equations may be identified
with the corresponding quantities in (54), (55) ; for a retardation of phase in
crossing and recrossing the stratum alike for all the waves might be dis-
regarded. We shall presently return to this question; but we will in the
meantime trace out the consequences which ensue when the double refraction,
if not extremely small in itself, is at least so small in relation to the
1888] AT A TWIN PLANE OF A CRYSTAL. 201
distances through which it acts (the thickness of the stratum), that the
relative changes of phase may be neglected. Then
,,, _2 i !» (Q ei_*ii»
~ - {**+ - - '
We have now to introduce certain relations derived from (37), (38). By
elimination of s, we get
Br.f + p{(A-C)i* + (D-C)pP}-Br(p> + r*) = 0 ....... (62)
If we here disregard the difference between pl and p» , we may treat it as a
quadratic, by which the two values of /* are determined ; and it follows that
-/*!/«. =!*+»* ............................... (63)
We might have arrived at this conclusion more quickly from the consideration
that in the limit the two directions of displacement (r, f*ltpi\ (r, v», p*) in
the reflected waves must be perpendicular to one another.
Again, from the general equation (37) we see that
Br(pl-ri+(pi*-p**)l> = 0,
whence approximately,
fr-ft - rB (64}
K-fr-ZpD' '
Introducing these relations into (60), (61), we find
These equations indicate that the intensity of the reflected light (J/'a+ S'*)
is proportional to that of the incident, without regard to the polarization of
the latter. Again, if the incident light be unpolarized (M and X equal, and
without permanent phase relation), so also is the reflected light. But what
is more surprising is, that if the incident light be polarized in or perpendicular
to the plane of incidence, the reflected light is polarized t'n the opposite
manner.
The intensity of reflexion may be expressed in terms of the angle of
incidence <f>, for
r2) = sin </>,
202 ON THE REFLEXION OF LIGHT [149
so that
When the angle of incidence is small, the intensity is proportional to its
square. And, as was to be expected, the reflexion is proportional to B*.
The laws here arrived at are liable to modification when, as must usually
happen in practice, the thickness of the plate cannot be neglected. The
incident light, on its way to the twin surface, and the reflected light on
its way back, is subject to a depolarizing influence, which in most cases
complicates the relation between the polarizations of the light before entering
and after leaving the crystal. One law, however, remains unaffected. If the
light impinging upon the crystal be unpolarized, it retains this character
upon arrival at the twin face. We have shown that it does not lose it in the
act of reflexion, neither can it lose it in the return passage after reflexion.
Hence, if the light originally incident upon the layer of crystal be unpolarized,
so is the re'flected light ultimately emergent from it.
If, on the other hand, the incident light be polarized, whether plane or
elliptically, the character of the emergent light must depend upon the precise
thickness of the crystalline layer, and will vary rapidly from one part of the
spectrum to another. The simplest case that we can consider is when the
polarization of the incident rays is such that one or other of ®i, ©2 vanish.
We will suppose that it is ®2 J so that after reflexion,
Intensity of ff = p* + rz + tf 8^
Intensity of 6" ~p2 + r*+t^' 6"*
= P-i ~ frfr 4/*iW -4^^
A*,1-/*!^ A*!1 0*1 +/*.)1 0*1 +**•)•
by (54), (55). This is the ratio of intensities that would be observed with
an analyzing nicol held so as to retain in succession & and 6". If the
crystalline layer be moderately thick, and the light be of mixed wave-lengths,
there will be no interference observable between & and 6", and thus the ratio
just found is the extreme ratio of intensities. By means of (62) we may
express it in terms of the angle of incidence (</>), and of the fundamental
optical constants of the crystal. Thus
(69)
(A - C)sin2 <f> + (D-C) cos8 <£ '
This expression shows that in general the emergent light will be fully
polarized only when <j> is very small. In this case we virtually fall back upon
1888] AT A TWIN PLAXE OF A CRYSTAL. 203
our original investigation where the thickness of the layer was neglected.
Since only 0x is present, there is no depolarization in the first passage ; and
when <f> = 0 there is no depolarization upon the return passage in consequence
of the disappearance of ff. The polarizations corresponding in this case to
BL 0, are obviously those in and perpendicular to the plane of incidence :
and we learn that, when the angle of incidence is small, polarizations in and
perpendicular to the plane of incidence are reversed in the reflected ray.
If the incident ray be polarized in other directions than these, the reflected
ray is in general not fully polarized, even though the angle of incidence be
small.
150.
ON THE REMARKABLE PHENOMENON OF CRYSTALLINE
REFLEXION DESCRIBED BY PROF. STOKES.
[Phil. Mag. xxvi. pp. 256—265, 1888.]
THE phenomenon in question is that exhibited by certain crystals of
chlorate of potash, consisting of a peculiar internal coloured reflexion. The
following, stated very briefly, are its leading features as described by
Stokes*:—
(1) If one of the crystalline plates be turned round in its own plane,
without alteration of the angle of incidence, the peculiar reflexion vanishes
twice in a revolution, viz. when the plane of incidence coincides with the
plane of symmetry of the crystal.
(2) As the angle of incidence is increased, the reflected light becomes
brighter and rises in refrangibility.
(3) The colours are not due to absorption, the transmitted light being
strictly complementary to the reflected.
(4) The coloured light is not polarized. It is produced indifferently
whether the incident light be common light or light polarized in any plane,
and is seen whether the reflected light be viewed directly or through
a Nicol's prism turned in any way.
(5) The spectrum of the reflected light is frequently found to consist
almost entirely of a comparatively narrow band. When the angle of
incidence is increased, the band moves in the direction of increasing
refrangibility, and at the same time increases rapidly in width. In many
cases the reflexion appears to be almost total.
Prof. Stokes has proved that the seat of the colour is a narrow layer,
about a thousandth of an inch in thickness, in the interior of the crystal ; and
* Proc. Roy. Soc. Feb. 1885.
1888] REMARKABLE PHENOMENON OF CRYSTALLINE REFLEXION. 205
he gives reasons for regarding this layer as a twin stratum. But the
phenomenon remains a mystery. "It is certainly very extraordinary and
paradoxical that light should suffer total or all but total reflexion at a
transparent stratum of the very same substance, merely differing in
orientation, in which the light had been travelling, and that, independently
of its polarization."
From the first reading of Prof. Stokes's paper, I have been much
impressed with the difficulty so clearly set forth. It seemed impossible that
a combination of two surfaces merely could determine either so copious or so
highly selected a reflexion. If light of a particular wave-length is almost
totally reflected, what hinders the reflexion when the wave-length is altered,
say, by one twentieth part ? Such a result may arise from the interference
of two streams under a relative retardation of many periods ; but in that case
there are necessarily a whole series of wave-lengths all equally effective.
The prism should reveal a number of bright bands and not merely a single
band. The selection of a particular wave-length reminds one rather of what
takes place in gratings ; and I was from the first inclined to attribute the
colours to a periodic structure, in which the twins alternate a large number
of times. Such a view explains not only the high degree of selection, but
also the copiousness of the reflexion.
Partly with a view to this question, I have discussed in a recent paper*
the propagation of waves in an infinite laminated medium (where, however,
the properties are supposed to vary continuously according to the harmonic
law), and have shown that, however slight the variation, reflexion is ultimately
total, provided the agreement be sufficiently close between the wave-length
of the structure and the half wave-length of the vibration. The number of
alternations of structure necessary in order to secure a practically perfect
reflexion will evidently depend upon the other circumstances of the case.
If the variation be slight, so that a single reflexion is but feeble, a large
number of alternations are necessary for the full effect, and a correspondingly
accurate adjustment of wave-lengths is then required. If the variation be
greater, or act to better advantage, so that a single reflexion is more powerful,
there is no need to multiply so greatly the number of alternations : and at
the same time the demand for precision of adjustment becomes less exacting.
The application of this principle to the case of an actual crystal, supposed to
include a given number of alternations, presents no difficulty. At perpen-
dicular incidence symmetry requires (and observation verifies) that the
reflexion vanish ; but, as the angle of incidence increases, a transition from
one twin to the other becomes more and more capable of causing reflexion.
* "On the Maintenance of Vibrations by Forces of Doable Frequency, and on the
Propagation of Waves through a Medium endowed with a Periodic Structure," Phil. Mag. Aug.
1887. [VoL m. p. 1.]
206 ON THE REMARKABLE PHENOMENON [150
Hence if the number of alternations be large, the spectrum of the reflected
light is at first limited to a narrow band (whose width determines in fact the
number of alternations). As the angle of incidence increases, the reflexion
at the centre soon becomes sensibly total, and at the same time the band
begins to widen*, in consequence of the less precise adjustment of wave-
lengths now necessary. At higher angles the reflexion may be sensibly
total over a band of considerable width. All this agrees precisely with
Prof. Stokes's description of the case considered by him to be typical. The
movement of the band towards the blue end of the spectrum is to be
attributed to the increasing obliquity within the crystal, as in the ordinary
theory of thin plates.
It thus appears that if we allow ourselves to invent a suitable crystalline
structure, there need be no difficulty in explaining the vigour and purity of
the reflexion ; but such an exercise of ingenuity is of little avail unless we
can at the same time render an account of the equally remarkable circum-
stances stated in (1) and (4). When the incidence is in the plane of
symmetry, no reflexion takes place. As Prof. Stokes remarks, this might be
expected as regards light polarized in the plane of symmetry ; but that there
should be no reflexion of the other polarized component is curious, to say
the least. Not less remarkable is it that when the incidence is in the
perpendicular plane, the reflected light should show no signs of polarization.
The phenomenon being certainly connected with the doubly refracting
property, we should naturally have expected the contrary.
The investigation of the reflexion from a twin-plane, contained in the
preceding paper [Vol. III. p. 194], shows, however, that the actually observed
results are in conformity with theory. In the plane of symmetry there
should be no reflexion of either polarized component, at least to the same
degree of approximation as is attained in Fresnel's well-known formulae for
isotropic reflexion. As regards light reflected in the perpendicular plane,
theory indicates that if the incident light be unpolarized, so also will be the
reflected light. Again, the intensity of the (unanalyzed) reflected light
should be independent of the polarization of the incident. So far there is
complete agreement with the observations of Prof. Stokes. But there is
a further peculiarity to be noticed. Theory shows that in the act of reflexion
at a twin plane, the polarization is reversed. If the incident light be
polarized in the plane of incidence, the reflected light is polarized in the
perpendicular plane, and vice versa. When I first obtained this result, I
* It should be observed that if the spectrum be a prismatic one, there is a cause of widening
which must be regarded as purely instrumental. According to Cauchy's law (fJ.=A +B\~'2),
d/j.= -2BX-33X;
so that if the band correspond in every position to a given relative range of X, its apparent width
(reckoned as proportional to 5/t) will vary as X"2. In a diffraction-spectrum this cause of
widening with diminishing X would be non-existent.
1888] OF CRYSTALLINE REFLEXION DESCRIBED BY PROF. STOKES. 207
thought it applicable without reservation in the actual experiment, and on
trial was disappointed to find that the reflected light was nearly unpolarized,
even when the incident light was fully polarized, whether in the plane of
incidence or in the perpendicular plane. When, however, the angle of
incidence was diminished, the expected phenomenon was observed, provided
that the original polarization were in, or perpendicular to, the plane of
incidence. If the original polarization were oblique, the reflected light was
not fully polarized, even though the angle of incidence were small*.
Further consideration appeared to show that the loss of polarization
usually observed could be explained by the depolarizing action of the
layer of crystal through which the light passes, both on its way to the
reflecting plane and on its return therefrom. As is shown in the preceding
paper, this depolarizing action does not occur when the angle of incidence is
small, and the polarization in, or perpendicular to, the plane of incidence. It
seems scarcely too much to say that the theory not only explains the laws
laid down by Stokes, but also predicts a very peculiar law not before
suspected f.
The theory, as so far developed, is indeed limited to incidences in the
two principal planes. It could probably be treated more generally without
serious difficulty ; but there seems no reason to suppose that anything very
distinctive would emerge. It is not unlikely that the intensity would prove
to be proportional to the square of the sine of the angle between the planes
of incidence and of symmetry. If this theory be accepted— and I see no
reason for distrusting it — the brilliant reflexion cannot be explained as due
to a single twin stratum. The simplest case which we can consider is when
the angle of incidence is small and the polarization in or perpendicular to
the plane of incidence. There is then sensibly but one wave reflected at the
first twin plane. On the arrival of the transmitted wave at the hinder
surface of the twin stratum, a second reflexion ensues, similar to the first,
except for the reversal of phase due to the altered circumstances. The
relation to one another of the two reflected waves is exactly the same as in
the ordinary theory of thin plates, and does not appear to admit of the
production of anything unusual. I think we may even go further, and
conclude that in conformity with our theory it is impossible to find an
* Whatever the angle of incidence, the arrangement of crossed nicols may sometimes be
conveniently applied in order to isolate the light under investigation from that reflected at the
front surface of the crystalline plate. In the observations described in the text the crystal was
mounted with Canada balsam between thick plates of glass, so that there was no difficulty in
observing separately the various reflexions. At small angles of incidence the coloured image is
at its brightest when the analyzing nicol is so turned that the white image (reflected from the
glass) vanishes, and rice versa, the incident light being polarized in, or perpendicularly to, the
plane of incidence.
t The wording of Prof. Stokes's description is perhaps a little ambiguous, but I gather that
he did not examine the result of a simultaneous operation of polarizer and analyzer.
208 ON THE REMARKABLE PHENOMENON [150
explanation of the brilliant and highly selected reflexion, unless upon the
supposition that there is a repeated alternation of structure.
The optical evidence in favour of the view that there is a large number
of twin planes thus appears to be very strong ; the difficulty is rather to
understand how such a structure can originate. And yet if we admit, as we
must, the possibility of the formation of one twin plane, and of two twin
planes at a very small distance asunder*, there seems nothing to forbid
a structure regularly periodic, which may perhaps be due to causes vibratory
in their nature.
It would undoubtedly be far more satisfactory to be able to speak of the
periodic structure as a matter of direct observation, and it is to be desired
that some practised microscopist should turn his attention to the subject.
Ex hypothesi, we could not expect to see the ruled pattern upon a section cut
perpendicularly to the twin planes, as it would lie upon, or beyond, the
microscopic limit. I have tried to detect it upon a surface inclined to the
planes at a very small angle, but hitherto without success t.
In the absence of complete evidence it is proper to treat the views here
put forward with a certain reserve ; but it is perhaps not premature to
consider a little further what may be expected to result from a structure
more or less regular. If the periodicity be nearly perfect, the bright central
band in the spectrum would be accompanied by subordinate bands of inferior
and decreasing brilliancy. If the angle of incidence be small, so that the
aggregate reflexion is but feeble, each stratum may be considered to act
independently, and the various reflected waves to be simply superposed.
The resultant intensity will depend of course upon the phase relations. At
the centre of the band the partial reflexions agree in phase, and the intensity
is a maximum. As we leave this point in either direction, the phases begin
to separate. When the alteration of wave-length is such that the phases of
the reflected waves range over a complete cycle, the resultant vanishes, and
a dark band appears in the spectrum. The same thing occurs whenever the
relative retardation of the extreme components amounts to a complete
number of periods. At points approximately midway between these, the
resultant is a maximum, but the values of the successive maxima diminish |.
Near the central band, where (when the number of alternations is great)
* This is the simplest supposition open to us, when, as in most of the coloured crystals, the
parts on either side of a very thin lamina are similarly oriented.
t [1U01. In Manchester Proceedings for 1889, Vol. in. p. 117, it is reported that "Dr Hodg-
kinson exhibited a specimen of iridescent chlorate of potash mounted in a special way in order
to demonstrate that the colour is produced, not by the interference of one thin plate, but by
numerous thin plates. The thin plates were readily seen in the specimen by means of a hand
magnifier, and the exhibit confirmed a prediction made several months since by Lord Rayleigh."]
% The case is similar to that of the distribution of brightness in the neighbourhood of a
" principal maximum," when light of given wave-length is diffracted by a grating.
1888] OF CRYSTALLINE REFLEXION DESCRIBED BY PROF. STOKES. 209
a considerable fraction of the incident light is reflected by the system of
layers, this way of regarding the matter may cease to be applicable, for then
the anterior and the posterior layers act under sensibly different conditions.
Apart from the magnitude of the complete linear period, something will
depend upon the manner in which it is divided between the twins. The
most favourable, as it is also perhaps the most probable, arrangement is that
in which the thicknesses are equal. In that case every partial reflexion may
agree in phase. If the thicknesses, though regular, are unequal, we may first
form the resultant for contiguous pairs, and then consider the manner in
which the partial resultants aggregate.
It will be seen that even if the thicknesses of the twins are equal, there
are still two ways in which a regularly laminated crystal may vary, as
compared with the single kind of variation open to a simple twin stratum.
These are the magnitude of the linear period, and the number of periods.
Comparison of a number of coloured crystals* seems to favour the view that
there are important differences of constitution, even when the colour is the
same at a given incidence.
In many cases the appearances are such as to suggest that the periodicity
is imperfect. A little irregularity might alter or obliterate the subordinate
bands, while leaving the central band practically unaffected. Sometimes
there is evidence of two or more distinct periods, each sustained through a
number of alternations. If the period were subject to a gradual change,
the central band in the spectrum of the reflected light would be diffused,
even at small angles of incidence. The mere broadening of the band might
be due to fewness of alternations ; but this case would be distinguished from
the other by the accompanying feebleness of illumination.
On the whole, the character of the reflected light appears to me to
harmonize generally with the periodical theory. One objection, however,
should be mentioned. It might be supposed that the total number of twin
planes was as likely to be odd as to be even. In the former case the layers
of crystal on either side of the thin lamina (which is the seat of the colour)
would be of opposite orientations. In many crystals the character of the
twinning is difficult of observation, but I have not noticed any instance of
brilliant coloration answering to this description. So far as it goes this
argument is in favour of the simple stratum theory; but, in view of our
ignorance as to how the twin planes originate, it can hardly be considered
decisive.
I have also examined a number of what appeared to be simply twinned
crystals, kindly sent me by Mr Stanford, of the North British Chemical
Works. The light reflected from the twin plane is not easily observed on
* For a rich collection of such crystals I am indebted to Mr Muspratt. He informs me that,
though the result of a second crystallization from comparatively pure liquids, the coloured
crystals are but rarely found when the chlorate is produced by the magnesium process.
B. in. H
210 ON THE REMARKABLE PHENOMENON [150
account of its feeble character, at least when, as in the experiments now
referred to, the incidence is limited by the requirement that the light must
enter the crystal at a face parallel to the twin plane. Using, however, the
method described by Prof. Stokes (§ 13), I was enabled to separate the
reflexions at the twin plane from those at the external surfaces of the crystal.
A narrow slit admitted sunlight into the dark room, and was focused upon
the crystal by a good achromatic object-glass*. When the obliquely reflected
light was examined with a hand magnifier, a ghost-like image corresponding
to the twin plane could usually be detected. As the crystal was rotated in
its own plane, this image vanished twice during the revolution.
It is worthy of notice that there is an evident difference both in the
brightness and quality of the reflected light obtained from different crystals,
even though apparently simply twinned. This suggests that, instead of
a single twin plane, there may sometimes be in reality 3, 5, or a higher odd
number of such in close juxtaposition. In other specimens, affording similar
reflexions, the principal thicknesses on either side of a very thin layer are
undoubtedly of the same kind, so that the number of twin planes must be
even. Here, again, the reflected light exhibited marked differences, when
various crystals were examined. In none of those now referred to could
the light reflected from the thin layer be observed without very special
arrangements.
In these experiments the light entered and left the crystal by a face
parallel to the twin planes. In one specially well-formed and apparently
simply twinned crystal I was able to observe a much more oblique reflexion
from the internal surface or surfaces. The light here entered and left the
crystal by cleavage faces making a large angle with the reflecting planes, and
thus under conditions widely different from those considered hitherto, and
in the latter part of the preceding theoretical discussion. Three reflected
images were seen, all completely polarized (the original light being un-
polarized), two in one direction and the third in the opposite direction.
These images are coloured, and present tolerably discontinuous spectra,
giving rise to a suspicion that the twin plane is not really single. These
observations were made without special arrangements by merely examining the
reflected images of a candle-flame, when the crystal was held close to the eye.
I have made many experiments on the crystallization of chlorate of potash
in the hope of tracing the genesis of the coloured crystals, but without
decisive results. Besides the usually small but highly coloured crystals,
found by Stokes, I have obtained many larger ones in which the reflexion is
feebler and less pure. These appear to be distinct from the exceedingly thin
plates which at the early stage of crystallization swim about in the solution.
Mounted in Canada balsam the crystals in question show colours of varying
degrees of brightness and purity ; and under these circumstances the effect
* I did not succeed in my first trials when I employed a common lens.
1888] OF CRYSTALLINE REFLEXION DESCRIBED BY PROF. STOKES. 211
can hardly be due to the action of the external surfaces (in contact with the
balsam). The light disappears twice during the revolution of the plates in
azimuth, just as in the case of the more highly coloured specimens. It seems
natural to suppose that the reflexion takes place from twin surfaces relatively
few in number, and perhaps less regular in disposition. Altogether the
existence of these crystals favours the view that fully formed colour is due
to a large number of regular alternations.
Some interesting observations bearing upon our present subject have
been recorded by Mr Madan*. Transparent crystals, free from twinning,
were heated on an iron plate to the neighbourhood of the fusion-point.
During the heating no change was observable, but " when the temperature
had sunk a few degrees a remarkable change spread quickly and quietly over
the crystal-plate causing it to reflect light almost as brilliantly as if a film
of silver had been deposited on it." Subsequently examined, the altered
crystals are found to " reflect little light at small angles of incidence, but at
all angles greater than about 10° they reflect light with a brilliancy which
shows that the reflexion must be almost total. . . .When the plate is turned
round in its own plane, two positions are found, differing in azimuth by 180°,
in which the crystal reflects no more light than an ordinary crystal under the
same conditions. In these cases the plane of incidence coincides with the
plane of crystallographic symmetry."
Mr Madan worked with comparatively thick (1 millim.) plates, from which
the associated twin had been removed by grinding. In repeating his
experiments I found it more convenient to use thin plates, such as may be
obtained without difficulty from crystallizations upon a moderate scale, and
which appear to be free from twinningf. There seems to be little doubt
that the altered crystals are composed of twinned layers. Except in respect
of colour, there is no difference between the behaviour of these crystals and
that of the brilliantly iridescent ones described by Stokes. If light be
incident at a small angle, and be polarized in or perpendicularly to the plane
of incidence, the polarization of the reflected light is the opposite to that of
the incident.
The only difference that I should suppose to exist between the con-
stitution of these crystals and that of the iridescent ones is, that in the
former case the alternations are irregular, and also probably more numerous.
Mr Madan conceives that there are actual cavities between the layers in the
heated crystals, comparing them to films of decomposed glass +. It is,
* "On the Effect of Heat in changing the Structure of Crystals of Potassium Chlorate,"
Nature, May 20, 1886.
t It is not clear why composite crystals free from included mother-liquor should suffer
disruption upon heating. A line drawn on the twin plane would tend to expand equally, to
whichever crystal it be considered to belong.
I "Although a large amount of light must escape reflexion at a single cavity, yet if the
transmitted rays encountered a large number of precisely similar and similarly situated cavities
H— 2
212 CRYSTALLINE REFLEXION. [150
however, certain that no closeness of contact could obviate the optical
discontinuity at a twin plane ; and there is besides a marked experimental
distinction between the cases in question. It is easy to observe, and was, I
think, observed by Brewster, that the application of water to a film of
decomposed glass destroys the effect. The water insinuates itself into the
cavities, and greatly attenuates the reflecting power. If a corresponding
experiment be tried, by wetting the edge of one of Mr Madan's crystals with
saturated solution of the salt, no change is observed to ensue.
Whether there are cavities or no, the fact that during the preparation the
silvery reflexion does not set in until the crystal has sensibly cooled is of
great interest. I have found that if a crystal in which the silvery lustre has
already been produced be reheated, the lustre disappears, to return again
upon a fall of temperature. The operation may be repeated any number of
times.
The existence of twin strata in Iceland spar was observed by Brewster*,
and Reuschf has shown that such strata can be induced artificially by
suitably applied pressure (Fig. 1) in rhombs
originally homogeneous. The planes of these
strata truncate the polar edges, i.e. the edges
which meet symmetrically at the obtuse trihedral
angle (0). Being desirous of examining whether
the reflexion from these strata would conform to
the law deduced from theory, I submitted a
rhomb to the treatment prescribed by Reusch
with the effect of developing several exceedingly
thin twin laminae (four or five at least) in close
juxtaposition. When light is reflected from these
strata in a plane perpendicular to the edge
(OD) which they truncate, the brilliancy is considerable. But the observation
which I wished to make required that the plane of incidence should be
perpendicular to this, so as to include the truncated edge and the optic axis.
Without much difficulty it was proved that in this plane the reflexion
vanished, reviving on either side as the plane of incidence deviated a little
from the plane of symmetry. The observation was facilitated by immersing
the crystal in a small cell containing water or bisulphide of carbon, the twin
strata being horizontal, and the plane of symmetry parallel to two of the
sides of the cell.
at slightly lower levels in the crystal, the sum of the partial reflexions would produce an effect
almost equivalent to a total reflexion of the original incident ray, and a corresponding deficiency
in the amount of light transmitted through the whole plate. The brilliancy of the colours in the
light reflected from the well-known films of decomposed glass is accounted for in precisely the
same way, and the successive separate films of glass can be easily seen under a microscope at
the edges of the compound film, where they only partially overlap."
* Treatise on Optics, 1853, p. 349. t Pogg. Ann. t. xn. p. 448 (1867).
151.
IS THE VELOCITY OF LIGHT IN AN ELECTROLYTIC LIQUID
INFLUENCED BY AN ELECTRIC CURRENT IN THE
DIRECTION OF PROPAGATION?
[Brit. Ass. Report, 1888, pp. 34-1—343.]
THE question here proposed has been considered by Roiti* and by
Zecherf. My experiments were made in ignorance of the work of these
observers, and the results would scarcely be worth recording were it not
that the examination seems to have been pushed further than hitherto. It
may be well to say at once that the result is negative.
The interference fringes were produced by the method of Michelson as
used in his important investigation respecting " The Influence of Motion of
the Medium upon the Velocity of Light}:." The incident ray ab meets a
half-silvered surface at b, by which part of the light is reflected and part is
transmitted. The reflected ray follows the course abcdefbg, being in all twice
reflected in 6. The transmitted ray takes the course abfedcbg, being twice
transmitted at b. These rays having pursued identical paths are in a
condition to form the centre of a system of fringes, however long and far
apart may be the courses cd, ef.
There is here nothing to distinguish the ray ab from a neighbouring
parallel ray. The incident plane wave-front perpendicular to ab gives rise
eventually to two coincident wave-fronts perpendicular to bg. With a wave
incident in another direction the case is different. The two emergent wave-
fronts remain, indeed, necessarily parallel, both having experienced an even
number of reflexions (four and six). But there will exist in general a
relative retardation, of amount (for wave-fronts perpendicular to the plane of
the diagram) proportional to the deviation from the principal wave-front.
* Pogg. Ann. CL. p. 164, 1873. t Rep. de Phys. xx. p. 151, 1884.
J Am. Jourtud, mi. p. 377, 1886.
214
VELOCITY OF LIGHT
[151
Hence, if the incident light comes in all directions, a telescope at g, focused
for indefinitely distant objects, reveals a system of interference bands, whose
direction should be vertical, if the adjustments could be perfectly carried out
in the manner intended.
\/
The success of the method does not require the complete symmetry of the
diagram. If the reflexions at d, e are effected by a right-angled prism, it is
necessary that cd, ef be parallel to one another but not that they be parallel
to the surface b. Supposing all the surfaces to remain vertical in any case,
the positions of b, f, and the incident ray ab, may be chosen arbitrarily. If the
distance de between the parallel courses is not closely prescribed, one adjust-
ment by rotation of the mirror c will suffice. In my experiments the optical
parts were mounted upon a large iron plate, so that the movable pieces c,
de could be shifted without loss of level. The incident ray ab was denned by
a small hole near the paraffin lamp which served as a source of light, and by
the centre of a moderately large circular aperture perforated in a screen and
illuminated when necessary with a candle. The mirror c was then rotated
until the rays cd,fe were parallel. This was tested by observing the equality
of their mutual distances near the extremities of their course.
If the distance between the parallel rays is prescribed, the adjustment is
more troublesome. The line/e being fixed, sights are laid down defining the
desired position of cd. These sights, as well as those before referred to
defining the incident ray, have now to be brought to apparent superposition
as seen by an eye looking along dc. For this purpose two conditions have to
be satisfied by, and two motions must be provided for, the mirror c. One of
these should be a movement of rotation, and the other of translation in a
direction nearly perpendicular to the plane of the mirror. Thus the mounting
may consist of a circular turntable resting upon an iron plate, the curved
edge of which is guided by the sides of a V, cut out of a flat piece of metal
1888] IN LIQUID TRAVERSED BY CURRENT. 215
and clamped to the plate. In each position of the V the angular motions
are easily swept over, and the double adjustment is effected without much
difficulty. When the parallelism of the rays is secured, the insertion of the
reflecting prism is all that remains. The adjustment of this is best effected
with the eye at the. observing telescope, which at this stage should be focused
upon the small aperture in the neighbourhood of the flame. By a motion of
the prism parallel to its hypothenuse the two images are brought to coinci-
dence*, and then the bands appear, if not at once, when the telescope is
accommodated for infinitely distant objects.
The half-silvered central plate would be at its best if it reflected light of
the same intensity as it transmits. I have generally found the reflexion on
the side next the air more powerful than upon the side next the glass; so
that the ideal would require the geometric mean of the two reflexions to be
equal to that of the two transmissions. A very slight silvering is all that is
wanted, such as from its want of coherence and brilliancy would be useless
for other purposes ; and the bands appear tolerable black, even though the
interfering lights are of decidedly unequal intensities. There is, of course,
a reflexion from the unsilvered surface of the plate. Owing to want of
parallelism in my apparatus, this image was distinctly separated from the
other. The two back reflectors were of flat glass, silvered by the milk sugar
process and used as specula.
The imperfections of the surfaces disturbed the formation of the bands
from full accordance with theory. The definition was usually better when
the pencils were limited, as by the screens employed to define the incident
ray, than when all obstruction was removed. The final adjustments for the
distinctness and desired width of bands were made with the eye at the
telescope by shifting the reflecting prism and occasionally by slight dis-
placements of one or other of the reflectors.
The tubes enclosing parts of cd, ef, and containing the electrolyte (diluted
sulphuric acid of nearly maximum conductivity), were closed at the ends by
plates of parallel glass. The current entered by lateral attachments, so
arranged that liquid (or gas) rising or falling from the platinum electrodes
would not at first enter the operative part of the tubes. The diameter of
the tubes was about f inch, and the effective length about 11 inches.
[inch = 2*54 cms.]
It will be sufficient to give details of one experiment. The two tubes
were connected in multiple arc, and of course in such a manner that the
* It should be noticed that if the object were at infinity, or if with the aid of a collimating lens
an image of it were thrown to infinity, the two images as seen focused through the telescope
would overlap in any case ; for it may be proved that, whatever may be the positions of the five
reflecting surfaces, the two emergent rays, corresponding to any incident ray, are necessarily
parallel.
216 VELOCITY OF LIGHT. [151
current travelled in opposite directions. The magnitude of the whole current
(say from eight Grove cells) was 1*5 ampere ; so that the current density, in
amperes per sq. cm., was
•75
7rx-382x2-542'
•26.
Now one of the interfering rays travelled 22 inches, or 56 centimetres,
with the current, and the other ray the same distance against the current.
On reversal of the current no shift of the bands could be perceived under
conditions where a shift of y1^ of a band* must have been evident. Hence
we may conclude that a current of the above-mentioned density does not
accelerate or retard the propagation, of light in the ratio of T'5 A, to 224 cms.
In the liquid we may take A = 4 x 10~5 cms., and if we reduce the result so as
to correspond to density unity, we may say that in dilute sulphuric acid a
current of one ampere per square centimetre does not alter the velocity of
light by 1 part in 13 millions, or by 15 metres per second.
It would probably be possible to carry the test ten or fifteen times further
by the use of much larger tubes and a more powerful battery, but there seems
to be no sufficient encouragement at present to make the attempt. The case
would, of course, be very different were anyone to show by a priori argument
a reason for expecting an effect of this order of magnitude.
* Probably I might say -gs, but it is best to be upon the safe side. When the contact was
maintained, a slight shift was observed, but in a direction independent of that of the current.
152.
OX THE BENDING AND VIBRATION OF THIN ELASTIC
SHELLS, ESPECIALLY OF CYLINDRICAL FORM.
[Proceedings of the Royal Society, XLV. pp. 105—123, 1888.]
IN a former publication* "On the Infinitesimal Bending of Surfaces of
Revolution," I have applied the theory of bending to explain the deformation
and vibration of thin elastic shells which are symmetrical about an axis, and
have worked out in detail the case where the shell is a portion of a sphere.
The validity of this application depends entirely upon the principle that
when the shell is thin enough and is vibrating in one of the graver possible
modes, the middle surface behaves as if it were inextensible. " When a thin
sheet of matter is subjected to stress, the force which it opposes to extension
is great in comparison with that which it opposes to bending. Under
ordinary circumstances, the deformation takes place approximately as if the
sheet were inextensible as a whole, a condition which, in a remarkable degree,
facilitates calculation, though (it need scarcely be said) even bending implies
an extension of all but the central layers." If we fix our attention upon one
of the terms involving sines or cosines of multiples of the longitude, into
which, according to Fourier's theorem, the whole deformation may be resolved,
the condition of inextensibility is almost enough to define the type. If
there are two edges, e.g., parallel to circles of latitude, the solution contains
two arbitrary constants ; but if a pole be included, as when the shell is in the
form of a hemisphere, one of the constants vanishes, and the type of defor-
mation is wholly determined, without regard to any other mechanical
condition, to be satisfied at the edge or elsewhere. It will be convenient
to restate, analytically, the type of deformation arrived at {equation (5)}.
If the point upon the middle surface, whose coordinates were originally
a, 6, <f>, moves to a + 8r, 9 + 88, </> + 8<f>, the solution is
s</> .................. (1)
Sr = Aa(s + cos 0) tan* $0 sin s$ }
London Math. Soc. Proc. Vol. xm. p. 4, November 1881. [Vol. i. Art. 78.]
218 ON THE BENDING AND VIBRATION [152
6 being the colatitude measured from the pole through which the shell is
complete. Any integral value higher than unity is admissible for s. The
values 0 and 1 correspond to displacements not involving strain.
In a recent paper* Mr Love dissents from the general principle involved
in the theory above briefly sketched, and rejects the special solutions founded
upon it as inapplicable to the vibration of thin shells. The argument upon
which I proceeded in my former paper, and which still seems to me valid,
may be put thus: It is a general mechanical principle! that, if given
displacements (not sufficient by themselves to determine the configuration)
be produced in a system originally in equilibrium by forces of corresponding
types, the resulting deformation is determined by the condition that the
potential energy of deformation shall be as small as possible. Apply this to
an elastic shell, the given displacements being such as not of themselves to
involve a stretching of the middle surface J. The resulting deformation will,
in general, include both stretching and bending, and any expression for the
energy will contain corresponding terms proportional to the first and third
powers respectively of the thickness. This energy is to be as small as
possible. Hence, when the thickness is diminished without limit, the actual
displacement will be one of pure bending, if such there be, consistent with
the given conditions. Otherwise the energy would be of the first order (in
thickness) instead of, as it might be, of the third order, in violation of the
principle.
It will be seen that this argument takes no account of special conditions
to be satisfied at the edge of the shell. This is the point at which Mr Love
concentrates his objections. He considers that the general condition necessary
to be satisfied at a free edge is in fact violated by such a deformation as (1).
But the condition in question§ contains terms proportional to the first and
to the third powers respectively of the thickness, the coefficients of the former
involving as factors the extensions and shear of the middle surface. It
appears to me that when the thickness is diminished without limit, the
fulfilment of the boundary condition requires only that the middle surface be
unstretched, precisely the requirement satisfied by solutions such as (1).
Of course, so long as the thickness is finite, the forces in operation will
entail some stretching of the middle surface, and the amount of this stretching
will depend on circumstances. A good example is afforded by a circular
cylinder with plane edges perpendicular to the axis. Let normal forces
locally applied at the extremities of one diameter of the central section cause
* " On the small free Vibrations and Deformation of a thin elastic Shell," Phil. Trans.
A, 1888.
t Phil. Mag. March 1875, [Vol. i. p. 23G] ; Theory of Sound, § 74.
£ There are cases where no displacement (involving strain at all) is possible without
stretching of the middle surface, e.g., the complete sphere.
§ See his equation (33).
1888] OF THIN ELASTIC SHELLS. 219
a given shortening of that diameter. That the potential energy may be a
minimum, the deformation must assume more and more the character of
mere bending as the thickness is reduced. The only kind of bending that
can occur in this case is the purely cylindrical one in which every normal
section is similarly deformed, and then the potential energy is proportional to
the total length of the cylinder. We see, therefore, that if the cylinder be
very long, the energy of bending corresponding to the given local contraction
of the central diameter may become very great, and a heavy strain is thrown
upon the principle that the deformation of minimum energy is one of pure
bending.
If the small thickness of the shell be regarded as given, a point will
at last be attained when the energy can be made least by a sensible local
stretching of the middle surface such as will dispense with the uniform
bending otherwise necessary over so great a length. But even in this
extreme case it seems correct to say that, when the thickness is sufficiently
reduced, the deformation tends to become one of pure bending.
At first sight it may appear strange that of two terms in an expression of
the potential energy, the one proportional to the cube of the thickness is to
be retained, while that proportional to the first power may be omitted. The
feet, however, is that the large potential energy which would accompany any
stretching of the middle surface is the very reason why such stretching will
not occur. The comparative largeness of the coefficient (proportional to the
first power of the thickness) is more than neutralised by the smallness of the
stretching itself, to the square of which the energy is proportional.
In general, if ^ be the coordinate measuring the violation of the tie
which is supposed to be more and more insisted upon by increasing stiffness,
and if the other coordinates be suitably chosen, the potential energy of the
system may be expressed
This follows from the general theorem that V and T may always be
reduced to sums of squares simply, if we suppose that T=|a1^r,s.
The equations of equilibrium under the action of external forces ¥,, ¥4, ...
are thus
*, = <vf,, ¥, = <*+„ &c,;
hence if the forces are regarded as given, the effect of increasing c, without
limit is not merely to annul ^r,, but also the term in Y which depends
upon it .
An example might be taken from the case of a rod clamped at one end A,
and deflected by a lateral force, whose stiffness from the end A up to a
neighbouring place B, is conceived to increase indefinitely. In the limit we
may regard the rod as clamped at B, and neglect the energy of the part AB,
in spite of, or rather in consequence of, its infinite stiffness.
220 ON THE BENDING AND VIBRATION [152
If it be admitted that the deformations to be considered are pure bendings,
the next step is the calculation of the potential energy corresponding thereto.
In my former paper, the only case for which this part of the problem was
attempted was that of the sphere. After bending, " the principal curvatures
differ from the original curvature of the sphere in opposite directions, and to
an equal amount*, and the potential energy of bending corresponding to any
element of the surface is proportional to the square of this excess or defect of
curvature, without regard to the direction of the principal planes." Though
he agrees with my conclusions, Mr Love appears to regard the argument as
insufficient. But clearly in the case of a given spherical shell, there are no
other elements upon which the energy of bending could depend. " Thus
the energy corresponding to the element of surface a2 sin 6 d0 d$ may be
denoted by
a?H(Sp-1)* sm0d0d<}>, ........................... (2)
where H depends upon the material and upon the thickness."
By the nature of the case H is proportional to the elastic constants and to
the cube of the thickness, from which it follows by the method of dimensions
that it is independent of a, the radius of the sphere. I did not, at the time,
attempt the further determination of H, not needing it for my immediate
purpose. Mr Love has shown that
H = ±nh*, ................................. (3)
where 2h represents the thickness, and n is the constant of rigidity. Why n
alone should occur, to the exclusion of the constant of compressibility, will
presently appear more clearly.
The application of (2) to the displacements expressed in (1) gave
(equation (18)}
(4)
0 being the colatitude of the (circular) edge. In the case of the hemisphere
of uniform thickness
V=i7rH2(s3-s)(2s*-l)A/ ...................... (5)
The calculation of the pitch of free vibration then presented no difficulty.
If <r denote the superficial density, and cos pt represent the type of vibration,
p2 corresponding to s = 2, pz to s = 3, and so on, it appeared that
p2 = V^x 5-2400, p3 = ^~ x 14726, p4 = — ? x 28'462 ;
so that
p3/p2 = 2-8102, p4/p3 = 5-4316,
determining the intervals between the graver notes.
* This is in virtue of Gauss's theorem that the product of the principal curvatures is
unaffected by bending.
1888] OF THIN ELASTIC SHELLS. 221
If the form of the shell be other than spherical, the middle surface is no
longer symmetrical with respect to the normal at any point, and the expression
of the potential energy is more complicated. The question is now not merely
one of the curvature of the deformed surface : account must also be taken of
the correspondence of normal sections before and after deformation*. A
complete investigation has been given by Love ; but the treatment of the
question now to be explained, even if less rigorous, may help to throw light
upon this somewhat difficult subject.
In the actual deformation of a material sheet of finite extent there will
usually be at any point not merely a displacement of the point itself, but a
rotation of the neighbouring parts of the sheet, such as a rigid body may
undergo. All this contributes nothing to the energy. In order to take the
question in its simplest form, let us refer the original surface to the normal
and principal tangents at the point in question as axes of coordinates, and let
us suppose that after deformation, the lines in the sheet originally coincident
with the principal tangents are brought back (if necessary) to occupy the
same positions as at first. The possibility of this will be apparent when it is
remembered that in virtue of the inextensibility of the sheet, the angles of
intersection of all lines traced upon it remain unaltered. The equation of
the original surface in the neighbourhood of the point being
that of the deformed surface may be written
In strictness (/>, + fyh)"1, 0*»+ fy*)"1 are the curvatures of the sections made
by the planes x = 0, y = 0 : but since principal curvatures are a maximum or
a minfmnm, they represent with sufficient accuracy the new principal cur-
vatures, although these are to be found in slightly different planes. The
condition of inextensibility shows that points which have the same JT and y
in (6) and (7) are corresponding points, and by Gauss's theorem it is further
necessary that
^ + ^ = 0. ................................. (8)
Pi P^
It thus appears that the energy of bending will depend upon two quantities,
one giving the alterations of principal curvature, and the other T depending
upon the shift (in the material) of the principal planes.
* An extreme case may serve as an illustration. Suppose that the bending is such that the
principal planes retain their positions relatively to the material surface, bat that the principal
curvatures are exchanged. The nature of the curvature at the point in question is the same after
deformation as before, and by a rotation through 90° round the normal the surfaces may be made
to fit ; nevertheless the energy of bending is finite.
222 ON THE BENDING AND VIBRATION [152
In calculating the energy we may regard it as due to the stretchings and
contractions under tangential forces of the various infinitely thin lamina; into
which the shell may be divided. The middle lamina, being unstretched,
makes no contribution. Of the other laminae, the stretching is in proportion
to the distance from the middle surface, and the energy of stretching is
therefore as the square of this distance. When the integration over the
whole thickness of the shell is carried out, the result is accordingly proportional
to the cube of the thickness.
The next step is to estimate more precisely the energy corresponding to
a small element of area of a lamina. The general equations in three
dimensions, as given in Thomson and Tait's Natural Philosophy, § 694, are
na = S, nb=T, nc=U, ..................... (9)
Mg = R- a (P + Q), ...(10)
where ff = T .................................. <n>*
The energy w, corresponding to the unit of volume, is given by
2w = (m + n) (e* +f* + #2) + 2 (m - n) (fg + ge + ef) + n (a2 + 62 + c2). (12)
In the application to a lamina, supposed parallel to xy, we are to take R — 0,
S = Q, T = 0; so that
g=-ae-±J-, a = 0, 6 = 0.
1 — <T
Thus in terms of the elongations e, f, parallel to x, y, and of the shear c,
we get
(13)
We have now to express the elongations of the various laminsB of a shell
when bent, and we will begin with the case where r = 0, that is, when the
principal planes of curvature remain unchanged. It is evident that in this
case the shear c vanishes, and we have to deal only with the elongations e
and / parallel to the axes. In the section by the plane of zx, let s, s' denote
corresponding infinitely small arcs of the middle surface and of a lamina
distant h from it. If T/T be the angle between the terminal normals,
s — pj-vjr, s' — (pi + h)^, s' —s = h -\IT. In the bending, which leaves s un-
changed,
Ss'
Hence
* M is Young's modulus, a is Poisson's ratio, n is the constant of rigidity, and (m-\ri) that
of cubic compressibility. In terms of Lame's constants (X, /A), m — \ + ^, n = fi.
1888] OF THIN ELASTIC SHELLS. 223
and in like manner /•=• h 8(l/p,). Thus for the energy U per unit of area
we have
and on integration over the whole thickness of the shell (2A)*
pi/ ps/ TO + n PI
This conclusion may be applied at once, so as to give the result applicable
to a spherical shell : for, since the original principal planes are arbitrary, they
can be taken so as to coincide with the principal planes after bending. Thus
T = 0 ; and b Gauss's theorem
so that
r=
where 8p~l denotes the change of principal curvature. Since e = — f. g = 0,
the various laminae are simply sheared, and that in proportion to their
distance from the middle surface. The energy is thus a function of the
constant of rigidity only.
The result (14) is applicable directly to the plane plate : but this case is
peculiar in that, on account of the infinitude of pl, pa, (8) is satisfied without
any relation between Bpt and Bp,. Thus for a plane plate
3 (PI* PJ* m + n Vpi pi' \
where p^1, p^1, are the two independent principal curvatures after bending.
We have thus far considered r to vanish: and it remains to investigate
the effect of the deformations expressed by
where £, 17 relate to new axes inclined at 45° to those of x, y. The curvatures
denned by (17) are in the planes of £, 17, equal in numerical value and opposite
in sign. The elongations in these directions for any lamina within the
thickness of the shell are AT, — hr, and the corresponding energy (as in the
case of the sphere just considered) takes the form
(18)
* It is here assumed that m and it are independent of h, that is, that the material is homo-
geneous. If we discard this restriction, we may form the conception of a shell of given thickness,
whose middle surface is physically inextensible, while yet the resistance to bending is moderate.
In this way we may realise the types of deformation discussed in the present paper, tcithout
svppotimg tke thiekmest to be infinitely tmaU ; and the independence of such types upon conditions
to be satisfied at a free edge is perhaps rendered more apparent.
224 ON THE BENDING AND VIBRATION [152
This energy is to be added* to that already found in (14); and we
get finally
, ...... (19)
p p m + n pl p '
as the complete expression of the energy, when the deformation is such that
the middle surface is unextended. We may interpret T by means of the
angle %, through which the principal planes are shifted ; thus
(20)
It will now be in our power to treat more completely a problem of great
interest, viz., the deformation and vibration of a cylindrical shell. In my
former paper [Art. 78] I investigated the types of bending, but without a
calculation of the corresponding energy. The results were as folio wsf.
If the cylinder be referred to columnar coordinates z, r, <£, so that the
displacements of a point whose equilibrium coordinates are z, a, $ are
denoted by Sz, Sr, a B<f), the equations expressing inextensibility take the
form
-0, + = 0 .......... (21)
<p d<f> dz
from which we may deduce
0 ............................... (22)
By (22), if 8<pcc cos s<p, we may take
aS<f> = (Asa + Bsz) cos s<}>, ........................ (23)
and then, by (21)
Sr = s (Asa + B8z) sin s<j>, Sz = - s^B^a sin s<f>. . . .(24, 25)
If the cylinder be complete, s is integral; Ag and Bg are independent
constants, either of which may vanish. In the latter case the displacement
is in two dimensions only*. It is unnecessary to stop to consider the
demonstrations of (21), inasmuch as these equations will present themselves
independently in the course of the investigation which follows.
It will be convenient to replace 8z, 8r, a %(f> by single letters, which,
however, it is difficult to choose so as not to violate some of the usual
conventions. In conformity with Mr Love's general notation, I will write
8z = u, a8(f> = v, Sr = w ................ (26)
* There are clearly no terms involving the products of T with the changes of principal
curvature 5 (ft"1), S (p2~l) ; for a change in the sign of T can have no influence upon the energy
of the deformation denned by (7).
t The method of investigation is similar to that employed by Jellet in his memoir ("On the
Properties of Inextensible Surfaces," Irish Acad. Trans. Vol. xxn. p. 179, 1855), to which
reference should have been made.
+ See Theory of Sound, § 233.
1888] OF THIX ELASTIC SHELLS. 225
The problem before us is the expression of the changes of principal curvature
and shifts of principal planes at any point P (z, £) of the cylinder in terms of
the displacements u, r, tr. As in (6), take as fixed coordinate axes the
principal tangents and normal to the undisturbed cylinder at the point P,
the axis of x being parallel to that of the cylinder, that of y tangential
to the circular section, and that of f normal, measured inwards. If, as it will
be convenient to do, we measure z and <f> from the point P, we may express
the undisturbed coordinates of a material point Q in the neighbourhood
of P, by
x=z, y = a£, r= Ja^. .................. (27)
During the displacement the coordinates of Q will receive the increments
w, w sin 4> + v cos tf>r — w cos <f> + r sin <f> :
so that after displacement
or if if, r, w be expanded in powers of the small quantities z. <f>.
Z+ + ~ .......... (28)
(29)
*,, v9, ... being the values of u, e at the point P.
These equations give the coordinates of the various points of the deformed
sheet. We have now to suppose the sheet moved as a rigid body so as to
restore the position (as far as the first power of small quantities is concerned)
of points infinitely near P. A purely translatory motion by which the
displaced P is brought back to its original position will be expressed by
the simple omission in (28), (29), (30) of the terms 11,, r», tr, respectively,
which are independent of z, <f>. The effect of an arbitrary rotation is
represented by the additions to x, y, f respectively of y03 — £0*, &i — -r^j,
*0* - y&i ; where for the present purpose 0,, 0,, 0, are small quantities of the
order of the deformation, the square of which is to be neglected throughout,
If we make these additions to (28), &c., substituting for j% y, f in the terms
. m 15
226 ON THE BENDING AND VIBRATION [152
containing 0 their approximate values, we find so far as the first powers
of z, <f>
du du . . a
x = z + ^j- z + -j-r- <p 4- a<bu3,
dzQ a<£0
dv dv , a
y = a<p + w0<j> + j- z + TT 9 — zVz ,
az0 U<<O
Now, since the sheet is assumed to be inextensible, it must be possible so
to determine 0,, #2, 03 that to this order x = z, y = a(j>, £= 0.
as-0'.'-' £+^°'
The conditions of inextensibility are thus (if we drop the suffixes as no
longer required)
du _ cfo C?M c?v /01.
-- = 0, w + ^-7^0, -jT+a;r = ()> ...... (31)
c?2r d0 d<f> dz
which agree with (21).
Returning to (28), &c., as modified by the addition of the translatory and
rotatory terms, we get
x = z + terms of 2nd order in z, </>,
, dzw ,„ dv . dv
-*5^*! + ^ + <
or since by (31) d2w/dz* = Q, and dv/d<f> = -w,
dv
The equation of the deformed surface after transference is thus
(1 dv 1 ftw ( 1 1 1 d*w
Comparing with (7) we see that
d*w\
' T--' ...... (33)
1888] OF THIJi ELASTIC SHELLS. 227
so that by (19)
_ do
=
This is the potential energy of bending reckoned per unit of area. It can
if desired be expressed by (31) entirely in terms of »*.
We will now apply (34) to calculate the whole potential energy of a
complete cylinder, bounded by plane edges z = ± I, and of thickness which, if
variable at all, is a function of z only. Since u, v, w are periodic when <f>
increases by 2ir, their most general expression in accordance with (31) is
(compare (23), &c.}
tr = 2 [(-4»a + Btz) cos s<f> — (A, a 4- Bt'z)siiis<f>], (35)
w = 2 [a (A.a + Btz)sms<f> + 8 (A.'a + Bjz) cos *£], (36)
« = 2 [-*-1 Bta sin s<j>-s-lBt'a cos *<£], (37)
in which the summation extends to all integral values of * from 0 to x .
But the displacements corresponding to a = 0, * = 1 are such as a rigid body
might undergo, and involve no absorption of energy. When the values of
u, v, w are substituted in (34) all the terms containing products of sines or
cosines with different values of s vanish in the integration with respect to <f>,
as do also those which contain cos &$> sin «£. Accordingly
^A.a + B.zY+(A,ta + Btfz^ + ^(^-iyi(B^+B^}}. ...
.(38)
Thus far we might consider A to be a function of z : but we will now treat it
as a constant. In the integration with respect to z the odd powers of z will
disappear, and we get as the energy of the whole cylinder of radius a; length
21, and thickness 2A,
V=
-io
...... (39)
in which * = 2, 3, 4, ....
* From Mr Love's general equations (12), (13), (18) m concordant result may be obtained by
introduction of the special conditions—
limiting the problem to the ease of the cylinder, and of those
ffl=ft=tg=0,
which express the ^extensibility of the middle surface.
15—2
228 ON THE BENDING AND VIBRATION [152
The expression (39) for the potential energy suffices for the solution of
statical problems. As an example we will suppose that the cylinder is
compressed along a diameter by equal forces F, applied at the points z = z^,
A) = 0, <f> — TT, although it is true that so highly localised a force hardly comes
within the scope of the investigation in consequence of the stretchings of the
middle surface, which will occur in the immediate neighbourhood of the points
of application*.
The work done upon the cylinder by the forces F during the hypothetical
displacement indicated by 8Ag, &c., will be by (36)
- F2s (aSAgf +.ZX&B/) (1 + cos sir),
so that the equations of equilibrium are
dv/dAs = 0, dv/dBg = 0.
dv/dAg' = — (1 + cos STT) saF, dv/dBs' = — (1 + cos STT) sz^F.
Thus for all values of s, As = Bs = 0 ; and for odd values of s, A, = Bs' = 0.
But when s is even,
3 . , 3sa*F ,
" ~ 32-2' "
*saz*F • (41)
s*- I)2' '
and the displacement w at any point (z, </>) is given by
w = 2 (A,' a + Bjz) cos 2<f> + 4 (At'a + Bt'z) cos 4<j> + . . . , ...(42)
where A,', B,', At', ... are determined by (40), (41).
If the cylinder be moderately long in proportion to its diameter, the
second term in the left-hand member of (41) may be neglected, so that
In this case (42) may be written
cos 40 + ...}, ...... (43)
showing that, except as to magnitude and sign, the curve of deformation
is the same for all values of zl and z-\.
If z = ±zly the amplitudes are in the ratio 1 ±3^12/^2J and if, further,
z,=l, i.e., if the force be applied at one of the ends of the cylinder, the
* Whatever the curvature of the surface, an area upon it may be taken so small as to behave
like a plane, and therefore bend, in violation of Gauss's condition, when subjected to a force which
is so nearly discontinuous that it varies sensibly within the area.
t That w is unaltered when z and zl are interchanged is an example of the general law
of reciprocity.
1888] OF THIX ELASTIC SHELLS. 22 •
amplitudes are as 2 : — 1. The section where the deformation (as represented
by IT) is zero, is given by 3zz, + P = 0, in which if z, = J, z = — £1.
When the condition as to the length of the cylinder is not imposed, the
ratio Bge : A, is dependent upon #, and therefore the curves of deformation
vary with z, apart from mere magnitude and sign. If, however, we limit
ourselves to the more important term * = 2, we have
ton Al
m+n a
—
so that w vanishes when
4m » } *
m+»3a* + l$ z, '
This equation may be applied to find what is the length of the cylinder when
the deformation just vanishes at one end if the force is applied at the other.
8^ J
For many materials a [equation (11)] is about \, or m = 2». In such
cases the condition is
* = fo.
It should not be overlooked that although w may vanish, u remains finite.
Reverting to (23), (24X (25) we see that, if the cylinder is open at both
ends, there are two types of deformation possible for each value of * If
we suppose the cylinder to be closed at z = 0 by a flat disk attached to it
round the circumference, the inextensibility of the disk imposes the con-
ditions, w = 8r=0, * = afy = 0, when z = 0*. Hence At = 0. and the only
deformation now possible is
v=afy = £,zcQS#l>, w = Sr = *£»* sin *£. (45)
Another disk, attached where z has a finite value, would render the cylinder
rigid.
Instead of a plane disk let us next suppose that the cylinder is closed at
z = 0 by a hemisphere attached to it round the circumference. By (1) the
three component displacements at the edge of the hemisphere (0 = ±v) are of
the form
r = a &£ = a cos *«/>, u = a&0 = — asmsf, w = &• = ga sin «£L
Equating these to the corresponding values for the cylinder, as given by (23),
(24), (25), we get At = 1, B, = s ; so that the deformation of the cylinder is now
limited to the type
», ...(46)
230 ON THE BENDING AND VIBRATION [152
in which we may, of course, introduce an arbitrary multiplier and an arbitrary
addition to <£. If the convexity of the hemisphere be turned outwards, z is to
be considered positive.
In like manner any other convex additions at one end of the cylinder
might be treated. There are apparently three conditions to be satisfied by
only two constants, but one condition is really redundant, being already
secured by the inextensibility of the edges provided for in the types of
deformations determined separately for the two shells. Convex additions,
closing both ends of the cylinder, render it rigid, in accordance with Jellet's
theorem that a closed oval shell cannot be bent.
It is of importance to notice how a cylinder, or a portion of a cylinder,
can not be bent. Take, for example, an elongated strip, bounded by two
generating lines subtending at the axis a small angle. Equations (31)
{giving d?w/dr'2 = 0} show that the strip cannot be bent in the plane con-
taining the axis and the middle generating line*. The only bending
symmetrical with respect to this plane is a purely cylindrical one which
leaves the middle generating line straight. There are two ways in which we
may conceive the strip altered so as to render it susceptible of the desired
kind of bending. The first is to take out the original cylindrical curvature,
which reduces it to a plane strip. The second is to replace it by one in
which the middle line is curved from the beginning, like the equator of a
sphere or ellipsoid of revolution. In this case the total curvature being
finite, the Gaussian condition can be satisfied by a change of meridional
curvature compensating the supposed change of equatorial curvature. It
is easy to calculate the actual stiffness from (8) and (14), for here r = 0.
We have
+n\ p
which expresses the work per unit of area corresponding to a given bending
Spf1 along the equator. If pt = oo , the cylindrical strip is infinitely stiff. If
the curvature be spherical, p2 = pt, and
and if p2 = oo ,
Whatever the equatorial curvature may be, the ratio of stiffnesses in the two
cases is equal to ra : m + n, or about 2 : 3, the spherically curved strip being
the stiffer.
The same principle applies to the explanation of Bourdon's gauge. In
this instrument there is a tube whose axis lies along an arc of a circle and
* This is the principle upon which metal is corrugated.
1888] OF THIN ELASTIC SHELLS. 231
whose section is elliptical, the longer axis of the ellipse being perpendicular
to the general plane of the tube. If we now consider the curvature at points
which lie upon the axial section, we learn from Gauss's theorem that a
diminished curvature along the axis will be accompanied by a nearer approach
to a circular section, and reciprocallv. Since a circular form has the larp-est
* • O
area for a given perimeter, internal pressure tends to diminish the eccentricity
of the elliptic section and with it the general curvature of the tube. Thus,
if one end be fixed, a pointer connected with the free end may be made to
indicate the internal pressure*.
We will now proceed with the calculation for the frequencies of vibration
of the complete cylindrical shell of length 2/. If the volume-density be pf ,
we have as the expression of the kinetic energy by means of (35), (36), (37)
T = £ . 2hp . (u2 + if
= 2-rphla 2 [a2 (1 + s5) (i,2 + A,'*) + [i P (1 + s5) + s~^] (B* + Bt'*) }. . . .(50)
From these expressions for V and T in (39), (50) the types and frequencies of
vibration can be at once deduced. The fact that the squares, and not the
products, of At , Bt, are involved, shows that these quantities are really the
principal coordinates of the vibrating system. If At,or At', van* as cosptt
we have
This is the equation for the frequencies of vibration in two dimensions^. For
a given material, the frequency is proportional to the thickness and inversely
as the square on the diameter of the cylinder^.
* Dec. 19.— It appears, however, that the bending of a curved tube of elliptical section cannot
be pure, since the parts of the walls which lie furthest from the [plane of the] circular axis are
necessarily stretched. The difficulty thus arising may be obviated by replacing the two halves of
the ellipse, which lie on either side of the major axis, by two symmetrical curves which meet on
the major axis at a .finite angle. [See Art. 171 below.]
According to the equations (in columnar coordinates) of my former paper, the conditions that
or, oz shall be independent of <f> lead to—
doz nfdr\* n
or=Cr, ^(-)=0,
where C is an absolute constant.
The case where the section is a rhombus (<fr/<fc= ± tana) may be mentioned.
The difficulty referred to above arises when dr,dz= x .
t This can scarcely be confused with the notation for the curvature in the preceding parts of
the investigation.
* See Thfory of Sound, § 233.
| There is nothing in these laws special to the cylinder. In the case of similar si
form, vibrating by pure bending, the frequency will be as the thicknesses and inversely as
corresponding areas. If the similarity extend also to the thickness, then the frequency is
inversely as the linear dimension, in accordance *ith the general law of Cauchy.
232 ELASTIC SHELLS. [152
In like manner if Bs, or B,'t vary as coaps't, we find
3a2 m + n
p;, = ^n^(^+**^r (52)
+ (s* + s2) P
If the cylinder be at all long in proportion to its diameter, the difference
between ps' and ps becomes very small. Approximately in this case
3a2
or if we take m = 2n, s = 2,
In my former paper I gave the types of vibration for a circular cone, of
which the cylinder may be regarded as a particular case. In terms of
columnar coordinates (z, r, <f>) we have
Sr = s tan 7 (A,z +- Bg) sin s<f>, ..................... (54)
Bz = tan2 7 [s-1 Bs - s (Asz + Bs)] sms<f>, ......... (55)
7 being the semi-vertical angle of the cone. For the calculation of the
energy of bending it would be simpler to use polar coordinates (r, 6, 0),
r being measured from the vertex instead of from the axis.
If the cone be complete up to the vertex, we must suppose, in (53) &c.,
Bg = 0. And if we proceed to calculate the potential energy, we shall find
it infinite, at least when the thickness is uniform. For since Ax is of no
dimensions in length, the square of the change of curvature must be
proportional to Ag2z~\ When this is multiplied by z dz, and integrated, a
logarithm is introduced, which assumes an infinite value when z = 0. The
complete cone must therefore be regarded as infinitely stiff, just as the
cylinder would be if one rim were held fast.
If two similar cones (bounded by circular rims) are attached so that the
common rim is a plane of symmetry, the bending may be such that the
common rim remains plane. If the distance of this plane from the vertex be
z1} the condition to be satisfied in (53) &c., is that &z = Q where z = z:.
Hence
(56)
8r = s tan 7 As \z— - — l— I sin s<f>, (57)
Sz = s tan2 7 Ag {z1 — z} sin s<f> (58)
153.
ON THE COMPOSITION OF WATER.
[Proceedings of the Royal Society, XLV. pp. 425—430, 1889.]
DURING the past year I have continued the work described in a former
communication on the relative densities of hydrogen and oxygen*, in the
hope of being able to prepare lighter hydrogen than was then found
possible. To this end various modifications have been made in the generating
apparatus. Hydrogen has been prepared from potash in place of acid. In
one set of experiments the gas was liberated by aluminium. In this case
the generator consisted of a large closed tube sealed to the remainder of the
apparatus ; and the aluminium was attached to an iron armature so arranged
that by means of an external electro-magnet it was possible to lower it into
the potash, or to remove it therefrom. The liberated gas passed through
tubes containing liquid potash f, corrosive sublimate, finely powdered solid
potash, and, lastly, a long length of phosphoric anhydride. But the result
was disappointing; for the hydrogen proved to be no lighter than that
formerly obtained from sulphuric acid.
I have also tried to purify hydrogen yet further by absorption in
palladium. In his recent important memoir*, " On the Combustion of
weighed Quantities of Hydrogen and the Atomic Weight of Oxygen."
Mr Keiser describes experiments from which it appears that palladium will
not occlude nitrogen — a very probable impurity in even the most carefully
prepared gas. My palladium was placed in a tube sealed, as a lateral
attachment, to the middle of that containing the phosphoric anhydride:
so that the hydrogen was submitted in a thorough manner to this reagent
both before and after absorption by the palladium. Any impurity that
* Roy. Soc. Proc. February, 1888 (Vol. XLIH. p. 356). [Vol. ra. Art. 146.]
t Of coarse this tube was superfluous in the present case, bat it was more convenient to
retain it.
± Amtr. Chem. Journ. VoL r. No. 4
234 ON THE COMPOSITION OF WATER. [153
might be rejected by the palladium was washed out of the tube by a current
of hydrogen before the gas was collected for weighing. But as the result
of even this treatment I have no improvement to report, the density of the
gas being almost exactly as before.
Hitherto the observations have related merely to the densities of hydrogen
and oxygen, giving the ratio 15'884, as formerly explained. To infer the
composition of water by weight, this number had to be combined with that
found by Mr Scott as representing the ratio of volumes*. The result was
2 x 15-884
1-9965
The experiments now to be described are an attempt at an entirely
independent determination of the relative weights by actual combustion of
weighed quantities of the two gases. It will be remembered that in Dumas's
investigation the composition of water is inferred from the weights of the
oxygen and of the water, the hydrogen being unweighed. In order to avoid
the very unfavourable conditions of this method, recent workers have made
it a point to weigh the hydrogen, whether in the gaseous state as in the
experiments of Professor Cooke and my own, or occluded in palladium as
in Mr Keiser's practice. So long as the hydrogen is weighed, it is not very
material whether the second weighing relate to the water or to the oxygen.
The former is the case in the work of Cooke and Keiser, the latter in the
preliminary experiments now to be reported.
Nothing could be simpler in principle than the method adopted. Globes
of the same size as those employed for the density determinations are rilled
to atmospheric pressure with the two gases, and are then carefully weighed.
By means of Sprengel pumps the gases are exhausted into a mixing chamber,
sealed below with mercury, and thence by means of a third Sprengel are
conducted into a eudiometer, also sealed below with mercury, where they
are fired by electric sparks in the usual way. After sufficient quantities of
the gases have been withdrawn, the taps of the globes are turned, the leading
tubes and mixing chamber are cleared of all remaining gas, and, after a final
explosion in the eudiometer, the nature and amount of the residual gas are
determined. The quantities taken from the globes can be found from the
weights before and after operations. From the quantity of that gas which
proved to be in excess, the calculated weight of the residue is subtracted.
This gives the weight of the two gases which actually took part in the
combustion.
In practice, the operation is more difficult than might be supposed from
the above description. The efficient capacity of the eudiometer being
* [1901. Dr Scott's final number (Proc. Roy. Soc. Vol. LIII. p. 133, 1893) was 2-00245.]
1889] ON THE COMPOSITION OF WATER. 235
necessarily somewhat limited, the gases must be fed in throughout in very
nearly the equivalent proportions; otherwise there would soon be such an
accumulation of residue that no further progress could be made. For this
reason nothing could be done until the intermediate mixing chamber was
provided. In starting a combustion, this vessel, originally full of mercury,
was charged with equivalent quantities of the two gases. The oxygen was
first admitted until the level of the mercury had dropped to a certain mark,
and subsequently the hydrogen down to a second mark, whose position
relatively to the first was determined by preliminary measurements of
volume. The mixed gases might then be drawn off into the eudiometer
until exhausted, after which the chamber might be recharged as before.
But a good deal of time may be saved by replenishing the chamber from the
globes simultaneously with the exhaustion into the eudiometer. In order to
do this without losing the proper proportion, simple mercury manometers
were provided for indicating the pressures of the gases at any time remaining
in the globes. But even with this assistance close attention was necessary to
obviate an accumulation of residual gas in the eudiometer, such as would
endanger the success of the experiment, or, at least, entail tedious delay. To
obtain a reasonable control, two sparking places were provided, of which the
upper was situate nearly at the top of the eudiometer. This was employed
at the close, and whenever in the course of the combustion the residual gas
chanced to be much reduced in quantity ; but, as a rule, the explosions were
made from the lower sparking point. The most convenient state of things
was attained when the tube contained excess of oxygen down to a point
somewhat below the lower sparking wires. Under these circumstances, each
bubble of explosive gas readily found its way to the sparks, and there was no
tendency to a dangerous accumulation of mixed gas before an explosion took
place. When the gas in excess was hydrogen, the manipulation was more
difficult, on account of the greater density of the explosive gas retarding its
travel to the necessary height.
In spite of all precautions several attempted determinations have failed
from various causes, such as fracture of the eudiometer and others which it is
not necessary here to particularise, leading to the loss of much labour. Five
results only can at present be reported, and are as follows : —
December 24, 1888 15-93
January 3, 1889 15'98
„ ' 21, „ 15-98
February 2, „ 15'93
13, „ 15-92
Mean 15'95
This number represents the atomic ratio of oxygen and hydrogen as deduced
immediately from the weighings with allowance for the unburnt residue. It
236 ON THE COMPOSITION OF WATER. [153
is subject to the correction for buoyancy rendered necessary by the shrinkage
of the external volume of the globes when internally exhausted, as explained
in my former communication*. In these experiments, the globe which
contained the hydrogen was the same (14) as that employed for the density
determinations. The necessary correction is thus four parts in a thousand,
reducing the final number for the atomic weight of oxygen to
15-89,
somewhat lower than that which I formerly obtained (15'91) by the use of
Mr Scott's value of the volume ratio. It may be convenient to recall that
the corresponding number obtained by Cooke and Richards (corrected for
shrinkage) is 15'87, while that of Reiser is 15*95.
In the present incomplete state of the investigation, I do not wish to lay
much stress upon the above number, more especially as the agreement of the
several results is not so good as it should be. The principal source of error,
of a non-chemical character, is in the estimation of the weight of the hydrogen.
Although this part of the work cannot be conducted under quite such
favourable conditions as in the case of a density determination, the error in
the difference of the two weighings should not exceed 0'0002 gram. The
whole weight of the hydrogen used is about O'l gram*f" ; so that the error
should not exceed three in the last figure of the final number. It is thus
scarcely possible to explain the variations among the five numbers as due
merely to errors of the weighings.
The following are the details of the determination of February 2, chosen
at random : —
Before combustion ... G14 + H + 0'2906 = Gu ... pointer 20'05
After „ . . . G14 + H + O4006 = Gu . . . pointer 20-31
Hydrogen taken = O'llOO - 0-00005 = 0'10995 gram.
Before combustion ... G13 + O = Gu + 2'237 . . . pointer 20'00
After „ ... Gu + 0 = Gu + 1-357 ... pointer 19'3
Oxygen taken = 0'8800 + O'OOOl = 0'8801 gram.
At the close of operations the residue in the eudiometer was oxygen,
occupying 7'8 c.c. This was at a total pressure of 29'6 — 16'2 = 13-4 inches
* The necessity of this correction was recognised at an early stage, and, if I remember rightly,
was one of the reasons which led me to think that a redetermination of the density of hydrogen
was desirable. In the meantime, however, the question was discussed by Agamennone (Atti
(Rendiconti) d. It. Accad. dei Lined, 1885), and some notice of his work reached me. When
writing my paper last year I could not recall the circumstances ; but since the matter has
attracted attention I have made inquiry, and take this opportunity of pointing out that the
credit of first publication is due to Agamennone.
f It was usual to take for combustion from two-thirds to three-fourths of the contents
of the globe.
1889] ON THE COMPOSITION OF WATER. 237
of mercury. Subtracting 0*4 inch for the pressure of the water vapour, we
get 13*0 as representing the oxygen pressure. The temperature was about
12° C. Thus, taking the weight of a cub. cm. of oxygen at 0° C. and under
a pressure of 76~0 cm. of mercury to be 0*00143 gram, we get as the weight
of the residual oxygen
+ !2
The weight of oxygen burnt was, therefore, 0'8801 - 0-0046 = 0'8755
gram.
Finally, for the ratio of atomic weights,
Oxygen _
£ Hydrogen
In several cases the residual gas was subjected to analysis. Thus, after
the determination of February 2, the volume was reduced by additions of
hydrogen to 1*2 c.c. On introduction of potash there was shrinkage to about
0'9, and, on addition of pyrogallic acid, to O'l or 0'2. These volumes of gas
are here measured at a pressure of ^ atmosphere, and are, therefore, to be
divided by 3 if we wish to estimate the quantities of gas under standard
conditions. The final residue of (say) OO5 c.c. should be nitrogen, and, even
if originally mixed with the hydrogen — the most unfavourable case — would
involve an error of only ^^ in the final result. The (H c.c. of carbonic
anhydride, if originally contained in the hydrogen, would be more important :
but this is very improbable. If originally mixed with the oxygen, or due to
leakage through india-rubber into the combustion apparatus, it would lead to
no appreciable error.
The aggregate impurity of O'lo, here indicated, is tolerably satisfactory in
comparison with the total quantity of gas dealt with — 2000 c.c. It is possible,
however, that nitrogen might be oxidised, and thus not manifest itself under
the above tests. In another experiment the water of combustion was examined
for acidity, but without definite indications of nitric acid. The slight reddening
observed appeared to be rather that due to carbonic acid, some of which, it
must be remembered, would be dissolved in the water. These and other
matters demand further attention.
The somewhat complicated glass blowing required for the combustion
apparatus has all been done at home by my assistant, Mr Gordon, on whom
has also Mien most of the rather tedious work connected with the evacuation
of globes and other apparatus, and with the preparation of the gases.
[1901. Further work upon this subject is recorded in Proc. Roy. Soc.
VoL L. p. 449, 1892. Vide infra.}
154.
THE HISTORY OF THE DOCTRINE OF RADIANT ENERGY.
[Philosophical Magazine, xxvu. pp. 265—270, 1889.]
IN his interesting Address* to the American Association for the Advance-
ment of Science, Prof. Langley sketches the development of the modern
doctrine of Radiant Energy, and deduces important lessons to be laid to heart
by all concerned in physical investigation. This is a most useful undertaking ;
but in the course of it there occur one or two statements which, in the
interest of scientific history, ought not to be allowed to pass without a
protest.
After quoting Melloni's very unequivocal conclusion of 1843, that " Light
is merely a series of calorific indications sensible to the organs of sight ; or,
vice versa, the radiations of obscure heat are veritable invisible radiations of
light," Prof. Langley goes on to say, " So far as I know, no physicist of
eminence reasserted Melloni's principle with equal emphasis till J. W. Draper,
in 1872. Only sixteen years ago, or in 1872, it was almost universally
believed that there were three different entities in the spectrum, represented
by actinic, luminous, and thermal rays."
These words struck me strangely as I first read them. My own scientific
ideas were formed between 1860 and 1866, and I certainly never believed in
the three entities. Having on a former occasion referred to this question f
as an illustration of the difference of opinion which is sometimes to be found
between the theoretical and experimental schools of workers, I was sufficiently
interested in the matter to look up a few references, with results which are,
I think, difficult to reconcile with Prof. Langley 's view.
In Young's Lectures I we read : — " Dr Herschel's experiments have shown
that radiant heat consists of various parts which are differently refrangible,
* Amer. Journ. Sci. Jan. 1889.
t Address to Section A, Brit. Assoc. Report, 1882. [Vol. n. p. 122.]
£ Vol. i. p. 638 (1807).
ON THE HISTORY OF THE DOCTRINE OF RADIANT ENERGY. 239
and that, in general, invisible heat is less refrangible than light This
discovery must be allowed to be one of the greatest that have been made
since the days of Newton —
"It was first observed in Germany by Bitter, and soon afterwards in
England by Dr Wollaston, that the muriate of silver is blackened by invisible
rays, which extend beyond the prismatic spectrum, on the violet side. It is
therefore probable that these black or invisible rays, the violet, blue, green,
perhaps the yellow, and the red rays of light, and the rays of invisible heat,
constitute seven different degrees of the same scale, distinguished from each
other into this limited number, not by natural divisions, but by their effects
on our senses : and we may also conclude that there is some similar relation
between heated and luminous bodies of different kinds."
And, again, on p. 654 : " If heat is not a substance, it must be a quality :
and this quality can only be motion. It was Xewton's opinion that heat
consists in a minute vibratory motion of the particles of bodies, and that this
motion is communicated through an apparent vacuum by the undulations of
an elastic medium, which is also concerned in the phenomena of light. If
the arguments which have been lately advanced in favour of the undulatory
theory of light be deemed valid, there will be still stronger reasons for
admitting this doctrine respecting heat: and it will only be necessary to
suppose the vibrations and undulations principally constituting it to be
larger and stronger than those of light, while at the same time the smaller
vibrations of light, and even the blackening rays, derived from still more
minute vibrations, may perhaps, when sufficiently condensed, concur in
producing the effects of heat. These effects, beginning from the blackening
rays, which are invisible, are a little more perceptible in the violet, which
still possess but a feint power of illumination : the yellow-green afford the
most light ; the red give less light, but much more heat : while the still
larger and less frequent vibrations, which have no effect upon the sense of
sight, may be supposed to give rise to the least refrangible rays, and to
constitute invisible heat."
It is doubtless true that Young's views did not at the time of the
publication of these lectures* command the authority which now attaches
to them. But when the undulatory theory gained acceptance, there was
no room left for the distinct entities.
J. B. Beade, one of the pioneers of photography, in a letter to B. Hunt *,
* I may remark, in panning, that Brougham knew a little of experimenting, as of everything
else, except law ! [190L The reference is to a contemporary gibe at Lord Chancellor Brougham
that -had he known a little law, he would have known a little of everything/' Young's views
were violently attacked by Brougham in the Ediubmryh Her it*.]
t Hunt's "Researches on Light," Lomfmau, 18$*, p. 374. Hunt himself, not being an
undulationist, was upon the other side.
240 ON THE HISTORY OF THE [154
of date Feb. 1854, thus speaks of Young : — " Dr Young's propositions are,
that radiant light consists in undulations of the luminiferous aether, that
light differs from heat only in the frequency of its undulations, that un-
dulations less frequent than those of light produce heat, and that undulations
more frequent than those of light produce chemical and photographic
action, — all proved by experiments."
Sir John Herschel's presentation of the matter* is not very explicit.
" The solar rays, then, possess at least three distinct powers : those of heating,
illuminating, and effecting chemical combinations or decompositions; and
these powers are distributed among the differently refrangible rays in such
a manner as to show their complete independence on each other. Later
experiments have gone a certain way to add another power to the list —
that of exciting magnetism." Although the marginal index runs " Calorific,
luminous, and chemical rays," the choice of words in the text, as well as the
reference to magnetism (for surely no one believed in a special magnetizing
entity), points to the conclusion that Herschel held the modern view.
For the decade between 1850 and 1860, the citation upon which I most
rely as indicative of the view held by the highest authorities, and by those
capable of judging where the highest authority was to be found, is from
Prof. Stokes's celebrated memoir upon Fluorescence f. On p. 465 we read : —
" Now according to the Undulatory Theory, the nature of light is defined by
two things, its period of vibration, and its state of polarization. To the
former corresponds its refrangibility, and, so far as the eye is a judge of
colour, its colour." And in a footnote here appended : —
"It has been maintained by some philosophers of the first eminence that light of
definite refrangibility might still be compound ; and though no longer decomposable by
prismatic refraction may still be so by other means. 1 am not now speaking of com-
positions and resolutions dependent upon polarization. It has been suggested by advocates
of the undulatory theory, that possibly a difference of properties in lights of the same
refrangibility might correspond to a difference in the law of vibration, and that lights of
given refrangibility may differ in tint, just as musical notes of given pitch differ in quality.
Were it not for the strong conviction I felt that light of definite refrangibility is in the
strict sense of the word homogeneous, I should probably have been led to look in this
direction for an explanation of the remarkable phenomena presented by a solution of
sulphate of quinine. It would lead rne too far from the subject of the present paper to
explain the grounds of this conviction. I will only observe that I have not overlooked the
remarkable effect of absorbing media in causing apparent changes of colour in a pure
spectrum ; but this I believe to be a subjective phenomenon depending upon contrast."
It can scarcely be necessary to insist that " light " is used here in the
wider sense, a large part of the memoir dealing with the transformation of
invisible into visible light.
* Art. Light, Enc. Met. 1830, § 1147.
t " On a Change of Eefrangibility of Light." Phil. Trans. 1852.
1889] DOCTRINE OF RADIANT ENERGY. 241
The allusion in the note is, of course, to Brewster. This distinguished
discoverer never accepted the wave-theory, and was thus insensible to the
repugnance with which his doctrine of three different kinds of luminous
radiation was regarded by every undulationist. The matter was not finally
set at rest until Helmholtz showed that Brewster's effects depended upon
errors of experiment not previously recognized.
The following, from W. Thomson*, is almost equally significant : —
"It is assumed in this communication that the undulatory theory of
radiant heat and light, according to which light is merely radiant heat, of
which the vibrations are performed in periods between certain limits of
duration, is true. ' The chemical rays ' beyond the violet end of the spectrum
consist of undulations of which the full vibrations are executed in periods
shorter than those of the extreme visible violet light, or than about the
eight hundred million millionth of a second. The periods of the vibrations
of visible light lie between this point and another, about double as great,
corresponding to the extreme visible red light. The vibrations of the obscure
radiant heat beyond the red end are executed in longer periods than this ;
the longest which has yet been experimentally tested being about the
eighty million millionth of a second."
Again, in Lloyd's "Wave Theory of Light "f, we find the following
passage: — "It appears, then, that sensibility of the eye is confined within
much narrower limits than that of the ear; the ratio of the times of the
extreme vibrations which affect the eye being only that of 1'58 to 1, which
is less than the ratio of the times of vibration of a fundamental note and its
octave. There is no reason for supposing, however, that the vibrations
themselves are confined within these limits. In fact, we know that there
are invisible rays beyond the two extremities of the spectrum, whose periods
of vibration (and lengths of wave) must fall without the limits now stated to
belong to the visible rays."
I believe that it would be not too much to say that during the decade
1850 — 1860 nearly all the leading workers in physics, with the exception of
Brewster, held the modern view of radiation. It would be quite consistent
with this that many chemists, photographers, and workers in other branches
of science, who trusted to more or less antiquated text-books for their
information, should have clung to a belief in the three entities. After I860,
and the discussions respecting the discoveries of Stewart and Kirchhoff, I
should have supposed that there were scarcely two opinions. Stewart's
Elementary Treatise on Heat was published in 1866, and was widely used
* "On the Mechanical Action of Radiant Heat or Light"; Ac. Proc. Boy. Soc. Edinb.
Feb. 1852.
t Longman*, 1857, p. 16.
B. III. 16
242 ON THE HISTORY OF THE [154
in schools and colleges. In book II. ch. II. he elaborately discusses the
whole question, summing up in favour of the view that " radiant light and
heat are only varieties of the same physical agent, and that when once the
spectrum of a luminous object has been obtained, the separation of the
different rays from one another is physically complete ; so that, if we take
any region of the visible spectrum, its illuminating and heating effects are
caused by precisely the same rays." What there was further for Draper or
any one else to say in 1872 I am at a loss to comprehend*.
To pass on to another point. I have followed the excellent advice to read
W. Herschel's original memoirs; but I must confess that the impression
produced upon my mind is different in some respects from that expressed by
Prof. Langley. It seems to me that Herschel fully established the diversity
of radiant heat. In the first memoir f a paragraph is headed " Radiant Heat
is of different Refrangibility" the question being fully discussed ; and from
the following memoir (p. 291) it is evident that this proposition extends to
invisible radiation. " The four last experiments prove that the maximum of
the heating power is vested among the invisible rays; and is probably not
less than half an inch beyond the last visible ones, when projected in the
manner before mentioned. The same experiments also show that the sun's
invisible rays, in their less refrangible state, and considerably beyond the
maximum, still exert a heating power fully equal to that of red-coloured
light — " Can it then be said of De la Roche that he, in 1811, before
anyone else, "derives the just and most important, as well as the then most
novel conception, that radiant heat is of different kinds " ? It was doubtless
a most important step when De la Roche and Melloni exhibited the diversity
of radiant heat by means of selective absorption ; but I do not see how we can
regard them as the discoverers of the fact.
It would take too long to establish by quotations, but it is pretty evident
that in his two earlier papers J Herschel leaned to the view that light was
not " essentially different from radiant heat." Why then, after laying hands
upon the truth, did he let it go, and decide that light and heat are not
occasioned by the same rays ?
"The question §, which we are discussing at present, may therefore at
once be reduced to this single point. Is the heat which has the refrangibility
of the red rays occasioned by the light of these rays ? For, should that be
the case, as there will be then only one set of rays, one fate only can attend
them, in being either transmitted or stopped, according to the power of the
glass applied to them. We are now to appeal to our prismatic experiment
* I have limited myself to citations from English writers, but I have no reason to think that
the course of opinion was different in France and Germany,
t Phil. Trans. 1800, p. 255.
J See pp. 272, 291, 292.
§ Third Memoir, p. 520.
DOCTRINE OF RADIAXT ENERGY. 243
upon the subject, which is to decide the question." The issue could not be
more plainly stated. The experiment is discussed, and this is the con-
clusion : — " Here then we have a direct and simple proof, in the case of the
red glass, that the rays of light are transmitted, while those of heat are
stopped, and that thus they have nothing in common but a certain equal
degree of refrangibility "
I am disposed to think that it was this erroneous conclusion from
experiment*, more, perhaps, than preconceived views about caloric, that
retarded progress in radiant heat for so many years. We are reminded of
Darwin's saying that a bad observation is more mischievous than unsound
theory. It would be interesting to inquire upon what grounds we now reject
the plain answer which Herschel thought himself to have received from
experiment. I do not recall a modern investigation in which the heat and
light absorptions are proved to be equal for the various parts of the visible
spectrum. Can it be that after all we have nothing but theory to oppose to
HerscheFs fects ?
I hope it will be understood that these criticisms, even if they are sound,
do not touch the substance of Prof Langley's address, which is doubly
interesting as coming from one who has done so much himself to enlarge our
knowledge of this branch of science.
* See WheweU's Hutory of the Imdmctive Sciemee*, Vol. n. p. 548 (1817).
16—2
155.
NOTE ON THE FREE VIBRATIONS OF AN INFINITELY
LONG CYLINDRICAL SHELL.
[Proceedings of the Royal Society, XLV. pp. 443—448, 1889.]
IN a recent memoir* Mr Love has considered this question among others;
but he has not discussed his result (equation (95)}, except in its application
to a rather special case involving the existence of a free edge. When the
cylinder is regarded as infinitely long, the problem is naturally of a simpler
character ; and I have thought that it might be worth while to express more
fully the frequency equation, as applicable to all vibrations, independent of
the thickness of the shell, which are periodic with respect both to the length
and the circumference of the cylinder.
In order to prevent misunderstanding, it may be well to premise that the
vibrations, whose frequency is to be determined, do not include the gravest
of which a thin shell is capable. If the middle surface be simply bent, the
potential energy of deformation is of a higher order of magnitude than in
the contrary case, and according to the present method of treatment the
frequency of vibration will appear to be zero. It is known, however, that
the only possible modes of bending of a cylindrical shell are such as are not
periodic along the length, or rather have the wave-length in this direction
infinitely longf. When the middle surface is stretched, as well as bent, the
potential energy of bending may be neglected, except in certain very special
* "On the small Free Vibrations and Deformation of a thin Elastic Shell," Phil. Trans. A,
Vol. CLXXIX. (1888), p. 491.
t "On the Bending and Vibration of thin Elastic Shells, especially of Cylindrical Form,"
Roy. Soc. Proc. supra, p. 105. [Vol. m. p. 217.]
1889] VIBRATIONS OF AN INFINITELY LONG CYLINDRICAL SHELL. 245
Taking cylindrical coordinates (r, <f>, z), and denoting the displacements
parallel to z, <f>, r by u, v, w respectively, we have for the principal elongations
and shear at any point (a, <f>, z)* —
du w 1 dv 1 du dv
0-!=^-, °"o=- + -T7> OT = ~ TT + J~ 5 (1)
dz a ad<f> a d<f> dz
and the energy per unit of area is expressed by
a + a-2a + iara+^ '^(<ri + <r*Y\, (2)
where 2h denotes the thickness of the shell, and m, n are the elastic constants
of Thomson and Tait's notation.
The functions u, v, w are to be assumed proportional to the sines, or
cosines, of /MZ and s<f>. These may be combined in various ways, but a
sufficient example is
M = U cos s<f> cos jj,z, v = V sin s<f> sin pz, w= TFcoss</> sin fj,z:...(3)
so that o-j = — fj. U cos S(j> sin p,z, (4)
<72 = ( W + s V) cos s<f> sinpz, (5)
«r = ( — sU+ p,V) sins<£ cos fj,z, (6)
unity being written for convenience in place of a. The energy per unit area
is thus
2»A Tcos* sd> sin*
+ ^ sin2 s<f> cos2 IJLZ (— s U + /* T7")2
Again, the kinetic energy per unit area is, if p be the volume density,
h .[(—}* cos? mkc 2 (dV\ ' " * ^dirV * <f> in" ^1 (8)
|_\ dt J \ dt i V dt J
In the integration of these expressions with respect to <£ and z, the mean
value of each sin2 or cos8 is ^f. We may then apply Lagrange's method. If
the type of vibration be cos pt, and pPp/n = &, the resulting equations may be
written
0, ...(9)
where ^.."Ll?.. (12)
m + ?i
* See a paper on the " Infinitesimal Bending of Surfaces of Revolution" (London Math. Soc.
Proc. Vol. xra. p. 4, Nov. 1881), and those already cited. [Vol. i. p. 551.]
t In the physical problem the range of integration for <j> is from 0 to 2w ; but mathematically
we are not confined to one revolution. We may conceive the shell to consist of several super-
posed convolutions, and then * is not limited to be a whole number.
246 NOTE ON THE FREE VIBRATIONS OF AN [155
The frequency equation is that expressing the evanescence of the deter-
minant of this triad of equations.
We will consider for a moment the simple case which arises when /* = 0,
that is, when the displacements are independent of z. The three equations
reduce to
(s2-^)^=0, ...................... ........ (13)
&2} V+2(M+I)sW=0, ............... (14)
I)sV+{2(M + !)-&} W = 0; ............ (15)
and they may be satisfied in two ways. First let V = W = 0 ; then U may
be finite, provided
s2-&2=0 .................................. (16)
The corresponding type for U is
U=coss<j> cospt, ........................... (17)
where ^2 = — a, ....................................... (18)
a being restored, as can be done at any moment by consideration of dimen-
sions. In this motion the material is sheared without extension, every
generating line of the cylinder moving along its own length. The frequency
depends upon the circumferential wave-length, and not upon the curvature of
the cylinder.
The second kind of vibrations are those in which U = 0, so that the
motion is strictly in two dimensions. The elimination of the ratio V/ W from
(14), (15) gives
*»{*»- 2 (4f +!)(!+ **)} = (), ..................... (19)
as the frequency equation. The first root is k2 = 0, indicating infinitely slow
motion. These are the flexural vibrations already referred to, and the
corresponding relation between V and W is by (14)
sV+ W = 0, ................................. (20)
giving by (4), (5), (6),
<TI = <r2 = tff = 0.
The other root of (19) gives, on restoration of a,
J«°= — ^ .... ...(22)
m + n a?p
while the relation between V and W is
- V + sW=0 ............................... (23)
It will be observed that when s is very large, the flexural vibrations tend
to become exclusively normal, and the extensional vibrations to become
1889] INFINITELY LONG CYLINDRICAL SHELL. 247
exclusively tangential, as might have been expected from the theory of plane
plates.
Returning now to the general case, the determinant of (9), (10), (11)
gives on reduction
0 ....... (24)
If fjL = 0, we have the three solutions already considered,
p = 0, V = s2, *» = 2 ( M + !)(** + 1).
If s = 0, that is, if the deformation be symmetrical about the axis, we
have
& = (**, or #[#-2(ir+l)(/i,'+l)] + 4(21f+l)/4s = 0. ...(25)
Corresponding to the first root we have U= 0, W= 0, as is readily proved
on reference to the original equations with * = 0. The vibrations are the
purely torsional ones represented by
v = sin fLZ cos pt, .............................. (26)
where pt = np?p ............ ........................ (27)
The frequency depends upon the wave-length parallel to the axis, and not
upon the radius of the cylinder.
The remaining roots of (25) correspond to motions for which Y= 0, or
which take place in planes through the axis. The general character of these
vibrations may be illustrated by the case where M is small, or the wave-length
a large multiple of the radius of the cylinder. We find approximately from
the quadratic (on restoration of a)
(J/ + 1)
The vibrations of (28) are nearly purely radial. If we suppose that
vanishes, we fell back upon
,
m + n a*p
as may be seen from (22), by putting « = 0.
On the other hand, the vibrations of (29) are nearly purely axial. In
terms of m and n,
?-&**=• ............................ (31)
p m
* This equation is given, in a slightly different form, by Love (foe. cit. p. 523).
248 VIBRATIONS OF AN INFINITELY LONG CYLINDRICAL SHELL. [155
Now, if q denote Young's modulus,
_ n (3m - ri) ,\
so that p2 = ........................................ (33)
This is the ordinary formula for the longitudinal vibrations of a rod, the fact
that the section is here a thin annulus not influencing the result to this order
of approximation.
Another extreme case worthy of notice occurs when s is very great.
Equation (24) then reduces to
^[F-/i2-s2][^-2(M + l)(/i2+s2)] = 0; ............ (34)
so that kz becomes a function of n and s only through (yu2 f s-), as might have
been expected from the theory of plane plates. The first root relates to
flexural vibrations; the second to vibrations of shearing, as in (18); the
third to vibrations involving extension of the middle surface, analogous to
those in (22).
It is scarcely necessary to add, in conclusion, that the most general
deformation of the middle surface can be expressed by means of a series
of such as are periodic with respect to 2 and </>, so that the problem con-
sidered is really the most general small motion of an infinite cylindrical shell.
Another particular case worth notice arises when s = l, so that (24)
assumes the form
k*(tf-P?-I)[k*-2(M+I)(^ + 2)] + W(k'>-^)(2M+l} = Q. ...(35)
As we have already seen, if //, be zero, one of the values of k2 vanishes. If yu,
be small, the corresponding value of fc2 is of the order ^4. Equation (35)
gives in this case
or in terms of p, q, and with restoration of a,
e-*g .................................. (37)
This agrees with the usual formula* for the transverse vibrations of rods.
* Theory of Sound, § 181.
156.
ON THE FREE VIBRATIONS OF AN INFINITE PLATE OF
HOMOGENEOUS ISOTROPIC ELASTIC MATTER.
[Proceedings of the London Mathematical Society, xx. pp. 225—234, 1889.]
THE solid here contemplated is that bounded by two infinite planes
parallel to xy ; and the vibrations are supposed to be periodic, not only with
respect to the time (eipt), but also with respect to x and y. The results, so
far as thin plates are concerned, have long been known ; but the method may
not be without interest in view of the difficulties which beset the rigorous
treatment of the theory of thin plates, and of the fact that it is not limited to
the case of small thickness. A former investigation*, " On Waves propagated
along the Plane Surface of an Elastic Solid," may be regarded as a particular
case of that now before us.
In conformity with the suppositions as to periodicity, we might assume
that all the functions concerned involve x and y only through the factors
(&*, (fr*. But, by a rotation of the axes, eit^x+ffy} may be replaced by e'S*
without loss of generality, and it will considerably simplify our equations if
we limit them to the latter form. Any function of x, y (e.g., the dilatation)
may be expanded in a series of such terms as cosfx cos gy, and this may be
resolved into two of the form
cos (fx + gy), cos (fx - gy).
But between these forms there is no essential difference, for on account of
the symmetry of the plane we shall have to deal in either case only with
V (f* + g*). The assumption of proportionab'ty with «#• is not, however,
equivalent to a limitation of the problem to two dimensions, as might at
first be supposed ; inasmuch as J3, the displacement parallel to y, is allowed
to remain finite.
* Proe. Lo*d. Math. Soc. VoL ivn. Nov. 1885. [Vol. n. p. 441.]
250 ON THE FREE VIBRATIONS OF AN INFINITE PLATE OF [156
If 0 be the dilatation, the usual equations are
and w, n denote the elastic constants of the material according to Thomson
and Tait's notation*.
If a, /9, 7 all vary as eipt, equations (1) become
7 j\
-r- 0, &c ...................... (3)
Differentiating equations (3) in order with respect to x, y, z, and adding, we
get
(V2 + A2)0 = 0, ................................. (4)
in which A2 = pp>/(m + n) .................................. (5)
Again, if we put k2 — pp2(n, .................................... (6)
equations (3) take the form
A particular solution of (7) isf
id0 ide idd
~h*dx> r""*^ 7=~A2^; ......... (8)
in order to complete which it is only necessary to add complementary terms
u, v, w satisfying the equations
(V2 + &2)«;=0, ......... (9)
(10)
dx dy dz
According to our present suppositions, x and y are involved only through tffx,
that is, y is not involved at all. Thus
d0/dy = Q, dv/dy=0.
The displacement (3 is thus identical with v, and satisfies the differential
equation
(V2-f&2)/3 = 0 ............................... (11)
Again, in virtue of (9) and (10), we may write
(12)
* Lamp's constants X, p are related to in, n according to \ + /* = m, /x = n.
t Lamb, " On the Vibrations of an Elastic Sphere," Math. Soc. Proc. May 1882.
1889] HOMOGENEOUS ISOTROPIC ELASTIC MATTER. 251
where ^ is a function of x and z, which satisfies
(V* + **)X = 0; .............................. (13)
1 dO d 1 dO d
We have not vet made use of the supposition that x occurs only in the factor
eP*. Under this condition we get from (4)
6 = P cosh rz + Q sinh rz, ........................ (15)
where r= = r--h*; .............................. (16)
and from (13), (11), * = 4 sinh *z + £ cosh *s, ........................ (17)
£= Ccosh sz + Dsunhsz, ........................ (18)
where «•=/»- If. .............................. (19)
The arbitrary quantities P, Q, A, B, C, D may be supposed to include the
factors &&, e*Sx, but are otherwise constants.
The evanescence of the three component stresses at the two bounding
surfaces gives in all six equations. The components of tangential stress are,
in general, proportional to
d0 + <fy Jy + fa
dz dy ' dx dz '
As regards the first of these, we have at present dy'dy = Q; so that the
condition to be satisfied at each surface is simply
dft/dz=0 .................................. (20)
The evanescence of the second tangential stress gives
d& d*
These equations are to hold good at both surfaces. If we take the origin at
the middle of the thickness, the bounding surfaces may be represented by
2=±^i; and equations (20), (21) must be satisfied by the odd and even
functions separately. Thus, from (18), (20),
CsaDhszl = Q, Dcoshs*, = 0 ................... (22)
a pair of equations which may be satisfied in two ways. We may suppose
D = 0, so that
0 = Cco8hsz, .............................. (23)
in conjunction with sinhszj = 0; .............................. (24)
or, on the other hand, @ = Dsinhsz, .............................. (25)
under the condition cosh*^, = 0 ............................
* Green, Camb. Trail*. 1837 ; Reprint of Green's Works, p. 261.
252 ON THE FREE VIBRATIONS OF AN INFINITE PLATE OF [156
During these vibrations the solid is simply sheared. In the vibrations of the
first class represented by (23), ft is an even function of z, a. and 7 vanishing.
In the vibrations of the second class, ft is an odd function of 2, and therefore
vanishes at the middle surface. The roots of (24) are
szl = iq-jr,
where q is an integer; so that, by (19),
# =/» + gV/*!2, ........................... (27)
and the stationary vibrations are of the type
ft=cospt cos/a cos^-, ..................... (28)
p being given by (6) and (27).
In like manner, for the vibrations of the second class,
ft = cospt cos/# sin , .................. (29)
z\
where ]f=f^ + ^L±&^. ............................ (30)
In (28), (29), we may of course replace cospt, or cosfx, by sinpt, or sin fa,
respectively*.
The kind of vibrations just considered are those for which ft is finite,
while a and 7 vanish. In the second kind of vibrations, ft vanishes, so that
the motion is strictly in two dimensions. There are four boundary conditions
to be satisfied, two derived from (21), and two expressive of the evanescence
of the normal stress. The latter condition is that
(m-n)0 + Zndy/dz = 0,
when z = ± zl ; or, in terms of Jc2 and hz,
(k*-2h*)e + 2h?dj{dz = 0 ......................... (31)
Substituting from (14), (15), (17), in (21), (31), we obtain, with use of
(16), (19),
Zifrhr* (P sinh rz + Q cosh rz} + (It? - 2/2) (A sinh sz + B cosh sz) = 0, . . .(32)
(&2 - 2/2) (P cosh rz + Q sinh rz) - 2h'ifs (A cosh sz + B sinh sz) = 0. . . .(33)
* In the present investigation the section of the solid perpendicular to y is an infinitely
elongated rectangle. It may be worth notice that the corresponding solutions (in which every
linear element parallel to the axis moves as a rigid body along its own length) may readily be
obtained for cylinders of other sections, e.g., the finite rectangle and the circle. There is
complete mathematical analogy with the vibrations of a stretched membrane having the form of
the section of the cylinder, under the condition that the boundary is free to move perpendicularly
to the plane of the membrane. (Theory of Sound, § 227.)
1889] HOMOGENEOUS ISOTBOFIC ELASTIC MATTER. 253
These equations are to hold when z = ± *,, and most therefore be true for the
odd and even parts separately. Thus
0, ............ (34)
0; ............ (35)
Zifrtr* Qcosh rzt + (& - 2/s)l? cosh sz^ = 0, ............ (36)
(1J - 2/») Q sinh rz, - 2Asi>* 5sinh **, = 0. ............ (37)
It will be seen that in these equations the constants P. A are separated from
Q, B. The system can therefore be satisfied in two distinct ways. For the
first class of vibrations Q =0, 5 = 0. Equations (36), (37) are thus disposed
of: while the first pair serve to determine the ratio P-.A. and in addition
impose a relation between the other quantities. Equations (14) show that 0
and a are even functions of z, but that 7 is an odd function. In this case
of vibrations, therefore, the middle surface remains plane, but undergoes
extension.
The frequency equation is found by elimination of P:A between (34)
(35):-
4/»r» sinh rz, cosh **, = (k* - 2/2)* cosh rz^ sinh **, ;
or, as it may be written,
4/*r» tanh rz^ = (**- 2/s)» tanh az, ................... (38)
Again, from (35),
(Jf - 2/») cosh rz* *
so that the type of vibration is, by (14),
a = «*V/* |2»fa cosh «, cosh r* + « (# - 2/s) cosh rz, cosh «} ....... (39)
7 = - j*jf* (2t/r* cosh ««, sinh rz + if (If - 2f*) cosh rz, sinh «}. ...(40)
We may apply these results to the case where the plate is thin, so that/ij
is smalL If rz,, sz,, in (38), be small, we find
- A*),
This equation determines L*, since the ratio A*/4? depends only upon the
elastic quality of the material In terms of m and H, from (5) and (6),
254 ON THE FREE VIBRATIONS OF AN INFINITE . PLATE OF [156
At the same time, (39), (40) give approximately
a = &WP« e**, 7 = - ifsz (& - 2A2
or, if we throw out the common factor T<?s,
a=eiPteifX} 7=-- '-ifz&W* ................... (44)
This gives the same relation between the principal strains as is obtained in
the ordinary theory of thin plates*, viz.,
dy _ _ m — n /da. dfi\
dz~ m + n \dx dy)'
A complete discussion of (38) would lead rather far, but we may easily find a
second approximation in which the square of zl is included. Thus, since
tanh rzt = rz^ (1 - ^r2^2 +...),
or 4/2r2 { 1 - ^ (r2 - s2)} = (k2 - 2/2)2 ;
whence, on substitution of the values of r2 and s2 from (16), (19),
/l2){l-^12(/2-^)} ................... (45)
From the first approximation we know that r2, or f2 — h2, is positive. Hence
A? diminishes with z?t or the pitch falls as the thickness increases. An
exception occurs when r2 = 0 ; but this can happen only when k2 = 2/2 = 2h2,
or the material is such that m = n. If the character of the material be of
this description, k2 = 2/2 satisfies (38), whatever may be the value of z^
Each lamina parallel to xy vibrates unconstrained by its neighbours, and
7 = 0 throughout.
If the material be incompressible, h? = 0, and (45) assumes the simplified
form
(46)
In any of these equations, if we suppose that the functions vary as
well as e^x, the generalized result is obtained by merely writing (
for/2.
We now pass on to consider the second class of vibrations, for which, in
(34), &c., P = 0, A = 0. Here 0 and a are odd functions of z1} while 7 is an
even function, so that the middle surface is bent without extension. As
regards the equations (36), (37), which involve Q and B, it will be seen that
* See, for example, Proc. Roy. Soc. Dec. 1888. [Vol. m. p. 222.]
1889] HOMOGENEOUS ISOTROPIC ELASTIC MATTER. 255
they differ from the first pair of equations involving P and A merely by the
interchange everywhere of cosh and sinh. We have, therefore, in place
of (38),
4/»r* coth rz, = (k2 - 2/2)2 coth sz, ; .................. (47)
and in place of (39), (40),
a = && eV* {2sf* sinh szl sinh rz + s (k2 - 2/i) sinh rzl sinh sz} , ...... (48)
7= - eipt <,</* fiifrs sinh sz, cosh rz + if(k2 - 2/2) sinh rzl cosh sz}. . . .(49)
If we now introduce the assumption that the plate is thin, we find, by
expanding the hyperbolic functions in (47),
4/2 (/» -*•){! + K (A? - A")} = (If - 2/7.
The first approximation gives A? = 0, signifying that the notes are
infinitely grave. The second approximation is
fr = K/4 (**-**). ........................... (50)
or, in terms of p, m, n, p,
mn 4/V _.
V-m^n,-^-'
Again, if we drop out a common factor (k^rz,), (48), (49) take the forms
a=f*zeiPteV*, y=ifeiPte^x ...................... (52)
Hence a = — zdy/da;, signifying that to this order of approximation every line
originally perpendicular to the middle surface retains its straightness and
perpendicularity during the vibrations.
The third approximation to the value of A;2 from (47) gives
......... <53>
so that, when the thickness is increased beyond a certain point, the rise of
pitch begins to be less rapid than according to the second approximation (51).
When zl is infinitely great, we get, from (38) or (47),
4/2rs=(&2-2/2)2, ........................... (54)*
the equation considered in the paper, already referred to, upon surface-waves.
From (43), (53) we learn that p2 is positive, or the equilibrium is stable,
so long as m is positive. On the other hand, it was proved by Green many
years ago that a solid body would be unstable if m were less than ^n, m — ^n
being in fact the dilatation modulus. The reconciliation of these apparently
contradictory results depends upon principles similar to those recently applied
* This is upon the supposition that r and « are real. In the contrary case the equation
would have no definite limit.
256 ON THE FREE VIBRATIONS OF AN INFINITE PLATE OF [156
by Sir W. Thomson*, to show that a solid, every part of the boundary of
which is held fixed, is stable, so long as m is greater than — n, and this in
spite of the fact that, if the boundary were freed, the solid would at once
collapse or expand indefinitely. In the present case of an infinite slab, the
assumption that the displacements are periodic with respect to x and y is
tantamount to the imposition of a constraint at infinity, rendering stability
possible under circumstances which would otherwise lead to indefinite collapse
or expansion of the medium.
The general expression for the energy of a strained isotropic solid is f
2w = (m + n) (ea + f2 + g2) + 2 (m - w) (fg + ge 4- ef ) + n (a2 + 62 + c2), . . .(55)
e, f, g being the principal extensions; a, b, c the shears, relatively to the
coordinate axes. Since e, f, g may vanish, it is clear that the stability of the
medium requires that n be positive ; and again, since a, b, c may all vanish,
the terms in e, f, g must of themselves be positive in all cases that may arise.
Thus, leaving out a, b, c, we write
2w -- (3m - n) (e2 + f2 + g2) + (n - m) {(e - f)2 + (f- g)2 + (g - e)2}, . . .(56)
from which it follows that, if n > m > \n, the equilibrium is stable. If,
however, m < %n, it will be possible to make w negative by taking e = f = g.
If m > n, the equilibrium is stable, as may be seen by writing 2w in the form
2w = (?rc-rO(e + f+g)2+2rc(e2 + f2 + g2) ............. (57)
Hence, if there be no limitation on the strains, the necessary and sufficient
conditions of stability are that n should be positive and m greater than £n.
But now suppose that the strains are limited to be in two dimensions, so
that (for example) g = 0. The supposition e = f=g is then not admissible,
and the criterion of stability is altered. We have
2w = (m + n) (e2 + f2) + 2 (m - n) ef
= (n-m)(e-f)2+2m(e2+f2) ................... (58)
This shows that there is stability if m be positive and less than n, and
instability if m be negative. That the equilibrium is stable if m be greater
than n is shown, as in (57), by putting 2w into the form
(e2 + f2). . ................. (59)
Hence, under the limitation g = 0, the necessary and sufficient conditions of
stability are that n and m be positive.
Comparing the results, we see that, as m diminishes, instability sets in
when m = %n, if the boundary be free ; when m = 0, if (as virtually in our
* Phil. Mag. Nov. 1888.
t Thomson and Tait's Natural Philosophy, § 695.
1889] HOMOGENEOUS ISOTROPIC ELASTIC MATTER. 257
present problem) the strains be limited to two dimensions ; when m = — n, if
the boundary be everywhere held fast.
I have endeavoured to investigate the two-dimensional free vibrations of
an infinitely long cylindrical shell directly from the fundamental equations,
as in the foregoing theory of the plane plate. The preliminary analysis is
simple, and there is no difficulty in obtaining the solutions analogous to (42).
If a be the radius of the cylinder, and the wave-length measured round the
circumference be 2-jr/f, we have
<60>
and a jnn_
pa2 m + n
But this solution is much more readily obtained by the special methods
applicable to thin plates, as to the legitimacy of which for this purpose there
can be no question. And if, in order to investigate the flexural vibrations of
the shell, we retain the lower powers of the thickness, the reduction of the
resulting determinant becomes a very complicated affair. I have not
succeeded in verifying by a rigorous application of this method the equation
analogous to (51), viz. :
mn 4/W (/*<*'-!)* (62)
P m + n Spa1 /2as + l '
2zl being the thickness, and as before fa the number of wave-lengths in the
circumference. Putting a = oo , we fall back, of course, upon the formulae for
the plane plate.
17
157.
ON THE LIMIT TO INTERFERENCE WHEN LIGHT IS
RADIATED FROM MOVING MOLECULES.
[Philosophical Magazine, xxvn. pp. 298—304, 1889.]
IN a recent number of Wiedemanns Annalen, Ebert* discusses the
application of Doppler's principle to the radiation from the moving molecules
of an incandescent gasf, and arrives at the conclusion that the widths of the
spectral lines, as calculated upon the basis of the principle, are much greater
than is consistent with experiments upon interference with a large relative
retardation. This is a matter of no small importance. Unless the discrepancy
can be explained, the dynamical theory of gases would, it appears to me, have
received a heavy blow from which it could with difficulty recover. If it be
true that a gas consists of molecules in irregular motion, and that for the
most part each molecule radiates independently, there seems no escape from
the conclusion that the character of the aggregate radiation must be governed
by Doppler's principle.
If v be the velocity of a molecule, 6 the inclination of its motion to the
line of sight, the natural wave-frequency N is changed by the motion into n,
where
(1)
and V is the velocity of light. If A, A. be the original and altered
wave-lengths, so that
A=F/#, \=V/n; ........................ (2)
then
x = A Y+ v cos 0 = A (* ~ VCOS 0) aPProximately> ...... (3)
when v/Vis small.
* Wied. Ann. MXVI. p. 466 (1889).
t Lippich, Pogg. Ann. cxxxix. p. 465 (1870). Bayleigh, Nature, vni. p. 474 (1873)
[Vol. i. p. 183].
OST THE LIMIT TO DTTERFEREXCE ETC. 259
As a first approximation,, Ebert supposes that the velocity e of every
molecule is the same. In this case the spectral hand, into which what
would otherwise be a mathematical line is dilated, has the limiting wave-
(4)
and between these limits is of uniform brightness. For the number of
molecules whose lines of motion lie between 0 and 6 + dff is proportional
to sin 0 d0F and this again bj (1) is proportional to da. It is here assumed
that the spectrum is formed upon a scale of wave-frequencies; but for the
present purpose the range concerned is so small that it becomes a matter of
indifference upon what principle the spectrum is disposed.
The typical case of interference arises when two streams of homogeneous
light are superposed, which differ in nothing but phase. If 8 denote this
difference of phase, the vibrations may be represented by cos ^r + cos (^fr -I- 5),
or by
and the intensity is
(6)
If the two streams are obtained by reflexion at the opposite feees of a parallel
plate, the circumstances are somewhat more complicated. But the simple
theory is applicable even here as a first approximation, which becomes more
and more rigorous as the difference of optical quality between the plate and
the medium in contact with it is supposed to diminish. If /* be the index of
the plate, A its thickness,
..,7)
If the plate be of air, ji= 1- In any case the variation of p, is small compared
to that of n; so that if A denote the equivalent thickness of air, we may take
/ = 4sina{2»iiA/F], ........................... (8)
a function of n — the frequency, as well as of A and V.
If now the light be heterogeneous, we have nothing further to do than to
integrate (8) with respect to n, after introduction of a factor i such that t dn
represents the illumination corresponding to d*,*. In the present case, where
the intensity is supposed to be uniform within limits if, and n~, and to vanish
outside them, we have
fid* = 4i f%m*(2«A/F) d*
J-.
It K here assumed thai the range included u too small to give rise to sensible chro
17—2
260 ON THE LIMIT TO INTERFERENCE [157
From this we fall back on (8), if we suppose that (n2 - n^ is infinitely small,
so that
// dn = Zfidn . [1 - cos (47rnA/F)].
The difference between (8) and (9) thus depends upon the factor
sin {27rA(n,-%)/F]
-^(n.-O/F-
which multiplies the second term of (9). If we introduce the special values
of HI, HZ from (4), and denote the angle in (10) by a,
(11)
So long as a is small, the mode of interference is nearly the same as if v = 0.
This will be the case when A is sufficiently small, so that at first the bands
are absolutely black. As A increases, the distinctness of the bands will
depend mainly upon the relative brightnesses of the least and most illumi-
nated parts. If we call this ratio h, and denote by a the numerical value
of (10), we have
A = (1 -»)/(!+ a), ........................... (12)
o = (l-A)/(l + fc) ............................ (13)
Now from (10) it appears that when a is equal to TT, or to any multiple of
TT, a = 0, and the field is absolutely uniform. Between values of a equal to TT
and 2?r, 2?r and STT, and so on, there are revivals of distinctness, the maxima
of which occur at values not far removed from |TT, \ TT, &c. Thus, between
TT and 2?r there is to be found a value of a at least equal to 2/37T, corresponding
to h = § nearly. At this stage the bands should certainly be visible.
In order to estimate at what point the interference-bands would first
disappear as A increases, we must make some supposition as to the largest
value of h indistinguishable in experiment from unity. Under favourable
circumstances in other respects we may perhaps assume for this purpose
h = '95, so that a = '025. Since a is small, a is nearly equal to TT. We may
take approximately sina = '0257r, or a = '975?r. In fact, so long as we take
h nearly equal to unity, the precise value makes very little difference to the
corresponding value of a, and for the purposes of such a discussion as the
present we may suppose with sufficient accuracy a = TT. In this case, by (11),
which gives the retardation (2A) measured in wave-lengths in the neighbour-
hood of which the bands would first disappear. This estimate differs widely
from that put forward by Ebert. The latter is equivalent to
-* ..................................
1889] WHEN LIGHT IS RADIATED FROM MOVDfG MOLECULES. 2«;l
According to my calculation the value of a corresponding to (15) would be
54°, a would be '86, and h would be -075 ; so that the bands should be hardly
distinguishable from those which occur when A = 0.
For the grounds of his estimate Ebert refers to an earlier paper*, in
which, however, the calculation seems to relate to a problem materially
different from the present, that, namely, in which the refrangibility of the
light is limited to two distinct values (as approximately in the case of the
soda lines), instead of being distributed equally over the same range. In this
case (9) is replaced by
4- 1-
so that, if o have the same form as in (11), and a' denote the numerical value
of cos o,
as before.
According to (16) the field is first uniform when a = ±ir, instead of TT. as
from (9). When a = -JT, the bands are again black, and as A fun her increases
there is a strictly periodic alternation between blackness and absolute dis-
appearance of the bands.
The substitution for a spectral band of uniform brightness of one in which
the illumination is all condensed at the edges erplains a large part of the
discrepancy between (14) and (15): but even in the latter problem (15)
seems to be a very small estimate of A. According to (15), a = 54:,
cos a = "59 ; so that from (17) h = '26. Bands of which the darkest parts
are of only one quarter of the illumination of the brightest parts could hardly
be invisible.
The more nearly correct formula (14) is itself, however, based upon the
assumption that all the vibrating molecules move with the same velocity.
This is the origin of the law expressed in (9), according to which the bands
should reappear at a retardation greater than that of first disappearance.
But the real law of the distribution of velocity is that discovered by Maxwell,
if there is any truth in the molecular theory t. That such is the case is
recognized by Ebert : and he argues that the broadening of the spectral band
due to velocities higher than the mean, will entail a further diminution in
the maximum retardation consistent with visible interference*. I proceed
* Witd. Amu. miT. p. 39 (1888).
t It is here assumed that we are dealing with a gas in approximate temperature equilibrium.
The ease of luminosity under electric discharge maj require further consideration.
* In the earlier memoir (Wi*d. Amu. mrr.) Ebert appears to regard the capability of inter-
ference (ImUiferemz-JSlugktit) of a ^pectral line as dependent upon other causes than the width
of the line and the distribution of brightness OTCT it. In this riew I cannot agree. •
narrowness of the bright Hue of light seen in the spectroscope, and the possibility of a large
262 ON THE LIMIT TO INTERFERENCE [157
to the actual calculation of the maximum retardation on the basis of
Maxwell's law.
If f> n> £ be the rectangular components of v, the number of molecules
whose component velocities lie at any time between £ and £ + dg, i) and
vj + dt), £ and £ + d£, will be proportional to
If f be the direction of the line of sight, the component velocities tj, % are
without influence in the present problem. All that we require to know is
that the number of molecules for which the component f lies between £ and
£ + d% is proportional to
<r**d1> (18)
The relation of /3 to the mean (resultant) velocity v* is
v=^ft (19)
If the natural frequency of the waves emitted by the molecules be N, the
actual frequency of the waves from a molecule travelling with component
velocity f is by Doppler's principle
n=N(l+?/V) (20)
Hence by (8) the expression to be investigated, and corresponding to (9), is
In (21) we have
-tirA
The last of the three terms, being of uneven order in f, vanishes when
integrated. The first and second are included under the well-known formula
f
Jo
o 2a
and we obtain
(22>
number of (interference) bands, depend npon precisely the same conditions; the one is in truth
as much an interference phenomenon as the other" (Enc. Brit. "Wave Theory," Vol. xxiv. p. 425
[Vol. in. p. 60]j. It is obvious that nothing could give rise in the spectroscope to a mathematical
line of light, but an infinite train of waves of harmonic type and of absolute regularity.
* This must be distinguished from the velocity of mean square, with which the pressure is
most directly connected.
1889] WHEN LIGHT IS RADIATED FROM MOVING MOLECULES. 263
In conformity with previous notation we may write
/ 4-7T2A2\
rexP(-£A^F>J;
or, if we introduce the value of ft from (19),
<23>
The ratio of the least and greatest brightnesses is then, as before,
A = (l-a")/(l+a") ............................ (24)
If we now assume as determining the limit of visibility h = '95, we find
a" = -025, and from (23)
^ = •690- ...(25)
J\ v
It appears therefore that the maximum admissible retardation is sensibly
greater than that calculated (14) upon the supposition that all the molecules
move with the mean velocity v, and as much as 4i times greater than that
(15) taken by Ebert as the basis of his comparison with observation.
Under these circumstances it would seem that there is no discrepancy
remaining to be explained. It is true that the width of spectral lines is not
wholly due to movement of the molecules ; but it is possible that this is the
principal cause of dilatation when the flames are coloured by the spray of
very dilute solutions, as in Ebert 's use of the method of Gouy *. Again, it is
true that interference-bands are often observed under conditions less favourable
than is supposed in the above estimate of h. In Michelson's method, however,
the bands may be very black at small retardations ; and it seems very probable
that at higher retardations bands involving even less than 5 per cent, of the
brightness might be visible f. The question is one of very great interest, and
I hope that Herr Ebert will pursue his investigations until it is thoroughly
cleared up.
* Ann. de Chim. xvra. p. 1 (1879).
t [See Art. 161 below.]
158.
IRIDESCENT CRYSTALS.
[Proc. Roy. Inst. xn. pp. 447—449, 1889 ; Nature, XL. pp. 227, 228, 1889.]
THE principal subject of the lecture is the peculiar coloured reflection
observed in certain specimens of chlorate of potash. Reflection implies a
high degree of discontinuity. In some cases, as in decomposed glass, and
probably in opals, the discontinuity is due to the interposition of layers of
air; but, as was proved by Stokes, in the case of chlorate crystals the
discontinuity is that known as twinning. The seat of the colour is a very
thin layer situated in the interior of the crystal and parallel to its faces.
The following laws were discovered by Stokes : —
(1) If one of the crystalline plates be turned round in its own plane,
without alteration of the angle of incidence, the peculiar reflection vanishes
twice in a revolution, viz. when the plane of incidence coincides with the
plane of symmetry of the crystal. [Shown.]
(2) As the angle of incidence is increased the reflected light becomes
brighter and rises in refrangibility. [Shown.]
(3) The colours are not due to absorption, the transmitted light being
strictly complementary to the reflected.
(4) The coloured light is not polarised. It is produced indifferently,
whether the incident light be common light or light polarised in any plane,
and is seen whether the reflected light be viewed directly or through a
Nicol's prism turned in any way. [Shown.]
(5) The spectrum of the reflected light is frequently found to consist
almost entirely of a comparatively narrow band. When the angle of incidence
is increased, the band moves in the direction of increasing refrangibility, and
at the same time increases rapidly in width. In many cases the reflection
appears to be almost total.
1889]
IRIDESCENT CRYSTALS.
265
In order to project these phenomena a crystal is prepared by cementing a
smooth face to a strip of glass, whose sides are not quite parallel. The white
reflection from the anterior face of the glass can then be separated from the
real subject of the experiment.
A very remarkable feature in the reflected light remains to be noticed.
If the angle of incidence be small, and if the incident light be polarised in or
perpendicularly to the plane of incidence, the reflected light is polarised in
the opposite manner. [Shown.]
Similar phenomena, except that the reflection is white, are exhibited by
crystals prepared in a manner described by Madan. If the crystal be heated
beyond a certain point the peculiar reflection disappears, but returns upon
cooling. [Shown.]
In all these cases there can be little doubt that the reflection takes place
at twin surfaces, the theory of such reflection* reproducing with remarkable
exactness most of the features above described. In order to explain the
vigour and purity of the colour reflected in certain crystals, it is necessary to
suppose that there are a considerable number of twin surfaces disposed at
approximate equal intervals. At each angle of incidence there would be a
particular wave-length for which the phases of the several reflections are in
agreement. The selection of light of a particular wave-length would thus
take place upon the same principle as in diffraction spectra, and might reach
a high degree of perfection.
BIRO CALL
Fig 1.
GENERAL SCHEME
Fig 2.
DETAIL Of LAZY-TON QS
NSITIVE FLAME
In illustration of this explanation an acoustical analogue is exhibited.
The successive twin planes are imitated by parallel and equidistant discs of
muslin (Figs. 1 and 2) stretched upon brass rings and mounted (with the aid
of three lazy-tongs arrangements), so that there is but one degree of freedom
* Phil. Mag. Sept. 1888. [Vol. in. Art. 149.]
266 IRIDESCENT CRYSTALS. [158
to move, and that of such a character as to vary the interval between the
discs without disturbing their equidistance and parallelism.
The source of sound is a bird-call, giving a pure tone of high pitch
(inaudible), and the percipient is a high pressure flame issuing from a
burner so oriented that the direct waves are without influence upon the
flame*. But the waves reflected from the muslin arrive in the effective
direction, and if of sufficient intensity induce flaring. The experiment
consists in showing that the action depends upon the distance between the
discs. If the distance be such that the waves reflected from the several discs
co-operate f, the flame flares, but for intermediate adjustments recovers its
equilibrium. For full success it is necessary that the reflective power of
a single disc be neither too great nor too small. A somewhat open fabric
appears suitable.
It was shown by Brewster that certain natural specimens of Iceland spar
are traversed by thin twin strata. A convergent beam, reflected at a nearly
grazing incidence from the twin planes, depicts upon the screen an arc of
light, which is interrupted by a dark spot corresponding to the plane of
symmetry. [Shown.] A similar experiment may be made with small rhombs
in which twin layers have been developed by mechanical force after the
manner of Reusch.
The light reflected from fiery opals has been shown by Crookes to possess
in many cases a high degree of purity, rivalling in this respect the reflection
from chlorate of potash. The explanation is to be sought in a periodic
stratified structure. But the other features differ widely in the two cases.
There is here no semicircular evanescence, as the specimen is rotated in
azimuth. On the contrary, the coloured light transmitted perpendicularly
through a thin plate of opal undergoes no change when the gem is turned
round in its own plane. This appears to prove that the alternate states are
not related to one another as twin crystals. More probably the alternate
strata are of air, as in decomposed glass. The brilliancy of opals is said to be
readily affected by atmospheric conditions.
* See Proc. Roy. Inst. Jan. 1888. [Vol. in. p. 28.]
t If the reflection were perpendicular, the interval between successive discs would be equal to
the half wave-length, or to some multiple of this.
159.
THE SAILING FLIGHT OF THE ALBATROSS.
[Nature, XL. p. 34, 1889.]
I HAVE been much interested by the letter of Mr A. C. Baines * upon this
subject. In the year 1883 ("The Soaring of Birds," Nature, Vol. xxvn. p. 534
[Vol. II. p. 194]) I suggested that the explanation of these puzzling per-
formances might be found in the increase of wind with height. To take
advantage of this, the bird must rise against the wind and fall with it ; but
at the time referred to, I had before me only the observations of Mr Peal, in
Assam, on the flight of pelicans, in which this feature is not alluded to. In
Mr Baines's observations the omission is supplied, and there seems little
reason to doubt that the true explanation of the flight of the albatross has
been arrived at. In the case of the pelican soaring to a great elevation,
it is less easy to understand how the differences of horizontal velocity can be
sufficient.
Reference may be made to a paper by Mr H. Airy (Nature, Vol. xxvu.
p. 590), in which the matter is further discussed. Similar views have also
been put forward more recently by an American Author, whose name I have
unfortunately forgotten. [1901. See further the Wilde Lecture on the
Mechanical Principles of Flight (Manchester Proceedings, 1900).]
* Nature, May 2, p. 9. ["The birds go through a series of movements which are related to the
direction of the wind. Starting from near the surface, they rise in a slanting direction against
the wind, to a height which varies with the direction and strength of the wind. The average
seemed to me about 20 feet. Then comes immediately a turn half round in a rather large circle,
followed at once by a rapid descent down the wind. They then take a longer or shorter flight in
various directions, almost touching the water. After that another ascent in the same manner,
and so on, repeating this series of movements ad libitum."]
160.
ON THE CHARACTER OF THE COMPLETE RADIATION AT A
GIVEN TEMPERATURE.
[Philosophical Magazine, xxvu. pp. 460 — 469, 1889.]
BY complete radiation is here meant the radiation which would ultimately
establish itself in an enclosure, whose walls are impervious, and are main-
tained at a uniform temperature*. It was proved by Stewart and Kirchhoff
that this radiation is definite, not only in the aggregate, but also in its
various parts; so that the energy radiated with wave-frequencies between
n and n + dn may be expressed by
F(n)dn, .................................... (1)
where, for a given temperature, F(n) is a definite function of n. The
reservation implied in the word ultimately is necessary in order to exclude
radiation due to phosphorescence or to chemical action within the enclosure.
The radiation commonly characterised, so far at any rate as its visible
elements are concerned, by the term white, is supposed to be approximately
similar to the complete radiation at a certain very high temperature.
As remarked by Kirchhoff, the function F, being independent of the
properties of any particular kind of matter, is likely to be of a simple form ;
and speculations have naturally not been wanting. Within the last two
years the subject has been considered by W. Michelsonf and by H. F. Weber J.
The former, on the basis of an d priori argument of a not very convincing
character, arrives at the conclusion that at temperature 0 the radiation
between the limits of wave-length \ and X + d\. may be expressed
(2)
* [1901. The radiation, here characterised as complete, is sometimes described as black.
To speak of a red-hot poker, or of the radiation from it, as black, does not seem happy.]
t Journal de Physique, t. vi. Oct. 1887; Phil. Mag. xxv. p. 425.
I Berlin. Sitz.-Ber. 1888.
1889] CHARACTER OF COMPLETE RADIATION. 2' '
According to Stephan the total radiation is proportional to 0*. In conformity
with this Michelson supposes that
so that (2) assumes the more special form
7A = W^X-*. .............................. (3)
If. as appears to be preferable, we take n as independent variable, F (*) dm is
of the form
Ae-******, ................................. (4)
A , a being functions of 0, but independent of n.
Weber's formula, so for as it here concerns us. is of a still simpler
character. Expressed in terms of », it differs from (4) merely by the
omission of the (actor it4, thus corresponding to p = — 1 in (2): so that
f(*)*i = Ar+*d* ............................ (5)
The agreement between (5) and the measurements by Langley of the
radiation at 178* C. is considered by Weber to be sufficient Iv good.
In contemplating such a formula as (5), it is impossible to refrain from
asking in what sense we must interpret it in accordance with the principles
of the Undulatory Theory, and whether we can form any distinct conception
of the character of the vibration indicated by it. My object in the present
paper is to offer some tentative suggestions towards the elucidation of these
questions.
The first remark that I would make is that the formula must not be
taken too literally. If there is one thing more certain than another, it is
that a definite wave-frequency implies an infinite and unbroken succession of
waves*. A good illustration is afforded by intermittent vibrations, as when a
sound, itself constituting a pure tone, is heard through a channel which is
periodically opened and closed. Such an intermittent vibration may be
represented byx
2(l+cos2™0cos2vn*, ........................ (6)
where it is the frequency of the original vibration, and m the frequency
of intermittence. By ordinary trigonometrical transformation (6) may be
written
2cos2Tii* + cos2v(ii + m)f + cos2*-(*-m)t: ............ (7)
which shows that in this case the intermittent vibration is equivalent to
three simple vibrations of frequencies n, n+m, n — m.
* " The pitch of a mioroas bodr vibrating freety cannot be defined with anj greater doaweai
than corresponds with the total number of vibrations which it is capable of executing." (Pnc.
Mmt. Auoc. Dee. 1878, p. 25.)
t "A«MBtiealODBeristioiis,m.'*P*i7.Jr^. April 1880. [YoL L p. 468.]
270 ON THE CHARACTER OF THE [160
In order to distinguish wave-frequencies, whose difference is small, a
correspondingly long series of waves is necessary; and of no finite train of
irregular vibrations can it be said that waves of a certain frequency are
present, and waves of a frequency infinitely little different therefrom absent.
Neither can the proportions in which the two are present be assigned. In
professing to assign these proportions, (5) and similar formulae make assertions
not directly supported by experiment. In a sense all the formulae of mathe-
matical physics are in this predicament ; but here the assertion is of such a
nature that it could not be tested otherwise than by experiments prolonged
over all time.
In practice it is not time that brings the limitation, but the resolving
power of our instruments. In gratings the resolving power is measured by
the product of the total number of lines and the order of the spectrum under
examination*. It will be allowing a good deal for the progress of experiment
if we suppose that in measurements of energy it may be possible to dis-
criminate wave-lengths (or frequencies) which differ by a millionth part.
But a million wave-lengths of yellow light would occupy only 60 cm., and the
waves would pass in 2 x 10"9 seconds ! Waves whose frequencies differ by
less than this are inextricably blended, even though we are at liberty to
prolong our observations to all eternity.
At any point in the spectrum of a hot body there are, therefore, mingled
waves of various frequencies lying within narrow limits. The resultant for
any very short interval of time may be identified with a simple train, whose
amplitude and phase, depending as they do upon the relative phases of the
components, must be regarded as matters of chance. The probability of
various amplitudes depends upon the principles explained in a former
communication, " On the Resultant of a large number of Vibrations of the
same Pitch and of Arbitrary Phase f." After an interval of time comparable
with 10~9 second the amplitude is again practically a matter of chance ; so
that during the smallest interval of time of which our senses or our instru-
ments could take cognizance, there are an immense number of independent
combinations. But, under these circumstances, as was shown in the place
referred to, we have to do merely with the sum of the individual intensities.
In his excellent memoir, Sur le mouvement lumineux\, M. Gouy suggests
that the nature of white light may be best understood by assimilating it to a
* Phil. Mag. VoL XLVII. p. 200 (1874). [Vol. i. p. 216.]
t Phil. Mag. Aug. 1880. [Vol. i. p. 491.]
% Journ. de Physique, 1886, p. 354. I observe that M. Gouy had anticipated me (Enc. Brit.
XMV. p. 425 [Vol. in. p. 60]) in the remark that the production of a large number of interference-
bands from originally white light is a proof of the resolving power of the spectroscope, and not of
the regularity of the white light. It would be instructive if some one of the contrary opinion
would explain what he means by regular white light The phrase certainly appears to me to be
without meaning— what Clifford would have called nonsense.
COMPLETE RADIATION AT A GIVES TEMPERATURE, 271
sequence of entirely irregular impulses. It was by means of this idea, that
Young* explained the action of gratings; and although J. Herschelf took
exception, there is no doubt that the method is perfectly sound. The
question that I wish to raise is whether it is possible to define the kind
of impulse of which an irregular sequence woukl represent the complete
radiation of any temperature.
The first thing to be observed is that it wffl not do to suppose the
impulses themselves to be arbitrary. In proof of this it may be sufficient to
point out that in that case there would be no room for distinguishing the
radiations of various temperatures. If the velocity [of disturbance] at every
point [along the line of propagation] were arbitrary, that is independent of
the velocity at neighbouring points however close, the radiation could have
no special relation to any finite wave-length or frequency. In order t«> avoid
this discontinuity we must suppose that the velocities at neighbouring points
are determined by the same causes., so that it is only when the interval
exceeds a certain amount that the velocities become independent of one
another. This independence enters gradually. When the interval is very
small, the velocities are the same. As the interval increases, th<e arbitrary
element begins to assert itself At a moderate distance the velocitv at the
second point is determined in part by agreement with the first, and in part
independently. With augmenting distance the arbitrary part gains in im-
portance until at last the common element is sensibly excluded*.
Now this is precisely the condition of things that would result from the
arbitrary distribution of a large number of impulses, in each of which the
medium is disturbed according to a defined law. A simple case would be
to suppose that each impulse is confined to a narrow region of given width,
and within that region communicates a constant velocity^ An arbitrary
distribution of such impulses over the whole length would produce a
disturbance having, in many respects, the character we wish. But it is easy
to see that this particular kind of impulse will not answer all requirements.
For in the result of each impulse, and therefore in the aggregate of all the
* PML Twmmt. 1801.
f Erne. Jblny., "I*ghV $ 70S (1830).
* The following may serve as an flhcslzation. Out of a iwy huge number of men (say an
aimj) let a ngiment of 1000 be chosen by lot* and let the iemtian of the mean hc%ht of
the ngiment fan that of the any be exhibited as the orfnate of a e*rre. If a •BMBH mt off
1000 be ehoeai by loC 0» w^ ordiM^ m J] b<« no rrf«i« to the oid. B«t if at each step b*t
272 ON THE CHARACTER OF THE [160
impulses, those wave-lengths would be excluded, which are submultiples of
the length of the impulse. The objection could be met by combining
impulses of different lengths ; but then the whole question would be again
open, turning upon the proportions in which the various impulses were
introduced. What I propose here to inquire is whether any definite type
can be suggested such that an arbitrary aggregation of them will represent
complete radiation. It will be evident that in the definition of the type a
constant factor may be left arbitrary. In other words, the impulses need
only to be similar, and not necessarily to be equal.
Probably the simplest type of impulse, </> (x), that could at all meet the
requirements of the case is that with which we are familiar in the theory of
errors, viz.
<j>(x) = e-™ (8)
It is everywhere finite, vanishes at an infinite distance, and is free from
discontinuities. A single impulse of this type may be supposed to be the
resultant of a very large number of localized infinitesimal simultaneous
impulses, all aimed at a single point (x = 0), but liable to deviate from it
owing to accidental causes. I do not at present attempt any physical
justification of this point of view, but merely note the mathematical fact.
The next step is to resolve the disturbance (8) into its elements in accordance
with Fourier's theorem. We have*
0 (x\ = - ( " t +°° cos k (v - x) 6 (t>) dkdv
7TJ o J -oo
cos kv cos kx (r** dkdv (9)
7TJ0
Now
(10)
J -00
so that
i r°°
(11)
This equation exhibits the resolution of (8) into its harmonic components ;
but it is not at once obvious how much energy we are to ascribe to each
value of k, or rather to each small range of values of k. As in the theory of
transverse vibrations of strings, we know that the energy corresponding to
the product of any two distinct harmonic elements must vanish ; but the
application of this, when the difference between two values of k is infini-
tesimal, requires further examination. The following is an adaptation of
Stokes's investigationf of a problem in diffraction.
* [1901. A slight change of notation is introduced.]
t Edinb. Trans, xx. p. 317 (1853) ; see also Enc. Brit. t. xxiv. p. 431. [Vol. in. p. 86.]
1889] COMPLETE RADIATION AT A GIVEX TEMPERATURE. 273
By Fourier's theorem (9) we have
where
In order to shorten the expressions, we will suppose that, as in (11),
We have
— . 6«
This equation is now to be integrated with respect to x from — oc to + oc ;
but, in order to avoid ambiguity, we will introduce the factor e***; where a is
a small positive quantity. The positive sign in the alternative is to be taken
when x is negative, and the negative sign when x is positive. The order of
integration is then to be changed, so as to take first the integration with
respect to x; and finally a is to be supposed to vanish. Thus
2»». {*(*)}'
= Lim.r* r
Now
so that
(19)
Of the right-hand member of (19) the second integral vanishes in the
limit, since k and it' are both positive quantities. But in the first integral
the denominator vanishes whenever k' is equal to k. If we put
k' = k = az, dk' = a.dz,
then, in the limit
r-
J.
= ^Mk\
J » «- -I- \K — KJ- j - x * T *
Thus
k. (20)
If /,(*) be finite, we have, in lieu of (20),
18
274 ON THE CHARACTER OF THE [160
In M. Gouy's treatment of this question, the function <j> (#) is supposed to
be ultimately periodic. In this case f(k) vanishes whenever k differs from
one or other of the terms of an arithmetical progression; and the whole
kinetic energy of the motion is equal to the sum of those of its normal com-
ponents, as in all cases of vibration. The comparison of this method with the
one adopted above, in which all values of k occur, throws light upon the
nature of the harmonic expansion.
It is scarcely necessary to point out that vibrations started impulsively
from rest divide themselves into two groups, constituting progressive waves
in the two directions, and that the whole energy of each of these waves is the
half of that communicated initially to the system in the kinetic form*.
The application of (21) to (11), where
gives
+~~ ..................... (22)
as may be easily shown independently. The intensity, corresponding to the
limits k and k + dk, is therefore
and this, since k and n are proportional, is of the form (5).
If an infinite number of impulses, similar (but not necessarily equal) to
(8), and of arbitrary sign, be distributed at random over the whole range from
— oo to + oo , the intensity of the resultant for an absolutely definite value of
n would be indeterminate. Only the probabilities of various resultants could
be assigned. And if the value of n were changed, by however little, the
resultant would again be indeterminate. Within the smallest assignable
range of n there is room for an infinite number of independent combinations.
We are thus concerned only with an average, and the intensity of each
component may be taken to be proportional to the total number of impulses
(if equal) without regard to their phase-relations. In the aggregate vibration,
the law according to which the energy is distributed is still for all practical
purposes that expressed by (5).
If we decompose each impulse (8) in the manner explained, we may
regard the whole disturbance as arising from an infinite number of simul-
taneous elementary impulses. These elementary impulses are distributed
not entirely at random ; for they may be arranged in groups such that the
members of each group are of the same sign, and are, as it were, aimed at
the same point under a law of error ; while the different groups are without
* Theory of Sound, Vol. n. § 245.
1889] COMPLETE RADIATION AT A GIVES! TEMPERATURE, 275
relation, except that the law of error is the same for alL It is obviously not
essential that the different groups should deliver their blows simultaneously.
Further, it would have come to the same thing had we supposed all the
impulses to be delivered at the same point in space, but to be distributed in
time according to a similar law. In comparing the radiations at various
temperatures, we should have to suppose that, as the temperature rises, not
only does the total number of elementary impulses (of given magnitude)
increase, but also the accuracy of aim of each group.
We have thus determined a kind of impulse such that a [random] aggre-
gation of them will represent complete radiation according to Weber's law (5 ).
One feature of this law is that F(n) approaches a finite limit as n decreases.
In this respect W. Michelson's special law (4) differs widely ; for, according
to it, J"(»> vanishes with it. This evanescence of F(m) implies that the
integrated value of each of our component impulses is zero. If we wish to
inquire further into the law of the impulse, we have to determine <j>(f) so that
........................... (23)
By successive differentiations of (10) with respect to L: it may be shown
that
. ..... . ...... (24)
Thus, if we take
*Or) = «^(l-2cV), ........................ (25)
ft (t) will be of the required form. The curve representative of (25), viz.
y=e-*(l-Z*), .............................. (26)
is symmetrical with respect to x = 0, vanishes when x = ± x and also when
x=±2-*. The positive area between the last-named limits is numerically
equal to the negative area lying outside them.
Other proposed forms for f(k), such as those included in (2), might be
treated in a similar way ; but the above examples may suffice. The simplicity
of (8) compared, e.y., with (25), may be regarded as an argument in its favour.
But we do not know enough of the mechanism of radiation to draw any
confident conclusion. What we most require at present is more complete
data from experiment, such as have been promised by Prof Langley. As
regards the radiation of very low frequency, a question may arise as to
whether it is included in our present measurements. Some authorities have
favoured the view that, when the frequency is sufficiently diminished, all
kinds of matter become transparent : but the electric theory seems to point
in the opposite direction. In comparing any theoretical formula with experi-
ment, we must not forget that what we learn directly from the latter is the
difference of radiations at two temperatures.
18—2
276 ON COMPLETE RADIATION AT A GIVEN TEMPERATURE. [160
One more remark in conclusion. If the complete radiation for a given
temperature be represented by (5), it follows that temperature may be
defined by the value of a. The contrary would imply that the law of distri-
bution is the same at all temperatures, and would be inconsistent with
ordinary observation respecting " red " and " white heats." Now the dimen-
sions of a are those of a time ; so that temperature may be defined by a time,
or (through the velocity of propagation) by a line. Thus in Prof. Langley's
curves, which represent the distribution of energy in a diffraction spectrum,
the wave-length corresponding to the maximum ordinate may be regarded as
a linear specification of the temperature to which the curve relates *.
* [1901. On some of the questions here discussed reference may be made to Schuster's paper
on "Interference Phenomena" (Phil. Mag. xxxvu. p. 509, 1894). The progress of knowledge
with respect to the law of complete radiation is not favourable to the idea that such radiation
can be represented as a random sequence of impulses of simple type.]
161.
OX THE VISIBILITY OF FAINT INTERFERENCE-BANDS.
[Philosophical Magazine, xxvn. pp. 484—486, 1889.]
Ix a recent paper on the limit to interference when light is radiated
from moving molecules*, it was necessary to form an estimate of the ratio
of illuminations (h) at the darkest and brightest parts of a system of bands
corresponding to the moment when they just cease to be visible from lack
of contrast. In the comparison of uniformly illuminated surfaces; brought
well into juxtaposition, h might be as great as '99 f: but in the case of
bands, where the transition is gradual, a higher degree of contrast between
the brightest and darkest parts may be expected to be necessary. In order
to allow for this, I supposed that h might be estimated at '95, the intensity
of the light and the angular magnitude of the bands being assumed to be
suitable. But since widely different estimates have been put forward by
others, I have thought it worth while to test the matter with bands that
are well under controL
In the first experiments light polarized by a Nicol fell upon a slit, against
which was held a somewhat stout selenite. Direct examination of the slit
through an analysing Nicol revealed no colour on account of the thickness
of the selenite; but when a dispersing-prism was added, the resulting
spectrum was marked out into bands, whose brightness and contrast
depended upon the relative orientations of the Nicols and of the selenite.
The theory of these bands is well known;;. If the Nicols be parallel, and
if the principal sections of the Nicols and the selenite be inclined at the
angle o, the expression for the brightness is
1 — sin* 2o sin* %p,
* PkiL Mag. April 1889. [Vol. ra. p. 258.]
t See Helmholtz' Physiological Optic*, § 21.
* See, for example, Enc. Brit. " Wave Theory," § 22. [Vol. ra. p. 156.]
278
ON THE VISIBILITY OF FAINT INTERFERENCE-BANDS.
[161
where p denotes the difference of retardations of the two rays to whose
interference the bands are due. At the brightest place p = 0, and at the
darkest sin2 %p = 1, so that
h = 1 - sin2 2a.
The bands are thus invisible when a = 0, and increase gradually in distinct-
ness with a. When a = 45°, the darkest place is absolutely black*.
The selenite was mounted upon a divided circle, and the observation
consisted in finding the two positions, on either side of a = 0, at which the
bands manifested themselves with the desired degree of distinctness. The
angular interval between the two positions was then taken as representing
the value of 2a. In order that the bands should be recognizable with
certainty it was found that 2a must be at least 14°. For a distinct and
continuous impression 2a = 17°. Corresponding to these, we have for 1 - h,
sin2 14° = -0585, sin2 17° = "0855.
In these observations the earliest recognition of the bands was somewhat
interfered with by a want of smoothness in the spectrum due to irregularities
in the selenite. Any irregularity, whether of this kind or caused by dust
upon the edges of the slit, gives rise to horizontal markings in the spectrum
which distract the eye. In a second set of experiments this difficulty was
obviated by the substitution for the selenite of an accurately worked plate
of quartz, cut parallel to the axis.
The following were the readings by myself (R) and by my assistant (G),
when the bands were but just recognizable with certainty.
82° 6'
72° 13'
81° 2'
71° 59'
81 0
71 40
80 43
72 16
82 2
72 40
81 7
72 9
81 41
72 0
80 31
Mean... 81 28
72 8
80 57
72 8
Hence
(R) 2a = 9°20',
(G)
so that, since sin2 9° = -0245, the bands are visible when 1-h is less than
half as great as before.
* This presupposes an infinitely narrow slit. In practice the width must be reduced until, in
this position, the bands are sensibly black.
OX THE VISIBILITY OF FADPT rVTERFEREXCE-RA3ffD8L 279
The following were the readings when die bands were considered to be
still distinct:—
,83*
15"
'
33*
~-_
•••
w»
83
i-
69
53
82
41
71 *
83
0
70
30
83
37
71 20
Me^.83
14
70
19
.83
5
71 5
Hence
{R) 2a=12*55', (G) 2a = 12°(T.
Here sin- 1 2* 30* = -0372: so that a difference of 4 per cent between the
darkest and brightest parts is sufficient to show the bands with distinctness.
It seems therefore that I was well within the mark in assuming that
bands involving 5 per cent, of the brightness might still be visible.
162.
ON THE UNIFORM DEFORMATION IN TWO DIMENSIONS OF
A CYLINDRICAL SHELL OF FINITE THICKNESS, WITH
APPLICATION TO THE GENERAL THEORY OF DEFORMA-
TION OF THIN SHELLS.
[Proceedings of the London Mathematical Society, xx. pp. 372—381, 1889.]
THE theory of a thin uniform shell of elastic isotropic material, slightly
deformed from an original curved condition, does not seem to be yet upon an
entirely satisfactory footing. If the middle surface be extended, it is clear*
that, to a first approximation, the potential energy per unit of area is
.(1)
where 2h denotes the thickness of the shell ; m, n the elastic constants of
Thomson and Tait's notation; a-l, <r2, OT the elongations and shear of the
middle surface at the place under consideration. Again, if the deforma-
tion be such that the middle surface remain unextended, so that (1) vanishes,
it is tolerably clear that the potential energy takes the form
+ 8-
p2 / m + n \ pl p2
where Bp^1, Sp2-1 are the changes of principal curvatures of the middle
surface, and T is determined by the angle (^) through which the principal
planes are shifted according to the equation
(3)t
« See Lamb, Proc. Math. Soc. Dec. 1882. Also Proc. Roy. Soc. XLV. (1888), p. Ill,
equation (13). [Vol. m. p. 222.]
t See Love, Phil. Trans. CLXXIX. (1888), A, pp. 505, 512; Rayleigh, loc. cit. p. 113.
[Vol. m. p. 224.]
1889] UNIFORM DEFORMATION OF A CYLINDRICAL SHELL, ETC. 281
But when the middle surface undergoes stretching, so that (1) is finite,
while yet the circumstances of the problem forbid us to remain satisfied with
terms involving the first power of h, it is a more difficult question to
determine the expression for the potential energy complete to the order h3.
An investigation of this problem has, however, been given by Mr Love, and
his result* is exhibited in terms of <rt, o-2, -or, and of quantities depending
upon these, and upon the alterations of curvature of the middle surface.
It may, indeed, be an under-statement of the case to speak of the
problem as difficult, for to all appearance it may well be impossible in the
fonii proposed. When the middle surface is plane, or when, though originally
curved, it remains unstretched, there is no difficulty in supposing that the
faces are exempt from imposed force. But when the middle surface of a
shell is originally curved, and undergoes extension, equilibrium cannot be
maintained without the cooperation of forces normal to the shell, and acting
either upon the interior or upon the faces. It is easy to understand that the
precise seat of these forces may be a matter of indifference, so far as the term
of the first order (1) is concerned; but is there any reason for anticipating
that there would be no effect upon the term of the third order ? Rather, it
would appear probable that there is no expression for the potential energy
complete to the order ft3, in the absence of more definite suppositions as to
the manner of application of the normal forces necessary in the general case.
These doubts led me to think an investigation desirable, which should be
based upon the general equations of elasticity, and conducted without the aid
of approximations of ill-defined significance. For this purpose I have chosen
the simplest problem involving the questions at issue — that namely of the
deformation in two dimensions of a shell originally cylindrical.
Taking polar coordinates, let u, vf be the displacements at the point (r, 6)
parallel to r and 0 respectively. The displacement w, parallel to the axis
of the cylinder, vanishes by hypothesis. The strains relative to these direc-
tions arej
The stresses P, Q, R, S, T, U corresponding to these strains are given by
P = (m + n)e + (m-n)f, Q = (m + n)f+(m-n)e, ...... (6)
flf = 0, T=0, U=nc ......................... (7)
* Loc. cit. p. 505.
t This notation differs from that employed in my former papers, where u denoted the
displacement parallel to the axis.
+ Ibbetson's Elastic Solids, 1887, p. 238.
282 ON THE UNIFORM DEFORMATION [162
If there be no internal impressed forces, the equations of equilibrium are
We will now limit the problem by the supposition that the strains and
stresses are independent of 0. Thus
and (8), (9) reduce to
From (12) it follows that Ur- is an absolute constant. Hence if, as we will
now suppose, U vanishes over the cylindrical faces of the shell, it necessarily
vanishes throughout the interior. Thus, by (7),
c = 0 (13)
throughout. From (5) and (13),
d ( d ( v\) d du de „
dr V dr Wj = ~d0dr=~d0= '
by hypothesis. Hence
v = C1 + C2r, (14)
where Gl} C2 are independent of r, but may be functions of 0. Again, from
(5) and (14),
du _ 2 d (v\ _ ~
dd dr \rj 1 '
so that, by (4),
But df/d0 = 0, by supposition. Accordingly,
or Ci=H cos0 + Ksin0, C2=C+D0, (15)
where H, K, C, D are absolute constants. Thus, by (14),
v = H cos 0 + K sin. 0 + (C + D6)r, (16)
and u = H sin 0 — K cos 0 + <f>(r\ (17)
where <£(r) is a function of r which is, so far, arbitrary. Again, by (4),
indicating that the strains are independent of the coefficients H, K, C. The
terms in H, K represent merely a displacement of the cylinder without
rotation or strain, and the term in C represents simple rotation of the cylinder
OF A CYLINDRICAL SHELL OF FINITE THICKNESS. 283
about its axis as a rigid body. They may be omitted without loss of
anything material to the present inquiry.
So far, we have made no use of the condition (11) that there is no
internal force in the radial direction. It is by means of this that the
form of <f> must be determined. From (6), (18),
; ............ (19)
; ............ (20)
so that, by (11),
_
rfr* dr m + n
— the differential equation which must be satisfied by <f>.
The solution of (21) is
rlogr, ..................... (22)
where A and B are arbitrary constants. Corresponding to (22),
e = A-Br-* + ^-n(\ogr + I), .................. (23)
f=D + A + Br-*+j^!^\ogr: .............. ....(24)
and from (16), (17), if H = K = C = 0,
v=Ar+Br-1 + ^-n rlogr, v=Dr6 .......... (25,26)
We have now to consider the potential energy of strain. The general
expression for the energy per unit of volume in a strained solid is
(27)
By (4), (5), (13), we have
a = 0, 6 = 0, c = 0,
so that (27) reduces to
In the present problem
l), .............. (29)
(30)
Before proceeding further, we will consider in detail the very simple case
which arises when D = 0. We have
e = A-Br-*, f=A+Br-*; .................. (31)
v = Q. ...................... (32)
284 ON THE UNIFORM DEFORMATION [162
These equations constitute the solution of the problem of the deformation of
a complete cylindrical shell (of finite thickness) under the action of hydro-
static pressures (or tractions) upon its inner and outer faces*. For the
radial stress at any point, we have
P = 2mA-2nBr~* ............................ (33)
Thus, if the stress upon the inner face r = rl be Hl, and upon the outer face
r = r2 be TI2,
n, = 2mA - 2n B rr2, n2 = 2mA - 2n£r2~2, ......... (34)
by which A and B are determined.
The expression for the energy becomes, by (28), (29), (30),
W=2mA*+2nBn-r-4 ............................ (35)
The whole potential energy per unit of length parallel to the axis is given by
2-n- I W rdr = 2ir {mA- (r2- - r?) - nB- (r^ - ?-,-*)} ....... (3r>)
In order to apply this result to a thin shell, we will write
r, = a — h, r.2 = a + h,
where 2/t denotes the thickness of the shell, and a the radius of the middle
surface. Thus
[
J
* 4-
approximately.
The extension of the middle surface is here, by (31),
<r = A+Ba-* ............................... (38)
Since there are two independent variables A, B, or Tll, n2, in (37), it is
clear that the potential energy cannot, in strictness, be determined by a only.
Let us, however, inquire to what order of approximation the energy is a
function of a-, when h is regarded as small.
If «r denote the ratio of surface forces by which the deformation is
maintained, we have, from (34),
mA (1 - «r) = nB (r2~n- - -srrr2) ;
from which, and (38),
equations giving A and B in terms of <r and -ST. Using these, we find,
on reduction,
mn<r* L 2m A2 4ran h2
1+- ^
m + n ( m + na? (m + w)2 a2 \1 - •
* Ibbetson, loc. cit. pp. 313, 314.
1889] OF A CYLINDRICAL SHELL OF FINITE THICKNESS. 285
the term containing the first power of h disappearing. Thus, for the potential
energy per unit of area of the shell, we obtain
The term in h agrees, as might have been expected, with (1)*. But, when
the approximation is carried so far as to include A*, (40) depends upon «r as
well as upon <r. If the normal forces are limited to one surface, « = 0,
or w = x . In either case
, ur ^
and a~l\Wrdr
f
\
J
4mn<r!A . 2m A* 4mn A
-1 7- ~
m + n \ w + na1 (m + n)1 a4
The energy involved in a given extension of the middle surface is thus
the same, whether the necessary normal force be an internal pressure or
an external traction; but the case is otherwise if the forces be distributed.
When the work is equally divided between the two surfaces, so that there is
(for example) a pressure upon the internal surface and a traction upon the
external surface, -or = — 1 : and
It will be seen that, in order to give rise to this discrepancy, it is not
necessary to suppose the introduction of surface forces more powerful than are
actually required to maintain the deformation. This instance is sufficient to
show that the potential energy of deformation cannot, in general, be expressed
in terms of extensions and changes of curvature of the middle surface, when
it is necessary to include terms of order AJ, without further information as to
the manner in which the surface forces are applied. According to Mr Love's
resultsf, the expression for the energy in the present problem should reduce
to its first term ; whereas (40) indicates that there is no manner of application
of the surface forces by which such a result could be brought about.
We will now abandon the restriction to D = 0. It will then be possible
to find a deformation such that, not only is there no impressed force upon the
interior of the shell, but also none upon either of the surfaces. Under these
circumstances the stresses between contiguous parts must reduce themselves
to a simple couple.
* or has there a different meaning from that belonging to it in (40). In (1) cr = 0, <r, = 0.
for the purposes of the present problem.
t Loc. cit. equations (12), (18). {December, 1889. I have been reminded by the Secretary
that in the investigation of Mr Lore it is expressly supposed (p. 504) that no surface tractions are
applied. But the absence of normal forces would, as it appears to me, be equivalent to a
limitation upon the generality of the middle surface extensions, «-,, «,.}
286 ON THE UNIFORM DEFORMATION [162
From (6), (29), (30), we find
P = m (e +/) + n (e -/) = 2mA - 2nBr^ + D \m + ^^p" j . . . .(43)
If P = 0, both when r = rt and when r = r2, the values of A, B, in terms
of D, are
„ jsi&Q^bfflp w
W-M ....(«>
These values, substituted in (23), (24), (25), (26), determine a definite type of
deformation, satisfying the conditions that there shall be no internal or
surface forces, and that the strains shall be independent of 0, and this
without any supposition limiting the thickness of the shell.
From the expression for Q in terms of e and f, or, more readily, by means
of (11), we may verify that
<*•-« (46)
In order to apply these results to a thin shell, we write, as before,
thus
Corresponding to these, from (29), (30),
..................... (49>
(50)
or, if r = a + p,
( }
The strains e,/both vanish approximately when r = a. By (6),
OF A CYLINDRICAL SHELL OF FINITE THICKNESS. '2x7
We will now calculate the potential energy of deformation. From (28),
(51), (52),
w= (2p> 2p* 2pfr) m*nD (2p> 6p» IQpfr).
(m + nyja' a» 3a« J (TO + n? ( a» a» 3a«j'
so that, for the potential energy per unit of area, we get
the next term involving A*.
In order to connect this with the change of curvature of the middle
surface, we require the expression for ti. From (25),
nD
so that the value of it at the middle surface (r = a) is, approximately,
ii = - aD .................................. (57)
Now a + u is the radius of curvature of the middle surface after deformation,
or Spj = u. Thus
I
The expression for the energy per unit area of surface is thus
,
n) V pi)
in agreement with (2) ; for in the present application
It is evident that the rigorous solution from which we started is available
for continuing the approximation, should it be thought desirable to retain
higher powers of h.
The solution of the problem of the bending of a cylindrical shell, here put
forward, favours then the idea that it is only when the middle surface of
a curved shell remains unextended that it is possible to express the potential
energy to the order h* in terms merely of the extensions and curvatures
of the middle surface.
163.
ON ACHROMATIC INTERFERENCE-BANDS.
[Philosophical Magazine, XXVIIL pp. 77— 91, 189—206, 1889.]
Introduction.
WHEN there is interference of light, the width of the resulting bands,
measured for example from darkness to darkness, is usually a function of the
colour of the light employed. Thus, in the case of Fresnel's well-known
interference-experiment, in which light reflected from two slightly inclined
mirrors illuminates a screen, the width of the bands is proportional to the
wave-length of the light. In order that a considerable number of bands may
be visible, it is necessary that the light be highly homogeneous ; otherwise it
is impossible that the various band-systems can fit one another over the
necessary range. If the light could be supposed to be absolutely homogeneous,
there would be no limit to the number of observable bands : and, what is
especially to be remarked, there would be nothing by which one band could
be distinguished from another, — in particular there could be no central band
recognizable. When, on the other hand, the light is white, there may be a
central band at which all the maxima of brightness coincide ; and this band,
being white, may be called the achromatic band. But the system of bands is
not usually achromatic. Thus, in Fresnel's experiment the centre of symmetry
fixes the position of the central achromatic band, but the system is far from
achromatic. Theoretically there is not even a single place of darkness, for
there is no point where there is complete discordance [opposition] of phase
for all kinds of light. In consequence, however, of the fact that the range of
sensitiveness of the eye is limited to less than an " octave," the centre of the
first dark band on either side is sensibly black ; but the existence of even one
band is due to selection, and the formation of several visible bands is favoured
by the capability of the retina to make chromatic distinctions within the
range of vision. After two or three alternations the bands become highly
OX ACHROMATIC nTTEBFERENCE-BAXDS. 289
coloured*: and, as the overlapping of the various elementary systems
increases, the colours fade away, and the field of view assumes a uniform
appearance.
There are, however, cases where it is possible to have systems of achromatic
bands. For this purpose it is necessary, not merely that the maxima of
illumination should coincide at some one place, but also that the widths of
the bands should be the same for the various colours. The independence of
colour, as we shall see, may be absolute : but it will probably be more
convenient not to limit the use of the term so closely. The focal length
of the ordinary achromatic object-glass is not entirely independent of colour.
A similar use of the term would justify us in calling a system of bands
achromatic, when the width of the elementary systems is a maximum or a
minimum for some ray very near the middle of the spectrum, or, which comes
to the same, has equal values for two rays of finitely different refrangibility.
The outstanding deviation from complete achromatism, according to the same
analog)-, may be called the secondary colour.
The existence of achromatic systems was known to Xewton+. and was
insisted upon with special emphasis by Fox Talbot*: but singularly little
attention appears to have been bestowed upon the subject in recent times.
In the article "Wave Theory" (Encyc. Brit. 1888 [Vol. m. p. 61]) I have
discussed a few cases, but with too great brevity. It may be of interest to
resume the consideration of these remarkable phenomena, and to detail some
observations which I have made, in part since the publication of the
"Encyclopaedia" article. A recent paper by M. Mascart§ will also be
referred to.
FresneVs Bands.
In this experiment the two sources of light which are regarded as
interfering with one another must not be independent : otherwise there
could be no fixed phase-relation. According to Fresnel's original arrangement
the sources (\, Ot are virtual images of a single source, obtained by reflexion
in two mirrors. The mirrors may be replaced by a bi-prism. Or, as in Lloyd s
form of the experiment, the second source may be obtained from the first by
reflexion from a mirror placed at a high degree of obliquity. The screen upon
which the bands are conceived to be thrown is parallel to 0i0lr at distance D.
• The series of colours thus arising are calculated, and exhibited in the form of a carve upon
the colour diagram, in a paper "On the Colours of Thin Plates, n Edimb. Tram. 1887. [VoL n.
p. 498.]
t Optics, Book n.
* Phil. Mag. [3] n. p. 401 (1836).
§ " On the Achromatism of Interference," Comptt* Bemdmt, March 1889 ; PkU. Mag. [5]
p. 519.
19
290 ON ACHROMATIC INTERFERENCE-BANDS. [163
If A be the point of the screen equidistant from Olt Oa, and P a neighbouring
point, then approximately
0lP - 02P =
where
0X02 = b, AP = u.
Thus, if X be the wave-length, the places where the phases are accordant are
determined by
u = n\D/b, ................................. (1)
n being an integer representing the order of the band. The linear width of
the bands (from bright to bright, or from dark to dark) is thus
(2)
The degree of homogeneity necessary for the approximate perfection of the
nth band may be found at once from (1) and (2). For, if da be the change in
u corresponding to the change d\, then
dujK = nd\l\ .................................. (3)
Now clearly du must be a small fraction of A, so that d\{\ must be many
times smaller than l/n, if the darkest places are to be sensibly black. But
the phenomenon will be tolerably well marked, if the proportional range of
wave-length do not exceed l/(2w), provided, that is, that the distribution of
illumination over this range be not concentrated towards the extreme parts.
So far we have supposed the sources at Olt 02 to be mathematically small.
In practice the source is an elongated slit, whose direction requires to be
carefully adjusted to parallelism with the reflecting surface, or surfaces. By
this means an important advantage is obtained in respect of brightness without
loss of definition, as the various parts of the aperture give rise to coincident
systems of bands.
The question of the admissible width of the slit requires careful con-
sideration. We will suppose in the first place that the lights issuing from
the various parts of the aperture are without permanent phase-relation, as
when the slit is backed immediately by a flame, or by the incandescent carbon
of an electric lamp. Regular interference can then only take place between
lights coming from corresponding parts of the two images ; and a distinction
must be drawn between the two ways in which the images may be situated
relatively to one another. In Fresnel's experiment, whether carried out with
mirrors or with bi-prism, the corresponding parts of the images are on the
same side ; that is, the right of one corresponds to the right of the other, and
the left of one to the left of the other. On the other hand, in Lloyd's
arrangement the reflected image is reversed relatively to the original source :
the two outer edges corresponding, as also the two inner. Thus in the first
arrangement the bands due to various parts of the slit differ merely by a
OX ACHROMATIC nfTERFEREXCE-BAXDS. 291
lateral shift, and the condition of distinctness is simply that the [projection
of the*] width of the slit be a small proportion of the width of the bands.
From this it follows as a corollaiy that the limiting width is independent
of the order of the bands under examination. It is otherwise in Lloyd's
method. In this case the centres of the systems of bands are the same,
whatever part of the slit be supposed to be operative, and it is the distance
apart of the images (6) that varies. The bands corresponding to the various
parts of the slit are thus upon different scales, and the resulting confusion
must increase with the order of the bands. From (1) the corresponding
changes in u and 6 are given by
so that
dujA = -ndb.b .................................. (4)
If db represents twice the width of the slit, (4) gives a measure of the
resulting confusion in the bands. The important point is that the slit
must be made narrower as n increases, if the bands are to retain the same
degree of distinctness.
If the various parts of the width of the slit do not act as independent
sources of light, a different treatment would be required. To illustrate the
extreme case, we may suppose that the waves issuing from the various
elements of the width are all in the same phase, as if the ultimate source
were a star situated a long distance behind- It would then be a matter
of indifference whether the images of the slit, acting as proximate sources
of interfering light, were reversed relatively to one another, or not. It is.
however, unnecessary to dwell upon this question, inasmuch as the conditions
supposed are unfavourable to brightness, and therefore to be avoided in
practice. The better to understand this, let us suppose that the slit is
backed by the sun, and is so narrow that, in spite of the sun's angular
magnitude, the luminous vibration is sensibly the same at all part* of the
width. For this purpose the width must not exceed ^ millim.x By
hypothesis,. the appearance presented to an eye close to the slit and looking
backwards towards the sun will be the same as if the source of light were
reduced to a point coincident with the sun's centre. The meaning of this is
that, on account of the narrowness of the aperture, a point would appear
dilated by diffraction until its apparent diameter became a large multiple
* [1901. Compare Walker, Phil. Mag. XLYI. p. 477, 1898.
In the case of the spectroscope, when resolTing power is important, the width of the slit
most evidently not exceed X/o, where X = ware-length and a = horizontal aperture (VoL i. p. 420).
This is the condition that the aperture of the instrument should just embrace the central
diffraction fringe (from darkness to darkness) formed bj light passing the slit aperture. Since
full resolTing power requires the cooperation of all parts of the aperture, we mar conclude that
an even narrower slit than that above specified is desirable.]
f Verdefs Lew** fOptiqme pkytiqme, 1. 1. p. 106.
19—2
292 ON ACHROMATIC INTERFERENCE-BANDS. [163
of that of the sun. Now it is evident that in such a case the brightness may
be enhanced by increasing the sun's apparent diameter, as can always be done
by optical appliances. Or, which would probably be more convenient in
practice, we may obtain an equivalent result by so designing the experiment
that the slit does not require to be narrowed to the point at which the sun's
image begins to be sensibly dilated by diffraction. The available brightness
is then at its limit, and would be no greater, even were the solar diameter
increased. The practical rule is that, when brightness is an object, slits
backed by the sun should not be narrowed to much less than half a
millimetre.
Lloyd's Bands.
Lloyd's experiment deserves to be more generally known, as it may be
performed with great facility and without special apparatus. Sunlight is
admitted horizontally into a darkened room through a slit situated in the
window-shutter, and at a distance of 15 or 20 feet is received at nearly
grazing incidence upon a vertical slab of plate glass. The length of the slab
in the direction of the light should not be less than 2 or 3 inches, and for some
special observations may advantageously be much increased. The bands are
observed on a plane through the hinder vertical edge of the slab by means of
a hand magnifying-glass of from 1 to 2 inch focus. The obliquity of the
reflector is of course to be adjusted according to the fineness of the bands
required.
From the manner of their formation it might appear that under no
circumstances could more than half the system be visible. But, according
to Airy's principle*, the bands may be displaced if examined through a
prism. In practice all that is necessary is to hold the magnifier some-
what excentrically. The bands may then be observed gradually to detach
themselves from the mirror, until at last the complete system is seen, as in
Fresnel's form of the experiment.
If we wish to observe interference under high relative retardation, we
must either limit the light passing the first slit to be approximately homo-
geneous, or (after Fizeau and Foucault) transmit a narrow width of the
band-system itself through a second slit, and subsequently analyse the light
into a spectrum. In the latter arrangement, which is usually the more
convenient when the original light is white, the bands seen are of a rather
artificial kind. If, apart from the heterogeneity of the light, the original
bands are well formed, and if the second slit be narrow enough, the spectrum
will be marked out into bands ; the bright places corresponding to the kinds
of light for which the original bands would be bright, and the black places
to the kinds of light for which the original bands would be black. The
* See below.
1889] OX ACHROMATIC INTERFERENCE-BAXDS. 293
condition limiting the width of the second slit is obviously that it be but a
moderate fraction of the width of a band (A).
If it be desired to pass along the entire series of bands up to those of a
high order by merely traversing the second slit in a direction perpendicular
to that of the light, a very long mirror is necessary. But when the second
slit is in the region of the bands of highest order (that is, near the external
limit of the field illuminated by both pencils), only the more distant part of
the mirror is really operative ; and thus, even though the mirror be small,
bands of high order may be observed, if the second slit be carried backwards,
keeping it of course all the time in the narrow doubly-illuminated field. In
one experiment the distance from the first slit to the (3-inch) reflector was
27 feet, while the second slit was situated behind at a further distance of 4 feet.
The distance (6) between the first slit and its image in the reflector (measured
at the window) was about 13 inches.
As regards the spectroscope it was found convenient to use an arrangement
with detached parts. The slit and collimating lens were rigidly connected,
and stood upon a long and rigid box, which carried also the mirror. The
narrowness of the bands in which this slit is placed renders it imperative to
avoid the slightest relative unsteadiness or vibration of these parts. The
prisms, equivalent to about four of 60", and the observing telescope were
upon another stand at a little distance behind the box which supported the
rest of the apparatus.
Under these conditions it was easy to observe bands in the spectrum whose
width (from dark to dark) could be made as small as the interval between the
D lines; but for this purpose the first slit had to be rather narrow, and the
direction of its length accurately adjusted, so as to give the greatest distinct-
ness. Since the wave-lengths of the two D lines differ by about ^^ part,
spectral bands of this degree of closeness imply interference with a retardation
of 1000 periods.
Much further than this it was not easy to go. When the bands were
rather more than twice as close, the necessary narrowing of the slits began
to entail a failing of the light, indicating that further progress would be
attained with difficulty.
Indeed, the finiteness of the illumination behind the first slit imposes of
necessity a somewhat sudden limit to the observable retardation. In this
respect it is a matter of indifference at what angle the reflector be placed.
If the angle be made small, so that the reflexion is very nearly grazing, the
bands are upon a larger scale, and the width of the second slit may be
increased, but in a proportional degree the width of the first slit must be
reduced.
The relation of the width of the second slit to the angle of the mirror
may be conveniently expressed in terms of the appearance presented to an
294 ON ACHROMATIC INTERFERENCE-BANDS. [163
eye placed close behind the former. The smallest angular distance which the
slit, considered as an aperture, can resolve, is expressed by the ratio of the
wave-length of light (X.) to the width (w2) of the slit. Now, in order that this
slit may perform its part tolerably well, w2 must be less than -| A ; so that,
b (2),
(5)
The width must therefore be less than the half of that which would just
allow the resolution of the two images (subtending the angle bjD) as seen by
an eye behind. In setting up the apparatus this property may be turned to
account as a test.
The existence of a limit to n, dependent upon the intrinsic brightness of
the sun, may be placed in a clearer light by a rough estimate of the illumi-
nation in the resulting spectrum; and such an estimate is the more interesting
on account of the large part here played by diffraction. In most calculations
of brightness it is tacitly assumed that the ordinary rules of geometrical optics
are obeyed.
Limit to Illumination.
The narrowness of the second slit would not in itself be an obstacle to
the attainment of full spectrum brightness, were we at liberty to make what
arrangements we pleased behind it. In illustration of this, two extreme cases
may be considered of a slit illuminated by ordinary sunshine. First, let the
width w2 be great enough not sensibly to dilate the solar image; that is, let w2
be much greater than \/s, where s denotes in circular measure the sun's
apparent diameter (about 30 minutes). In this case the light streams through
the slit according to the ordinary law of shadows, and the pupil (of diameter p)
will be filled with light if situated at a distance exceeding d*, where
p/d = s ..................................... (6)
At this distance the apparent width of the slit is w2/d, or w^s/p ; and the
question arises whether it lies above or below the ocular limit \jp, that is,
the smallest angular distance that can be resolved by an aperture p. The
answer is in the affirmative, because we have already supposed that w2s
exceeds X. The slit has thus a visible width, and it is seen backed by
undiffracted sunshine. If a spectrum be now formed by the use of dispersion
sufficient to give a prescribed degree of purity, it is as bright as is possible
with the sun as ultimate source, and would be no brighter even were the
solar diameter increased, as it could in effect be by the use of a burning-glass
throwing a solar image upon the slit. The employment of a telescope in the
formation of the spectrum gives no means of escape from this conclusion.
The precise definition of the brightness of any part of the resulting spectrum
* About 30 inches [76 cm.].
03T ACHROMATIC DTTERFEREXCE-BAXDS. 295
would give opportunity for a good deal of discussion: but for the present
purpose it may suffice to suppose that, if the spectrum is to be divided into
N distinguishable parts, so that its angular width is n times the angular width
of the slit, the apparent brightness is of order I/it as compared with that of
the sun.
Under the conditions above supposed the angular width of the slit is in
excess of the ocular limit, and the distance might be increased beyond d
without prejudice to the brilliancy of the spectrum. As the angular width
decreases, so does the angular dispersion necessary to attain a given degree
of purity. But this process must not be continued to the point where w^Jd
approaches the ocular limit. Beyond that limit it is evident that no accession
of purity would attend an increase in d under given dispersion. Accordingly
the dispersion could not be reduced, if the purity is to be maintained ; and
the brightness necessarily suffers. It must always be a condition of full
brightness that the angular width of the slit exceed the ocular limit-
Let us now suppose, on the other hand, that w* is so small that the image
of the sun is dilated to many times *, or that in, is much less than X s. The
divergence of the tight is now not *. but X w^: and, if the pupil be just
immersed,
The angular width of the slit wjd is thus equal to X Jp, that is. it coincides
with the ocular limit. The resulting spectrum necessarily Mis short of full
brightness, for it is evident that, further brightness would attend an augmen-
tation of the solar diameter, up to the point at which the dilatation due to
diffraction is no longer a sensible fraction of the whole. In comparison with
full brightness the actual brightness is of order «v*/X: or, if the purity
required is represented by «, we may consider the brightness of the spectrum
relatively to that of the sun to be of order M>s*/(jiX).
In the application of these considerations to Lloyd's bands we have to
regard the narrow slit w± as illuminated, not by the sun of diameter *, but
by the much narrower source allowed by the first slit, whose angular width is
Wi/D. On this account the reduction of brightness is at least icy/ (*D). If IT,
be so narrow as itself to dilate the solar image, a further reduction would
ensue ; but this could always be avoided, either by increase of D, or by the
use of a burning-glass focusing the sun upon the first slit. The brightness
of the spectrum of purity n from the second slit is thus of order
We have now to introduce the limitations upon tr-, and «,. By (4) », must
not exceed 6/(4n); and by (2) M^ must not exceed Xl>/(26). Hence the
brightness is of order
296 ON ACHROMATIC INTERFERENCE-BANDS. [163
independent of s, and of the linear quantities. The fact that the brightness
is inversely as the square of the number of bands to be rendered visible
explains the somewhat sudden failure observed in experiment. If n = 2000,
the original brightness of the sun is reduced in the spectrum some 30 million
times, beyond which point the illumination could hardly be expected to
remain sufficient for vision of difficult objects such as narrow bands.
In Fresnel's arrangement, where the light is reflected perpendicularly
from two slightly inclined mirrors, interference of high order is obtained by
the movement of one of the mirrors parallel to its plane. The increase of n
does not then entail a narrowing of wl ; and bands of order n may be observed
in the spectrum of light transmitted through w2) whose brightness is propor-
tional to w1, instead of, as before, to n~2.
Achromatic Interference-Bands.
We have already seen from (3) that in the ordinary arrangement, where
the source is of white light entering through a narrow slit, the heterogeneity
of the light forbids the visibility of more than a few bands. The scale of the
various band-systems is proportional to X. But this condition of things, as
we recognize from (2), depends upon the constancy of b, that is, upon the
supposition that the various kinds of light all come from the same place.
Now there is no reason why such a limitation should be imposed. If we
regard b as variable, we recognize that we have only to take b proportional
to X, in order to render the band-interval (A) independent of the colour. In
such a case the system of bands is achromatic, and the heterogeneity of the
light is no obstacle to the formation of visible bands of high order.
These requirements are very easily met by the use of Lloyd's mirror, and
of a diffraction-grating with which to form a spectrum. White light enters
the dark room through a slit in the window-shutter, and falls in succession
upon a grating and upon an achromatic lens, so as to form a real diffraction-
spectrum, or rather series of such, in the focal plane. The central image, and
all the lateral coloured images, except one, are intercepted by a screen. The
spectrum which is allowed to pass is the proximate source of light in the
interference experiment; and since the deviation of any colour from the
central white image is proportional to X, it is only necessary so to arrange
the mirror that its plane passes through the white image in order to realize
the conditions for the formation of achromatic bands.
There is no difficulty in carrying out the experiment practically. I have
used the spectrum of the second order, as given by a photographed grating of
6000 lines in an inch, and a photographic portrait lens of about 6 inches
focus. At a distance of about 7 feet from the spectrum the light fell upon a
vertical slab of thick plate-glass 3 feet in length and a few inches high. The
ON ACHROMATIC INTERFERENCE-BANDS. 297
observer upon the further side of the slab examines the bands through a
Coddington lens of somewhat high power, as they are formed upon the plane
passing through the end of the slab. It is interesting to watch the appear-
ance of the bands as dependent upon the degree in which the condition of
achromatism is fulfilled. A comparatively rough adjustment of the slab in
azimuth is sufficient to render achromatic, and therefore distinct, the first
20 or 30 bands. As the adjustment improves, a continually larger number
becomes visible, until at last the whole of the doubly illuminated field is
covered with fine lines.
In these experiments the light is white, or at least becomes coloured only
towards the outer edge of the field. By means of a fine slit in the plane of
the spectrum we may isolate any kind of light, and verify that the band-
systems corresponding to various wave-lengths are truly superposed.
When the whole spectrum was allowed to pass, the white and black bands
presented so much the appearance of a grating under the microscope that
I was led to attempt to photograph them, with the view of thus forming a
diffraction-grating. Gelatine plates are too coarse in their texture to be very
suitable for this purpose ; but I obtained impressions capable of giving
spectra. Comparison with spectra from standard gratings showed that the
lines were at the rate of 1200 to the inch. A width of about half an inch
(corresponding to 600 lines) was covered, but the definition deteriorated in
the outer half A similar deterioration was evident on direct inspection
of the bands, and was due to some imperfection in the conditions — perhaps
to imperfect straightness of the slab. On one occasion the bands were seen
to lose their sharpness towards the middle of the field, and to recover in the
outer portion.
With respect to this construction of a grating by photography of
interference-bands, a question may be raised as to whether we are not
virtually copying the lines of the original grating used to form the spectrum.
More may be said in favour of such a suggestion than may at first appear.
For it would seem that the case would not be essentially altered if we
replaced the real spectrum by a virtual one, abolishing the focusing lens, and
bringing Lloyd's mirror into the neighbourhood of the grating. But then
the mirror would be unnecessary, since the symmetrical spectrum upon the
other side would answer the purpose as well as a reflexion of the first
spectrum. Indeed, there is no escape from the conclusion that a grating
capable of giving on the two sides similar spectra of an}- one order, without
spectra of other orders or central image, must produce behind it, without
other appliances and at all distances, a system of achromatic interference-
fringes, which could not fail to impress themselves upon a sensitive photo-
graphic plate. But a grating so obtained would naturally be regarded as
merely a copy of the first.
298 ON ACHROMATIC INTERFERENCE-BANDS. [163
Another apparent anomaly may be noticed. It is found in practice that,
to reproduce a grating by photography, it is necessary that the sensitive
plate be brought into close contact with the original ; whereas, according to
the argument just advanced, no such limitation would be required.
These discrepancies will be explained if, starting from the general theory,
we take into account the actual constitution of the gratings with which
we can experiment. If plane waves of homogeneous light (\) impinge
perpendicularly upon a plane (z = 0) grating, whose constitution is periodic
with respect to x in the interval <r, the waves behind have the general
expression
A0 cos (kat — kz) + Al cos ( px +fi) cos (kat — ^z)
+ Bl cos (px + #1) sin (kat — ^z)
+ A2 cos (2pa? +/2) cos (kat - ^z} + . . . ; (8)
where
2> = 27T/eT, k=<2TT/\,
and
tf = A;2 - p\ f^=kz- V, &c.,
the series being continued as long as /i is real*. Features in the wave-form
for which //. is imaginary are rapidly eliminated. For the present purpose we
may limit our attention to the first three terms of the series, which represent
the central image and the two lateral spectra of the first order.
When the first term .occurs, as usually happens, the phenomena are
complicated by the interaction of this term with the following ones, and the
effect varies with z in a manner dependent upon \. This is the ordinary
case of photographic reproduction, considered in the paper referred to. If A0
vanish, there is no central image ; but various cases may still be distinguished
according to the mutual relations of the other constants. If only Alt or only
J5i, occur, we have interference-fringes. The intensity of light is (in the
first case)
41«cofiC(jw+/), (9)
vanishing when
p* +/=!(» +!)*•;
and these fringes may be regarded as arising from the interference of the two
lateral spectra of the first order,
$A! cos (kat — ^2 +px +/),
\ A1 cos (kat — fJ-iZ — px— /i).
As an example of only one spectrum, we may suppose
B^A,, 9l=f, -ITT,
* Phil. Mag. March 1881 [Vol. i. p. 510]; Enc. Brit. "Wave Theory," p. 440 [Vol. in.
p. 122].
OX ACHROMATIC DiTERFEREXCB-BAXDS.
giving
(10)
A photographic plate exposed to this would yield no impression, since the
intensity is constant.
In order, then, that a grating may be capable of giving rise to the
ideal system of interference-fringes, and thus impress itself upon a sensitive
plate at any distance behind, the vibration due to it must be of the form
Acos(j^+f)cos(tat-f^}. ..................... (11)
It does not appear how any actual grating could effect this*. Supposing
z = 0, we see that the amplitude of the vibration immediately behind the
grating must be a harmonic function of x, while the phase is independent
of x, except as regards the reversals implied in the variable sign of the
amplitude. Gratings may act partly by opacity and partly by retardation,
but the two effects would usually be connected: whereas the requirement
here is that at two points the transmission shall be the same while the phase
is reversed.
We can thus hardly regard the interference-bands obtained from a grating
and Lloyd's mirror as a mere reproduction of the original ruling. As will be
seen in the following paragraphs, much the same result may be got from a
prism, in place of a grating: and if the light be sufficiently homogeneous
to begin with, both these appliances may be dispensed with altogether.
Prism instead of Grating.
If we are content with a less perfect fulfilment of the achromatic condition,
the diffraction-spectrum may be replaced by a prismatic one, so arranged
that d(X/6) = 0 for the most luminous rays. The bands are then achromatic
in the same sense that the ordinary telescope is so. In this case there is no
objection to a merely virtual spectrum, and the experiment may be very
simply executed with Lloyd's mirror and a prism of (say) 20 ' held just in
front of it.
The number of black and white bands to be observed is not so great
as might perhaps have been expected. The lack of contrast which soon
supervenes can only be due to imperfect superposition of the various com-
ponent systems. That the fact is so is at once proved by observation
according to the method of Fizeau ; for the spectrum from a slit at a very
moderate distance out is seen to be traversed by bands. If the adjustment
has been properly made, a certain region in the yellow-green is uninterrupted,
* [1901. It would seem that the required conditions are satisfied by a grating composed of
equal transparent puts, giving alternately a relative retardation of J V. and too fine to allow the
formation of spectra of the second and higher orders.]
300 ON ACHROMATIC INTERFERENCE-BANDS. [163
while the closeness of the bands increases towards either end of the spectrum.
So far as regards the red and blue rays, the original bands may be considered
to be already obliterated, but so far as regards the central rays, to be still
fairly defined. Under these circumstances it is remarkable that so little
colour should be apparent on direct inspection of the bands. It would seem
that the eye is but little sensitive to colours thus presented, perhaps on
account of its own want of achromatism.
It is interesting to observe the effect of coloured glasses upon the
distinctness of the bands. If the achromatism be in the green, a red or
orange glass, so far from acting as an aid to distinctness, obliterates all the
bands after the first few. On the other hand, a green glass, absorbing rays
for which the bands are already confused, confers additional sharpness. With
the aid of a red glass a large number of bands are seen distinctly, if the
adjustment be made for this part of the spectrum.
A still better procedure is to isolate a limited part of the spectrum by
interposed screens. For this purpose a real spectrum must be formed, as in
the case of the grating above considered.
We will now inquire to what degree of approximation A/6 may be made
independent of A with the aid of a prism, taking Cauchy's law of dispersion
as a basis. According to it the value of b for any ray may be regarded as
made up of two parts — one constant, and one varying inversely as X2. We
therefore write
where A is to be so chosen that Xfb is stationary when A, has a prescribed
value, \0. This condition gives
A\0*=3B; .................................... (13)
so that
As an example, let us suppose that the disposition is achromatic for
the immediate neighbourhood of the line D, so that A0 = \D, and inquire
into the proportional variation of A/ 6, when we consider the ray C. Assuming
XD = -58890, \c= -65618,
we obtain from (14)
The meaning of this result will be best understood if we inquire for what
order (n) the bands of the (7-system are shifted relatively to those of the
Z)-system through half the band-interval. From (1)
- A0/&0} = |
1889] OX ACHROMATIC DSTERFEREXCE-BAXDS. 3Q1
bj hypothesis: so that
Thus, in the case supposed, n = 32. After 32 periods the black places of
the C-system will coincide with the bright places of the D-system, and
conversely. If no prism had been employed (6 constant), a similar condition
of things would have arisen when
If (X- X.), or, as we may call it, SX, be small,
x/6-yfe,
is of the second order in $X. An analytical expression is readily obtained
from (14). We have
X/fe = 1 + 3SX X« + 3(SX
X, + |(5X
approximately ; so that, by (15),
if X* i
(16)
This gives the order of the band at which complete discrepance first occurs
for A, and X^+SX, the adjustment being made for X». It is, of course.
inversely proportional to the square of Sx, when SX is small.
The corresponding value of H, if no prism be used, so that 6 is constant, is
Tne effect of the prism is thus to increase the number of bands in
the ratio
A try's Theory of the White Centre.
If a system of interference-bands be examined through a prism, the
central white band undergoes an abnormal displacement, which has been
supposed to be inconsistent with theory. The explanation has been shown
by Airy* to depend upon the peculiar manner in which the white band is
* Airy, "Bemarks on Mr Potter's Experiment on Interference," Pkil. Mmy. a. p. 161 (1853).
302 ON ACHROMATIC INTERFERENCE-BANDS. [163
in general formed. Thus, " Any one of the kinds of homogeneous light
composing the incident heterogeneous light will produce a series of bright
and dark bars, unlimited in number so far as the mixture of light from the
two pencils extends, and undistinguishable in quality. The consideration,
therefore, of homogeneous light will never enable us to determine which is
the point that the eye immediately turns to as the centre of the fringes.
What, then, is the physical circumstance that determines the centre of
the fringes ?
" The answer is very easy. For different colours the bars have different
breadths. If, then, the bars of all colours coincide at one part of the mixture
of light, they will not coincide at any other part ; but at equal distances on
both sides from that place of coincidence they will be equally far from a state
of coincidence. If, then, we can find where the bars of all colours coincide,
that point is the centre of the fringes.
" It appears, then, that the centre of the fringes is not necessarily the
point where the two pencils of light have described equal paths, but is
determined by considerations of a perfectly different kind ____ The distinction
is important in this and other experiments."
The effect in question depends upon the dispersive power of the prism.
If v be the linear shifting, due to the prism, of the originally central band,
v must be regarded as a function of X. Measured from the original centre,
the position of the nth bar is now
v+n\D/b.
The coincidence of the various bright bands occurs when this quantity is as
independent as possible of X ; that is, when n is the nearest integer to
or, as Airy expresses it, in terms of the width of a band (A),
n=-dv/dA ............................... (19)
The apparent displacement of the white band is thus not v simply, but
v-Adv/dA ............................... (20)
The signs of dv and dA being opposite, the abnormal displacement is in
addition to the normal effect of the prism. But, since dv/dA, or dv/dX,, is not
constant, the achromatism of the white band is less perfect than when no
prism is used.
If a grating were substituted for a prism, v would vary as A, and the
displacement (20) would vanish.
1889] 03f ACHBOMATIC DTTERFEBEXCE-BAXDS. 303
More recently the matter has engaged the attention of Gornu*, who thus
formulates the general principle : — " Dans un systeme de /ranges tf interference
produites a faide dune I u mitre heterogene ayant nit spectre continu, U exists
toujours une /range achromatique qui joue le role de /range centrale et q>ti se
troure au point de champ ou ties radiations les plus intense* presentent une
difference de phase majcimum ou minimum."
In Fresnel's experiment, if the retardation of phase due to an interposed
plate, or to any other cause, be F(\), the whole relative retardation of the
two pencils at the point u is
and the situation of the central, or achromatic, band is determined, not by
<£ = 0. but by <fyrfX = 0, or
It is scarcely necessary to say that although the nth band may be
rendered achromatic, the system is no more achromatic than if the prism
had been dispensed with. The width of the component systems being
unaltered, the manner of overlapping remains as before. The present use
of the prism is of course entirely different from that previously discussed,
in which by a suitable adjustment the system of bands could be achro-
matized.
Thin Plates.
The series of tints obtained by nearly perpendicular reflexion from thin
plates of varying thickness is the same as that which occurs in Lloyd's
interference experiment, or at least it would be the same if the material
of the plates were non-dispersive and the reflecting power small. If t be the
thickness, p. the index, a the inclination of the ray within the plate to the
normal, the relative retardation of the two rays (reckoned as a distance)
is 2/dcosa', and is sensibly independent of X.
" This state of things may be greatly departed from when the thin plate
is rarer than its surroundings, and the incidence is such that a' is nearly
equal to 90^ ; for then, in consequence of the powerful dispersion, cos a' mav
vary greatly as we pass from one colour to another. Under these circum-
stances the series of colours entirely alters its character, and the bands
(corresponding to a graduated thickness) may even lose their coloration,
becoming sensibly black and white through many alternations^. The general
explanation of this remarkable phenomenon was suggested by Newton, but
* Jomrm. d. Pkytiyut, L p. 293 (1883).
t Erne. Brit., " W*TC Theoiy." XDT. p. 425 (To!, m. p. 6SJ.
: Newton's Optic*, Book n. ; Fox Talbot, Phil. May. n. p. 401 (1836).
304 ON ACHROMATIC INTERFERENCE-BANDS. [163
it does not appear to have been followed out in accordance with the wave
theory.
" Let us suppose that plane waves of white light travelling in glass are
incident at angle a upon a plate of air, which, is bounded again on the other
side by glass. If p be the index of the glass, a.' the angle of refraction, then
sin a' = fj, sin a ; and the retardation, expressed by the equivalent distance in
air, is
2t sec a! — p, *2t tan a' sin a = 2t cos a' ;
and the retardation in phase is 2£cosa'/X, X being as usual the wave-length
in air.
"The first thing to be noticed is that, when a approaches the critical
angle, cos a' becomes as small as we please, and that, consequently, the
retardation corresponding to a given thickness is very much less than at
perpendicular incidence. Hence the glass surfaces need not be so close
as usual.
" A second feature is the increased brilliancy of the light. But the
peculiarity which most demands attention is the lessened influence of a
variation in X upon the phase retardation. A diminution of X of itself
increases the retardation of phase, but since waves of shorter wave-length are
more refrangible, this effect may be more or less perfectly compensated by
the greater obliquity, and consequent diminution in the value of cos a!. We
will investigate the conditions under which the retardation of phase is
stationary in spite of a variation of X.
" In order that X"1 cos a' may be stationary, we must have
X sin a' da' + cos a' d\ = 0,
where (a being constant)
cos of da' = sin a. d/j,.
Thus
.(23)
giving a' when the relation between //, and X is known.
"According to Cauchy's formula, which represents the facts very well
throughout most of the visible spectrum,
fi = A + B\~*, .............................. (24)
so that
If we take, as for Chance's ' extra-dense flint,'
B = -984 x 10-10,
1889] OX ACHROMATIC INTERFERENCE-BANDS. 305
and, as for the soda-lines,
p = 1-65, X = 5-89 x 10-5,
we get
a' =79° 30'.
At this angle of refraction, and with this kind of glass, the retardation of
phase is accordingly nearly independent of wave-length, and therefore the
bands formed, as the thickness varies, are approximately achromatic."
Perfect achromatism would be possible only under a law of dispersion*
ft = A'+R\-* (26)
The above investigation, as given in the Enc. Brit., was intended to apply
to Talbot's manner of experimenting, and it affords a satisfactory explanation
of the formation of achromatic bands. In order to realize the nearly grazing
incidence, the plate of air must be bounded on one side by a prism (Fig. 1).
Fig. i.
Upon this fall nearly parallel rays from a "radiant point of solar light,"
obtained with the aid of a lens of short focus. The bands may be observed
upon a piece of ground glass held behind the prism in the reflected light, or
they may be received directly upon an eyepiece.
These bands undoubtedly correspond to van-ing thicknesses of the plate
of air, just as do the ordinary Newton's rings formed at nearly perpendicular
incidence. For theoretical purposes we have the simplest conditions, if we
suppose the thickness uniform, and that all the rays incident upon the plate
are strictly parallel. Under these suppositions the field is uniform, the
brightness for any kind of light depending upon the precise thickness in
operation. If the thickness be imagined to increase gradually from zero, we
are presented with a certain sequence of colours. When, however, the
relation (23) is satisfied, the formation of colour is postponed, and the series
commences with a number of alternations of black and white. In actual
experiment it would be difficult to realize these conditions. If the surfaces
> [1901. The above formula was given in Enc. Brit. 1888, hot at the time of publication of
the present paper it was thought to be erroneous. The correctness of the original version was
pointed oat by Mr Preston.]
R. HI. 20
306 ON ACHROMATIC INTERFERENCE-BANDS. [163
bounding the plate are inclined to one another, the various parts of the field
correspond to different thicknesses ; and, at any rate if the inclination be
small, there is presented at one view a series of colours, constituting bands,
the same as could only be seen in succession were the parallelism maintained
rigorously.
The achromatism secured by (23) not being absolute, it is of interest to
inquire what number of bands are to be expected. The relative retardation
of phase, with which we have to deal, is 2t cos a'/\, or
If this be stationary for extra-dense glass and for the line D, we have, as
already mentioned, of = 79° 30', and corresponding thereto a = 36° 34'. Taking
this as a prescribed value of a, we may compare the values of (27) for the
lines C, D, E, using the data given by Hopkinson*, viz.: —
C, p = 1-644866, X = -65618 x 10~4,
D, p = 1-650388, \ = -58890 x 1Q-4,
E, fi = 1-657653, X = '52690 x 10~4.
We find
for C (27) = 3036-9 x 2 1,
D (27) = 3094-5 x2£,
E (27) = 2984-3 x 2 1.
These retardations are reckoned in periods. If we suppose that the retarda-
tion for the (7-system is just half a period less than for the .D-system, we
have 57*6 x2£ = ^; so that t « 5$^ centim. Thus about 27 periods of the
D-bands correspond to 26|- of the (7-bands.
If the range of refrangibility contemplated be small, the calculation may
conveniently be conducted algebraically. According to Cauchy's law we may
replace (27) by
2gV(l-/*«Bin«a)fc-4)
V£
Setting p = /JLQ + 8fi, we have approximately
(1 - ^ sin2 a) (p, - A ) = (1 - ^ sin2 a) (^-A)
+ Bfi {(I - /V sin2 a - 2ya0 sin2 a) (/A, - A)}
-A sin2a+....
If a be so chosen that the value of (28) is stationary for /*0, the term of the
first order in fyt vanishes, and we obtain finally as the approximate value
of (28)
2< sin « (p.-
* Proc. Roy. Soc. June 1877.
OX ACHROMATIC DTTERFEREXCE-BAXDS. 307
If now the circumstances be such that n periods of the /*, system correspond
to n — ^ of the n system,
1 _<».-.*) (fry
'
in which the ratio of (Sp^ — A) to 2m does not differ much from unity.
In the application to extra-dense flint the simplified formula
n = (/t.-Ar:(ji-rf ........................ (31)
gives very nearly the same result as that previously found. The number of
bands which approximately coincide is inversely as the square of the range of
refrangibility included.
It must not be overlooked that the preceding investigation, though
satisfactory so far as it goes, is somewhat special on account of the assumption
that the angle of incidence (a) upon the plate of air is the same for the
various colours. If the ravs are parallel before they fall upon the prism, they
cannot remain parallel unless the incidence upon the first surface be perpen-
dicular. There is no reason why this should not be the case : but it is
tantamount to a restriction upon the angle of the prism, since a is determined
by the achromatic condition. If the angle of the prism be other than a. the
required condition will be influenced by the separation of the colours upon
first entering the glass. Although the general character of the phenomenon
is not changed, it may be well to give the calculation applicable to all angles
of prism, as was first done by M. Mascart.
Denoting, as before, by o, o' the angles of incidence and refract k»n upon
the plate of air, let f¥, 0 be the angles of incidence and refraction at the first
surface of the prism (Fig. 2), whose angle is A. Then, if A, equal to nX, be
the retardation,
A = nX=2f cos of, ........................... (32)
before : while the relations among the angular quantities are : —
sin a' = /* sin a, a + @=A, sin £" = /* sin /3. ...(33, 34, 35)
Kg. 2.
as
20—2
308 ON ACHROMATIC INTERFERENCE-BANDS. [163
We have now to inquire under what conditions A/\, or n, will be stationary,
in spite of a variation of X, if ft' be constant. Thus
A, sin a da + cos a'd\ = 0,
while
cos a.' da? = dfj, sin a + /* cos a da,
0 = da + d/3,
0 = dp sin ft + fi cos ft dft.
Accordingly,
cot afdX, , , j
cos a = a/* sin a + fi cos a da
= d[i sin a + cos a tan ft dp = sin A dp/ cos B ;
so that
,. , \dp smA /QfJ.
cot2a = -£• -. — Q (36)
pd\ sin a cos ft
is the condition that n should be stationary. In the more particular case
considered above, ft'=0, /3 = 0, a = A.
These bands, which I should have been inclined to designate after Talbot,
were it not that his name is already connected with another very remarkable
system of bands, are readily observed. For the " radiant point of solar light "
we may substitute, if more convenient, that of the electric arc. A small hole
in a piece of metal held close to the arc allows sufficient light to pass if the
bands are observed without the intervention of a diffusing-screen. At a
distance of say 20 feet the nearly parallel rays fall upon the prism* and
plate, which should be mounted in such a fashion that the pressure may be
varied, and that the whole may be readily turned in azimuth. The coloured
bands are best seen when the surfaces are nearly parallel and pretty close.
It is best to commence observations under these conditions. When the
achromatic azimuth has been found, the interval may be increased. If it is
desired to see a large number of bands, a strip of paper may be interposed
between the surfaces along one edge, so as to form a plate of graduated
thickness. Talbot speaks of from 100 to 200 achromatic bands ; but I do not
think any such large number can be even approximately achromatic. The
composition of the light may be studied with the aid of a pocket spectroscope,
and the appearances correspond with what has been already described under
the head of interference-bands formed from a prismatic spectrum in place of
the usual line of undecomposed light. As has been already remarked, the
colours of fine bands are difficult to appreciate ; and indistinctness is liable to
be attributed to other causes when really due to insufficient achromatism.
The use of a wedge-shaped layer of air is convenient in order to obtain a
simultaneous view of a large number of bands ; but it must not be overlooked
* A right-angled isosceles prism (^ = 45°) answers very well. The plate should be blackened
at the hind surface ; or it may be replaced by a second prism.
1889] ON ACHROMATIC INTERFERENCE-BANDS. 309
that it involves some departure from theoretical simplicity. The proper
development of the light due to any thickness requires repeated reflexions to
and fro within the layer, and at a high degree of obliquity this process
occupies a considerable width. If the band-interval be too small, complica-
tions necessarily ensue, which are probably connected with the fact that the
appearance of the bands changes somewhat according to the distance from
the reflecting combination at which they are observed.
Herschel's Bands.
In the system of bands above discussed, substantially identical (I believe)
with those observed by Talbot, all the rays of a given colour are refracted
under constant angles, the variable element being the thickness of the
plate of air. A system in many respects quite distinct was described by
W. Herschel, and has recently been discussed by M. Mascart*. In this case
the combination of prism and plate remains as before, but the thickness
of the film of air is considered to be constant, the alternations constituting the
bands being dependent upon the varying angles at which the light (even
though of given colour) is refracted. In order to see these bands all that is
necessary is to view a source of light presenting a large angle, such as the
sky, by reflexion in the layer of air. They are formed a little beyond the
limit of total reflexion. They are broad and richly coloured if the layer
of air be thin, but as the thickness increases they become finer, and the
colour is less evident.
The theoretical condition of constant thickness is better satisfied if (after
Mascart) we place the layer of air in the focus of a small radiant point
(e.g. the electric arc) as formed by an achromatic lens of wide angle. In this
case the area concerned may be made so small that the thickness in operation
can scarcely vary, and the ideal Herschel's bands are seen depicted iipon
a screen held in the path of the reflected light. It will of course be under-
stood that bands may be observed of an intermediate character, in the
formation of which both thickness and incidence vary. Herschel's observa-
tions relate to one particular case — that of constant thickness; Talbot 's to
the other especially simple case of constant angle of incidence.
From our present point of view there is, however, one very important
distinction between the two systems of bands. The one system is achromatic,
and the other is not. In order to understand this, it is necessary to follow in
greater detail the theory of Herschel's bands.
We will commence by supposing that the light is homogeneous (\ con-
stant), and inquire into the law of formation of the bands, t being given.
The same equations, (32) &c., apply as before, and also Fig. 2, if we suppose
* Loc. cit. ; also Traitl d'Optiqut, torn. i. Paris, 1889.
310 ON ACHROMATIC INTERFERENCE-BANDS. [163
the course of the rays reversed, so that the direction of the emergent ray is
determined by ft'. The question to be investigated is the relation of ft' to w,
and to the other data of the experiment.
The band of zero order (n = 0) occurs when a = 90°, that is at the limit of
total reflexion. The corresponding values of a, ft, and ft' may be determined
in succession from (33), (34), (35). The value of o' for the nth band is given
immediately by (32). For the width of the band, corresponding to the
change of n into n + 1, we have
and from the other equations,
cos a da = /* cos a da.,
da + dft = 0,
cos ft'dft' = /* cos £d/S ;
so that the apparent width of the nth band is given by
' nV
- per - : - ;
cos ft cos a sin a
In the neighbourhood of the limit of total reflexion sin a' is nearly equal
to unity, and the factors cos ft, cos ft', cos a vary but slowly with the order of
the band and also with the wave-length. Hence the width of the nth band
is approximately proportional to the order, to the square of the wave-length,
and to the inverse square of the thickness.
This series of bands, commencing at the limit of total reflexion, and
gradually increasing in width, are easily observed with Herschel's apparatus
by the aid of a soda-flame. In order to increase the field of view, the flame
may be focused upon the layer of air by a wide-angled lens. The eye
should be adjusted for distant objects, and the thickness of the layer should
be as uniform as possible. For the latter purpose the glass surfaces may
be pressed against two strips of rather thin paper, interposed along two
opposite edges.
We have now to consider what happens when the source of light is white.
According to Airy's principle the centre of the system is to be found where
there is coincidence of bands of order n, in spite of a variation of A, This is
precisely the question already dealt with in connexion with the other system
of bands, and the answer is embodied in (36). About the achromatic centre
thus determined will the visible bands be grouped.
And now the question arises, Are these bands achromatic ? Certainly
not. M. Mascart, to whom is due equation (37), appears to me to mis-
OX ACHBOMATIC
interpret it when he eonchides that the '
At the central band m is the same far the
widths of the various systems of Ait place
to X2. It will be seen that, so fer from the system being
much less so than the ordinary system of mter&rence-banday or of Newton's
rings, in which the width is proportional to the jtnl power of X. And this
theoretical conclusion appears to me to be in haimmiy with observation.
At first sight it may appear strange that an achromatic centre JM^M be
possible with bands proportional to X* The ""fr1"— *•—• Ay**** upon the
tact that the limit of total reflexion, where the bands «•••••»••«. is itself a
function of X
The apparent width of the visible bands depends upon I, but in not,
as might erroneously be supposed, proportional to t~*. For this purpose a in
{37) must be regarded as a function of t In feet, bj (32f|, if a" be givenv
n varies as t X : so that, in estimating the influence of tf other arenmscances
remaining unaltered, the width is proportional to f^. Hence, as the interval
between the surfaces increases, the bands become finer, but the centre does
not shift, nor is there any change in their number as limited by the advent
of chromatic confusion.
Effect of a Priam «JNM Jforfw" * Rin.pt,
If Newton's rings are examined through a prism, some very remarkable
phenomena are exhibited, described in his 24fch observation*.
u When the two object-glasses were laid upon one another, so as to. make
the rings of the colours appear, though with my naked eye I coold not
discern above 8 or 9 of these rings, yet by viewing them through a prism I
have seen a tar greater multitude, insomuch that I eoold number moire than.
40, besides many others which were so very small and close together that
I could not keep my eye steady on them severally so as to number them, but
by their extent I have sometimes estimated them to be more than a hundred.
And I believe the experiment may be improved to the dJEeovery of fiur
greater numbers ; for they seem to be really TmHmitedt though risible only so
far as they can be separated by the refaction, as I shall hereafter rrplahi
"• Bat it was bat one side of these rings — namely, that towards which the
refraction was made — which by that refraction was rendered distinct: add
the other side became more confined than when viewed by the naked eye,
* Ifcritf dTOptafMc. t- L PL 45L "O*
ToiamMS de la frange adbnMlifK ant i
ouverture angnlaire awtekle at foToB ea dBkmpae VB 0
t Oftiekt. B»al»llMfc,.lfaMp. ^aa-onr. juSM
312 ON ACHROMATIC INTERFERENCE-BANDS. [163
insomuch that there I could not discern above 1 or 2, and sometimes none of
those rings, of which I could discern 8 or 9 with my Fig. 3.
naked eye. And their segments or arcs, which on
the other side appeared so numerous, for the most
part exceeded not the third part of a circle. If the
refraction was very great, or the prism very distant
from the object-glasses, the middle part of those arcs
became also confused, so as to disappear and constitute
an even whiteness, while on either side their ends,
as also the whole arcs furthest from the centre, be-
came distincter than before, appearing in the form as you see them designed
in the fifth figure [Fig. 3]."
" The arcs, where they seemed distinctest, were only black and white
successively, without any other colours intermixed. But in other places
there appeared colours, whose order was inverted by the refraction in such
manner that if I first held the prism very near the object-glasses, and then
gradually removed it further off towards my eye, the colours of the 2nd, 3rd,
4th, and following rings shrunk towards the white that emerged between
them, until they wholly vanished into it at the middle of the arc, and
afterwards emerged again in a contrary order. But at the ends of the arcs
they retained their order unchanged."
" I have sometimes so laid one object-glass upon the other, that to the
naked eye they have all over seemed uniformly white, without the least
appearance of any of the coloured rings ; and yet, by viewing them through
a prism, great multitudes of these rings have discovered themselves. And
in like manner, plates of Muscovy glass, and bubbles of glass blown at a
lamp-furnace, which were not so thin as to exhibit any colours to the naked
eye, have through the prism exhibited a great variety of them ranged
irregularly up and down in the form of waves. And so bubbles of water,
before they began to exhibit their colours to the naked eye of a bystander,
have appeared through a prism, girded about with many parallel and
horizontal rings ; to produce which effect it was necessary to hold the prism
parallel, or very nearly parallel, to the horizon, and to dispose it so that the
rays might be refracted upwards."
Newton was evidently much struck with these "so odd circumstances,"
and he explains the occurrence of the rings at unusual thicknesses as due to
the dispersing power of the prism. The blue system being more refracted
than the red, it is possible, under certain conditions, that the nth blue ring
may be so much displaced relatively to the corresponding red ring as at one
part of the circumference to compensate for the different diameters. White
and black stripes may thus be formed in a situation where, without the
prism, the mixture of colours would be complete, so far as could be judged by
the eye.
1889] ON ACHROMATIC INTERFERENCE-BANDS. 313
The simplest case that can be considered is when the "thin plate" is
bounded by plane surfaces inclined to one another at a small angle. Without
the prism, the various systems coincide at the bar of zero order. The width
of the bands is constant for each system, and in passing from one system to
another is proportional to X, Regarded through a prism of small angle
whose refracting edge is parallel to the intersection of the bounding surfaces
of the plate, the various systems no longer coincide for zero order; but by
drawing back the prism, it will always be possible so to adjust the effective
dispersing power as to bring the nth bars to coincidence for any two assigned
colours, and therefore approximately for the entire spectrum. The formation
of the achromatic band, or rather central black bar, depends indeed upon
precisely the same principles as the fictitious shifting of the centre of a
system of Fresnel's bands when viewed through a prism.
In this example the formation of visible rings at unusual thicknesses is
easily understood; but it gives no explanation of the increased numbers
observed by Newton. The width of the bands for any colour is proportional
to X, as well after the displacement by the prism as before. The manner of
overlapping of two systems whose nth bars have been brought to coincidence
is unaltered ; so that the succession of colours in white light, and the number
of perceptible bands, is much as usual.
In order that there may be an achromatic system of bands, it is necessary
that the width of the bands near the centre be the same for the various
colours. As we have seen, this condition cannot be satisfied when the plate
is a true wedge; for then the width for each colour is proportional to X.
If, however, the surfaces bounding the plate be curved, the width for each
colour varies at different parts of the plate, and it is possible that the blue
bands from one part, when seen through the prism, may fit the red bands
from another part of the plate. Of course, when no prism is used, the
sequence of colours is the same whether the boundaries of the plate be
straight or curved.
For simplicity we will first suppose that the surfaces are still cylindrical,
so that the thickness is a function of but one coordinate x, measured in the
direction of refraction. If we choose the point of nearest approach as the
origin of x, the thickness may be taken to be
t=a + ba?, .................................... (38)
a being thus the least distance between the surfaces. The black of the nth
order for wave-length X occurs when
£nX=a+fa*; ................................. (39)
so that the width (&r) of the band at this place (a-) is given by
X/4fer. .................................... (40)
314 ON ACHROMATIC INTERFERENCE-BANDS. [163
Substituting for x from (38), we obtain, as the width of the band of nth order
for any colour,
It will be seen that, while at a given part of the plate the width is
proportional to X, the width for the nth order is a different function depen-
dent upon a. It is with the latter that we are concerned when, by means of
the prism, the nth bars have been brought to coincidence.
If the glasses be in contact, as is usually supposed in the theory of
Newton's rings, a=0; and therefore, by (41), &coc\>, or the width of the
band of the nth order varies as the square root of the wave-length, instead of
as the first power. Even in this case the overlapping and subsequent
obliteration of the bands is much retarded by the use of the prism ; but the
full development of the phenomenon demands that a should be finite. Let
us inquire what is the condition in order that the width of the band of the
nth order may be stationary, as X varies. By (41) it is necessary that the
variation of \*/(%n\ — a) should vanish. Hence
2X(£nX-a)-|nX2=0,
(42)
The thickness of the plate where the nth band for X is formed being £nX,
equation (42) may be taken as signifying that the thickness must be half due
to curvature and half to imperfect contact at the place of nearest approach.
If this condition be satisfied, the achromatism of the nth band, effected
by the prism, carries with it the achromatism of a large number of
neighbouring bands*.
We will return presently to the consideration of the spherically curved
glasses used by Newton, and to the explanation of some of the phenomena
which he observed ; but in the meantime it will be convenient to state the
theory of straight bands in a more analytical form.
Analytical Statement.
If the coordinate £ represent the situation of the nth band, of wave-length
X, then, in any case of straight bands, f may be regarded as a function of n
and X, or, conversely, n (not necessarily integral) may be regarded as a
function of f and X. If we write
n = 0(£\), (43)
* Enc. Brit., "Wave Theory," xxiv. p. 428 (1888). [Vol. in. p. 72.]
OS ACHROMATIC DSTERFEREXCE-BAXDS. 315
and expand by Taylor's theorem,
where
*»*(&, 1.X ........................... (45)
The condition for an achromatic band at £„, X* is
and, farther, the condition for an achromatic system at this place is
* 0. ...................... ...(47)
If these conditions are both satisfied, » becomes very approximately a
fnnct ion of f only throughout the region in question.
In several cases considered in the present paper, the functional relation is
such that
^r (X) denoting a function of X only. The expansion may then be written
] ....... (49)
The line £ = 0 is here of necessity perfectly achromatic. If there be an
achromatic system,
and when this condition is satisfied., the whole field is achromatic, so long as
(fix? can be neglected.
If the width of the bands be a function of X only, N is of the form
(50)
more general than that just considered (48 K though of course less general
than (43). The condition for an achromatic line is
0, .................... (51)
and for an achromatic system,
^=+'(X.) = 0; ...(52)
so that, for an achromatic system, +' and % must both vanish.
316 ON ACHROMATIC INTERFERENCE-BANDS. [163
Curved Interference-Bands.
If the bands are not straight, n must be regarded as a function of a
second coordinate rj, as well as of £ and X. In the equation
n = 0(fciy,X), .............................. (53)
if we ascribe a constant value to X, we have the relation between £, 77
corresponding to any prescribed values of n — that is, the forms of the
interference-bands in homogeneous light. If the light be white, the bands
are in general confused ; but those points are achromatic for which
This is a relation between £ and r\ defining a curve, which we may call the
achromatic curve, corresponding in some respects to the achromatic line of
former investigations, where n is independent of TJ. There is, however, a
distinction of some importance. When n is a function of £ and X only, the
achromatic line is also an achromatic band ; that is, n remains constant as we
proceed along it. But under the present less restricted conditions n is not
constant along (54). The achromatic curve is not an achromatic band ; and,
indeed, achromatic bands do not exist in the same development as before.
They must be regarded as infinitely short, following the lines n = constant,
but existent only at the intersection of these with (54). Practically a small
strip surrounding (54) may be regarded as an achromatic region in which are
visible short achromatic bands, crossing the strip at an angle dependent upon
the precise circumstances of the case.
The application of this theory to the observations of Newton presents no
difficulty. The thickness of the layer of air at the point x, y, measured from
the place of closest approach, is
t = a+b(x* + f); ........................... (55)
and since t = £ n\, the relation of n to x, y, and X is
%n = a\-l + b\-l(z? + y*) ......................... (56)
This equation defines the system of bands when the combination is viewed
directly. The achromatic curve determined by (54) is
and is wholly imaginary if a and b are both positive and finite. Only when
a = 0, that is when the glasses touch, is there an achromatic point x — Q,
y-0,
When a prism is brought into operation, we may suppose that each
homogeneous system is shifted as a whole parallel to x by an amount
OX ACHBOMATIC IXTERFEREXCE-RAXDS. 317
variable from one homogeneous system to another. If the apparent coordi-
nates be f , if. we may write
f=J— /(X), ,r=y. ........................ (57)
Using these in (56), we obtain as the characteristic equation of the rings
viewed through a prism,
a+blS+fWF + l* ™
^2X
The equation of the achromatic curve is then, by (54),
:f+/(A,)-X,Ax,)J« + y=VLr(X)F-a*! ......... (59)
which represents a cinrfr. whose centre is situated upon the axis of £.
If the glasses are in contact (a = 0). the locus is certainly real, and passes
through the point
£+/(*.) = <>, if=0;
that i&. the image with rays of wave-length X, of the point of contact
(x = 0, y = 0> The radius of the circle is \*f\\t), and increases with the
dispersive power of the prism. The other point where the circle meei<s
the axis,
*=2X,/'(A.), 5=0,
marks the place where the bands, being parallel to the achromatic curve.
attain a special development. It is that which we should have found bv an
investigation in which the curvature of the band-systems is ignored.
If a be supposed to increase from zero, other conditions remaining
unaltered, the radius of the achromatic circle decreases, while the centre
maintains its position. The two places where the circle crosses the axis are
thus upon the same side of the image of x= 0, y = 0. When a is such that
4»-Vir(US*. ........................... (60)
the circle shrinks into a point, whose situation is defined by
Since there are two coincident achromatic points upon the axis, the
condition is satisfied for an achromatic system, By (60 i. (61).
n/fe = j?,
so that
f = a + kfa=2a. .............................. (62)
This is the same result as was found before (42) by the simpler treatment of
the question in which points along the axis were alone considered.
If a exceed the value specified in (60), the achromatic curve becomes
wholly imaginary*.
* Compare Uucait. TrmUf fOfti^me. 1. 1. p. 455.
164.
ON BELLS*.
[Philosophical Magazine, xxix. pp. 1 — 17, 1890.]
THE theory of the vibrations of bells is of considerable difficulty. Even
when the thickness of the shell may be treated as very small, as in the case
of air-pump receivers, finger-bowls, claret glasses, &c., the question has given
rise to a difference of opinion. The more difficult problem presented by
church bells, where the thickness of the metal in the region of the sound-bow
(where the clapper strikes) is by no means small, has not yet been attacked.
A complete theoretical investigation is indeed scarcely to be hoped for ; but
one of the principal objects of the present paper is to report the results of an
experimental examination of several church bells, in the course of which some
curious facts have disclosed themselves.
In practice bells are designed to be symmetrical about an axis, and we
shall accordingly suppose that the figures are of revolution, or at least differ
but little from such. Under these circumstances the possible vibrations
divide themselves into classes, according to the number of times the motion
repeats itself round the circumference. In the gravest mode, where the
originally circular boundary becomes elliptical, the motion is once repeated,
that is it occurs twice. The number of nodal meridians, determined by the
points where the circle intersects the ellipse, is four, the meridians corre-
sponding (for example) to longitudes 0° and 180° being reckoned separately.
In like manner we may have 6, 8, 10... nodal meridians, corresponding to
3, 4, 5... cycles of motion. A class of vibrations is also possible which are
symmetrical about the axis, the motion at any point being either in or
perpendicular to the meridional plane. But these are of no acoustical
importance.
* [1901. Some of the results of this investigation had been communicated to the British
Association. (See Report for 1889, p. 491.)]
1890] ON BELLS. 319
The meaning here attached to the word nodal must be carefully observed.
The meridians are not nodal in the sense that there is no motion, but only
that there is no motion normal to the surface. This can be best illustrated
by the simplest case, that of an infinitely long thin circular cylinder vibrating
in two dimensions*. The graver vibrations are here purely flexural, the
circumference remaining everywhere unstretched during the motion. If we
fix our attention upon one mode of vibration of n cycles, the motion at the
surface is usually both radial and tangential. There are, however, 2n points
distributed at equal intervals where the motion is purely tangential, and other
2n points, bisecting the intervals of the former, where the motion is purely
radial. There are thus no places of complete rest; but the first set of points,
or the lines through them parallel to the axis, are called nodal, in the sense
that there is at these places no normal motion.
The two systems of points have important relations to the place where
the vibrations are excited. " When a bell-shaped body is sounded by a blow,
the point of application of the blow is a place of maximum normal motion of
the resulting vibrations, and the same is true when the vibrations are excited
by a violin-bow, as generally in lecture-room experiments. Bells of glass,
such as finger-glasses, are, however, more easily thrown into regular vibration
by friction with the wetted finger carried round the circumference. The
pitch of the resulting sound is the same as that elicited by a tap with the
soft part of the finger ; but inasmuch as the tangential motion of a vibrating
bell has been very generally ignored, the production of sound in this manner
has been felt as a difficulty. It is now scarcely necessary to point out that
the effect of the friction is in the first instance to excite tangential motion,
and that the point of application of the friction is the place where the
tangential motion is greatest, and therefore where the normal motion
vanishes f."
When the symmetry is complete, the system of nodal meridians has no
fixed position, and may adapt itself so as to suit the place at which a normal
blow is delivered. If the point of application of the blow be conceived to
travel round a circle symmetrical with respect to the axis (say, for brevity,
a circle of latitude) the displacement will make no difference to the vibration
considered as a whole, but the effect upon an observer who retains a fixed
position will vary. If the bell be situated in an open space, or if the ear of
the observer be so close that reflexions are relatively unimportant, the sound
disappears as nodes pass by him, swelling to a maximum when the part
nearest to the ear is one of the places of maximum normal motion, which for
brevity we will call loops. In listening to a particular note it would thus be
* Theory of Sound, § 232.
t Theory of Sound, § 234. That the rubbing finger and the violin-bow most be applied at
different points in order to obtain the same vibration was known to Chladni.
320 ON BELLS. [164
possible to determine the number of nodal meridians by watching the
variations of intensity which occur as the place of the blow travels round
a circle of latitude.
In practice the symmetry is seldom so complete that this account of the
matter is sufficient. Theoretically the slightest departure from symmetry
will in general render determinate the positions of the nodal systems. For
each number n of cycles, there is one determinate mode of vibration
with 2n nodes and 2n intermediate loops, and a second determinate mode
in which the nodes and loops of the first mode exchange functions.
Moreover the frequencies of the vibrations in the two modes are slightly
different.
In accordance with the general theory, the vibrations of the two modes, as
dependent upon the situation and magnitude of the initiating blow, are to be
considered separately. The vibrations of the first mode will be excited, unless
the blow occur at a node of this system ; and in various degrees, reaching
a maximum when the blow is delivered at a loop. The intensity, as ap-
preciated by an observer, depends also upon the position of his ear, and will
be greatest when a loop is immediately opposite. As regards the vibrations
of the second mode, they reach a maximum when those of the first mode
disappear, and conversely.
Thus in the case of n cycles, there are 2n places where the first vibration
is not excited and 2n places, midway between the former, where the second
vibration is not excited. At all 4n places the resulting sound is free from
beats. In all other cases both kinds of vibration are excited, and the sound
will be affected by beats. But the prominence of the beats depends upon
more than one circumstance. The intensities of the two vibrations will be
equal when the place of the blow is midway between those which give no
beats. But it does not follow that the audible beats are then most distinct.
The condition to be satisfied is that the intensities shall be equal as they
reach the ear, and this will depend upon the situation of the observer as well
as upon the vigour of the vibrations themselves. Indeed, by suitably choosing
the place of observation it would be theoretically possible to obtain beats with
perfect silences, wherever (in relation to the nodal systems) the blow may be
delivered.
There will now be no difficulty in understanding the procedure adopted
in order to fix the number of cycles corresponding to a given tone. If, in
consequence of a near approach to symmetry, beats are not audible, they are
introduced by suitably loading the vibrating body. By tapping cautiously
round a circle of latitude the places are then investigated where the beats
disappear. But here a decision must not be made too hastily. The in-
audibility of the beats may be favoured by an unsuitable position of the ear,
or of the mouth of the resonator in connexion with the ear. By travelling
1890] ox BELLS. 321
round, a situation is soon found where the observation can be made with the
best advantage. In the neighbourhood of the place where the blow is being
tried there is a loop of the vibration which is most excited and a (coincident)
node of the vibration which is least excited. When the ear is opposite to a
node of the first vibration, and therefore to a loop of the second, the original
inequality is redressed, and distinct beats may be heard even although the
deviation of the blow from a nodal point may be very smalL The accurate
determination in this way of two consecutive places where no beats are
generated is all that is absolutely necessary. The ratio of the entire
circumference of the circle of latitude to the arc between the points repre-
sents 4n. that is four times the number of cycles. Thus, if the arc between
consecutive points proved to be 451*, we should infer that we are dealing with
a vibration of two cycles — the one in which the deformation is ellipticaL As
a greater security against error, it is advisable in practice to determine a
larger number of points where no beats occur. Unless the deviation from
symmetry be considerable, these points should be uniformly distributed along
the circle of latitude *.
In the above process for determining nodes we are supposed to hear
distinctly the tone corresponding to the vibration under investigation. For
this purpose the beats are of assistance in directing the attention : but with
the more difficult subjects, such as church bells, it is advisable to have recourse
to resonators. A set of Helmholtz's pattern, manufactured by Koenig, are
very convenient. The one next higher in pitch to the tone under examination
is chosen and tuned by advancing the finger across the aperture. Without
the security afforded by resonators, the determination of the octave is in my
experience very uncertain. Thus pure tones are often estimated by musicians
an octave too low.
Some years ago I made observations upon the tones of various glass bells.
of which the walls were tolerably thin. A few examples may be given : —
L c, e"b, c'"$-
IL a, c"«. b".
ILL /% V.
The value of a for the gravest tone is 2, for the second 3, and for the
third 4. On account of the irregular shape and thickness only a very rough
comparison with theory is possible ; but it may be worth mention that for
a thin uniform hemispherical bell the frequencies of the three slowest
vibrations should be in the ratios
1 : 2-8102: 5-4316;
* The bells, or gongs, as they are sometimes called, of striking docks often giro disagreeable
beats. A remedy may be found in a suitable rotation of the bell about its axis.
21
322 ON BELLS. [164
so that the tones might be
c> f'$> f"> approximately.
More recently, through the kindness of Messrs Mears and Stainbank,
I have had an opportunity of examining a so-called hemispherical metal bell,
weighing about 3 cwt. A section is shown in Fig. 1. Four tones could be
plainly heard,
e\>, ft, e", b",
the pitch being taken from a harmonium. The gravest tone has a long
duration. When the bell is struck by a hard body, the higher tones are
Fig. 1.
at first predominant, but after a time they die away, and leave el? in
possession of the field. If the striking body be soft, the original pre-
ponderance of the higher elements is less marked.
By the method above described there was no difficulty in showing that
the four tones correspond respectively to n = 2, 3, 4, 5. Thus for the gravest
tone the vibration is elliptical with 4 nodal meridians, for the next tone
there are 6 nodal meridians, and so on. Tapping along a meridian showed
that the sounds became less clear as the edge was departed from, and this in
a continuous manner with no suggestion of a nodal circle of latitude.
A question, to which we shall recur in connexion with church bells, here
suggests itself. Which of the various coexisting tones characterizes the pitch
of the bell as a whole ? It would appear to be the third in order, for the
founders give the pitch as E nat.
My first attempts upon church bells were made in September 1879, upon
the second bell (reckoned from the highest) of the Terling peal ; and I was
much puzzled to reconcile the pitch of the various tones, determined by
resonators, with the effective pitch of the bell, when heard from a distance in
conjunction with the other bells of the peal. There was a general agreement
that the five notes of the peal were
/8, g$, a$, b, c$,
according to harmonium pitch, so that the note of the second bell was b.
A tone of pitch att could be heard, but at that time nothing coincident with
b or its octaves. Subsequently, in January 1880, the b was found among the
1890] ox BELLS. 323
tones of the bell, but at much higher pitch than had been expected. The
five gravest tones were determined to be
df, a'S, d", tf'S+, 6";
so that the nominal note of the bell agreed with the fifth component tone,
and with no graver one. The octaves are here indicated by dashes in the
usual way. the c immediately below the d' being the middle c of the musical
scale.
Attempts were then made to identify the modes of vibration corresponding
to the various tones, but with only partial success. By tapping round the
sound-bow it appeared that the minima of beats for d' occurred at intervals
equal to | of the circumference, indicating that the deformation in this mode
was elliptical (« = 2 ), as had been expected. In like manner g"Z gave n = 3 ;
but on account of the difficulty of experimenting in the belfry, the results
were not wholly satisfactory, and I was unable to determine the modes for
the other tones. One observation, however, of importance could be made.
All five tones were affected with beats, from which it was concluded that
none of them could be due to symmetrical vibrations, as. till then, had been
thought not unlikely.
Nothing further worthy of record was effected until last year, when I
obtained from Messrs Hears and Stainbank the loan of a 6-cwt. bell. Hung
in the laboratory at a convenient height, and with freedom of access to all
parts of the circumference, this bell afforded a more convenient subject for
experiment, and I was able to make the observations by which before I had
been baffled. Former experience having shown me the difficulty of estimating
the pitch of an isolated bell, I was anxious to have the judgment of the
founders expressed in a definite form, and they were good enough to supply
me with a fork tuned to the pitch of the bell. By my harmonium the
fork is d".
Bv tapping the bell in various places with a hammer or mallet, and
listening with resonators, it was not difficult to detect 6 tones. They were
identified with the following notes of the harmonium * : —
e', c", f"+, b"t, d'", /'".
(4) (4) (6) (6) (8)
As in the former case, the nominal pitch is governed by the fifth
component tone, whose pitch is, however, an octave higher than that of
the representative fork. It is to be understood, of course, that each of the
6 tones in the above series is really double, and that in some cases the
components of a pair differ sufficiently to give rise to somewhat rapid beats.
* In comparisons of this kind the observer most bear in mind the highly compound character
of the notes of a reed instrument. It is usually a wise precaution to ascertain that a similar
effect is not produced by the octave (or twelfth) above.
21—2
324 ON BELLS. [164
The sign + affixed to /" indicates that the tone of the bell was decidedly
sharp in comparison with the note of the instrument.
I now proceeded to determine, as far as possible, the characters of the
various modes of vibration by observations upon the dependence of the sounds
upon the place of tapping in the manner already described. By tapping
round a circle of latitude it was easy to prove that for (each of the approxi-
mately coincident tones of) e' there were 4 nodal meridians. Again, on
tapping along a meridian to find whether there were any nodal circles of
latitude, it became evident that there were none such. At the same time
differences of intensity were observed. This tone is more fully developed
when the blow is delivered about midway between the crown and rim of the
bell than at other places.
The next tone is c". Observation showed that for this vibration also
there are four, and but four, nodal meridians. But now there is a well-defined
nodal circle of latitude, situated about a quarter of the way up from the rim
towards the crown. As heard with the resonator, this tone disappears when
the blow is accurately delivered at some point of this circle, but revives with
a very small displacement on either side. The nodal circle and the four
meridians divide the surface into segments, over each of which the normal
motion is of one sign.
To the tone /" correspond 6 nodal meridians. There is no well-defined
nodal circle. The sound is indeed very faint, when the tap is much removed
from the sound-bow ; it was thought to fall to a minimum when the tap was
about halfway up.
The three graver tones are heard loudly from the sound-bow. But the
next in order, 6"b, is there scarcely audible, unless the blow be delivered to
the rim itself in a tangential direction. The maximum effect occurs at about
halfway up. Tapping round the circle, we find that there are 6 nodal
meridians.
The fifth tone, d'", is heard loudly from the sound-bow, but soon falls off
when the locality of the blow is varied, and in the upper three-fourths of
the bell it is very faint. No distinct circular node could be detected.
Tapping round the circumference showed that there were here 8 nodal
meridians.
The highest tone recorded,/'", was not easy of observation, and I did not
succeed in satisfying myself as to the character of the vibration. The tone
was perhaps best heard when the blow was delivered at a point a little below
the crown.
All the above tones, except f", were tolerably close in pitch to the
corresponding notes of the harmonium.
1890] ox BELLS. 325
Although the above results seemed perfectly unambiguous, I was glad to
have an opportunity of confirming them by examination of another bell.
This was afforded by a loan of a bell cast by Taylor, of Loughborough, and
destined for the church of Ampton, Suffolk, where it now hangs. Its weight
is somewhat less than 4 cwt., and the nominal pitch is rf. The observations
were entirely confirmatory of the results obtained from Messrs Mears's bell
The tones were
e't-2, d"-Q, f"+4, b"V—b", d'", g'";
(4) (4) (6) (6) (8)
the correspondence between the order of the tone and the number of nodal
meridians being as before. In the case of d" there was the same well-
defined nodal circle. The highest tone, g'", was but imperfectly heard,
and no investigation could be made of the corresponding mode of vibration.
In the specification of pitch the numerals following the note indicate bv
how much the frequency for the bell differed from that of the harmonium.
Thus the gravest tone e>}9 gave 2 beats per second, and was flat. When the
number exceeds 3, it is the result of somewhat rough estimation and cannot
be trusted to be quite accurate. Moreover, as has been explained, there are
in strictness two frequencies under each head, and these often differ sensiblv.
In the case of the 4th tone, b"b — b" means that, as nearly as could be judged,
the pitch of the bell was midway between the two specified notes of the
harmonium.
The sounds of bells may be elicited otherwise than by blows. Advantage
may often be taken of the response to the notes of the harmonium, to the
voice, or to organ-pipes, sounded in the neighbourhood. In these cases the
subsequent resonance of the bell has the character of a pure tone. Perhaps
the most striking experiment is with a tuning-fork. A massive e'b (e' on
the c'=256 scale) fork, tuned with wax. and placed upon the waist of the
Ampton bell, called forth a magnificent resonance, which lasted for some
time after removal and damping of the fork. The sound is so utterly
unlike that usually associated with bells that an air of mystery envelops the
phenomenon. The fork may be excited either by a preliminary blow upon
a pad (in practice it was the bent knee of the observer), or by bowing when
in contact with the bell. In either case the adjustment of pitch should
be very precise, and it is usually necessary to distinguish the two nearly
coincident tones of the bell. One of these is to be chosen, and the fork is to
be held near a loop of the corresponding mode of vibration. In practice the
simplest way to effect the tuning is to watch the course of things after the
vibrating fork has been brought into contact with the bell. When the tuning
is good the sound swells continuously. Any beats that are heard must be
gradually slowed down by adjustment of wax, until they disappear.
326
ON BELLS.
[164
ac s «o
<N (N Tt< CO
+ + + +
CM (N ^ CO
** _c^ jr^ ?»5 ^>
IF
CO O 00 d- <N
1 1 + + +
CO O 00 CO <N
1 1 + + +
6C >>§5
CO CD "* CO
+ + + + J*
« CO -* CO
•so1"1
H
g
1
O
1
« 8 « »
0)
1
ifi
+ T + T T
8 It* 2o ** **
CO •* CO ** CO
+ 1+^1
« -o j^ ^^
.
p1
3
co •* CD n CM
1 J + J 1
^i **5i "\5 »>& ^^
S-,
a
1
CO "^ CD CO CM
' ' + 1 1
** ** _C^ £k 0
•i?
t
Tf *S 1-H CO
J i :*• J i
^ 5^ X S
&
4-
2 J + * *j ^
(N CO Tt< 8
1 i + '
oT
I2
k % -* - ^ S
"""" :
1890] ON BELLS. 327
Observations upon the two bells in the laboratory having settled the
modes of vibration corresponding to the five gravest tones, other bells of the
church pattern can be sufficiently investigated by simple determinations of
pitch. I give in tabular form results of this kind for a Belgian bell, kindly
placed at my disposal by Mr Haweis, and for the five bells of the Terling
peal. For completeness' sake the Table includes also the corresponding
results for the two bells already described.
It will be seen that in every case where the test can be applied, it is the
fifth tone in order which agrees with the nominal pitch of the bell. The
reader will not be more surprised at this conclusion than I was, but there
seems to be no escape from it. Even apart from estimates of pitch, an
examination of the tones of the bells of the Terling peal proves that it is
only from the third and fifth tones that a tolerable diatonic scale can be
constructed. Observations in the neighbourhood of bells do not suggest any
special predominance of the fifth tone, but the effect is a good deal modified
by distance.
It has been suggested, I think by Helmholtz, that the aim of the original
designers of bells may have been to bring into harmonic relations tones which
might otherwise cause a disagreeable effect. If this be so, the result cannot
be considered very successful. A glance at the Table shows that in almost
every case there occur intervals which would usually be counted intolerable,
such as the false octave. Terling (5) is the only bell which avoids this false
interval between the two first tones ; but the improvement here shown in
this respect still leaves much to be desired, when we consider the relation
of these two tones to the fifth tone, and the nominal pitch of the bell. Upon
the assumption that the nominal pitch is governed by that of the fifth tone,
I have exhibited in the second part of the above Table the relationship in
each case of the various tones to this one.
One of my objects in this investigation having been to find out, if possible,
wherein lay the difference between good and bad bells, I was anxious to
interpret in accordance with my results the observations of Mr Haweis, who
has given so much attention to the subject. The comparison is, however,
not free from difficulty. Mr Haweis says*: — "The true Belgian bell when
struck a little above the rim gives the dominant note of the bell ; when
struck two-thirds up it gives the third ; and near the top the fifth ; and the
' true ' bell is that in which the third and fifth (to leave out a multitude
of other partials) are heard in right relative subordination to the dominant
note."
If I am right in respect of the dominant note, the third spoken of by
Mr Haweis must be the minor third (or, rather, major sixth) presented by
• Times, October 29, 1878.
328 ON BELLS. [164
the tone third in order, which it so happens is nearly the same interval in
all cases. The only fifth which occurs is that of the tone fourth in order.
Thus, according to Mr Haweis's views, the best bell in the series would be
Terling (1), for which the minor chord of the last three tones is nearly true.
It must be remarked, however, that the tone fourth in order is scarcely heard
in the normal use of the bell, so that its pitch can hardly be of importance
directly, although it may afford a useful criterion of the character of the bell
as a whole. It is evident that the first and second tones of Terling (1) are
quite out of relation with the higher ones. If the first could be depressed
a semitone and the second raised a whole tone, harmonic relations would
prevail throughout.
Judging from the variety presented in the Table, it would seem not a
hopeless task so to construct a bell that all the important tones should be
brought into harmonic relation ; but it would require so much tentative work
that it could only be undertaken advantageously by one in connexion with
a foundry. As to what advantage would be gained in the event of success,
I find it difficult to form an opinion. All I can say is that the dissonant
effect of the inharmonious intervals actually met with is less than one would
have expected from a musical point of view ; although the fact is to a great
extent explained by Helmholtz's theory of dissonance.
One other point I will touch upon, though with great diffidence. If there
is anything well established in theoretical acoustics it is that the frequencies
of vibration of similar bodies formed of similar material are inversely as the
linear dimensions — a law which extends to all the possible modes of vibration.
Hence, if the dimensions are halved, all the tones should rise in pitch by an
exact octave. I have been given to understand, however, that bells are not
designed upon this principle of similarity, and that the attempt to do so
would result in failure. It is just possible that differences in cooling may
influence the hardness, and so interfere with the similarity of corresponding
parts, in spite of uniformity in the chemical composition of the metal ; but
this explanation does not appear adequate. Can it be that when the scale
of a bell is altered it is desirable at the same time to modify the relative
intensities, or even the relative frequencies, of the various partials ?
Observations conducted about ten years ago upon the manner of bending
of bell-shaped bodies — waste-paper baskets and various structures of flexible
material — led me to think that these shapes were especially stiff as regards
the principal mode of bending (with four nodal meridians) to forces applied
normally and near the rim, and that possibly one of the objects of the
particular form adopted for bells might be to diminish the preponderance
of the gravest tone. To illustrate this I made calculations, according to the
theory of the paper already alluded to, of the deformation by pure bending
of thin shells in the form of hyperboloids of revolution, and in certain
1890]
ON BELU3.
329
composite forms built up of cylinders and cones so as represent approximately
the actual shape of bells. In the case of the hyperboloid of one sheet
(Fig. 2), completed by a crown in the form of a circular disk through the
centre, and extending across the aperture, it appeared that there was no
nodal circle for ;* = 2. The investigation is appended to this paper.
The composite forms, Figs. 4 and 5, represent the actual bell (Fig. 3*)
as nearly as may be. At the top is a circular disk, and to this is attached
Fig. 2.
Fig. 3.
a cylindrical segment. The expanding part of the bell is represented by
one (Fig. 4), or with better approximation by two (Fig. 5), segments of cones.
The calculations are too tedious to be reproduced here, but the results are
shown upon the figures. In both cases there is a circular node N for n = 2,
not far removed from the rim, and in Fig. 5 very nearly at the place which
represents the sound-bow of an actual bell. In the latter case there is a
node N' for n = 3 near the middle of the intermediate conical segment.
The nodal circle for n = 2 has been verified experimentally upon a bell
constructed of thin sheet zinc in the form of Fig. 5. The gravest note, G«>
and the corresponding mode of vibration, could be investigated exactly in the
Fig. 4.
Fig. 5.
manner already described. In each mode of this kind there were four nodal
meridians, and a very well defined nodal circle. The situation of this circle
was not quite so low as according to calculation ; it was almost exactly in the
middle of the lower conical segment. By merely handling the model it was
* Copied from Zamminer, Die Musik und die musikalischen Instrumente. Giessen, 1855.
330 ON BELLS. [164
easy to recognize that it was stiff to forces applied at N, but flexible higher
up, in the neighbourhood of N'.
It is clear that the actual behaviour of a church bell differs widely from
that of a bell infinitely thin ; and that this should be the case need not
surprise us when we consider the actual ratio of the thickness at the sound-
bow to the interval between consecutive nodal meridians. I think, however,
that the form of the bell does really tend to render the gravest tone less
prominent.
APPENDIX.
On the Bending of a Hyperboloid of Revolution.
The deformation of the general surface of revolution was briefly treated in
a former paper*. The point whose original cylindrical coordinates are z, r, </>,
is supposed to undergo such a displacement that its coordinates become
z + Bz, r + Br, <£ + B(f>.
The altered value (ds + dBs) of the element of length traced upon the
surface is given by
(ds + dBs)* = (dz + dBz)* + (r + Br)2 (d<f> + dB<f>)2 + (dr + dSr)2.
Hence, if the displacement be such that the element is unextended,
dz dBz + r*d(f> d8<f> + rBr (dtyf + dr dBr = 0.
Now
,., dBz , dBz ,
ddz = -3- dz + ^— d<b,
dz d<f>
„,.'* • '*
dz
and by the equation to the surface
, dr , dr , .
dr = -j- dz + -j-. dd>,
dz d<> '
in which, by hypothesis, dr/d(f> = 0. Thus
., {dSz drdSr) ., f 9dSd> ,
(dzY \-j-+-j-- + (d$f \ r- ~-yf + rSr
{ dz dz dz } ' { d<(>
* " On the Infinitesimal Bending of Surfaces of Kevolution," Proc. Math. Soc. xin. p. 4 (1881).
[Vol. i. p. 551.]
1890] ox BELLS. 331
If the displacement be of such a character that no tine traced upon the
surface is altered in length, the coefficients of (dzf, (<fyf, dsd4>.m the above
equation, must vanish separately, so that
d&z .d&4 A-dSr
From these, by elimination of Sr,
dSz dr d
•<»
.(2)
' ...............
from which again, by elimination of Sr,
For the purposes of the present problem we may assume that c<£ varies
as cos*<6. or as sin#$: thus,
is the equation by which the form of £<£ as a function of r is to be
determined.
When application is made to the hyperboloid of one sheet
we find, since
The solution of this equation is expressed by an auxiliary variable
such that
2 = 6 tang, r = asecg (10)
in the form
fy = A cos«g + fisin«x. (11)
In order to verify this it is only necessary to observe that by (10)
d a'd
332 ON BELLS. [164
We will now apply this solution to an inextensible surface formed by half
the hyperboloid and a crown stretching across in the plane of symmetry z — 0
(Fig. 2). The deformation of this crown can take place only in the direction
perpendicular to its plane, so that Br = 0, S(j> = 0. These conditions must
apply also to the hyperboloid at the place of attachment to the crown.
Hence 8</> must vanish with z, or, which is the same, with ^. Accordingly
.4 = 0 in (11); and dropping the constant multiplier we may take as the
solution
S(f> = sms% coss<£, ........................... (12)
and in correspondence therewith by (2) and (3)
s<f> ....................................... (13)
X + s tan X
It is evident from these equations that, whatever may be the value of s, there
is no circle of latitude over which both £</> (or Sr) and Bz vanish*. Hence
there can be no circular nodal line in the absolute sense. But just as there
are meridians (sin s<f> = 0) on which the normal motion vanishes, so there may
be nodal circles in this more limited sense. The condition to be satisfied
is obviously
&r/Sz = dr/dz ;
or in the present case
sin 2^ + 2s tan s% (sin2 x + &2/a2) = ° ................ (1 5)
In this equation the range of ^ is from 0 to \TT\ and thus there can be
solutions only when tan s% is negative.
In the case s = 2 the equation reduces to
which can never be satisfied.
When s = 3, the roots, if any, must lie between ^ = 30° and ^ = 60°.
A more detailed consideration shows that there is but one root, and that
it occurs when % is a little short of 60°.
* A corresponding proposition may be proved more generally, that is without limitation
to the hyperboloid.
165.
THE CLARK STANDARD CELL.
[The Electrician, Jan. 1890, p. 285.]
IN order to expedite the settlement of any open questions respecting
Clark cells. I send the following remarks upon the Paper of Prof. Carhart.
reprinted in The Electrician, p. 271, from the Philosophical Magazine, entitled
"An Improved Standard Clark Cell with Low Temperature Coefficient.
[Phil. Mag. XXVHI. p. 420, 1889.]
In the first place Prof. Carhart appears to me rather to exaggerate the
inconvenience arising from temperature changes in a Clark cell of ordinary
construction. The coefficient is about "00077 per degree cent., so that an
uncertainty of a whole degree, affecting the E.M.F. by less than 77^. would
hardly be of practical importance. The sensitiveness to temperature is in
fact only about the double of that of German-silver resistance coils. In a
suitable situation, and with the most ordinary care, the temperature would
not be uncertain to more than one or two tenths. I have found it possible
to work even closer than this in a room (next the roof), far from specially
suitable, without any particular precautions: but if desired, it is easy to
reduce the uncertainty under this head by some such plan as embedding the
cell, with a thermometer bulb, in a vessel of sand.
The really serious question is whether the temperature coefficient itself is
liable to important variation without assignable cause. If it be uncertain
whether the proper coefficient is "00077, or, as in Prof. Carhart's cells, -00039,
the utility of the standard would indeed be seriously compromised.
Undoubtedly the lower coefficient would be an advantage in itself, if it
could be obtained without loss in other respects. The principal feature
insisted upon by Prof. Carhart is the separation of the zinc from the
mercurous salt; but my experience is totally opposed to the view that the
lower coefficient can thus be secured. The separation actually occurred in a
large number of the cells upon which I experimented, especially in those of
the H pattern*, where the mercury and mercurous salt occupied one leg. and
* Phil. Tnuu. 1884. [Vol. n. p. 315.]
334
THE CLARK STANDARD CELL.
[165
an amalgam of zinc the other. The arrangement is shown in the figure.
That these cells have practically the same temperature coefficient as others
in which the paste touches the zinc, is proved
by Table XIII. of my second Paper*. I am
thus at a loss to explain the low temperature
coefficient of Prof. Carhart's cells, unless indeed
upon the supposition that his solutions were
not throughout saturated with zinc sulphate. In
this case the coefficient is just what might have
been expected, for I found from two cells of
this description the coefficient '00038. It may
be remarked that the H form is safer in this
respect than those in which the zinc is at the
top of the liquid, especially when removed from
the paste ; for the part of the liquid where
saturation is of importance is that in contact
with the zinc. At the top of the column the
salt may easily become deficient, when the
temperature rises, even though there be plenty
of undissolved crystals below. The objections
to unsaturated solutions are discussed in my
Papers. They turn upon the difficulty of pre-
H-PATTERN CELL.
B, Amalgam of Zinc ; C, Pure
Mercury; D, Mercurous Sulphate;
E, Saturated Solution of Zinc
Sulphate; F, Corks. At the bottom
of each leg of the cell a platinum
wire, sealed through the glass, is
paring a standard solution, and upon the liability shown,
to change with evaporation.
I quite agree with Prof. Carhart as to the importance of pure mercury.
And there is undoubtedly something yet to be done in respect of the
mercurous sulphate. I may remark that the sample used by me did not turn
yellow when treated simply with zinc solution, but only when rubbed up also
with zinc carbonate. In the cells made by Mr M. Evans the paste was quite
white, but this did not prevent the variation with temperature being the
same as in other cases. (Table XIV., Tlt T3.)
I once came across a sample of mercurous sulphate with which it was
difficult to prepare satisfactory cells. When rubbed up with zinc carbonate
and zinc sulphate solution it turned dark green instead of yellow. Until the
question is further elucidated I should be disposed to avoid a sample which
behaved in this way.
* Phil. Tram. 1885,
[Vol. ii. p. 453.]
166.
ON THE VIBRATIONS OF AN ATMOSPHERE.
[Philosophical Magazine, xxix. pp. 173—180, 1890.]
IN order to introduce greater precision into our ideas respecting the
behaviour of the Earth's Atmosphere, it seems advisable to solve anv
problems that may present themselves, even though the search for simplicity
may lead us to stray rather far from the actual question. It is proposed here
to consider the case of an atmosphere composed of gas which obeys Boyle's
law, viz. such that the pressure is always proportional to the density. And
in the first instance we shall neglect the curvature and rotation of the earth,
supposing that the strata of equal density are parallel planes perpendicular
to the direction in which gravity acts.
If p, a be the equilibrium pressure and density at the height z, then
and by Boyle's law,
p = a2*, .................................... (2)
where a is the velocity of sound. Hence
*„. .................. (8)
*
and
a- = <r0e-^<*2, .............................. (4)
where <rc is the density at z = 0. According to this law, as is well known,
there is no limit to the height of the atmosphere.
Before proceeding further, let us pause for a moment to consider how the
density at various heights would be affected by a small change of tempera-
ture, altering a to a', the whole quantity of air and therefore the pressure p0
336 ON THE VIBRATIONS OF AN ATMOSPHERE. [166
at the surface remaining unchanged. If the dashes relate to the second state
of things, we have
- a- =
while
a2o-0 = a'2<r0'.
If a'2 — a2 = Sa2, we may write approximately
** Sin**
p, a? a2
The alteration of pressure vanishes when z — 0, and also when z = oc . The
maximum occurs when gz/a? = 1, that is when p=p0/e. But relatively to tr,
(p'—po) increases continually with z.
Again, if p denote the proportional variation of density,
-1).
If a'2 > a2, p is negative when z = 0, and becomes + oo when z = oo . The
transition p = 0 occurs when gz\o?= 1, that is at the same place where p — p
reaches a maximum.
In considering the small vibrations, the component velocities at any point
are denoted by u, v, w, the original density cr becomes (a + ap), and the
increment of pressure is 8p. On neglecting the squares of small quantities
the equation of continuity is
dp du dv dw da-
dt dx dy dz dz
or by (3),
dp du dv dw gw
-r: + -j h -j p j r = V 1 0 1
dt dx dy dz a2
The dynamical equations are
dSp du d&p dv d8p dw
~j — — °" ~r , —r^ — — 0" -Ji > ~j — — <7°"P — <r -j ',
dx dt dy dt dz dt
or by (3), since Sp = tfcrp,
dp du dp dv ndp dw
adx = ~dt> ady=~dt' a~dz = ~df (°
We will consider first the case of one dimension, where u, v vanish,
while p, w are functions of z and t only. From (5) and (6),
dp , dw gw_.
,__,
adz~ dt'
1890] OX THE VIBRATIONS OF AX ATMOSPHERE.
or by elimination of p,
a dm
— " — — — _ y. ^z tn\
as dP d*> a'dz'
The right-hand member of (9) may be written
(i_xy «,_.£«
and in this the latter term may be neglected when the variation of w with
respect to z is not too slow. If X be of the nature of the wave-length, dvc dz
is comparable with w \: and the simplification is justifiable when a* is large
in comparison with g\t that is when the velocity of sound is great in
comparison with that of gravity-waves (as upon water) of wave-length X.
The equation then becomes
d*w
or, if
,c= We*****, .............................. (10)
o1.«i*W7«fo': ........................ (11)
the ordinary equation of sound in a uniform medium. Waves of the kind
contemplated are therefore propagated without change of type except for the
effect of the exponential factor in (10), indicating the increase of motion as
the waves pass upwards. This increase is necessary in order that the same
amount of energy may be conveyed in spite of the growing attenuation of the
medium. In feet w*<r must retain its value, as the waves pass on.
If w vary as j**, the original equation (9) becomes
Let M,, w, be the roots of
g n-
TO2-- iii+-
so that
then the solution of (12) is
w = A€f** + B«r*^ (14)
A and B denoting arbitrary constants in which the foctor e**1 may be sup-
posed to be included.
The case already considered corresponds to the neglect of g3 in the radical
of (13), so that
2a»
and
wtrW = A***+***+B *-«-*** (15)
R. m. --
338 ON THE VIBRATIONS OF AN ATMOSPHERE. [166
A wave propagated upwards is thus
Cosn(t-z/a), ........................ (16)
and there is nothing of the nature of reflexion from the upper atmosphere.
A stationary wave would be of type
w = e\gzia? COBnt sin(w.z/a), ..................... (17)
w being supposed to vanish with z. According to (17), the energy of the
vibration is the same in every wave-length, not diminishing with elevation.
The viscosity of the rarefied air in the upper regions would suffice to put
a stop to such a motion, which cannot therefore be taken to represent
anything that could actually happen.
When 2na <g, the values of ra from (13) are real, and are both positive.
We will suppose that m^ is greater than m.2. If w vanish with z, we have
from (14) as the expression of the stationary vibration
w = cosnt(em>z-em'z), ........................... (18)
which shows that w is of one sign throughout. Again by (8)
(19)
m2
Hence dpfdz, proportional to w, is of one sign throughout ; p itself is negative
for small values of z, and positive for large values, vanishing once when
e(»h-«j* = mi/TOa ............................ (20)
When n is small, we have approximately
so that p vanishes when
or by (4) when
(7/0-0= naaV0" ............................... (23)
Below the point determined by (23) the variation of density is of one sign
and above it of the contrary sign. The integrated variation of density,
roo
represented by I <rp dz, vanishes, as of course it should do.
Jo
It may be of interest to give a numerical example of (23). Let us
suppose that the period is one hour, so that in c.G.S. measure rc = 27r/3600.
We take a = 33 x 104, g = 981. Then o-/o-0 = ?fa ; showing that even for
this moderate period the change of sign does not occur until a high degree
of rarefaction is reached.
In discarding the restriction to one dimension, we may suppose, without
real loss of generality, that v = 0, and that u, w, p are functions of x and z
1890] OX THE VIBRATIONS OF AX ATMOSPHERE. 339
only. Further, we may suppose that x occurs only in the factor &**: that is,
that the motion is periodic with respect to x in the wave-length 2,ir-k: and
that as before t occurs only in the factor e1**. Equations (5), (6) then become
Q, (24)
(25)
a*dp dz = -inw; (26)
from which if we eliminate n, IP, we get
an equation which may be solved in the same form as (12).
One obvious solution of (27) is of importance. If dp dz = Q, so that
w = Q, the equations are satisfied by
n'= &a*. ................................. (28)
Every horizontal stratum moves alike, and the proportional variation of
density (p) is the same at all levels. The possibility of such a motion is
evident beforehand, since on account of the assumption of Boyle's law the
velocity of sound is the same throughout.
In the application to meteorology, the shortness of the more important
periods of the vertical motion suggests that an " equilibrium theory " of this
motion may be adequate. For vibrations like those of (28) there is no
difficulty in taking account of the earth's curvature. For the motion is
that of a simple spherical sheet of air, considered in my book upon the
Theory of Sound, § 333. If r be the radius of the earth, the equation
determining the frequency of the vibration corresponding to the harmonic
of order h is
»»r3 = A(A + l)aa? .............................. (29)
the actual frequency being n/2ir. If T be the period, we have
-
For h = 1, corresponding to a swaying of the atmosphere from one side of the
earth to the opposite
and in like manner for h = 25
(32)
To reduce these results to numbers we may take for the earth's quadrant
\-irr • = 10* cm.; and if we take for a the velocity of sound at 0~ as ordinarily
observed, or as calculated upon Laplace's theory, viz. 33 x 10* cm. 'sec., we
shall find
32—2
340 ON THE VIBRATIONS OF AN ATMOSPHERE. [166
On the same basis,
T2 = 137 hours.
It must, however, be remarked that the suitability of this value of a is very
doubtful, and that the suppositions of the present paper are inconsistent with
the use of Laplace's correction to Newton's theory of sound propagation.
In a more elaborate treatment a difficult question would present itself as to
whether the heat and cold developed during atmospheric vibrations could be
supposed to remain undissipated. It is evidently one thing to make this
supposition for sonorous vibrations, and another for vibrations of about
24 hours period. If the dissipation were neither very rapid nor very slow
in comparison with diurnal changes — and the latter alternative at least
seems improbable — the vibrations would be subject to the damping action
discussed by Stokes*.
In any case the near approach of TJ to 24 hours, and of r.2 to 12 hours,
may well be very important. Beforehand the diurnal variation of the
barometer would have been expected to be much more conspicuous than
the semi-diurnal. The relative magnitude of the latter, as observed at most
parts of the earth's surface, is still a mystery, all the attempted explanations
being illusory. It is difficult to see how the operative forces can be mainly
semi-diurnal in character; and if the effect is so, the readiest explanation
would be in a near coincidence between the natural period and 12 hours.
According to this view the semi-diurnal barometric movement should be the
same at the sea-level all round the earth, varying (at the equinoxes) merely
as the square of the cosine of the latitude, except in consequence of local
disturbances due to want of uniformity in the condition of the earth's surface.
* Phil. Mag. [4] i. p. 305, 1851. Theory of Sound, § 247.
167.
OX THE TENSION OF RECENTLY FORMED LIQUID SURFACES.
[Proceedings of the Royal Society, XLYH. pp. 281—287, 1890.]
IT has long been a mjsteij why a few liquids, such as solutions of soap
and saponine, should stand so far in advance of others in regard to their
capability of extension into large and tolerably durable laminae. The subject
was specially considered by Plateau in his valuable researches, but with
results which cannot be regarded as wholly satisfactory. In his view the
question is one of the ratio between capillary tension and superficial viscosity.
Some of the facts adduced certainly favour a connexion between the pheno-
mena attributed to the latter property and capability .of extension: but the
superficial viscosity™ is not clearly defined, and itself stands in need of
explanation.
It appears to me that there is much to be said in favour of the suggestion
of Marangoni* to the effect that both capability of extension and so-called
superficial viscosity are due to the presence upon the body of the liquid of
a coating, or pellicle, composed of matter whose inherent capillary force is less
than that of the mass. By means of variations in this coating. Marangoni
explains the indisputable fact that in vertical soap films the effective tension
is different at various levels. Were the tension rigorously constant, as it is
sometimes inadvertently stated to be, gravity would inevitably assert itself.
and the central parts would fall 16 feet in the first second of time. By a
self-acting adjustment the coating will everywhere assume such thickness as
to afford the necessary tension, and thus any part of the film, considered
without distinction of its various layers, is in equilibrium. There is nothing,
however, to prevent the interior layers of a moderately thick film from
draining down. But this motion, taking place as it were between two fixed
walls, is comparatively slow, being much impeded by ordinary fluid viscosity.
In the case of soap, the formation of the pellicle is attributed by
Marangoni to the action of atmospheric carbonic acid, liberating the fatty
» 3Tm*n Cime*to, Yob. T. TT. 1*71— TS, p- 339.
342 ON THE TENSION OF RECENTLY FORMED LIQUID SURFACES. [167
acid from its combination with alkali. On the other hand, Sondhauss*
found that the properties of the liquid, and the films themselves, are better
conserved when the atmosphere is excluded by hydrogen ; and I have myself
observed a rapid deterioration of very dilute solutions of oleate of soda when
exposed to the air. In this case a remedy may be found in the addition
of caustic potash. It is to be observed, moreover, that, as has long been
known, the capillary forces are themselves quite capable of overcoming weak
chemical affinities, and will operate in the direction required.
A strong argument in favour of Marangoni's [general] theory is afforded
by his observation-}-, that within very wide limits the superficial tension of
soap solutions, as determined by capillary tubes, is almost independent of the
strength. My purpose in this note is to put forward some new facts tending
strongly to the same conclusion.
It occurred to me that, if the low tension of soap solutions as compared
with pure water was due to a coating, the formation of this coating would be
a matter of time, and that a test might be found in the examination of the
properties of the liquid surface immediately after its formation. The experi-
mental problem here suggested may seem difficult or impossible ; but it was,
in fact, solved some years ago in the course of researches upon the Capillary
Phenomena of Jets*. A jet of liquid issuing under moderate pressure from
an elongated, e.g., elliptical, aperture perforated in a thin plate, assumes a
chain-like appearance, the complete period (X), corresponding to two links of
the chain, being the distance travelled over by a given part of the liquid in
the time occupied by a complete transverse vibration of the column about
its cylindrical configuration of equilibrium. Since the phase of vibration
depends upon the time elapsed, it is always the same at the same point in
space, and thus the motion is steady in the hydrodynamical sense, and the
boundary of the jet is a fixed surface. Measurements of X under a given
head, or velocity, determine the time of vibration, and from this, when the
density of the liquid and the diameter of the column are known, follows in its
turn the value of the capillary tension (T) to which the vibrations are due.
Cceteris paribus, T oc X~2 ; and this relation, which is very easily proved, is
all that is needed for our present purpose. If we wish to see whether a
moderate addition of soap alters the capillary tension of water, we have only
to compare the wave-lengths X in the two cases, using the same aperture
and head. By this method the liquid surface may be tested before it is
YJ^ second old.
Since it was necessary to be able to work with moderate quantities of
liquid, the elliptical aperture had to be rather fine, about 2 mm. by 1 mm.
* Pogg. Ann. Erganzungsband vm. 1878, p. 266.
t Pogg. Ann. Vol. CXLHI. 1871, p. 342. The original pamphlet dates from 1865.
* Boy. Soc. Proc. May 15, 1879. [Vol. i. p. 377.]
1890] ON THE TENSION OF RECENTLY FORMED LIQUID SURFACES.
343
The reservoir was an ordinary flask, 8 cm. in diameter, to which was sealed
below as a prolongation a (1 cm.) tube bent at right angles (Figs. 1, 2). The
Figs. 1 and 2.
aperture was perforated in thin sheet brass, attached to the tube by cement.
It was about 15cm. below the mark, near the middle of the flask, which
defined the position of the free surface at the time of observation.
The arrangement for bringing the apparatus to a fixed position, designed
upon the principles laid down by Sir W. Thomson, was simple and effective.
344 ON THE TENSION OF RECENTLY FORMED LIQUID SURFACES. [167
The body of the flask rested on three protuberances from the ring of a retort
stand, while the neck was held by an india-rubber band into a V-groove
attached to an upper ring. This provided five contacts. The necessary sixth
contact was effected by rotating the apparatus about its vertical axis until
the delivery tube bore against a stop situated near its free end. The flask
could thus be removed for cleaning without interfering with the comparability
of various experiments.
The measurements, which usually embraced two complete periods, could
be taken pretty accurately by a pair of compasses with the assistance of a
magnifying glass. But the double period was somewhat small (16 mm.),
and the little latitude admissible in respect to the time of observation was
rather embarrassing. It was thus a great improvement to take magnified
photographs of the jet, upon which measurements could afterwards be made
at leisure. In some preliminary experiments the image upon the ground
glass of the camera was utilised without actual photography. Even thus a
decided advantage was realised in comparison with the direct measurements.
Sufficient illumination was afforded by a candle flame situated a few
inches behind the jet. This was diffused by the interposition of a piece of
ground glass. The lens was a rapid portrait lens of large aperture, and the
ten seconds needed to produce a suitable impression upon the gelatine plate
was not so long as to entail any important change in the condition of the
jet. Otherwise, it would have been easy to reduce the exposure by the
introduction of a condenser. In all cases the sharpness of the resulting
photographs is evidence that the sixth contact was properly made, and
thus that the scale of magnification was .strictly preserved. Fig. 3 is a
Fig. 3.
reproduction on the original scale of a photograph of a water jet taken upon
9th November. The distance recorded as 2X is between the points marked
A and B, and was of course measured upon the original negative. On each
occasion when various liquids were under investigation, the photography of
the water jet was repeated, and the results agreed well.
After these explanations it will suffice to summarise the actual measure-
ments upon oleate of soda in tabular form. The standard solution contained
1890] ON THE TENSION OF RECENTLY FORMED LIQUID SURFACES. 345
1 part of oleate in 40 parts of water, and was diluted as occasion required*.
All lengths are given in millimetres.
Water
s
Oleate
1/40
Oleate
180
Oleate
1400
Oleate
1/4000
2X 40O
45-5
44-0
39O
390
A 31-5
11-0
HO
11-0
23-0
In the second row A is the rise of the liquid in a capillary tube, carefully
cleaned before each trial with strong sulphuric acid and copious washing.
In the last case, relating to oleate solution jj^nj, the motion was sluggish and
the capillary height but ill-defined. It will be seen that even when the
capillary height is not much more than one-third of that of water, the wave-
lengths differ but little, indicating that, at any rate, the greater part of the
lowering of tension due to oleate requires time for its development. According
to the law given above, the ratio of tensions of the newly -formed surfaces for
water and oleate (?^) would be merely as 6 : 5f .
Whether the slight differences still apparent in the case of the stronger
solutions are due to the formation of a sensible coating in less than -^ second,
cannot be absolutely decided: but the probability appears to lie in the
negative. No distinct differences could be detected between the first and
second wave-lengths; but this observation is, perhaps, not accurate enough
to settle the question. It is possible that a coating may be formed on the
surface of the glass and metal, and that this is afterwards carried forward.
As a check upon the method, I thought it desirable to apply it to the
comparison of pure water and dilute alcohol, choosing for the latter a mixture
of 10 parts by volume of strong (not methylated) alcohol with 90 parts water.
The results were as follows : —
2 X (water) = 38-5, 2 X (alcohol) = 46'5,
h (water) = 30-0, A (alcohol) = 22-0 :
but it may be observed that they are not quite comparable with the pre-
ceding for various reasons, such as displacements of apparatus and changes
of temperature. It is scarcely worth while to attempt an elaborate reduction
of these numbers, taking into account the differences of specific gravity in
the two cases; for, as was shown in the former paper, the observed values
of X are complicated by the departure of the vibrations from isochronism,
* Although I can find no note of the fact, I think I am right in saving that large babbles
could be blown with the weakest of the solutions experimented upon.
t Curiously enough, I find it already recorded in my note-book of 1879, that X is not
influenced by the addition to water of soap sufficient to render impossible the rebound of
colliding jets. [VoL i. p. 375.]
346 ON THE TENSION OF RECENTLY FOEMED LIQUID SURFACES. [167
when, as in the present experiments, the deviation from the circular section
is moderately great. We have —
(46-5 /38-5)2 = 1-46, 30/22 = T36 ;
and these numbers prove, at any rate, that the method of wave-lengths is
fully competent to show a change in tension, provided that the change really
occurs at the first moment of the formation of the free surface.
In view of the great extensibility of saponine films it seemed important
to make experiments upon this material also. The liquid employed was an
infusion of horse chestnuts of specific gravity T02, and, doubtless, contained
other ingredients as well as saponine. It was capable of giving large bubbles,
even when considerably diluted (6 times) with water. Photographs taken on
November 23rd gave the following results : —
2 X (water) - 39'2, 2X (saponine) = 39'5,
h (water) = 30'5, h (saponine) = 207.
Thus, although the capillary heights differ considerably, the tensions at
the first moment are almost equal. In this case then, as in that of soap,
there is strong evidence that the lowered tension is the result of the
formation of a pellicle.
Though not immediately connected with the principal subject of this
communication, it may be well here to record that I find saponine to have no
effect inimical to the rebound after mutual collision of jets containing it.
The same may be said of gelatine, whose solutions froth strongly. On the
other hand, a very little soap or oleate usually renders such rebound
impossible, but this effect appears to depend upon undissolved greasy matter.
At least the drops from a nearly vertical fountain of clear solution of soap
were found not to scatter*. The rebound of jets is, however, a far more
delicate test than that of drops. A fountain of strong saponine differs in
appearance from one of water ; but this effect is due rather to the superficial
viscosity, which retards, or altogether prevents, the resolution into drops.
The failure of rebound when jets or drops containing milk or undissolved
soap come into collision has not been fully explained; but it is probably
connected with the disturbance which must arise when a particle of grease
from the interior reaches the surface of one of the liquid masses.
P.S. — I have lately found that the high tension of recently formed
surfaces of soapy water was deduced by A. Dupref, as long ago as 1869,
from some experiments upon the vertical rise of fine jets. Although this
method is less direct than that of the present paper, M. Dupre must be
considered, I think, to have made out his case. It is remarkable that so
interesting an observation should not have attracted more attention.
* Roy. Soc. Proc. June 15, 1882. [Vol. n. p. 103.]
t Theorie Mecanique de la Chaleur, Paris, 1869.
168.
MEASUREMENTS OF THE AMOUNT OF OIL NECESSARY IN
ORDER TO CHECK THE MOTIONS OF CAMPHOR UPON
WATER.
[Proceedings of the Royal Society, XLVII. pp. 364—367, March, 1890.]
THE motion upon the surface of water of small camphor scrapings, a
phenomenon which had puzzled several generations of inquirers, was satis-
factorily explained by Van der Mensbrugghe * as due to the diminished
surface-tension of water impregnated with that body. In order that the
rotations may be lively, it is imperative, as was well shown by Mr Tomlinson,
that the utmost cleanliness be observed. It is a good plan to submit the
internal surface of the vessel to a preliminary treatment with strong
sulphuric acid. A touch of the finger is usually sufficient to arrest the
movements by communicating to the surface of the water a film of grease.
When the surface-tension is thus lowered, the differences due to varying
degrees of dissolved camphor are no longer sufficient to produce the effect.
It is evident at once that the quantity of grease required is excessively
small, so small that under the ordinary conditions of experiment it would
seem likely to elude our methods of measurement. In view, however, of the
great interest which attaches to the determination of molecular magnitudes,
the matter seemed well worthy of investigation ; and I have found that by
sufficiently increasing the water surface the quantities of grease required may
be brought easily within the scope of a sensitive balance.
In the present experiments the only grease tried is olive oil. It is
desirable that the material which is to be spread out into so thin a film
should be insoluble, involatile, and not readily oxidised, requirements which
greatly limit the choice.
* Mfmoires Couronnes (4to) of the Belgian Academy, Vol. xxxiv. 1869.
348 AMOUNT OF OIL NECESSAEY TO [168
Passing over some preliminary trials, I will now describe the procedure
by which the density of the oil film necessary for the purpose was determined.
The water was contained in a sponge-bath of extra size, and was supplied to
a small depth by means of an india-rubber pipe in connexion with the tap.
The diameter of the circular surface thus obtained was 84 cm. (33"). A short
length of fine platinum wire, conveniently shaped, held the oil. After each
operation it was cleaned by heating to redness, and counterpoised in the
balance. A small quantity of oil was then communicated, and determined by
the difference of readings. Two releasements of the beam were tried in each
condition of the wire, and the deduced weights of oil appeared usually to
be accurate to ^ milligram at least. When all is ready, camphor scrapings
are deposited upon the water at two or three places widely removed from
one another, and enter at once into vigorous movement. At this stage the
oiled extremity of the wire is brought cautiously down so as to touch the
water. The oil film advances rapidly across the surface, pushing before it
any dust or camphor fragments which it may encounter. The surface of the
liquid is then brought into contact with all those parts of the wire upon
which oil may be present, so as to ensure the thorough removal of the latter.
In two or three cases it was verified by trial that the residual oil was
incompetent to stop camphor motions upon a surface including only a few
square inches.
The manner in which the results are exhibited will be best explained
by giving the details of the calculation for a single case, e.g., the second
of December 17. Here 0'81 milligram of oil was found to be very nearly
enough to stop the movements. The volume of oil in cubic centimetres is
deduced by dividing 0'00081 by the sp. gr., viz., 0'9. The surface over which
this volume of oil is spread is
|TT x 842 square centimetres ;
so that the thickness of the oil film, calculated as if its density were the same
as in more normal states of aggregation, is
0-00081 _ 1-63
0-9x^x842~ 107 Cm''
or 1'63 micro-millimetres. Other results, obtained as will be seen at
considerable intervals of time, are collected in the Table. For convenience
of comparison they are arranged, not in order of date, but in order of densities
of film.
The sharpest test of the quantity of oil appeared to occur when the
motions were nearly, but not quite, stopped. There may be some little
uncertainty as to the precise standard indicated by " nearly enough," and it
may have varied slightly upon different occasions. But the results are quite
distinct, and under the circumstances very accordant. The thickness of oil
1890]
CHECK THE MOTIONS OF CAMPHOR UPON WATER.
349
required to take the life out of the camphor movements lies between one and
two millionths of a millimetre, and may be estimated with some precision at
1*6 micro-millimetre. Preliminary results from a water surface of less area
are quite in harmony.
A Sample of Oil, somewhat decolourised by exposure.
Date
Weight of
oil
Calculated
thickness
of film
Effect upon camphor fragments
Dec. 17...
0-40 mg.
0-81
No distinct effect.
Jan. 11...
0-52
1-06
Barely perceptible.
Jan. 14...
0-65
•32
Not quite enough.
Dec. 20...
0-78
•58
Nearly enough.
Jan. 11...
0-78
•58
Just enough.
Dec. 17...
0-81
•63
Just about enough.
Dec. 18...
0-83
1-68
Nearly enough.
Jan. 22...
0-84
•70
About enough.
Dec. 18...
0-95
1-92
Just enough.
Dec. 17...
0-99
2OO
All movements very nearly stopped.
Dec. 20...
1-31
2-65
Fully enough.
A fresh
Sample.
Jan. 28...
0-63
1-28
Barely perceptible.
Jan. 28...
1-06
2-14
Just enough.
For purposes of comparison it will be interesting to note that the
thickness of the black parts of soap films was found by Messrs Reinold and
Riicker to be 12 micro-millimetres.
An important question presents itself as to how far these water surfaces
may be supposed to have been clean to begin with. I believe that all
ordinary water surfaces are sensibly contaminated ; but the agreement of the
results in the Table seems to render it probable that the initial film was not
comparable with that purposely contributed. Indeed, the difficulties of the
experiments proved to be less than had been expected. Even a twenty-four
hours' exposure to the air of the laboratory* does not usually render a water
surface unfit to exhibit the camphor movements.
The thickness of the oil films here investigated is of course much below
the range of the forces of cohesion ; and thus the tension of the oily surface
* In the country.
350 MOTIONS OF CAMPHOR UPON WATER. [168
may be expected to differ from that due to a complete film, and [calculated] by
addition of the tensions of a water-oil surface and of an oil-air surface. The
precise determination of the tension of oily surfaces is not an easy matter.
A capillary tube is hardly available, as there would be no security that the
degree of contamination within the tube was the same as outside. Better
results may be obtained from the rise of liquid between two parallel plates.
Two such plates of glass, separated at the corners by thin sheet metal, and
pressed together near the centre, dipped into the bath. In one experiment
of this kind the height of the water when clean was measured by 62. When
a small quantity of oil, about sufficient to stop the camphor motions, was
communicated to the surface of the water, it spread also over the surface
included between the plates, and the height was depressed to 48. Further
additions of oil, even in considerable quantity, only depressed the level to 38.
The effect of a small quantity of oleate of soda is much greater. By this
agent the height was depressed to 24, which shows that the tension of a
surface of soapy water is much less than the combined tensions of a water-oil
and of an oil-air surface. According to Quincke, these latter tensions are
respectively 2'1 and 3'8, giving by addition 5'9 ; that of a water-air surface
being 8'3. When soapy water is substituted for clean, the last number
certainly falls to less than half its value, and therefore much below 5'9.
169.
FOAM.
[Proceedings of the Royal Institution, xm. pp. 85 — 97, March, 1890.]
WHEN I was turning over in my mind the subject for this evening, it
occurred to me to take as the title of the lecture, " Froth." But I was told
that a much more poetical title would be " Foam," as it would so easily lend
itself to appropriate quotations. I am afraid, however, that I shall not be
able to keep up the poetical aspect of the subject very long; for one of the
things that I shall have most to insist upon is that foaming liquids are
essentially impure, contaminated — in fact, dirty. Pure liquids will not foam.
If I take a bottle of water and shake it up, I get no appreciable foam. If,
again, I take pure alcohol, I get no foam. But if I take a mixture of water
with 5 per cent, of alcohol there is a much greater tendency. Some of the
liquids we are most familiar with as foaming, such as beer or ginger- beer,
owe the conspicuousness of the property to the development of gas in the
interior, enabling the foaming property to manifest itself; but of course the
two things are quite distinct. Dr Gladstone proved this many years ago by
showing that beer from which all the carbonic acid had been extracted in
vacuo still foamed on shaking up. I now take another not quite pure but
strong liquid, acetic acid, and from it we shall get no more foam than we
did from the alcohol or the water. The bubbles, as you see, break up
instantaneously. But if I take a weaker acid, the ordinary acid of commerce,
there is more, though still not much, tendency to foam. But with a liquid
which for many purposes may be said to contain practically no acetic acid at
all, seeing that it consists of water with but one-thousandth part of acid, the
tendency is far stronger; and we get a very perceptible amount of foam.
These tests with the alcohol and acetic acid are sufficient to illustrate the
principle that the property of foaming depends on contamination. In pure
ether we have a liquid from which the bubbles break even more quickly than
from alcohol or water. They are gone in a moment. In some experiments
I made at home I found that water containing a small proportion of ether
352 FOAM. . [169
foamed freely; but on attempting two or three days ago to repeat the
experiment, I was surprised to find a result very different. I have here some
water containing a very small fraction of ether, about 1 /240th part. If I
shake it up, it scarcely foams at all ; but another mixture made in the same
proportion from another sample shows more tendency to foam. This is rather
curious, because both ethers were supposed to be of the same quality; but
one had been in the laboratory longer than the other, and perhaps contained
more greasy matter in solution.
Another liquid which foams freely is water impregnated with camphor.
Camphor dissolves sparingly; but a minute quantity of it quite alters the
characteristics of water in this respect. Another substance, very minute
quantities of which communicate the foaming property to water, is glue or
gelatine. This liquid contains only 3 parts in 100,000 of gelatine, but it
gives a froth entirely different from that of pure water. Not only are
there more bubbles, but the duration of the larger bubbles is quite out of
proportion. This sample contains- 5 parts in 100,000, nearly double as much ;
but even with but 1 part in 100,000, the foaming property is so evident as
to suggest that it might in certain cases prove valuable for indicating the
presence of minute quantities of impurities. I have been speaking hitherto
of those things which foam slightly. They are not to be compared with, say,
a solution of soap in water, which, as is well known to everybody, froths very
vigorously. Another thing comparable to soap, but not so well known, is
saponine. It may be prepared from horse chestnuts by simply cutting them
in small slices and making an infusion with water. A small quantity of this
infusion added to water makes it foam strongly. The quantity required to do
this is even less than in the case of soap ; so the test is more delicate. It
is well known that rivers often foam freely. That is no doubt due to the
effect of saponine or some analogous substance*. Sea- water foams, but not,
I believe, on account of the saline matter it contains ; for I have found that
even a strong solution of pure salt does not foam much. I believe it has
been shown that the foaming of sea-water, often so conspicuous, is due to
something extracted from seaweeds during the concussion which takes place
under the action of breakers.
Now let us consider for a moment what is the meaning of foaming. A
liquid foams when its films have a certain durability. Even in the case
of pure water, alcohol, and ether, these films exist. If a bubble rises, it is
covered for a moment by a thin film of the liquid. This leads us to consider
the properties of liquid films in general. One of their most important and
striking properties is their tendency to contract. Such surfaces may be
regarded as being in the condition of a stretched membrane, as of india-
rubber, — only with this difference, that the tendency to contract never ceases.
* [1901. Compare " Experiments upon Surface-Films," Phil. May. xxxin. p. 370, 1892.]
1890] FOAM. 353
We may show that by blowing a small soap bubble, and then removing the
mouth. The air is forced back again by the pressure exerted on the bubble
by the tension of the liquid. This ancient experiment suffices to prove
conclusively that liquid films exercise tension.
A prettier form of the same experiment is due to Van der Mensbrugghe,
who illustrated liquid tension by means of a film in which he allowed to float
a loop of fine silk, tied in a knot. As long as the interior of the loop, as well
as the exterior, is occupied by the liquid film, it shows no tendency to take
any particular shape : but if, by insertion of, say, a bit of blotting paper, the
film within the loop be ruptured, then the tension of the exterior film is
free to act, and the thread flies instantaneously into the form of a circle, in
consequence of the tendency of the exterior surface to become as small as
possible. The exterior part is now occupied by the soap film, and the interior
is empty [shown]. Many other illustrations of this property of liquids might
be given, but time does not permit.
In the soap film, as in the films which constitute ordinary foam, each
thin layer of liquid has two surfaces ; each tends to contract ; but in many
cases we have only one such surface to consider, as when a drop of rain falls
through the air. Again, suppose that we have three materials in contact
with one another, — water, oil, and air. There are three kinds of surfaces
separating the three materials, one separating water and oil, another oil and
air, and a third surface separating the water from the air. These three
surfaces all exert a tension, and the shape of the mass of oil depends upon
the relative magnitudes of the tensions. As I have drawn it here (Fig. 1),
it is implied that the tension of the water-air surface is less than the sum of
the other two tensions — those of the water-oil surface and the air-oil surface ;
because the two latter acting obliquely balance the former. It is only
under such conditions that the equilibrium of the three materials as there
drawn in contact with one another is possible. If the tension of the surface
separating water and air exceeded the sum of the other two, then the
equilibrium as depicted would be impossible. The water-air tension, being
greater, would assert its superiority by drawing out the edge of the lens, and
the oil would tend to spread itself more and more over the surface.
Fig. 1.
AIR
And that is what really happens. Accurate measurements, made by
Quincke and others, show that the surface tension separating water and air
is really greater than the sum of the two others. So oil does tend to spread
upon a surface of water and air. That this is the fact, we can prove by
K. in. 23
354 FOAM. [169
a simple experiment. At the feet of our chairman is a large dish, containing
water which at present is tolerably clean. In order to see what may happen
to the surface of the water, it is dusted over with fine sulphur powder, and
illuminated with the electric light. If I place on the surface a drop of water,
no effect ensues ; but if I take a little oil, or better still a drop of saponine, or
of soap-water, and allow that to be deposited upon the middle of the surface,
we shall see a great difference. The surface suddenly becomes dark, the
whole of the dust being swept away to the boundary. That is the result
of the spread of the film, due to the presence of the oil.
How then is it possible that we should get a lens-shaped mass of oil, as
we often do, floating upon the surface of water ? Seeing that the general
tendency of oil is to spread over the surface of water, why does it not do so in
this case ? The answer is that it has already spread, and that this surface is
not really a pure water surface at all, but one contaminated with oil. It is in
fact only after such contamination that an equilibrium of this kind is possible.
The volume of oil necessary to contaminate the surface of the water is very
small, as we shall see presently; but I want to emphasise the point that,
so far as we know, the equilibrium of the three surfaces in contact with
one another is not possible under any other conditions. That is a fact not
generally recognised. In many books you will find descriptions of three
bodies in contact, and a statement of the law of the angles at which they
meet; that the sides of a triangle, drawn parallel to the three intersecting
surfaces, must be in proportion to the three tensions. No such equilibrium,
and no such triangle, is possible if the materials are pure ; when it occurs, it
can only be due to the contamination of one of the surfaces. These very thin
films, which spread on water, and, with less freedom, on solids also, are of
extreme tenuity ; and their existence, alongside of the lens, proves that the
water prefers the thin film of oil to one of greater thickness. If the oil were
spread out thickly, it would tend to gather itself back into drops, leaving over
the surface of the water a film of less thickness than the molecular range.
One experiment by which we may illustrate some of these effects I owe to
my colleague, Professor Dewar. It shows the variation in the surface tension
of water, due to the presence on it of small quantities of ether. I hold in
my hand masses of charcoal, which can be impregnated with ether. The
greater part of the surface of the charcoal is covered with paraffin wax, and, in
consequence, the ether which has already penetrated the charcoal can only
escape from it again on one side. The result is that the water in the rear
of this boat of charcoal will be more impregnated with ether than the part
in front, so the mass of charcoal will enter into motion, and the motion will
extend over a considerable interval of time. As long as the ether remains in
sufficient quantity to contaminate the water in the rear, so long is there a
tendency to movement of the mass. The water covered with the film of ether
has less tension than the pure water in front, and the balance of tensions
1890] FOAM. 355
being upset, the mass is put in motion. If the nature of the case is such
that the whole surface surrounding the solid body is contaminated, then
there is no tendency to movement, the same balance in fact obtaining as
if the water were pure.
Another body which we may use for this purpose is camphor. If we
spread some camphor scrapings on a surface of pure water, they will, if the
surface is quite clean, enter into vigorous movement, as you now see. This
is because the dissolved camphor diminishes the surface tension of the water.
But if I now contaminate the water with the least possible quantity of
grease, the movements of the camphor will be stopped. I merely put my
finger in, and you observe the effect. There is not much poetry about that !
A very slight film, perfectly invisible by ordinary means, is sufficient so to
contaminate the water that the effect of the dissolved camphor is no longer
visible.
I was very desirous to ascertain, if possible, the actual thickness of oil
necessary to produce this effect, because all data relating to molecules are, in
the present state of science, of great interest. From what I have alreadv
said, you may imagine that the quantity of oil required is very small, and
that its determination may be difficult. In my experiments*, I used the
surface of water contained in a large sponge bath three feet in diameter.
By this extension of the surface, I was able to bring the quantity of oil
required within the range of a sensitive balance. In Diagram 2 [see Vol. HI.
p. 340], I have given a number of results obtained at various dates, showing
the quantity of oil required to produce the effects recorded in the fourth
column. Knowing the weight of the oil deposit, and the area of the water
surface upon which it was uniformly spread, it was easy to calculate the
thickness of the film. It is seen that a film of oil about 1£ millionth of a
millimetre thick is able to produce this change. I know that large numbers
are not readily appreciated, and I will therefore put the matter differently.
The thickness of the oil film thus determined as sufficient to stop the motions
of the camphor is ^ of the wave-length of yellow light. Another way
of saying the same thing is that this thickness of oil bears to one inch the
same ratio that one second of time bears to half a year.
When the movement of the camphor has been stopped by the addition
of a minute quantity of oil, it is possible, by extending the water surface
enclosed within the boundary, without increasing the quantity of oil, to
revive the movements of the camphor; or, again, by contraction, to stop
them. I can do this with the aid of a flexible boundary 3f thin sheet brass ;
and you see that the camphor recovers its activity, though a moment ago
it was quite dead. It would be an interesting subject for investigation to
determine what is the actual tension of an oily surface contaminated to an
* Proe. Soy. Soe. March, 1890.
23—2
356 FOAM. [169
extent just sufficient to stop the camphor movements ; but it is not an easy
problem. Usually we determine surface tensions by the height to which
the liquids will rise in very fine tubes. Here, however, that method is not
available, because if we introduce a tube into such a surface, there is no proof
that the contamination of the inner surface in the tube is the same as that
prevailing outside. Another method, however, may be employed which is
less open to the above objection, and that is to substitute for the very fine or
capillary tube, a combination of two parallel plates open at their edges. We
have here two such plates of glass, kept from absolutely closing by four pieces
of thin metal inserted at the corners, the plates being held close against these
distance-pieces by suitable clamps. If such a combination be inserted in
water, the liquid will rise above the external level, and the amount of the rise
is a measure of the surface tension of the water. You see now the image on
the screen. A is the external water surface ; B is the height of the liquid
contained between the glass plates, so that the tension may be said to be
measured by the distance AB. If a little oil be now deposited upon the
surface, it will find its way between the plates. The fall which you now see
shows that the surface tension has been diminished by the oil which has
found its way in. A very minute quantity will give a great effect. When
the height of the pure water was measured by 62, a small quantity of oil
changed the 62 into 48, and subsequent large additions of oil could only lower
it to 38. But after oil has done its worst, a further effect may be produced
by the addition of soap. If Mr Gordon now adds some soap, we shall find
that there is a still further fall in the level, showing that the whole tension
now in operation is not much more than one-third of what it was at first.
This is an important point, because it is sometimes supposed that the effect
of soap in diminishing the tension of water is due merely to the formation
upon the surface of a layer of oil by decomposition of the soap. This
experiment proves the contrary, because we find that soap can do so much
more than oil. There is, indeed, something more or less corresponding to the
decomposition of the soap and the formation of a superficial layer of oil. But
the decomposition takes place in a very peculiar manner, and under such
conditions that there is a gradual transition from the soapy liquid in the
interior to the oily layer at the top, and not, as when we float a layer of oil on
water, two sudden transitions, first from water to oil, and secondly from oil
to air. The difference is important, because, as I showed some years ago
[Vol. i. p. 234], capillary tension depends on the suddenness of change. If we
suppose that the change from one liquid to another takes place by slow
stages, though the final change may be as before, the capillary tension would
absolutely disappear.
There is another very interesting class of phenomena due to oil films,
which I hope to illustrate, though I am conscious of the difficulty of the
task, — namely, the action of oil in preventing the formation of waves. From
1890] FOAM. 357
the earliest times we haTe records of the effect of oil in stilling waves, and all
through the Middle Ages the effect was recognised, though connected with
magic and fanciful explanations. Franklin, than whom, I suppose, no soberer
inquirer ever existed, made the thing almost a hobby. His attention was
called to it accidentally on board ship from noticing the effect on the waves
caused by the ,greasy debris of a dinner. The captain assured him that
it was due to the oil spread on the water, and for some time afterwards
Franklin used to carry oil about with him, so as never to miss a chance
of trying an experiment. A pond is necessary to illustrate the phenomena
properly, but we shall get an idea of it by means of this trough six feet
long, containing water*. Along the surface of the water we shall make
an artificial wind by means of a fan4", driven by an electro-motor. In my
first experiments I used wind from an organ bellows, which is not here
available. Presently we shall get up a ripple, and then we will try the effect
of a drop of oil put on to windward. I have now put on the drop, and you
see a smooth place advancing along. As soon as the waves come up again,
I will repeat the experiment While the wind is driving the oil away. I
may mention that this matter has been tested at Peterhead. Experiments
were there made on a large scale to demonstrate the effect of oil in facilitating
the entrance of ships into harbour in rough weather. Much advantage was
gained. But here a distinction must be observed. It is not that the large
swell of the ocean is damped down. That would be impossible. The action
in the first instance is upon the comparatively small ripples. The large
waves are not directly affected by the oil : but it seems as if the power of the
wind to excite and maintain them is due to the small ripples which form on
their backs, and give the wind, as it were, a better hold of them. It is only
in that way that large waves can be affected. The immediate effect is on the
small waves which conduce to that breaking of the large waves which from
the sailor s point of view is the worst danger. It is the breaking waters
which do the mischief and these are quieted by the action of the oiL
I want to show also, though it can only be seen by those near, the return
of the oil when the wind is stopped. The ofl is at present driven to one end
of the trough*; when the wind stops, it will come back, because the oil film
tends to spread itself uniformly over the surface. As it comes back, there
will be an advancing wave of oil : and as we light the surface very obliquely
by the electric lamp, there is visible on the bottom of the trough a white line,
showing its progress.
* The width is 8 inches, and the depth 4 inches. The sides are of glass ; the bottom and
ends of wood, punted white.
t For this fen and its fittings the Institution la indebted to die liberality of the Blackman
Ventilating Company.
* May, 1890. Any moderate quantity of ofl may be driven off to leewmrd; bat if okate
of soda be applied, the
358 FOAM. [169
Now, as to the explanation. The first attempt on the right lines was
made by the Italian physicist, Marangoni. He drew attention to the
importance of contamination upon the surface of the water, and to its
tendency to spread itself uniformly, but for some reason which I cannot
understand, he applied the explanation wrongly. More recently Reynolds
and Aitken have applied the same considerations with better success. The
state of the case seems to be this: — Let us consider small waves as
propagated over the surface of clean water. As the waves advance, the
surface of the water has to submit to periodic extensions and contractions.
At the crest of a wave the surface is compressed, while at the trough it is
extended. As long as the water is pure there is no force to oppose that,
and the wave can be propagated without difficulty; but if the surface be
contaminated, the contamination strongly resists the alternate stretching and
contraction. It tends always, on the contrary, to spread itself uniformly ; and
the result is that the water refuses to lend itself to the motion which is
required of it. The film of oil may be compared to an inextensible membrane
floating on the surface of the water, and hampering its motion; and under
these conditions it is not possible for the waves to be generated, unless the
forces are very much greater than usual. That is the explanation of the
effect of oil in preventing the formation of waves.
The all-important fact is that the surface has its properties changed, so
that it refuses to submit to the necessary extensions and contractions. We
may illustrate this very simply by dusting the surface of water with sulphur
powder, only instead of dispersing the sulphur, as before, by the addition
of a drop of oil, we will operate upon it by a gentle stream of wind projected
downwards on the surface, and of course spreading out radially from the
point of impact. If Mr Gordon will blow gently on the surface in the middle
of the dusty region, a space is cleared*; if he stops blowing, the dust comes
back again. The first result is not surprising, but why does the dusty
surface come back ? Such return is opposed to what we should expect from
any kind of viscosity, and proves that there must be some force directly
tending to produce that particular motion. It is the superior tension of the
clean surface. No oil has been added here, but then no water surface is
ever wholly free from contamination; there may be differences of degree,
but contamination is always present to some extent. I now make the
surface more dirty and greasy by contact of the finger, and the experiment
no longer succeeds, because the jet of wind is not powerful enough to cleanse
the place on which it impinges ; the dirty surface refuses to go away, or if it
goes in one direction it comes back in another.
I want now to bring to your notice certain properties of soap solutions,
which, however, are not quite so novel as I thought when I first came upon
* This experiment is due to Mr Aitkeii. [It was exhibited by projection.]
1890] FOAM. 359
them in my own inquiries*. If we measure by statical, or slow, methods
the surface tension of soapy water, we find it very much less than that
of clean water. We can prove this in a very direct manner by means of
capillary tubes. Here, shown upon the screen, are two tubes of the same
diameter, in which, therefore, if the liquids were the same, there would be
the same elevation ; one tube dips into clean water, and the other into soapy
water, and the clean water rises much (nearly three times) higher than the
soapy water.
Although the tension of soapy water is so much less than that of pure
water when measured in this way, I had some reason to suspect that the case
might be quite different if we measured the tensions immediately after the
formation of the surfaces. I was led to think so by pondering on Marangoni's
view that the behaviour of foaming liquids was due to the formation of a
pellicle upon their surfaces; for if the change of property is due to the
formation of a pellicle, it is reasonable to suppose that it will take time, so
that if we can make an observation before the surface is more than sav -^ of
a second old, we may expect to get a different result. That mar seem an
impossible feat, but there is really no difficulty about it ; all that is necessary
is to observe a jet of the substance in question issuing from a fine orifice.
If such a jet issues from a circular orifice it will be cylindrical at first, and
afterwards resolve itself into drops. Ifj however, the orifice is not circular,
but elongated or elliptical, the jet undergoes a remarkable transformation
before losing its integrity. As it issues from the elliptical orifice, it is in
vibration, and trying to recover the circular form ; it does so, but afterwards
the inertia tends to carry it over to the other side of equilibrium. The
section oscillates between the ellipse in one direction and the ellipse in the
perpendicular direction. The jet thus acquires a sort of chain-like appearance,
and the period of the movement, represented by the distance between
corresponding points [A, B, Fig. (3), Art. 167], is a measure of the capillary
tension to which these vibrations of the elliptical section about the circular
form are due. A measure, then, of the wave-length of the recurrent pattern
formed by the liquid gives us information as to the tension immediately after
escape : and if we wish to compare the tensions of various liquids, all we have
to do is to fill a vessel alternately with one liquid and another, and compare
the wave-lengths in the various cases. The jet issues from a flask, to which
is attached below a tubular prolongation; the aperture is made small in
order that we may be able to deal with small quantities of liquid. You now
see the jet upon the screen. As it issues from the orifice, it oscillates, and
we can get a comparative measure of the tension by observing the distance
between corresponding points (A, B).
* I here aflnde to the experiments of Dopre, and to the masted? theoretical discussion of
liquid films by Professor Wfllard Gibbs.
360 FOAM. [169
If we were now to take out the water, and substitute for it a moderately
strong solution of soap or saponine, we should find but little difference, show-
ing that in the first moments the tension of soapy water is not very different
from that of pure water. It will be more interesting to exhibit a case in
which a change occurs. I therefore introduce another liquid, water containing
10 per cent, of alcohol, and you see that the wave-length is different from
before. So this method gives us a means of investigating the tensions of
surfaces immediately after their formation. If we calculate by known
methods how long the surface has been formed before it gets to the point B,
at which the measurement is concluded, we shall find that it does not exceed
jfo of a second.
Another important property of contaminated surfaces is what Plateau and
others have described as superficial viscosity. There are cases in which the
surfaces of liquids — of distilled water, for example — seem to exhibit a special
viscosity, quite distinct from the ordinary interior viscosity, which is the
predominant factor in determining the rate of flow through long narrow
tubes. Plateau's experiment was to immerse a magnetised compass needle
in water ; the needle turns, as usual, upon a point, and the water is contained
in a cylindrical vessel, not much larger than the free rotation of the needle
requires (Fig. 4). The observation relates to
the time occupied by the needle in returning
to its position of equilibrium in the meridian,
after having been deflected into the east and
west positions, and Plateau found that in the
case of water more time was required when the
needle was just afloat than when it was wholly
immersed, whereas in the case of alcohol the
time was greater in the interior. The longer
time occupied when the needle is upon the
surface of water is attributed by Plateau to
an excessive superficial viscosity of that body.
Instead of a needle, I have here a ring of brass wire (Fig. 5), floating on
the surface of the water. You see upon the screen the image of the ring,
as well as the surface of the water, which has been made visible by sulphur.
The ring is so hung from a silk fibre that it can turn upon itself, remaining
all the while upon the surface of the water. Attached to it is a magnetic
needle, for the purpose of giving it a definite set, and of rotating it as
required by an external magnet. On this water, which is tolerably clean,
when the ring is made to turn, it leaves the dust in the interior entirely
behind. That shows that the water inside the ring offers no resistance to
the shearing action brought into play. The part of the surface of water
immediately in contact with the ring no doubt goes round ; but the move-
ment spreads to a very little distance. The same would be observed if we
1890] FOAX. 361
added soap. Bat if I add some saponine. we shall find a different result, and
that the behaviour of the dost in the interior of the ring is materially altered.
Kg. 5. Rg-6L
The saponine has stiffened the surface, so that the ring turns with more
difficulty: and when it turns, it carries round the whole interior with it.
The surface has now got a stiffness from which before it was free : but the
point upon which I wish to fix jour attention is that the surface of pure
water does not behave in the same way. If, however, we substitute for the
simple hoop another provided with a material diameter (Fig. 6). bring also
in the surface of the water, then we shall find, as was found by Plateau in his
experiment, that the water is carried round. In this case, it is no longer
possible for the surface to be left behind., as it was with the simple hoop,
unless it is willing to undergo local expansions and contractions of area.
The difference of behaviour proves that what a water surface resists is not
shearing, but expansions and contractions; in fact, it behaves just as a
contaminated surface should do. On this supposition, it is easy to explain
the effects observed by Plateau; but the question at once arises, can we
believe that all water surfaces hitherto experimented upon are sensibly
contaminated ? and if yes, is there any means by which the contamination
may be removed ? I cannot in the time at my disposal discuss this question
fully, but I may say that I have succeeded in purifying the surface of the
water in Plateau's experiment, until it behaved like alcohol It is therefore
certain that Plateau's superficial viscosity is due to contamination, as was
conjectured by Marangoni.
I must now return to the subject of foam, from which I may seem to
have digressed, though I have not really done so. Why does surface
contamination enable a film to exist with greater permanence than it
otherwise could? Imagine a vertical soap film. Could the film continue
362 FOAM. [169
to exist if the tension were equal at all its parts ? It is evident that the
film could not exist for more than a moment ; for the interior part, like the
others, is acted on by gravity, and, if no other forces are acting, it will fall
16 feet in a second. If the tension above be the same as below, nothing can
prevent the fall. But observation proves that the central parts do not fall,
and thus that the tension is not uniform, but greater in the upper parts than
in the lower. A film composed of pure liquid can have but a very brief life.
But if it is contaminated, there is then a possibility of a different tension
at the top and at the bottom, because the tension depends on the degree of
contamination. Supposing that at the first moment the film were uniformly
contaminated, then the central parts would begin to drop. The first effect
would be to concentrate the contamination on the parts underneath and
to diminish it above. The result of that would be an increase of tension on
the upper parts. So the effect would be to call a force into play tending to
check the motion, and it is only in virtue of such a force that a film can have
durability. The main difference between a material that will foam and one
that will not, is in the liability of the surface to contamination from the
interior.
I find my subject too long for my time, and must ask you to excuse the
hasty explanations I have given in some parts. But I was anxious above
all to show the principal experiments upon which are based the views that
I have been led to entertain.
170.
ON THE SUPERFICIAL VISCOSITY OF WATER
[Proceedings of the Royal Society, XLvm. pp. 127—140, 1890.]
THE idea that liquids are endowed with a viscosity peculiar to the surface
is to he found in the writings of Descartes and Rumford : but it is to Plateau
that its general acceptance is due. His observations related to the behaviour
of a compass needle, turning freely upon a point, and mounted in the centre
of a cylindrical glass vessel of diameter not much more than sufficient to allow
freedom of movement. By means of an external magnet the needle was
deflected 90° from the magnetic meridian. When all had come to rest the
magnet was suddenly removed, and the time occupied by the needle in
recovering its position of equilibrium, or rather in traversing an arc of 85%
was noted. The circumstances were varied in two ways : first, by a change
of liquid, e#., from water to alcohol : and, secondly, by an alteration in the
level of the liquid relatively to the needle. With each liquid observations
were made, both when the needle rested on the surface, so as to be wetted
only on the under side, and also when wholly immersed to a moderate depth.
A comparison of the times required in the two cases revealed a remarkable
dependence upon the nature of the liquid. With water, and most aqueous
solutions, the time required upon the surface was about double of that in the
interior: whereas, with liquids of Plateau's second category, alcohol, ether,
oil of turpentine, &c., the time on the surface was about half of the time in
the interior. Of liquids in the third category (from which bubbles may be
blown), a solution of soap behaved in much the same manner as the distilled
water of the first category. On the other hand, solutions of albumen, and
notably of saponine, exercised at their surfaces an altogether abnormal
resistance.
These experiments of Plateau undoubtedly establish a special property
of the surfaces of liquids of the first and third categories : but the question
remains open whether the peculiar action upon the needle is to be attributed
364 ON THE SUPERFICIAL VISCOSITY OF WATER. [170
to a viscosity in any way analogous to the ordinary internal viscosity which
governs the flow through capillary tubes.
In two remarkable papers*, Marangoni attempts the solution of this
problem, and arrives at the conclusion that Plateau's superficial viscosity may
be explained as due to the operation of causes already recognised. In the
case of water and other liquids of the first category, he regards the resistance
experienced by the needle as mainly the result of the deformation of the
meniscuses developed at the contacts on the two sides with the liquid surface.
This view does not appear to me to be sound ; for a deformation of a meniscus
due to inertia would not involve any dissipation of energy, nor permanent
resistance to the movement. But the second suggestion of Marangoni is
of great importance.
On various grounds the Italian physicist concludes that "many liquids,
and especially those of Plateau's third category, are covered with a superficial
pellicle ; and that it is to this pellicle that they owe their great superficial
viscosity." After the observations of Dupref and myself |, supported as they
are by the theory of Professor Willard Gibbs§, the existence of the superficial
pellicle cannot be doubted; and its mode of action is thus explained by
Marangoni||: — "The surface of a liquid, covered by a pellicle, possesses two
superficial tensions ; the first, which is the weaker and in constant action, is
due to the pellicle ; the second is in the latent state, and comes into operation
only when the pellicle is ruptured. Since the latter tension exceeds the
former, it follows that any force which tends to rupture the superficial pellicle
upon a liquid encounters a resistance which increases with the difference
of tensions between the liquid and the pellicle." In Plateau's experiment
the advancing edge of the needle tends to concentrate the superficial
contamination, and the retreating edge to attenuate it ; the tension in front
is thus inferior to the tension behind, and a force is called into operation
tending to check the vibration. On a pure surface it is evident that nothing
of this sort can occur, unless it be in a very subordinate degree, as the result
of difference of temperature.
There is an important distinction, discussed by Willard Gibbs, according
as the contamination, to which is due the lowering of tension, is merely
accidentally present upon the surface, or is derived from the body of the
liquid under the normal operation of chemical and capillary forces. In the
latter case, that, for example, of solutions of soap and of camphor, the changes
* Nuovo Cimento, Ser. 2, Vol. v. vi. Apr. 1872 ; Nuovo Cimento, Ser. 3, Vol. in. 1878.
t Theorie Mecanique de la Chaleur, Paris, 1869, p. 377.
J " On the Tension of Recently Formed Liquid Surfaces," Roy. Soc. Proc. Vol. XLVII. 1890,
p. 281 (supra). [Vol. m. p. 341.]
§ Connecticut Acad. Trans. Vol. in. Part 2, 1877 — 78. In my former communication I over-
looked Prof. Gibbs's very valuable discussion on this subject.
|| Nuovo Cimento, Vol. v. vi. 1871—72, p. 260 (May, 1872).
1890] ON THE SUPERFICIAL VISCOSITY OF WATER. 365
of tension which follow an extension or contraction of the surface may be
of very brief duration. After a time, dependent largely upon the amount of
contaminating substance present in the body of the liquid, equilibrium is
restored, and the normal tension is recovered. On the other hand, in the
case of a surface of water contaminated with a film of insoluble grease, the
changes of tension which accompany changes of area are of a permanent
character.
It is not perfectly clear how far Marangoni regarded his principle of
surface elasticity as applicable to the explanation of Plateau's observations
upon distilled water ; but, at any rate, he applied it to the analogous problem
of the effect of oil in calming ripples. It is unfortunate that this attempt at
the solution of a long-standing riddle cannot be regarded as successful. He
treats the surface of the sea in its normal condition as contaminated, and
therefore elastic, and he supposes that, upon an elastic surface, the wind will
operate efficiently. When oil is scattered upon the sea, a non-elastic surface
of oil is substituted for the elastic surface of the sea, and upon this the
wind acts too locally to generate waves. It is doubtless true that an excess
of oil may render a water surface again inelastic; but I conceive that the
real explanation of the phenomenon is to be found by a precisely opposite
application of Marangoni's principle, as in the theories of Reynolds* and
Aitkenf. Marangoni was, perhaps, insufficiently alive to the importance
of varying degrees of contamination. An ordinary water surface is indeed
more or less contaminated ; and on that account is the less, and not the more,
easily agitated by the wind. The effect of a special oiling is, in general, to
increase the contamination and the elasticity dependent thereupon, and stops
short of the point at which, on account of saturation, elasticity would again
disappear. The more elastic surface refuses to submit itself to the local
variations of area required for the transmission of waves in a normal manner.
It behaves rather as a flexible but inextensible membrane would do, and,
by its drag upon the water underneath, hampers the free production and
propagation of waves.
The question whether the effects observed by Plateau upon the surface
of distilled water are, or are not, due to contamination must, I suppose, be
regarded as still undecided. Oberbeck, who has experimented on the lines
of Plateau, thus sums up his discussion: — " Wir miissen daher schliessen,
entweder, dass der freien Wasseroberflache ein recht bedeutender Ober-
flachenwiderstand zukommt, oder dass eine reine Wasseroberflache in
Beriihrung mit der Luft uberhaupt nicht existirtj."
Postponing for the moment the question of the origin of "superficial
viscosity," let us consider its character. A liquid surface is capable of two
* Brit. Assoc. Rep. 1880.
t Edinburgh Roy. Soc. Proc. 1882—83, Vol. xn. p. 56.
J Wied. Ann. Vol. xi. 1880, p. 650.
366 ON THE SUPERFICIAL VISCOSITY OF WATER. [170
kinds of deformation, dilatation (positive or negative) and shearing ; and the
question at once presents itself, is it the former or the latter which evokes
the special resistance ? Towards the answer of this question Marangoni
himself made an important contribution in the earlier of the memoirs cited.
He found (p. 245) that the substitution for the elongated needle of Plateau
of a circular disc of thin brass turning upon its centre almost obliterated the
distinction between liquids of the two first categories. The ratio of the
superficial to the internal viscosity was now even greater for ether than for
water. From this we may infer that the special superficial viscosity of water
is not called into play by the motions of the surface due to the rotation of the
disc, which are obviously of the nature of shearing.
A varied form of this experiment is still more significant. I have reduced
the metal in contact with the water surface to a simple (2") ring ABGD
of thin brass wire [Fig. (5), p. 361]. This is supported by a fine silk fibre,
so that it may turn freely about its centre. To give a definite set, and to
facilitate forced displacements, a magnetised sewing needle, NS, is attached
with the aid of wax. In order to make an experiment, the ring is adjusted
to the surface of water contained in a shallow vessel. When all is at rest, the
surface is dusted over with a little fine sulphur* and the suspended system
is suddenly set into rotation by an external magnet. The result is very
distinct, and contrasts strongly with that observed by Plateau. Instead of
the surface enclosed by the ring being carried round with it in its rotation,
not the smallest movement can be perceived, except perhaps in the immediate
neighbourhood of the wire itself. It is clear that an ordinary water surface
does not appreciably resist shearing.
A very slight modification of the apparatus restores the similarity to that
of Plateau. This consists merely in the addition to the ring of a material
diameter of the same brass wire, CD [Fig. (6), p. 361]. If the experiment be
repeated, the sulphur indicates that the whole water surface included within
the semicircles now shares in the motion. In general terms the surface may
be said to be carried round with the ring, although the motion is not that
of a rigid body.
Experiments of this kind prove that what a water surface resists is not
shearing, but local expansions and contractions of area, even under the
condition that the total area shall remain unchanged. And this is precisely
what should be expected, if the cause of the viscosity were a surface
contamination. A shearing movement does not introduce any variation in
the density of the contamination, and therefore does not bring Marangoni's
principle into play. Under these circumstances there is no resistance.
* Sulphur seems to be on the whole the best material, although it certainly communicates
some impurity to the surface. Freshly heated pumice or wood-ashes sink immediately; and
probably all powders really free from grease would behave in like manner.
1880] ox
It remains to consider liquids of the third category in Plateau's nomen-
clature. The addition of a fittfe ofeate of soda does not alter the behaviour
of water, at least if the surface he tolerably fresh. On the other hand, a ray
small quantity of saponine suffices to render the surface almost rigid. In the
experiment with the simple ring the whole interior surface is carried round
as if rigidly attached. A similar effect is produced by gelatine, though in
a less marked degree.
In the case of saponine, therefore, it must be rally admitted that there
is a superficial viscosity not to be accounted for on HarangonFs principle
by the tendency of contamination to spread itself uniformly. It seems not
improbable that the pellicle formed upon the surface may have the properties
of a solid, rather than of a liquid. However this may be, the fact is certain
that a contracting saponine surface has no definite tension alike in all
directions. A Ji«IKi'jiiMii proof is to be found in the well-known experiment
in which a saponine bubble becomes wrinkled when the internal air is
removed.
The quasi-solid pellicle on the surface of saponine would be of extreme
thinness, and, even if it exist, could hardly be recognisable by ordinary
methods of examination. It would moreover be capable of re-abso*pti<G*i into
the body of liquid if unduly concentrated by contraction of surface, differing
in this respect from the gross, and undoubtedly solidr pellicles which form on
the surface of hard water on exposure to the atmosphere.
Two further observations relative to saponine may here find a place. The
wrinkling of a bubble when the contained gas is exhausted occurs aba in an
atmosphere (of coal gas) from which oxygen and carbonic acid are excluded.
In Plateau's experiment a needle which is held stiffly upon the surface
of a saponine solution is to a great extent released, when the surface is
contaminated by grease from the finger or by a minute drop of pemno4enm.
To return to the case of water, it is a question of the utmost importance
to decide whether the superficial viscosity of even distilled water is. or is nos.
due to contamination with a film of foreign matter capable of lowering the
IIIMJHII The experiments of Oberbeck would appear to render the former
alternative very improbable : but* on the other hand, if the existence of the
film be once admitted, the observed facts can be very readily explained. The
ijlMElinii is thus reduced to this: Can we believe that the water surface in
Plateau's apparatus is almost of necessity contaminated with a greasy film ?
The argument which originally weighed most with me, in favour of the
affirmative answer, is derived from the experiments of Qnincke upon mercury.
It is known that, contrary to all analogy, a drop of water does not ordinarily
spread upon the surface of mercury. This is certainly due to contamination
with a greasy film; for Professor Qnincke* found that it was possible so to
* P*». A**. Yd. ODEOX. 1810. p. 6*.
368
ON THE SUPERFICIAL VISCOSITY OF WATER.
[170
prepare mercury that water would spread upon it. But the precautions
required are so elaborate that probably no one outside Professor Quincke's
laboratory has ever witnessed what must nevertheless be regarded as the
normal behaviour of these two bodies in presence of one another. The
bearing of this upon the question under discussion is obvious. If it be so
difficult to obtain a mercury surface which shall stand one test of purity, why
may it not be equally difficult to prepare a water surface competent to pass
another ?
The method by which I have succeeded in proving that Plateau's super-
ficial viscosity is really due to contamination consists in the preparation of a
pure surface exhibiting quite different phenomena ; and it was suggested to
me by an experiment of Mr Aitken*. This observer found that, if a gentle
stream of air be directed vertically downwards upon the surface of water
dusted over with fine powder, a place is cleared round the point of impact.
It may be added that on the cessation of the wind the dust returns, showing
that the tension of the bared spot exceeds that of the surrounding surface.
The apparatus, shown in Figs. 3, 4, is constructed of sheet brass. The
circular part, which may be called the well, has the dimensions given by
Fig. 3.
PLAN
Fig. 4.
Scale =|.
Plateau. The diameter is 11 cm., and the depth 6 cm. The needle is
10 cm. long, 7 mm. in breadth at the centre, and about 0'3 mm. thick. It is
* Loc. cit. p. 69.
1890] OX THE SUPERFICIAL VISCOSITY OF WATER. 369
suspended at a height of 2£ cm. above the bottom of the vessel. So far there
is nothing special ; but in connexion with the well there is a rectangular
trough, or tail-piece, about 2£ cm. broad and 20 cm. long. Between the two
parts a sliding door may be inserted, by which the connexion is cut off, and
the circular periphery of the well completed. The action of the apparatus
depends upon a stream of wind, supplied from an acoustic bellows, and
discharged from a glass nozzle, in a direction slightly downwards, so as to
strike the water surface in the tail-piece at a point a little beyond the door.
The effect of the wind is to carry any greasy film towards the far end, and
thus to purify the near end of the tail-piece. When the door is up, this
effect influences also the water surface in the well upon which the jet does
not operate directly. For, if the tension there be sensibly less than that
of the neighbouring surface in the tail-piece, an outward flow is generated,
and persists as long as the difference of tensions is sensible. The movements
of the surface are easily watched if a little sulphur be dusted over ; when the
water in the well has been so far cleansed that but little further movement
is visible, the experiment may be repeated without changing the water by
contaminating the surface with a little grease from the finger or otherwise.
In this way the surface may be freed from an insoluble contamination anv
number of times, the accumulation of impurity at the far end of the tail-piece
not interfering with the cleanness of the surface in the well.
Another device that I have usually employed facilitates, or at any rate
hastens, the cleansing process. When the operation is nearly complete, the
movement of the surface becomes sluggish on account of the approximate
balance of tensions. At this stage the movement may be revived, and the
purification accelerated, by the application of heat to the bottom of the
well at the part furthest removed from the tail-piece. It may, perhaps, be
thought that convection currents might be substituted altogether for wind ;
but in my experience it is not so. Until a high degree of purity is attained,
the operation of convection currents does not extend to the surface, being
resisted by the film according to Marangoni's principle.
When the apparatus was designed, it was hoped that the door could be
made a sufficiently good fit to prevent the return of the greasy film into the
well ; but experience showed that this could not be relied upon. It was thus
necessary to maintain the wind during the whole time of observation. The
door was, however, useful in intercepting mechanical disturbance.
A very large number of consistent observations have been recorded. The
return of the needle, after deflection to 90°, is timed over an arc of 60°, viz.,
from 90° to 30°, and is assisted by a fixed steel magnet acting in aid of the
earth's magnetism. A metronome, beating three times per second, facilitates
the time measurement. As an example, I may quote some observations
made on April 11.
24
370 ON THE SUPERFICIAL VISCOSITY OF WATER. [170
The apparatus was rinsed and carefully filled with distilled water. In
this state the time was 12 (beats). After blowing for a while there was
a reduction to 10, and after another operation to 8. The assistance of
convection currents was then appealed to, and the time fell to 6f , and after
another operation to 6. This appeared to be the limit. The door was then
opened, and the wind stopped, with the result that the time rose again to 12.
More water was then poured in until the needle was drowned to the depth
of about half an inch. Under these conditions the time was 6f .
It will be seen, that while upon the unprepared surface the time was
nearly twice as great as in the interior, upon the purified surface the time
was somewhat less than in the interior.
For the sake of comparison, precisely similar observations were made
upon the same day with substitution for water of methylated alcohol. Before
the operation of wind the time was 5 ; after wind, 5 ; on repetition, still 5.
Nor with the aid of convection currents could any reduction be effected.
When the needle was drowned, the time rose to 7J. The alcohol thus
presents, as Plateau found, a great contrast with the unprepared water; but
comparatively little with the water after treatment by wind and heat.
An even more delicate test than the time of vibration is afforded by the
behaviour of the surface of the liquid towards the advancing edge of the
needle. In order to observe this, it is necessary to have recourse to motes,
but all superfluity should be avoided. In a good light it is often possible
to see a few motes without any special dusting over. In my experience,
an unprepared water surface always behaves in the manner described by
Plateau ; that is, it takes part in rotation of the needle, almost from the first
moment. Under the action of wind a progressive change is observed. After
a time the motes do not begin their movement until the needle has described
a considerable arc. At the last stages of purification, a mote, situated upon
a radius distant 30° or 40° from the initial direction of the needle, retains its
position almost until struck ; behaving, in fact, exactly as Plateau describes
for the case of alcohol. I fancied, however, that I could detect a slight
difference between alcohol and water even in the best condition, in favour
of the former. With a little experience it was easy to predict the " time "
from observations upon motes; and it appeared that the last degrees of
purification told more upon the behaviour of the motes than upon the time
of describing the arc of 60°. It is possible, however, that a different range
from that adopted might have proved more favourable in this respect.
The special difficulties under which Plateau experimented are well known,
and appealed strongly to the sympathies of his fellow workers ; but it is not
necessary to refer to them in order to explain the fact that the water surfaces
that he employed were invariably contaminated. Guided by a knowledge
of the facts, I have several times endeavoured to obtain a clean surface
1890] OX THE SUPERFICIAL VISCOSITY OF WATER. 371
without the aid of wind, but have never seen the time less than 10. More
often it is 12, 13, or 14. It is difficult to decide upon the source of the
contamination. If we suppose that the greasy matter is dissolved, or at any
rate suspended in the body of the liquid in a fine state of subdivision, it is
rather difficult to understand the comparative permanence of the cleansed
surfaces. In the case of distilled water, the condition will usually remain
without material change for several minutes. On the other hand, with
tap water (from an open cistern), which I have often used, although there
is no difficulty in getting a clean surface, there is usually a more rapid
deterioration on standing. The progressive diminution of the tension
of well-protected water surfaces observed by Quincke* is most readily
explained by the gradual formation of a greasy layer composed of matter
supplied from the interior, and present only in minute quantity; although
this view did not apparently commend itself to Quincke himself. If we
reject the supposition that the greasy layer is evolved from the interior of the
liquid, we must admit that the originally clean free surface, formed as the
liquid issues from a tap, is practically certain to receive contamination from
the solid bodies with which it comes into contact. The view, put forward
hypothetically by Oberbeck, that contamination is almost instantly received
from the atmosphere is inconsistent with the facts already mentioned.
Some further observations, made in the hope of elucidating this question,
may here be recorded. First, as to the effect of soap, or rather oleate of soda.
A surface of distilled water was prepared by wind and heat until the time
was 5£, indicating a high degree of purity. The door being closed, so as to
isolate the two parts of the surface, and the wind being maintained all the
while, a few drops of solution of oleate were added to the water in the
tail-piece. With the aid of gentle stirring, the oleate found its way, in a few
minutes, under the door, and reached the surface of the water in the well.
The time gradually rose to 13, 14, 15; and no subsequent treatment with
wind and heat would reduce it again below 12. In this case there can be no
doubt that the contamination comes from the interior, and is quickly renewed
if necessary; not, however, so quickly that the tension is constant in spite
of extension, or the surface would be free from superficial viscosity.
In like manner, the time upon the surface of camphorated distilled water
could not be reduced below 10, and the behaviour of motes before the
advancing needle was quite different to that observed upon a clean surface.
A nearly saturated solution of chloride of sodium could not be freed from
superficial viscosity ; while, on the other hand, an addition of £ per cent, of
alcohol did not modify the behaviour of distilled water.
The films of grease that may be made evident in Plateau's apparatus
are attenuated in the highest degree. In a recent paperf I have estimated
* Fogg. Ann. Vol. CLX. 1877, p. 580. t Supra, p. 36*. [Vol. m. p. 347.]
24—2
372 ON THE SUPERFICIAL VISCOSITY OF WATER. [170
the thickness of films of olive oil competent to check the movements of
camphor fragments as from one to two micro-millimetres; but these films
are comparatively coarse. For example, there was never any difficulty in
obtaining from tap-water surfaces upon which camphor was fully active
without the aid of wind or special arrangements. I was naturally desirous
of instituting a comparison between the quantities necessary to check
camphor movements and the more minute ones which could be rendered
manifest by Plateau's needle; but the problem is of no ordinary difficulty.
A direct weighing of the contamination is out of the question, seeing that
the quantity of oil required in the well of the apparatus, even to stop
camphor, would be only ^ mg.
The method that I have employed depends upon the preparation of an
ethereal solution of olive oil, with which clean platinum surfaces are con-
taminated. It may be applied in two ways. Either we may rely upon the
composition of the solution to calculate the weight of oil remaining upon the
platinum after evaporation of the solvent, or we may determine the relative
quantities of solution required to produce the two sorts of effects. In the
latter case we are independent of the precise composition of the solution,
and more especially of the question whether the ether may be regarded as
originally free from dissolved oil of an involatile character. In practice,
both methods have been used.
The results were not quite so regular as had been hoped, the difficulty
appearing to be that the oil left by evaporation upon platinum was not
completely transferred to the water surface when the platinum was immersed,
even although the operation was performed slowly, and repeated two or three
times. On the other hand, there was no difficulty in cleansing a large surface
of platinum by ignition in the flame of a spirit lamp, so that it was absolutely
without perceptible effect upon the movement of the needle over a purified
water surface.
The first solution that was used contained 7 mgs. of oil in 50 c.c. of ether.
The quantities of solution employed were reckoned in drops, taken under
conditions favourable to uniformity, and of such dimensions that 100 drops
measured 0'6 c.c. The following is an example of the results obtained : — On
April 25, the apparatus was rinsed out and recharged with distilled water.
Time = 13. After purification of surface by wind and heat, 5^, rising, after
a considerable interval, to 6. After insertion of a large plate of platinum,
recently heated to redness, time unchanged. A. narrow strip of platinum,
upon which, after a previous ignition, three drops of the ethereal solution had
been evaporated, was then immersed, with the result that the time was at
once increased to 8|. In subsequent trials two drops never failed to produce
a distinct effect. Special experiments, in which the standard ether was
tested after evaporation upon platinum, showed that nearly the whole of the
effect was due to the oil purposely dissolved.
1890] OS THE SUPERFICIAL VISCOSITY OF WATER. 373
The determination of the number of drops necessary to check the move-
ments of camphor upon the same surface seemed to be subject to a greater
irregularity. In some trials 20 drops sufficed : while in others 40 or 50 drops
were barely enough. There seems to be no doubt that the oil is left in
a rather unfavourable condition*, very different from that of the compact
drop upon the small platinum surface of former experiments: and the
appearance of the platinum on withdrawal from the water often indicates
that it is still greasy. Under these circumstances it is clearly the smaller
number that should be adopted : but we are safe in saying that -fe of the oil
required to check camphor produces a perceptible effect upon the time in
Plateau's experiment, and still more upon the behaviour of the surface
before the advancing needle., as tested by observation of motes. At this
rate the thickness at which superficial viscosity becomes sensible in Plateau's
apparatus is about ^ of a micro-millimetre, or about ^^ of the wave-length
of yeDow light.
A tolerably concordant result is obtained from a direct estimate of the
smaller quantity of oil, combined with the former results for camphor, which
were arrived at under more favourable conditions. The amount of oil in two
drops of the solution is about 0-0017 mg. This is the quantity which suffices
to produce a visible effect upon the needle. On the large surface of water
of the former experiments the oil required to check camphor was about 1 mg.
In order to allow for the difference in area, this must be reduced 64 times,
or to 0*016 mg. According to this estimate the ratio of thicknesses for the
two classes of effects is about as 10 : 1.
Very similar results were obtained from experiments with an ethereal
solution of double strength, one drop of which, evaporated as before, upon
platinum, produced a distinct effect upon the time occupied by the needle in
traversing the arc from 90° to 30°.
I had expected to find a higher ratio than these observations bring out
between the thicknesses required for the two effects. The ratio 15 : 1 does
not give any too much room for the surfaces of ordinary tap water, such
as were used in the bath observations upon camphor, between the purified
surfaces on the one side and those oiled surfaces upon the other which do not
permit the camphor movements.
It thus became of interest to inquire in what proportion the film
originally present upon the water in the bath experiments requires to be
concentrated in order to check the motion of camphor fragments. This
information may be obtained, somewhat roughly it is true, by dusting over
a patch of the water surface in the centre of the bath. When a weighed
drop of oil is deposited in the patch, it drives the dust nearly to the edge,
* It should be stated that the evaporation of the ether, and of the dew which was often
by the application of a gentle warmth.
374 ON THE SUPERFICIAL VISCOSITY OF WATER. [170
and the width of the annulus is a measure of the original impurity of the
surface. When the deposited oil is about sufficient to check the camphor
movements, we may infer that the original film bears to the camphor
standard a ratio equal to that of the area of the annulus to the whole area
of the bath. Observations of this kind indicated that a concentration of about
six times would convert the original film into one upon which camphor would
not freely rotate.
Another method by which this problem may be attacked depends upon
the use of flexible solid boundary. This was made of thin sheet brass,
and is deposited upon the bath in its expanded condition, so as to enclose
a considerable area. Upon this surface camphor rotates, but the movement
may be stopped by the approximation of the walls of the boundary. The
results obtained by this method were of the same order of magnitude.
If these conclusions may be relied upon, it will follow that the initial
film upon the water in the bath experiments is not a large multiple of that
at which superficial viscosity tends to disappear. At the same time, the
estimate of the total quantity of oil which must be placed upon a really
pure surface in order to check the movements of camphor must be somewhat
raised, say, from 1'6 to 1'9 micro-millimetre. It must be remembered,
however, that on account of the want of definiteness in the effects, these
estimates are necessarily somewhat vague. By a modification of Plateau's
apparatus, or even in the manner of taking the observations, such as would
increase the extent of surface from which the film might be accumulated
before the advancing edge of the needle, it would doubtless be possible to
render evident still more minute contaminations than that estimated above
at one-tenth of a micro-millimetre.
Postscript, June 4. — In order to interpret with safety the results obtained
by Plateau, I thought it necessary to follow closely his experimental arrange-
ments ; but the leading features of the phenomenon may be well illustrated
without any special apparatus. For this purpose, the needle of the former
experiments may be mounted upon the surface of water contained to a depth
of 1 or 2 inches in a large flat bath. Ordinary cleanliness being observed, the
motes lying in the area swept over by the needle are found to behave much
as described by Plateau. Moreover, the motion of the needle, under the
action of the magnet used to displace it, is decidedly sluggish. In order to
purify the surface, a hoop of thin sheet brass is placed in the bath, so as to
isolate a part including the needle. The width of the hoop must of course
exceed the depth of the water, and that to an extent sufficient to allow of
manipulation without contact of the fingers with the water. If the hoop
be deposited in its contracted state, and be then opened out, the surface
contamination is diminished in the ratio of the areas. By this simple device
1890] ON THE SUPERFICIAL VISCOSITY OF WATER. 375
there is no difficulty in obtaining a highly purified surface, upon which motes
lie quiescent, almost until struck by the oscillating needle. In agreement
with what has been stated above, an expansion of three or four times usually
sufficed to convert the ordinary water surface into one upon which superficial
viscosity was tending to disappear.
I propose to make determinations of the actual tensions of surfaces
contaminated to various degrees ; but in the meantime it is evident that the
higher degrees of purity do not imply much change of tension*. In the last
experiment upon a tolerably pure surface, if we cause the needle to oscillate
rapidly backwards and forwards through a somewhat large angle, we can
clear away the contamination from a certain area. This contamination will
of course tend to return, but observation of motes shows that the process is
a rather slow one.
The smallness of the forces at work must be the explanation of the failure
to clean the surface in Plateau's apparatus by mere expansion. For this
experiment the end wall was removed from the tail-piece (Fig. 3), and a large
flexible hoop substituted. By this means, it was hoped that when the whole
was placed in the bath it would be possible, by mere expansion of the hoop,
to obtain a clean surface in the well. The event proved, however, that the
purification did not proceed readily beyond the earlier stages, unless the
passage of the contamination through the long channel of the tail-piece was
facilitated by wind.
* [1901. Miss Pockels' experiments (Nature, xioii. p. 437, 1891) show that this is an
understatement. See also Bayleigh, Phil. Mag. u.vm. p. 331, 1899.]
171.
ON HUYGENS'S GEARING IN ILLUSTRATION OF THE
INDUCTION OF ELECTRIC CURRENTS*.
[Philosophical Magazine, xxx. pp. 30—32, 1890.]
As a mechanical model of the electric machinery at work in the induction
of currents, Maxwell employed differential gearing ; and an apparatus on this
principle, designed by him, is in use at the Cavendish Laboratory. Wishing
to show something similar in a recent course of lectures, and not having
differential gearing at my disposal, I designed more than one combination of
pulleys, the action of which should be analogous to that of electric currents.
These eventually resolved themselves into Huygens's gearing, invented, I
believe, in connexion with the winding of clocks. As this apparatus is easier
to understand than differential gearing, and the parts of which it is composed
are more likely to be useful for general purposes in a laboratory, I have
thought that it might be worth while to give a description, accompanied by
an explanation of the mode of action.
Two similar pulleys, A, B, turn upon a piece of round steel fixed
horizontally f. Over these is hung an endless cord, and the two bights
carry similar pendent pulleys, G, D, from which again hang weights, E, F.
The weight of the cord being negligible, the system is devoid of potential
energy; that is, it will balance, whatever may be the vertical distance
between C and D.
Since either pulley A, B may turn independently of the other, the system
is capable of two independent motions. If A, B turn in the same direction
and with the same velocity, one of the pendent pulleys G, D rises, and the
other falls. If, on the other hand, the motions of A, B are equal and opposite,
the axes of the pendent pulleys and the attached weights remain at rest.
* Read before the Physical Society on May 16, 1890.
t Light wooden laths, variously coloured and revolving with the pulleys, render the move-
ments evident at a distance.
1890]
HUYGENS'S GEARING.
377
In the electrical analogy the rotatory velocity of A corresponds to a
current in a primary circuit, that of B to a current in a
secondary. If when all is at rest the rotation of A be
suddenly started, by force applied at the handle or other-
wise, the inertia of the masses, E, F, opposes their sudden
movement, and the consequence is that the pulley B turns
backwards, i.e., in the opposite direction to the rotation
imposed upon A. This is the current induced in a secondary
circuit when an electromotive force begins to act in the
primary. In like manner, if A having been for some time
in uniform movement suddenly stops, B enters into motion in
the direction of the former movement of A. This is the
secondary current on the break of the current in the primary
circuit.
It must be borne in mind that in the absence of friction
there is nothing to correspond with electrical resistance, so
that the conductors must be looked upon as perfect. The
frictions which actually enter do not follow the same laws as
electrical resistances, and only very imperfectly represent
them. However, the frictions which oppose the rotations of
A and B have a general effect of the right sort; but the
rotations of C and D, corresponding to dielectric machinery, should be as
free as
The effect of a condenser, to which the terminals of one of the circuits is
joined, would be represented by a spiral spring (as in a watch) attached to
the corresponding pulley, the stiffness of the spring being inversely as the
capacity of the condenser. The absence of the spring, or (which comes to
the same thing) the indefinite decrease of its stiffness, corresponds to infinite
electrical capacity, or to a simply closed circuit.
The equations which express the mechanical properties of the system are
readily found, and are precisely the same as those applicable in the electrical
problem. Since the potential energy vanishes, everything turns upon the
expression for the kinetic energy. If x and y denote the circumferential
velocities, in the same direction, of the pulleys A, B where the cord is in
contact with them, \ (x + y) is the vertical velocity of the pendent pulleys.
Also 4 (x — y) is the circumferential linear velocity of C, D, due to rotation, at
the place where the cord engages. If the diameter be here 2a, the angular
velocity is (x — y)!2a. Thus, if J/ be the total mass of each pendent pulley
and attachment, JHk* the moment of inertia of the revolving parts, the whole
kinetic energy corresponding to each is
378 HUYGENS'S GEARING. [171
For the energy of the whole system we should have the double of this, and,
if it were necessary to include them, terms proportional to a? and y2 to
represent the energy of the fixed pulleys. The reaction between the pulleys
A, B depends upon the presence of a term xy in the expression of the energy.
We see that this would disappear if fc2 = a2 ; as would happen if the whole
mass of the pendent pulleys and attachments were concentrated in the circles
where the cord runs. The case discussed above, as analogous to electric
currents, occurs when Id1 < a2, a condition that will be satisfied, even without
non-rotating attachments, if the cord run near the circumference of the
rotating pulleys. The opposite state of things, in which k? > a2, would be
realized by carrying out masses beyond the groove, and thus increasing the
rotatory in comparison with the translatory inertia. In this case the mutual
action between A and B is reversed. If when all is at rest A be suddenly
started, B moves forward in the same direction. Otherwise C and D would
have to rotate, and this in their character of fly-wheels they oppose.
Generally, if L, N be the coefficients of self-induction, and M of mutual
induction, we have (constant factors being omitted)
In order to imitate the case of two circuits coiled together in close proximity
throughout, we must have in the mechanical model k2 = 0 ; that is, the
rotatory inertia of the pendent pulleys must be negligible in comparison with
the translatory inertia. Also the energy of the fixed pulleys, not included in
the above expressions, must be negligible. If these conditions be satisfied, a
sudden rotation imposed upon A generates an equal and opposite motion in B.
172.
THE BOURDON GAUGE.
[.Voter*, XLH. p. 197, 1890*]
ALLOW me to suggest; to such of jour readers as are interested in this
subject the following experiment. Cat out of cardboard two annular strips,
each of somewhat more than a quadrant, the inner radius being, say, 7 inches,
and the outer radius 9 inches. Along the middle of each strip — that is,
along the circle of 8 inches radius — cut the boards half through, so as to
render them flexible, and then join the two strips together with gum paper
at the inner and outer edges. In this way we obtain a curved tube whose
section is a rhombus, and whose curvature is connected with the magnitudes
of the angles of the rhombus. The manipulation of such a tube gives
to one's ideas, and enables one to recognize that internal
tending to augment the included volume, and therefore to make
the section square, must also cause the curvature of the axis to approach
a definite associated value. In this case the deformations are practically by
bending, principally indeed at the hinges; and I cannot doubt that in its
main features the mechanism of an ordinary Bourdon gauge may be looked
at in the same light.
* [1901. See YoL m. p. 230. Hie present note had reference to a I'm IIHJIIB upon the theory
of the Bondon gauge which appealed
173.
ON DEFECTIVE COLOUR VISION.
[British Association Report (Leeds), pp. 728—729, 1890.]
THE existence of a defect is probably most easily detected in the first
instance by Holmgren's wool test ; but this method does not decide whether
the vision is truly dichromic. For this purpose we may fall back upon
Maxwell's colour discs. Dichromic vision allows a match between any four
colours, of which black may be one. Thus we may find 64 green + 36 blue
= 61 black -f 39 white — a neutral matched by a green-blue. But this is
apparently not the most searching test. The above match was in fact made
by an observer whose vision I have reason to think is not truly dichromic, for
he was unable to make a match among the four colours red, green, blue, black.
The nearest approach appeared to be 100 red = 8 green + 7 blue + 85 black,
but was pronounced far from satisfactory. An observer with dichromic vision,
present at the same time, made without difficulty 82 red + 18 blue = 22 green
+ 78 black — a bright crimson against a very dark green.
It would usually be very unsafe to conclude that a colour-blind person is
incapable of making a match because he thinks himself so. But, in the
present instance, repeated trials led to the same result, while other matches,
almost equally forced in my estimation, were effected without special difficulty.
It looked as though the third colour sensation, presumably red, was defective,
but not absolutely missing. When a large amount of white was present,
matches could be made in spite of considerable differences in the red
component, but when red light was nearly isolated its distinctive character
became apparent.
This view of the matter was confirmed by experiments with my colour
box, in which, by means of double refraction, a mixture of spectral red and
green can be exhibited in juxtaposition with spectral yellow (Nature, Nov. 17,
1881) [Vol. I. p. 545]. A match to normal vision requires, of course, that
(by rotation of the nicol) the red and green should be mixed in the right
1890]
OX DEFECTIVE COLOUR VTSIOX.
proportions : and secondly, that (by adjustment of gas) the brightness of the
spectral yellow should be brought to the right point. An observer whose
vision is dichromic does not require the first adjustment ; any mixture of red
and green, or even the red and green unmixed, can be matched against the
yellow. In the present case, however, although the green could be matched
satisfactorily against the yellow, the red could not. The construction of the
instrument allowed the point to be investigated at which the match began
to fail Pure green corresponding to 0, and pure red to 25, the match with
yellow began to foil when the setting reached about 17. Normal vision
required a setting of about 11.
Truly dichromic vision may be thus exhibited in a diagram. If we take
red, green, blue, as angular points of a triangle, there is a point upon the
plane which represents darkness. Any colours which lie upon a line through
this point differ only in brightness. Maxwell determined the point by
comparison of colour-blind matches with his own normal ones. It seems
preferable to use the colour-blind matches only, as may be done as follows : —
From the match between red, green, blue, and black, the position of black
on the diagram may be at once determined, and for most purposes would
represent darkness sufficiently welL A match between white and the
principal colours will then fix its [white's] position relatively to the funda-
mental points. A line joining black and white is the neutral line : all colours
that lie on one side of it are warm, like yellow : all that lie upon the other
side are cold. The point representing darkness will lie upon the neutral
line and a little beyond black. The diagram sketched depends upD»n the
following matches obtained from an observer, whom Holmgren would call
green-blind : —
Bed
Green
Blue Black White
rdkn
-»C
+21-8
-180 +78-2 0
0
+57-2
-100
| +4-8
0
+380
0
0
+960
+ 40
-53-0
-47-0
0
-100
0
+ 5-0
+78-7
0
+ 16-3
174.
INSTANTANEOUS PHOTOGRAPHS OF WATER JETS.
[British Association Report (Leeds), p. 752, 1890.]
THESE photographs were taken by the light of the electric spark. A
battery of Leyden jars was charged by a Wimshurst machine, and discharged
itself between brass balls, held half an inch apart, in the optical lantern. By
means of a large condenser a good proportion of light was concentrated upon
the lens of the camera. The jet of water, regularized by a tuning-fork, fell
in front of the condenser, and was focused upon the photographic plate.
In the absence of anything to diffuse the light, the pictures are simple
shadows, such as have been obtained without any optical appliances by
Mr Bell and Mr Boys. The only detail is due to the lens-like action of the
jets and the drops into which it is resolved. This arrangement is quite
sufficient to illustrate the behaviour of electrified jets. But the interposition
of a plate of ground glass close to the condenser effects a great improvement
in the pictures by bringing out half-tones, and the results printed on
aristotype paper are now very good. The only difficulty is that due to loss
of light. In some of the experiments it was found advantageous to diminish
the diffusion by slightly oiling the ground glass.
The degree of instantaneity required depends upon circumstances. In
some cases the outlines would have lost their sharpness had the exposure
exceeded -ij^wu second. It is probable that the actual duration of the
principal illumination was decidedly less than this.
175.
ON THE TENSION OF WATER SURFACES, CLEAN AND
CONTAMINATED, INVESTIGATED BY THE METHOD OF
RIPPLES*.
[Philosophical Magazine, xxx. pp. 386—400, Nov. 1890.]
NUMEROUS and varied phenomena prove that the tension of a water
surface is lowered by the presence of even a trace of grease. In the case
of olive-oil, a film whose calculated thickness is as low as 2 micro-millimetres
is sufficient to entirely alter the properties of the surface in relation to
fragments of camphor floating thereupon. It seemed to me of importance
for the theory of capillarity to ascertain with some approach to precision
the tensions of greasy surfaces ; and in a recent paperf I gave some results
applicable to the comparison of a clean surface with one just greasy enough
to stop the camphor movements and also with one saturated with olive-oil.
The method employed was that depending upon the rise of liquid between
parallel plates of glass ; and I was not satisfied with it, not merely on account
of the roughness of the measurement, but also because the interpretation
of the result depends upon the assumption that the angle of contact with
the glass is zero. In the opinion of Prof. Quincke, whose widely extended
researches in this field give great weight to his authority, this assumption
is incorrect even in the case of pure liquids, and, as it seemed to me, is still
less to be trusted in its application to contaminated surfaces, the behaviour
of which is still in many respects obscure. I was thus desirous of checking
my results by a method independent of the presence of a solid body.
The solution of the problem was evidently to be found in the observation
of ripples, as proposed by Prof. Tait, upon the basis of Sir W. Thomson's
theory. Thomson has shown that when the wave-length is small, the
* Bead September 6 before Section A of the British Association at Leeds. [Brit. Att. Rep.
p. 746, 1890.]
t Proc. Roy. Soe. March 1890, Vol. XLVIL p. 367. [VoL in. p. 350.]
384 ON THE TENSION OF WATER [175
vibration depends principally upon capillary tension; so that a knowledge
of corresponding wave-lengths and periods will lead to a tolerably accurate
estimate of tension.
Besides some early observations of my own*, made for the most part
for another purpose, I had before me the work of Matthiessenf, who has
compared Thomson's formulae with observation over a wide range of wave-
length. The results are calculated on the basis of an assumed surface-tension,
and are exhibited as a comparison of calculated and observed wave-lengths.
On the whole the agreement is fair ; but the accuracy attained seemed to be
insufficient for the purpose which I had in view. As will presently appear,
an error in the wave-length is multiplied about three times in the tension
deduced from it, so that, in a reversal of Matthiessen's calculations, the errors
would appear much magnified.
Quite recently Mr Michie Smith has published an account of experiments
made by Thomson's method for the determination of the tension of mercury.
Some anomalies were met with ; and it seems not improbable that the
vibrations observed were in some cases an octave below those of the vibrating
source J.
When it is remembered that Thomson's theory is one of infinitely small
vibrations, it will be seen that for my purpose it was necessary above all
things that the amplitude of vibration should be very moderate. The sub-
octave vibrations of Faraday are especially to be avoided as almost necessarily
of large amplitude. At the same time the limitation is not without its
inconvenience. One of the great difficulties of the experiment is to see the
waves properly, and this is much increased when the vibrations are extremely
small.
In considering the problem thus presented, it occurred to me that it was
essentially the same as that so successfully solved by Foucault in relation to
the figuring of optical surfaces. The undisturbed surface of liquid is an
accurate plane. The waves upon it may be regarded as deviations from
optical truth, and may be made evident in the same way as any other
deviations from truth in a reflecting surface. Guided by this idea, I was
able to work with waves of which nothing whatever was to be seen by
ordinary observation of the surface over which they were travelling.
In the application of Foucault's method it is necessary that light from a
radiant point, after reflexion from the surface under test, should be brought
to a focus, in the immediate neighbourhood of which is placed the eye of the
* "On the Crispations of Fluid resting upon a Vibrating Support," Phil. Mag. July 1883.
[Vol. ii. p. 212.]
t Wied. Ann. xxxvin. p. 118 (1889).
t Faraday, Phil. Trans. 1831. See also Rayleigh, Phil. Mag. April and July 1883. [Vol. n.
pp. 188, 212.]
SURFACES, CLEAN ASTD COXTAMDfATED. 385
observer. Any small irregularities in the surface then render themselves
conspicuous to the eje focused upon it. In the present case the reflector is
plane, and the formation of a real image of the radiant requires the aid of
a lens. In mj experiments this was usually a large single lens of 6 inches
diameter and 34 inches focus. On one occasion an achromatic telescope-lens
was substituted.,, but the aperture was too small to include the number of
waves necessary for accuracy. Although the want of achromatism was
prejudicial to the appearance of the image, it is not certain that the accuracy
of the determinations was impaired, at least after experience in observation
had been acquired. The lens was fixed horizontally near the floor, a few
inches above the surface of the water under examination. The radiant point,
a very small gas-flame, was situated in the principal focal plane, but a little
on one side of the axis of symmetry, so that the image formed after reflexion
from the water and a double passage through the lens might be a little
separated from the source. For greater convenience reflecting strips of
looking-glass were introduced at angles of 45°, or thereabouts,, so that the
initial and final directions of the rays were horizontal.
The smalmess of the disturbance is not the only obstacle to its visibility.
Even with Foucault's arrangement for viewing minute departures from
planeness, nothing could usually be seen of the waves here employed without
a further device necessary on account of the rapidity with which all phases
are presented in succession. A clear view of the waves most be an inter-
mittent one, isoperiodic with the vibrations themselves, and may be obtained
in the manner first described by Plateau. In the present case it was found
simplest to render the light itself intermittent. Close in front of the small
gas-flame was placed a vibrating blade of tin-plate rigidly attached t-> the
extremity of the prong of a large tuning-fork, and so situated that once
during each vibration the light was intercepted by the interposition of the
Made. The vibrations of the fork were maintained electroinagnetically in the
usual manner, and the intermittent current furnished by the interrupter fork
was utilized, as in HeJmholtzs vowel-sound experiments, to excite a second,
in unison with itself. The second fork generated waves in the dish of water
by means of a dipper attached to its lower prong.
When the action is regular, the vibrations of the two forks are strictly
isochronous, even though the natural periods may differ somewhat*. The
view presented to the observer is then perfectly steady, and corresponds to
one particular phase of the vibration, or rather, since the illumination is not
A dirty condition of the
tte. Di
in their natural period. In this waj a
;-
386 ON THE TENSION OF WATER [175
instantaneous, to an average of phases in the neighbourhood of a par-
ticular one.
Even in the case of a perfectly regular train of waves, the appearance will
depend upon the precise position occupied by the eye. It is evident that the
light most diverted from its course is that reflected from the shoulders of the
waves — the points midway between the troughs and crests, for it is here that
the slope of the surface is greatest. Thus if the eye be moved laterally
outwards from the focal point, until all light has nearly disappeared, the
residual illumination will mark out the instantaneous positions of one set of
shoulders, all other parts of the complete wave remaining dark. This is one
of the most favourable positions for observation. If the deviation from the
focal point be in the opposite direction, the other set of shoulders will be
seen bright.
The aspect of the waves was not always equally pleasing. Sometimes the
formation of stationary waves, due to reflexions, interfered with regularity.
A readjustment of the walls of the vessel relatively to the dipper would then
often effect an improvement. The essential thing is that there should be no
ambiguity in the wave pattern over the measured part of the field. It would
occasionally happen that in certain positions of the eye a change of phase
would occur in the middle of the field, so that the bright bands in one part
were the continuation of the dark bands of another part. Near the transition
Fig. l.
the bands would appear confused, a sufficient indication that no measurement
must be attempted. On the other hand, it is not necessary that the contrast
1890] SURFACES, CLEAN AND CONTAMINATED. 387
between the dark and bright parts should be very great. Indeed the
measuring marks were better seen when no part of the field was very dark.
Fig. 1 gives a general idea of the appearance of the field. On the right
is seen a paper with a notched edge, the use of which was to facilitate the
counting. The measuring arrangement was something like a beam compass.
Stout brass wires, attached to a bar of iron, were shaped at their ends like
bradawls, and the edges were placed parallel to the crests of the waves. In
order to avoid residual parallax, the rod was so supported that the edges were
in close proximity to the water surface.
In many of the experiments the distance between the edges was set
beforehand, e.g. to 10 cm., and was not altered when the wave-lengths varied
with the deposition of grease. The number of wave-lengths included was
determined by counting, and estimation of tenths. Usually the discrepancy
between Mr Gordon's estimation and my own did not exceed a single tenth,
and in a large proportion of cases there was no difference. Probably the
mean of our readings would rarely be wrong by more than ^V °f a wave-
length, when the pattern was well seen. In the experiments specially-
directed to the determination of the tension of a clean surface, it was found
advisable to work with an unknown distance ; otherwise the recollection of
previous results interfered with the independence of the estimates.
It is probable that somewhat greater accuracy in single measurements
might have been attained had the distance been adjustable by a smooth
motion within reach of the observer. Each measuring edge might then have
been set to the most favourable position, that is, to the centre of a bright
band. The frequent removal of the apparatus for comparison with a scale
would, however, be rather objectionable; and it was thought doubtful
whether any final gain would accrue in the mean of several observations.
Some trouble was experienced from the communication of vibration
through unintended channels. In order to prevent the direct influence of
the interrupter fork upon the liquid surface, it was found advantageous to
isolate it from the floor by supporting it upon a shelf carried upon the walls
across a corner of the laboratory. On one occasion it was noticed that the
waves were visible without the aid of the arrangement for making the light
intermittent. This was traced to a tremor of one of the mirrors, supported
upon the same shelf as the interrupter fork. Such a method of rendering
the waves visible is objectionable, since it destroys the definition of the
measuring points. The tremor was eliminated by the introduction of rubber
tubing under the stand of the interrupter.
During the experiments on greasy surfaces one pair of forks only was
employed. The frequency of the interrupter was about 42 per second, so
that the intermittent current could be used to excite a fork of about 126.
The beats between this and a standard Koenig fork of 128 were counted at
25—2
388 ON THE TENSION OF WATER [175
intervals, and found to be sufficiently constant. The pitch of the standard
has been verified by myself*, and at the temperature of the laboratory may
be taken with sufficient accuracy to be 128. If we take the number of beats
per minute at 98, we have for the frequency of the interrupter
8- HJ) =4212.
In the case of clean water another pair of forks of about 128 was employed as
a check. The number of beats was 184 per minute, and
/= 128 -^ = 124-9.
The water was contained in a shallow 12" x 10" porcelain dish; and before
commencing observations its surface was purified with the aid of an expansible
hoop of thin sheet brass. The width of the hoop is greater than the depth of
water, and it is deposited in the dish so as to include the dipper, but otherwise
in as contracted a condition as possible. It is then opened out to its maximum
area with the effect of attenuating many times the thickness of the greasy
film, which no amount of preliminary cleaning seems able to obviate. It not
unfrequently happened that the first attempt to get a clean surface was a
partial failure, but a repetition of the operation was usually successful. It
seems as if impurity attaches itself to the brass so obstinately that only
contact with a clean water surface will remove it.
In the earlier experiments the waves were generated by a dipper of
circular section, a closed tube of glass, somewhat like a test-tube. The
measurements were quite satisfactory, but I felt doubts as to a possible
influence of curvature upon wave-length. In order to avoid any risk of this
kind, and to render the waves straight from the commencement, a straight
horizontal edge of glass plate, about 2^ inches long, was afterwards sub-
stituted, and worked very satisfactorily. It is not necessary or desirable
that the dipper should pass in and out of the water. In most cases the
vibrations employed were very small, and the edge of the dipper was
immersed throughout.
The purity of the water surface could be judged by the result of the
observation of the number of wave-lengths; the smallest number corre-
sponding to the purest surface. But it soon became apparent that a more
delicate test was to be found in the general appearance of the wave pattern.
Upon a clean surface there is a strong tendency to irregularity, dependent no
doubt upon reflexions, which become more important when the propagation
is very free. In order to meet this, it was often found necessary to weaken
the vibrations of the secondary fork, either by putting it more out of tune
with the primary, or by shifting its magnet to a less favourable position, or,
* Phil. Trans, p. 316, 1883. [Vol. n. p. 177.]
1890] SURFACES, CLEAN AND CONTAMINATED. 389
finally, by shunting the current across. A slight trace of grease would then
render itself evident by a damping down of the waves before any change
could be observed in the wave-length. After a little experience with the
forks in a given state of adjustment, a momentary glance at the pattern was
sufficient to enable one to recognize the condition of the surface.
The interpretation of the observations depends upon the following formula,
due to Thomson : —
Let U= velocity of propagation, A. = wave-length, T = periodic time,
p = density, T = superficial tension, h = depth of water; then (Basset's
Hydrodynamics, Vol. n. p. 177)
so that to find T we have
In the present experiments the effect of the limitation of depth is
negligible. We have h =1-8 cm., and for the greatest value of X about
•7 cm. Now
approximately, when h is relatively large ; so that
coth (27rA/X) = 1 + 2 e-30 = 1 ,
with abundant accuracy. Again, in the case of water we have p = 1 ; and
thus
T= K 9*
27TT2 47T2'
which is the formula by which the calculation of T is to be made. The
second term will be found to be small in comparison with the first, so that
approximately T varies as X3. A one-per-cent. error in the estimation of X
will therefore involve one of three per cent, in the deduced value of T. In
many of the experiments about 15 waves were included between the marks.
An error of ^ of a wave is thus 1 in 150, leading to a two-per-cent. error in
T. We may expect the final mean value to be correct to less than one per
cent., but we must not be surprised if individual results show discrepancies of
two per cent.
An example (August 2) will now be given in which the surface of clean
water was greased with oleic acid. The dish after rinsing was filled with
water drawn from a tap in connexion with a cistern supplied mainly by rain
water, and placed in position. On expansion of the brass hoop, the number of
waves included between the measuring points was estimated to be 137, 13*8
390 ON THE TENSION OF WATER [175
by the two observers. A piece of paper was then greased with oleic acid,
and with this a platinum wire, previously cleaned by ignition, was wiped.
On introduction of part of the wire into the water contained within the hoop,
the number of waves rose to 15'4, 15'3. Upon this surface camphor scrapings
were found to be quite dead, so that the mark had been overshot.
The dish was then refilled. Upon expansion the number of waves upon
the clean surface was 13'7, 13'7. On contamination with a little oleic acid,
14'8, 14'8. Camphor was now moderately active. More oleic was added.
Readings were now 15*4, 15'4, and camphor was quite dead.
The point to be fixed evidently lay between 14'8 and 15*4. A fresh
surface was taken, and on addition of a little oleic the readings were 14'8,
14*8. Camphor was then tried and found moderately active. Reading still
14*8. A little more oleic added; readings 15*1, 15'1 ; camphor scrapings
were now "nearly dead." More oleic; 15'2, 15*2; camphor "very nearly
dead." More oleic; 15*4, 15'4; "not absolutely dead." More oleic; 15'5,
15'5 ; camphor " absolutely dead." The temperature of the water was 63° F.
On a previous occasion (July 29) accordant results had been obtained.
Clean water 13'7, 13'7. Oleic added; 15'0, 15*0; camphor nearly dead.
More oleic; 15'2, 15'25; camphor very nearly dead. Oleic; 15'55, 15'6 :
camphor dead. On both days the distance over which the waves were
measured was 9 '20 cm.
It may be well to exhibit in full the calculation for the clean water : —
log 9-2 = -9638 logg = 2'9917
log 13-7 = 1-1367 logX =1-8271
logX =1-8271 1-8271
3 2^6459
log\3 =1-4813 Iog47r2 = 1-5962
log 42-12 = 1-6245 log 11 -2 = 1-0497
1-6245
2-7303
Iog2-7r = -7981
log 85-5 = 1-9322
Finally, T= 85'5- ll-2 = 74'3.
If we take as the reading when the camphor is nearly dead 15'2, we find
in like manner
T= 62-7 -9-1 = 53-6.
After this example a summary of results may suffice. The interest
attaching to the determination of the tension of a clean surface led me to
strive after a higher degree of accuracy than perhaps would otherwise have
1890]
SURFACES, CLEAN AND
CONTAMINATED. 39]
been necessar
y. The following table contains the results obtained with botl
forks:—
Date
Distance Frequency
Tension
Water Temp.F. ; Remarks
>>•
Jane 23
frOo i 40*
72-3
Tap JL '
25
412 40*
74-5
73 Teleaeopelera
26
11-70 j 40*
73-7
66
30
11-27 42-12
740
61
J«iy i
9-96 4i-12
73-2
61
2
9*6
42-12
74-7
„ 62 Strip dipper
4
9*6
42-12
74-7
Distilled 64 introduced
8
9*6
42-12
74-7
60
25 !
1000
42-12
74-2
Tap 65
25 j
9-20
42-12
75-2
65
28
9-20
42-12
74-3
„
28 j
9-20
42-12
743
Distilled 63
29 1
9-20
42-12
74-3
„ ;
Aug. 2
9-20
42-12
74-3
Tap 63
July 23
1000
124*
74-1
Tap 65
23 |
9-49
124*
73-2
„ 66
23
8-13
124*
73-5
Distilled 66
i
The mean result with the graver fork is T = 74'2 : and with the quicker
one T=73~6. The discrepancy of nearly one per cent, marks the limit of
accuracy. It should be remarked that some of the consecutive results where
no variation occurred in the distance between the points cannot be regarded
as quite independent.
On several occasions distilled water proved a less satisfactory subject than
tap water. The surface seemed more unwilling to become and remain clean.
Sometimes after expansion a notable increase of readings would occur in the
course of a few minutes without assignable cause.
I was very anxious to satisfy mvself that in the surfaces experimented
upon by the wave method a high degree of purity was really attained. In
the experiments of July 28 a Plateau needle vibrating upon a portable stand
was introduced. After the examination by the method of waves, the dish
was brought out into a good light, and the quality of the surface tested by
observation of the motion of motes when the needle lying upon it was caused
to vibrate by an external magnet*. In making the necessary arrangements
* "On the Superficial Viscosity of Water," Proc. Roy. Soe. Jane 1§90, Vol. U-TTIL p. 139.
[YoL m. p. 374.]
392 ON THE TENSION OF WATER [175
there was some risk of introducing contamination, so that the discovery of an
unclean surface would prove nothing definite. If, however, the behaviour of
the surface under the needle test was good, it could be inferred with
confidence that the measured waves were not affected by impurity. On two
occasions the test succeeded fairly well.
The observations with the 128 fork were rather difficult, the waves being
about twice as close as in the other case. In the calculation of results it
appears, as was to be expected, that the importance of the second term, due
to gravity, is diminished. Thus for July 22,
T= 76-5 -2-4 = 74-1.
The general result that at temperatures such as 65° (18° C.) the tension of
clean water surfaces is about 74'0 c.G.s. absolute units of force per centimetre
seems entitled to considerable confidence. It agrees with some former obser-
vations* of my own upon the transverse vibrations of jets, as has been
remarked by Mr Worthingtonf. Some interesting experiments upon the
vibrations of falling drops by LenardJ point also in the same direction. On
the other hand it deviates largely from the higher value, about 81, which
Prof. Quincke thinks the most probable. The deviation from 81 is certainly
not due to contamination. It has been explained that great care was taken
in this respect during the present experiments; and in the jet method the
surfaces are probably the purest attainable. The method favoured by
Quincke depends upon the measurement of large flat bubbles confined under
the horizontal surface of a solid body. In default of experience I must leave
it to others to judge whether a systematic error due to optical or other causes
could enter here. Mr Worthington contends that some of Quincke's
deductions from his measurements require correction for curvature perpen-
dicular to the meridional plane. To this and other criticisms Prof. Quincke
has replied §.
Experimenters upon capillary tubes have generally been led to adopt the
lower value, but here the interpretation involves an assumption that the
angle of contact 6 is zero. What these measurements give in the first
instance is Tcosd; so that if 0 = 30°, or thereabouts, the higher value of T
is the one really indicated. This is the view adopted by Quincke, who in an
important series of observations || has shown that the edge angle between
water and glass has frequently a considerable value dependent upon the
impurity of glass surfaces, even when carefully cleaned by ordinary methods.
But I confess that the argument does not appear to me conclusive. The
angles recorded are maximum angles. If after a drop has been deposited
* Proc. Roy. Soc. Vol. xxix. p. 71, 1879. [Vol. i. p. 387.]
t Phil. Mag. Vol. xx. p. 51, 1885.
£ Wied. Ann. Bd. xxx. (1887). § Ibid. xxvu. p. 219 (1886).
|| Ibid. Vol. n. p. 145, 1877.
1890]
SURFACES, CLEAN AND CONTAMINATED.
some of the liquid is drawn off, the angle may be diminished almost to zero.
Observations upon capillary heights correspond surely to the latter condition
of things, for no experimenter measures the gradual rise of liquid in a dry
tube. I am disposed to think that the assumption 0 = 0 is legitimate, and
thus that the lower value of I1 is really supported by experiments of
this class.
Leaving now the results for pure surfaces, let us pass on to those found
for water contaminated with grease up to the point where the camphor
scrapings were judged to be " very nearly dead." It must be remembered
that the additions of oil were discontinuous, and that the point could not
always be hit with precision. On any one day it is possible to set up a
fairly precise standard of what one means by " very nearly dead " ; but the
standard is liable to vary in one's own mind, and is of course impossible to
communicate to another. Too much importance therefore must not be
ascribed to exact agreement or the failure of it. On one day experiments
were made by varying the areas enclosed within the hoop. Thus, if the
motions were a little too lively, they could be deadened to the required point
by contraction of the area and consequent concentration of grease. This
procedure was not so convenient as had been hoped, in consequence of the
mechanical disturbance attending a motion of the hoop. In all cases an
observation, for the most part recorded in the previous table, was made first
upon a clean surface, so as to ensure that the contamination was all of the
kind intended. The results are collected in the annexed table : —
Date
Water
Oil
Tension
Remarks
June 30
Tap
Olive-oil
537
July 1
n
n
5M
2
n
n
52-1
4
Distilled
53O
7
„ „
53-0
Not quite independent
11
•
n
530
i» » „ n
29
Tap
Oleic Acid
536
Aug. 2
»
n
53-6
2
»»
Olive-oil
52-4
2
„
52-4
Another sample
The tension of the surface when the camphor movements are just stopping
may thus be reckoned at 53'0 C.G.S., or about 72 per cent, of that of a clean
surface. There is some reason to infer that the tension is the same whatever
kind of grease be used. In the last experiment the sample of oil was one of
which it was necessary to take decidedly more than usual (in the ratio of
about 3 : 2) in order to stop the motions. This was proved by the balance in
394 ON THE TENSION OF WATER [175
the manner described in a former paper. I have other grounds for thinking
it probable that the tension does not depend upon the kind of oil, and hope
to investigate the matter further by a more appropriate method*.
On several occasions the effect of large additions of oil was tried. The
limit did not appear to be very definite ; for a second and even a third drop
gave a sensible indication. The results were June 30, 38'8; July 1, 40'3;
July 7, 41-0 ; July 8, 41'7 ; July 26, 38'9 ; mean 401. They relate to olive-
oil ; and it is possible that the largeness of the quantity required to approach
the limit depends upon the heterogeneous character of the substance.
Two observations were made of the effect of additions of oleate of soda to
distilled water. When the limit seemed nearly attained final readings were
taken with the results : July 8, 25'3 ; July 11, 24'6 ; mean 25'0. It will be
seen that the tension is lowered very much further by soap than by oil.
The principal results of the present experiments may be thus summarized.
The tension of a water surface, reckoned in c.G.s. measure, is in the various
Clean
74-0
100
Greasy to the point where the camphor motions nearly cease
53-0
72
Saturated with olive-oil
41 '0
54
Saturated with oleate of soda
25-0
34
In the last column the tensions are exhibited as fractions of that of a
clean surface.
POSTSCRIPT, Sept. 19.
It appeared probable that the tension of otherwise pure water saturated
with camphor would be the same as that of greasy water upon which
camphor fragments were just dead ; and before the above paper was written
I had already attempted to examine this point. The experiment, however,
did not succeed. The camphorated water had decidedly too much tension
(wave-number 14'7 instead of 15'5), but on the other hand the liquid was
clearly not saturated, inasmuch as fresh camphor scrapings were lively upon
an expanded surface. I have recently returned to the subject with water
which has stood in contact with excess of camphor for more than a month.
Sept. 15. Fresh clean water. Expanded 137, 13'7. Motes still. Olive-
oil added, 15'2, 15'2 ; camphor fragments moderately active. More oil, 15'3,
15'3; camphor nearly dead. More oil, 15'5, 15'5 ; camphor dead. Fragments
of camphor and motes quite still.
* [Phil. Mag. xxxrn. p. 366, 1892.]
1890] SURFACES, CLEAN AND CONTAMINATED. 395
The saturated solution of camphor was now substituted. Surface ex-
panded ; 15'5, 15*5. Expanded, 15'5. This number could not be reduced by
any number of expansions of the surface.
It was observed that the surface was usually in motion, as evidenced by
an irregular drift of motes and camphor fragments. The latter had no
individual motion, all neighbouring particles moving together. The effect
is probably due to local evaporation of camphor and accompanying increase of
tension. Associated with this was a fluctuation backwards and forwards of
the number of waves, such as was never observed with pure, or simply greasy,
water.
We are thus justified in the conclusion that saturated solution of camphor
has the same tension as is found for greasy water when camphor fragments
are just dead. When the saturated solution was diluted with about an equal
volume of water, the wave-number was reduced to 14'7. In these experiments
the distance between the points was 9'20 cm., and the frequency was 42'12,
so that the observations are directly comparable with those in the example
calculated at length.
The comparison of tensions for clean and camphorated water may also be
effected by the method of capillary heights. Some observations by Mr Gordon
gave the following : —
Clean water 7'94, 7'91, 7'92
Water changed 7'92, 7'90, 7'90
Saturated camphor 5'63, 5'68, 5'65
Clean water 7*97, 7'90, 7*92
Water changed 7'94, 7'96, 7'93
Saturated camphor 5'62, 5'63, 5'66
Thus, as a mean, capillary height for clean water is 7 '93 cm., and for water
saturated with camphor 5'64 cm. The ratio of these is '71.
Observations by myself upon the same tube, but read in a somewhat
different manner, gave
Clean water 8*04, 8'03, 8'04, 8'05.
Water changed 8'02, 8'02.
Camphorated water... 577, 5'80, 5'79, 5'80, 5'80, 5'83.
As means we may take 8'03 cm. and 5'80 cm., giving for the ratio '71, as
before.
The ratio of tensions thus found agrees remarkably well with that deduced
from the observations upon ripples, viz. '72. It will be remembered that the
latter might be expected to be somewhat higher, as corresponding with a
condition of things where camphor fragments were nearly, but not quite, dead.
396 ON THE TENSION OF WATER SURFACES. [175
October 8. — I take this opportunity of recording that a film of grease,
insufficient to check the motion of camphor fragments, exercises a marked
influence upon the reflexion of light from the surface of water in the neigh-
bourhood of the polarizing angle. In the case of a clean surface and at the
Brewsterian angle, the reflexion of light polarized perpendicularly to the
plane of incidence appears to vanish, in accordance with the formula of
Fresnel.
[1901. This subject is further treated in Phil. Mag. xxxm. p. 1, 1892 ;
Vol. ill. of present collection, Art. 185 below.]
176.
ON THE THEORY OF SURFACE FORCES.
[Philosophical Magazine, xxx. pp. 285—298, 456—475, 1890.]
SINCE the time of Young the tendency of a liquid surface to contract has
always been attributed to the mutual attraction of the parts of the liquid,
acting through a very small range, — to the same forces in fact as those by
which the cohesion of liquids and solids is to be explained. It is sometimes
asserted that Laplace was the first to look at the matter from this point of
view, and that Young contented himself with calculations of the consequences
of superficial tension. Such an opinion is entirely mistaken, although the
authority of Laplace himself may be quoted in its favour*. In the in-
troduction to his first paper f, which preceded the work of Laplace, Young
writes : — " It will perhaps be more agreeable to the experimental philosopher,
although less consistent with the strict course of logical argument, to proceed
in the first place to the comparison of this theory [of superficial tension] with
the phenomena, and to inquire afterwards for its foundation in the ultimate
properties of matter." This he attempts to do in Section VI., which is
headed Physical Foundation of the Law of Superficial Cohesion. The
argument is certainly somewhat obscure; but as to the character of the
" physical foundation " there can be no doubt. " We may suppose the
particles of liquids, and probably those of solids also, to possess that power
of repulsion, which has been demonstrably shown by Newton to exist in
aeriform fluids, and which varies in the inverse ratio of the distance of the
particles from each other. In air and vapours this force appears to act
uncontrolled; but in liquids it is overcome by a cohesive force, while the
particles still retain a power of moving freely in all directions.... It is simplest
* Mec. Cel. Supplement au Xe livre, 1805 :— " Mais il n'a pas tente, comme Segner, de driver
ces hypotheses, de la loi de 1'attraction des molecules, decroissante avec une extreme rapidite ; ce
qui e*tait indispensable pour les realiser."
t " On the Cohesion of Fluids," Phil. Trans. 1805.
398 ON THE THEORY OF SURFACE FORCES. [176
to suppose the force of cohesion nearly or perfectly constant in its magnitude,
throughout the minute distance to which it extends, and owing its apparent
diversity to the contrary action of the repulsive force which varies with the
distance."
Although nearly a century has elapsed, we are still far from a satisfactory
theory of these reactions. We know now that the pressure of gases cannot
be explained by a repulsive force varying inversely as the distance, but that
we must appeal to the impacts of colliding molecules*. There is every
reason to suppose that the molecular movements play an important part in
liquids also; and if we leave them out of account, we can only excuse
ourselves on the ground of the difficulty of the subject, and with full
recognition that a theory so founded is probably only a first approximation
to the truth. On the other hand, the progress of science has tended to
confirm the views of Young and Laplace as to the existence of a powerful
attraction operative at short distances. Even in the theory of gases it is
necessary, as Van der Waals has shown, to appeal to such a force in order to
explain their condensation under increasing pressure in excess of that
indicated by Boyle's law, and explicable by impacts. Again, it would appear
that it is in order to overcome this attraction that so much heat is required
in the evaporation of liquids.
If we take a statical view of the matter, and ignore the molecular
movements!, we must introduce a repulsive force to compensate the
attraction. Upon this point there has been a good deal of confusion, of
which even Poisson cannot be acquitted. And yet the case seems simple
enough. For consider the equilibrium of a spherical mass of mutually
attracting matter, free from external force, and conceive it divided by an
ideal plane into hemispheres. Since the hemispheres are at rest, their total
action upon one another must be zero, that is, no force is transmitted across
the interface. If there be attraction operative across the interface, it must
be precisely compensated by repulsion. This view of the matter was from
the first familiar to Young, and he afterwards gave calculations, which we
shall presently notice, dependent upon the hypothesis that there is a constant
attractive force operative over a limited range and balanced by a repulsive
force of suitable intensity operative over a different range. In Laplace's
theory, upon the other hand, no mention is made of repulsive forces, and it
would appear at first as if the attractive forces were left to perform the
impossible feat of balancing themselves. But in this theory there is in-
troduced a pressure which is really the representative of the repulsive forces.
* The argument is clearly set forth in Maxwell's lecture " On the Dynamical Evidence of the
Molecular Constitution of Bodies " (Nature, Vol. xi. p. 357, 1875. [Maxwell's Scientific Papers,
Vol. n. p. 418]).
t Compare Worthington, " On Surface Forces in Fluids," Phil. Mag. xvui. p. 334 (1884).
1890] Olf THE THEORY OF SURFACE FORCES. 399
It may be objected that if the attraction and repulsion must be supposed
to balance one another across any ideal plane of separation, there can be
no sense, or advantage, in admitting the existence of either. This would
certainly be true if the origin and law of action of the forces were similar,
but such is not supposed to be the case. The inconclusiveness of the
objection is readily illustrated. Consider the case of the earth, conceived
to be at rest. The two halves into which it may be divided by an ideal
plane do not upon the whole act upon one another: otherwise there could
not be equilibrium. Nevertheless no one hesitates to say that the two halves
attract one another under the law of gravitation. The force of the objection
is sometimes directed against the pressure, denoted by K, which Laplace
conceives to prevail in the interior of liquids and solids. How. it is asked,
can there be a pressure, if the whole force vanishes ? The best answer to
this question may be found in asking another — Is there a pressure in the
interior of the earth ?
It must no doubt be admitted that in availing ourselves of the conception
of pressure we are stopping short of a complete explanation. The mechanism
of the pressure is one of the things that we should like to understand. But
Laplace's theory, while ignoring the movements and even the existence of
molecules, cannot profess to be complete ; and there seems to be no incon-
sistency in the conception of a continuous, incompressible liquid, whose parts
attract one another, but are prevented from undergoing condensation by
forces of infinitely small range, into the nature of which we do not further
inquire. All that we need to take into account is then covered by the
ordinary idea of pressure. However imperfect a theory developed on these
lines may be, and indeed must be, it presents to the mind a good picture of
capillary phenomena, and, as it probably contains nothing not needed for
the further development of the subject, labour spent upon it can hardly be
thrown away.
Upon this view the pressure due to the attraction measures the cohesive
force of the substance, that is the tension which must be applied in order to
cause rupture. It is the quantity which Laplace denoted by K, and which
is often called the molecular pressure. Inasmuch as Laplace's theory is not
a molecular theory at all, this name does not seem very appropriate. Intrinsic
pressure is perhaps a better term, and will be employed here. The simplest
method of estimating the intrinsic pressure is by the force required to break
solids. As to liquids, it is often supposed that the smallest force is adequate
to tear them asunder. If this were true, the theory of capillarity now under
consideration would be upset from its foundations, but the fact is quite
otherwise. Berthelot* found that water could sustain a tension of about
* Am. de Ckimit, zxx. p. 232 (1830). See also Worthington, Brit. Aaoe. Report, 1888,
p. 583.
400 ON THE THEORY OF SURFACE FORCES. [176
50 atmospheres applied directly, and the well-known phenomenon of retarded
ebullition points in the same direction. For if the cohesive forces which tend
to close up a small cavity in the interior of a superheated liquid were less
powerful than the steam-pressure, the cavity must expand, that is the liquid
must boil. By supposing the cavity infinitely small, we see that ebullition
must necessarily set in as soon as the steam* pressure exceeds that intrinsic
to the liquid. The same method may be applied to form a conception of the
intrinsic pressure of a liquid which is not superheated. The walls of a
moderately small cavity certainly tend to collapse with a force measured by
the constant surface-tension of the liquid. The pressure in the cavity is
at first proportional to the surface-tension and to the curvature of the walls.
If this law held without limit, the consideration of an infinitely small cavity
shows that the intrinsic pressure would be infinite in all liquids. Of course
the law really changes when the dimensions of the cavity are of the same
order as the range of the attractive forces, and the pressure in the cavity
approaches a limit, which is the intrinsic pressure of the liquid. In this way
we are forced to admit the reality of the pressure by the consideration of
experimental facts which cannot be disputed.
The first estimate of the intrinsic pressure of water is doubtless that of
Young. It is 23,000 atmospheres, and agrees extraordinarily well with
modern numbers. I propose to return to this estimate, and to the remarkable
argument which Young founded upon it.
The first great advance upon the theory of Young and Laplace was the
establishment by Gauss of the principle of surface-energy. He observed that
the existence of attractive forces of the kind supposed by his predecessors
leads of necessity to a term in the expression of the potential energy
proportional to the surface of the liquid, so that a liquid surface tends
always to contract, or, what means precisely the same thing, exercises a
tension. The argument has been put into a more general form by Boltzmamrf .
It is clear that all molecules in the interior of the liquid are in the same
condition. Within the superficial layer, considered to be of finite but very
small thickness, the condition of all molecules is the same which lie at the
same very small distance from the surface. If the liquid be deformed without
change in the total area of the surface, the potential energy necessarily
remains unaltered ; but if there be a change of area the variation of potential
energy must be proportional to such change.
A mass of liquid, left to the sole action of cohesive forces, assumes a
spherical figure. We may usefully interpret this as a tendency of the surface
* If there be any more volatile impurity (e.g., dissolved gas) ebullition must occur much
earlier.
t Pogg. Ann. CXLI. p. 582 (1870). See also Maxwell's Theory of Heat, 1870 ; and article
"Capillarity," Enc. Brit. [Maxwell's Scientific Papers, Vol. n. p. 541.]
1890] ON THE THEORY OF SURFACE FORCES. 401
to contract ; but it is important not to lose sight of the idea that the spherical
form is the result of the endeavour of the parts to get as near to one another
as is possible*. A drop is spherical under capillary forces for the same
reason that a large gravitating mass of (non-rotating) liquid is spherical.
In the following sketch of Laplace's theory we will commence in the
manner adopted by Max well f. If / be the distance between two particles
m, m', the cohesive attraction between them is denoted in Laplace's notation
by mm'<f>(f), where <f>(f) is a function of f which is insensible for all
sensible values of f, but which becomes sensible and even enormously great,
when / is exceedingly small.
" If we next introduce a new function of f and write
.(1)
then mm'H(f) will represent (1) the work done by the attractive force on
the particle m, while it is brought from an infinite distance from m' to the
distance f from m' ; or (2) the attraction of a particle m on a narrow straight
rod resolved in the direction of the length of the rod, one extremity of the
rod being at a distance f from m, and the other at an infinite distance, the
mass of unit of length of the rod being m'. The function II(/) is also
insensible for sensible values of f, but for insensible values off it may become
sensible and even very great."
" If we next write
/"n </)/<*/*=*(*). (2)
then 2imwr^r(z) will represent (1) the work done by the attractive force
while a particle m is brought from an infinite distance to a distance z from
an infinitely thin stratum of the substance whose mass per unit of area is a ;
(2) the attraction of a particle m placed at a distance z from the plane surface
of an infinite solid whose density is <r."
The intrinsic pressure can now be found immediately by calculating the
mutual attraction of the parts of a large mass which lie on opposite sides of
an imaginary plane interface. If the density be a, the attraction between
the whole of one side and a layer upon the other, distant z from the plane
and of thickness dz, is 27nr!^-(z)<iz, reckoned per unit of area. The expression
for the intrinsic pressure is thus simply
(3)
• See Sir W. Thomson's lecture on "Capillary Attraction" (Proe. Roy. Iiut. 1886), reprinted
in Papular Lecture* and Addreue*.
t Enc. Brit., " Capillarity." [Maxwell's Scientific Papert, Vol. n. p. 541.]
K. ill. 26
402 ON THE THEORY OF SURFACE FORCES. [176
In Laplace's investigation <r is supposed to be unity. We may call the value
which (3) then assumes K0, so that
i;-$*P*f*)& ............................... (4)
Jo
The expression for the superficial tension is most readily found with the
aid of the idea of superficial energy, introduced into the subject by Gauss.
Since the tension is constant, the work that must be done to extend the
surface by one unit of area measures the tension, and the work required for
the generation of any surface is the product of the tension and the area.
From this consideration we may derive Laplace's expression, as has been
done by Dupre* and Thomson f. For imagine a small cavity to be formed
in the interior of the mass and to be gradually expanded in such a shape that
the walls consist almost entirely of two parallel planes. The distance
between the planes is supposed to be very small compared with their
ultimate diameters, but at the same time large enough to exceed the range
of the attractive forces. The work required to produce this crevasse is twice
the product of the tension and the area of one of the faces. If we now
suppose the crevasse produced by direct separation of its walls, the work
necessary must be the same as before, the initial and final configurations
being identical ; and we recognize that the tension may be measured by half
the work that must be done per unit of area against the mutual attraction
in order to separate the two portions which lie upon opposite sides of an ideal
plane to a distance from one another which is outside the range of the forces.
It only remains to calculate this work.
If <rlt o-2 represent the densities of the two infinite solids, their mutual
attraction at distance z is per unit of area
-di; .............................. (5)
or 27ro-1o-2#(.2), if we write
J*f(*),il«*(er) ............................... (6)
The work required to produce the separation in question is thus
27ro-1o-2Ja00<»^; .............................. (7)
and for the tension of a liquid of density a- we have
T=7r<r°-r8(z)dz... ...(8)
Jo
The form of this expression may be modified by integration by parts. For
Theorie Mecanique de la Chaleur (Paris, 1869).
t "Capillary Attraction," Proc. Boy. Inst., Jan. 1886. Reprinted, Popular Lectures and
Addresses, 1889.
1890] OX THE THEORY OF SURFACE FORCES. 403
Since 0(0) is finite, proportional to K, the integrated term vanishes at both
limits, and we have simply
•(9)
and
(10)
In Laplace's notation the second member of (9), multiplied by 2ir, is repre-
sented by H.
As Laplace has shown, the values for K and T inav also be expressed in
terms of the function <f>, with which we started. Integrating by parts, we get
by means of (1) and (2),
(z) +
In all cases to which it is necessary to have regard the integrated terms
vanish at both limits, and we may write
J»
to
so that
(12)
A few examples of these formulae will promote an intelligent comprehen-
sion of the subject. One of the simplest suppositions open to us is that
*(/) = *-# .............................. (13)
From this we obtain
(14)
(15)
The range of the attractive force is mathematically infinite, but practically of
the order ft~\ and we see that T is of higher order in this small quantity
than K. That K is in all cases of the fourth order and T of the fifth order in
the range of the forces is obvious from (12) without integration.
An apparently simple example would be to suppose <f>(z) = z*. From (1),
(2), (4) we get
v
The intrinsic pressure will thus be infinite whatever n may be. If n + 4
be positive, the attraction of infinitely distant parts contributes to the result ;
while if n + 4 be negative, the parts in immediate contiguity act with infinite
26 _ 2
404 ON THE THEORY OF SURFACE FORCES. [176
power. For the transition case, discussed by Sutherland*, of n + 4 = 0, K> is
also infinite. It seems therefore that nothing satisfactory can be arrived at
under this head.
As a third example we will take the law proposed by Young, viz.
£(*) = ! from 2 =
<f> (z) = 0 from
and corresponding therewith,
Il(z) = a — z from z = 0 to z = a, .
II* = 0 from z = a to * =
2 = 0 to z = a, \ .^
z = atoz=ao; ) '
from 2 = 0 to z = a, V ............ (19)
^ (z) = 0 from z = a to z = <x>
Equations (12) now give
The numerical results differ from those of Young-f, who finds that " the con-
tractile farce is one-third of the whole cohesive force of a stratum of particles,
equal in thickness to the interval to which the primitive equable cohesion extends,"
viz. T=^aK; whereas according to the above calculation T = ^aK. The
discrepancy seems to depend upon Young having treated the attractive force
as operative in one direction only.
In his Elementary Illustrations of the Celestial Mechanics of Laplace £,
Young expresses views not in all respects consistent with those of his earlier
papers. In order to balance the attractive force he introduces a repulsive
force, following the same law as the attractive except as to the magnitude of
the range. The attraction is supposed to be of constant intensity C over a
range c, while the repulsion is of intensity R, and is operative over a range r.
The calculation above given is still applicable, and we find that
(22)
In these equations, however, we are to treat K as vanishing, the specification
of the forces operative across a plane being supposed to be complete. Hence,
as Young finds, we must take
c*C=r*R, ................................. (23)
and accordingly
(24)
At this point I am not able to follow Young's argument, for he asserts (p. 490)
that " the existence of such a cohesive tension proves that the mean sphere of
* Phil. Mag. DIT. p. 113 (1887). t Erne. Brit.; Collected Wort*, Vol. i. p. 461.
J 1821. Collected Works, Vol. i. p. 485.
1890] ON THE THEORY OF SURFACE FORCES. 405
action of the repulsive force is more extended than that of the cohesive : a
conclusion which, though contrary to the tendency of some other modes of
viewing the subject, shows the absolute insufficiency of all theories built upon
the examination of one kind of corpuscular force alone." According to (24)
we should infer, on the contrary, that if superficial tension is to be explained
in this way, we must suppose that or.
My own impression is that we do not gain anything by this attempt to
advance beyond the position of Laplace. So long as we are content to treat
fluids as incompressible, there is no objection to the conception of intrinsic
pressure. The repulsive forces which constitute the machinery of this pressure
are probably intimately associated with actual compression, and cannot advan-
tageously be treated without enlarging the foundations of the theory. Indeed
it seems that the view of the subject represented by (23), (24), with c greater
than r, cannot consistently be maintained. For consider the equilibrium of a
layer of liquid at a free surface A of thickness AB equal to r. If the void
space beyond A were filled up with liquid, the attractions and repulsions
across B would balance one another ; and since the action of the additional
liquid upon the parts below B is wholly attractive, it is clear that in the
actual state of things there is a finite repulsive action across B, and a
consequent failure of equilibrium.
I now propose to exhibit another method of calculation, which not only
leads more directly to the results of Laplace, but allows us to make a not
unimportant extension of the formulae to meet the case where the radius of a
spherical cavity is neither very large nor very small in comparison with the
range of the forces.
The density of the fluid being taken as unity, let V be the potential of the
attraction, so that
U(f)dxdydz, ........................ (25)
jjj
/denoting the distance of the element of the fluid dxdydz from the point at
which the potential is to be reckoned. The hydrostatic equation of pressure
is then simply dp = dV; or, if A and B be any two points,
P*-PA=VB-VA ............................ (26)
Suppose, for example, that A is in the interior, and B upon a plane surface
of the liquid. The potential at B is then exactly one half of that at A, or
VB = VA\ so that
'0 J 0
Now pA —ps is the intrinsic pressure K0', and thus
, 27T
as before.
K0 =
406
ON THE THEORY OF SURFACE FORCES.
[176
Again, let us suppose that the fluid is bounded by concentric spherical
surfaces, the interior one of radius r being either large or small, but the
exterior one so large that its curvature may be neglected. We may suppose
that there is no external pressure, and that the tendency of the cavity to
collapse is balanced by contained gas. Our object is to estimate the necessary
internal pressure.
Fig. 1.
In the figure BDCE represents the cavity, and the pressure required is
the same as that of the fluid at such a point as B. [A is supposed to lie
upon the external surface.] Since j^ = 0, ps= Vjj — VA. Now VA is equal
to that part of VB which is due to the infinite mass lying below the plane
BF. Accordingly the pressure required (PB) is the potential at B due to the
fluid which lies above the plane BF. Thus
where the integrations are to be extended through the region above the
plane BF which is external to the sphere BDCE. On the introduction of
polar coordinates the integral divides itself into two parts. In the first from
f = 0 to f= 2r the spherical shells (e.g. DH) are incomplete hemispheres,
while in the second part from /= 2r to/= oo the whole hemisphere (e.g.
IGF) is operative. The spherical area DH, divided by f'2,
= 2vr
The area IGF= 27r/2.
Thus, dropping the suffix B, we get the unexpectedly simple expression
. ............... (27)
If 2r exceed the range of the force, the second integral vanishes and the first
may be supposed to extend to infinity. Accordingly
(28)
1890] ON THE THEORY OF SURFACE FORCES. 407
in accordance with the value (12) already given for Tg. We see then that,
if the curvature be not too great, the pressure in the cavity can be calculated
as if it were due to a constant tension tending to contract the surface. In
the other extreme case where r tends to vanish, we have ultimately
P =
In these extreme cases the results are of course well known ; but we may
apply (27) to calculate the pressure in the cavity when its diameter is of the
order of the range. To illustrate this we may take a case already suggested,
in which $ (/) = g-fS, !!(/) = /6M e~V. Using these, we obtain on reduction,
(29)
From (29) we may fall back upon particular cases already considered. Thus,
if r be very great,
and if r be very small, p = 4nrft~4, in agreement with (15).
In a recent memoir* Fuchs investigates a second approximation to the
tension of curved surfaces, according to which the pressure in a cavity would
consist of two terms ; the first (as usual) directly as the curvature, the second
subtractive, and proportional to the cube of the curvature. This conclusion
does not appear to harmonize with (27), (29), which moreover claim to be
exact expressions. It may be remarked that when the tension depends upon
the curvature, it can no longer be identified with the work required to
generate a unit surface. Indeed the conception of surface-tension appears to
be appropriate only when the range is negligible in comparison with the
radius of curvature.
The work required to generate a spherical cavity of radius r is of course
readily found in any particular case. It is expressed by the integral
(30)
As. a second example we may consider Young's supposition, viz. that the
force is unity from 0 to a, and then altogether ceases. In this case by (18),
II (/) absolutely vanishes, if/> a ; so that if the diameter of the cavity at all
exceed a, the internal pressure is given rigorously by
* Wien. Ber. Bd. xcvin. Abth. n. a, Mai
408 ON THE THEORY OF SURFACE FORCES. [176
When, on the other hand, 2r < a, we have
p = I f\a -/)/• df+ 27T J V -/)/' df
•• .............................. <32)
coinciding with (31) when 2r = a. If r = 0, we fall back upon K0 = ?ra4/6.
We will now calculate by (30) the work required to form a cavity of radius
equal to \a. We have
The work that would be necessary to form the same cavity, supposing the
pressure to follow the law (31) applicable when 2r > a, is
/•*« 2 Tra5 . 7r2a7
I - .-jf- . 4an* dr = -rf- .
Jo r 40 40
The work required to generate a cavity for which 2r > a is therefore less than
if the ultimate law prevailed throughout by the amount
lO 18 35~ 47977
We may apply the same formulae to compare the pressures at the centre
and upon the surface of a spherical mass of fluid, surrounded by vacuum. If
the radius be r, we have at the centre
F=47rJ>II(/)d/,
and at the surface
so that the excess of pressure at the centre is
47r (>n(/)d/-27r[Vn (/)<*/+- P>n(/)#: ...... (34)
Jo Jo r J o
If r exceed the range of the forces, (34) becomes
, ........... (35)
o i J o
as was to be expected. As the curvature increases from zero, there is at first
a rise of pressure. A maximum occurs when r has a particular value, of the
order of the range. Afterwards a diminution sets in, and the pressure
approaches zero, as r decreases without limit.
If the surface of fluid, not acted on by external force, be of variable
curvature, it cannot remain in equilibrium. For example, at the pole of an
oblate ellipsoid of revolution the potential will be greater than at the equator,
1890] ON THE THEOKY OF SUBFACE FORCES. 409
so that in order to maintain equilibrium an external polar pressure would be
needed. An extreme case is presented by a rectangular mass, in which the
potential at an edge is only one half, and at a corner only one [quarter], of that
general over a face.
When the surface is other than spherical, we cannot obtain so simple a
general expression as (34) to represent the excess of internal over superficial
pressure ; but an approximate expression analogous to (35) is readily found.
The potential at a point upon the surface of a convex mass differs from
that proper to a plane surface by the potential of "the meniscus included
between the surface and its tangent plane. The equation of the surface
referred to the normal and principal tangents is approximately
jRj, Rj being the radii of curvature. The potential, at the origin, of the
meniscus is thus
where /"* = a3 + y* ; and
Accordingly
The excess of internal pressure above that at the superficial pjint in question
is thus
in agreement with (35).
For a cylindrical surface of radius r, we have simply
K+T/r (37)
Returning to the case of a plane surface, we know that upon it V = K,
and that in the interior V=2K. At a point P
(Fig. 2) just within the surface, the value of V Fig- 2-
cannot be expressed in terms of the principal quan-
tities K and T, but will depend further upon the
precise form of the function II. We can, however,
express the value of / Vdz, where z is measured in-
wards along the normal, and the integration extends A
over the whole of the superficial layer where V
differs from 2K.
It is not difficult to recognize that this integral
must be related to T. For if Q be a point upon the
normal equidistant with P from the surface AB, the potential at Q due to
410
ON THE THEORY OF SURFACE FORCES.
[170
fluid below AB is the same as the potential at P due to imaginary fluid
above AB. To each of these add the potential of the lower fluid at P. Then
the sum of the potentials at P and Q due to the lower fluid is equal to the
potential at P due to both fluids, that is to the constant 2K. The deficiency
of potential at a point P near the plane surface of a fluid, as compared with
the potential in the interior, is thus the same as the potential at an external
point Q, equidistant from the surface. Now it is evident that J VQ dz inte-
grated upwards along the normal represents the work per unit of area that
would be required to separate a continuous fluid of unit density along the
plane AB and to remove the parts beyond the sphere of influence, that is,
according to the principle of Dupre, 2T. We conclude that the deficiency in
jVpdz, integrated along the normal inwards, is also 2^T; or that
Fj.cfc~2jr.s-221;
z being large enough to include the whole of the superficial stratum,
pressure p at any point P is given by p = VP — K, so that
(38)
The
.(39)
We may thus regard 2T as measuring the total deficiency of pressure in the
superficial stratum.
The argument here employed is of course perfectly satisfactory ; but it is
also instructive to investigate the question directly, without the aid of the
idea of superficial tension, or energy, and this is easily done.
In polar coordinates the potential at any point P is expressed by
vp = 27T jjn (/)/2 sin 0 de df,
the integrations extending over the whole space
ACB (Fig. 3). If the distance EP, that is z,
exceed the range of the forces, every sphere of
radius / under consideration, is complete, and
VP = %K. But in the integration with respect
to z incomplete spheres have to be considered,
such as that shown in the figure. The value of
the potential, corresponding to a given infinitely
small range of/, is then proportional to
sin 6 dO = 1 + cos 0 = 1 + z/f.
If now we effect first the integration with respect to z, we have as the
element of the final integral,
1890] ON THE THEORY OF SURFACE FORCES. 411
and thus, on the whole,
= z.2K-2T, as before.
An application of this result to a calculation of the pressure operative
between the two halves of an isolated sphere will lead us to another inter-
pretation of T. The pressure in the interior is K+ 2T[r, r being the radius ;
and this may be regarded as prevailing over the whole of the diametral
dividing plane, subject to a correction for the circumferential parts which are
near the surface of the fluid. If the radius r increase without limit, the
correction will be the same per unit of length as that investigated for a plane
surface. The whole pressure between the two infinite hemispheres is thus
irr»(J5T+2T/r)-2r.2wr, or -m*K - T . 2irr (40)
This expression measures equally the attraction between the two hemi-
spheres, which the pressure is evoked to balance. If the fluid on one side of
the diametral plane extended to infinity, the attraction upon the other
hemisphere, supposed to retain its radius r, would
be "irr*K simply : so that the second term T . 2irr
may be considered to represent the deficiency of
attraction due to the absence of the fluid external
to one hemisphere. Regarding the matter in two
dimensions, we recognize T as the attraction per ^~
unit of length perpendicular to the plane of the
paper of the fluid occupying (say) the first quadrant
XOY (Fig. 4) upon the fluid in the third quadrant
X'OY', the attraction being resolved in one or other
of the directions OX, OF. In its actual direction, bisecting the angle XOY
the attraction will be of course V2 . T.
Fig. 4.
Y
Fig. 5.
We will now suppose that the sphere is divided by a plane AB (Fig. 5),
which is not diametral, but such that the angle BAO — 0\
AO = r, AB = 2p. In the interior of the mass, and gene-
rally along the section AB, V=2K. On the surface of
the sphere, and therefore along the circumference of AB,
V=K — 2T/r. When V was integrated along the normal,
from a plane surface inwards, the deficiency was found to
be 221 In the present application the integration is along
the oblique line AB, and the deficiency will be 2Tsec0.
Hence when r and p increase without limit, we may
take as the whole pressure over the area AB
Trp* (K + 2T/r) - 2-irp . 2Tsec 0 = -irfK - 2irp (2T sec 0 - Tcoa 0).
412 ON THE THEORY OF SURFACE FORCES. [176
The deficiency of attraction perpendicular to AB is thus for each unit of
perimeter
2Tsec0-Tcos0, (41)
and this we may think of as applicable in two dimensions (Fig. 6) to each
unit of length. When 6 = 0, (41) reduces to T.
The term T cos 6 in the expression for the total pressure appears to
have its origin in the curvature of the surface, only not
disappearing when the curvature vanishes, in consequence
of the simultaneous increase without limit of the area over
which the pressure is reckoned. If we consider only a
distance AB, which, though infinite in comparison with the
range of the attraction, is infinitely small in comparison
with the radius of curvature, T cos 0 will disappear from
the expression for the pressure, though it must necessarily
remain in the expression for the attraction. The pressure acting across
a section AB proceeding inwards from a plane surface AE of a fluid is
thus inadequate to balance the attraction of the two parts. It must be
aided by an external force perpendicular to AB of magnitude T cos 6 ; and
since the imaginary section AB may be made at any angle, we see that the
force must be T and must act along AE.
An important class of capillary phenomena are concerned with the
spreading of one liquid upon the surface of another, a subject investigated
experimentally by Marangoni, Van der Mensbrugghe, Quincke, and others.
The explanation is readily given in terms of surface-tension ; and it is
sometimes supposed that these phenomena demonstrate in a special manner
the reality of surface-tension, and even that they are incapable of explanation
upon Laplace's theory, which dealt in the first instance with the capillary
pressures due to curvature of surfaces*.
In considering this subject, we have first to express the dependence of the
tension at the interface of two bodies in terms of the forces exercised by the
bodies upon themselves and upon one another, and to effect this we cannot
do better than follow the method of Dupre. If T12 denote the interfacial
tension, the energy corresponding to unit of area of the interface is also
Ti2, as we see by considering the introduction (through a fine tube) of one
body into the interior of the other. A comparison with another method of
generating the interface, similar to that previously employed when but one
body was in question, will now allow us to evaluate T12.
The work required to cleave asunder the parts of the first fluid which lie
on the two sides of an ideal plane passing through the interior, is per unit
* Van der Mensbrugghe, " Essai sur la Theorie Mecanique de la Tension Snperficielle, Ac."
Bulletins de VAcad. roy. de Belgique, 3me serie, t. ix. No. 5, 1885, p. 12. Worthington, Phil. Mug.
Oct. 1884, p. 364.
1890] OX THE THEORY OF SURFACE FORCES. 413
of area 27*,, and the free surface produced is two units in area. So for the
second fluid the corresponding work is 2Tj. This having been effected, let
us now suppose that each of the units of area of free surface of fluid (1) is
allowed to approach normally a unit of area of (2) until contact is established.
In this process work is gained which we mar denote by 42"B, 2T'B for each
pair. On the whole, then, the work expended in producing two units of
interface is 27, + 2TS — 42",., and this, as we have seen, may be equated to
2Te. Hence
TH=ri+T2-2rH ............................ (42)
If the two bodies are similar, Z\ = T*=T^\ and TK = 0, as it should da
Laplace does not treat systematically the question of interfaced tension,
but he gives incidentally in terms of his quantity H a relation analogous
to (42).
If 2TfK > Tx + T,, TK would be negative, so that the interface would of
itself tend to increase. In this case the fluids must mix. Conversely, if two
fluids mix, it would seem that T'K must exceed the mean of Tt and T, :
otherwise work would have to be expended to effect a close alternate
stratification of the two bodies, such as we may suppose to constitute a first
step in the process of mixture*.
The value of T'K has already been calculated (7). We may write
.(43)
and in general the functions 0, or £, must be regarded as capable of assuming
different forms. Under these circumstances there is no limitation upon the
values of the interfacial tensions for three fluids, which we may denote by
2*18, 2»» T*. If the three fluids can remain in contact with one another,
the sum of any two of the quantities must
exceed the third, and by Neumann's rule the
directions of the interfaces at the common
edge must be parallel to the sides of a 3
triangle, taken proportional to TH, Ta, Tn. If ^ i
the above-mentioned condition be not satis-
fied, the triangle is imaginary, and the three
fluids cannot rest in contact, the two weaker
tensions, even if acting in full concert, being incapable of balancing the
strongest. For instance, if T*>Tn+T*, the second fluid spreads itself
indefinitely upon the interface of the first and third fluids.
The experimenters who have dealt with this question, Marangoni, Van
der Mensbrugghe, Qnincke, have all arrived at results inconsistent with the
reality of Neumann's triangle. Thus Marangoni saysf : — " Die gemeinschanV
* Duprf, loc. at. p. 372. Thomson, Aytdrr L^tum, p. 53.
+ PO&. Am*, crun. p. 348, 1871 (1865). It wms toh«r«i»nitly Aon bj Qnneke that
mercury is not icaDr an exception.
414 ON THE THEORY OF SURFACE FORCES. [176
liche Oberflache zweier Fliissigkeiten hat eine geringere Oberflachenspannung
als die Differenz der Oberflachenspannung der Fliissigkeiten selbst (mit
Ausnahme des Quecksilbers)." Three pure bodies (of which one may be air)
cannot accordingly remain in contact. If a drop of oil stands in lenticular
form upon a surface of water, it is because the water-surface is already
contaminated with a greasy film.
On the theoretical side the question is open until we introduce some
limitation upon the generality of the functions. By far the simplest
supposition open to us is that the functions are the same in all cases, the
attractions differing merely by coefficients analogous to densities in the
theory of gravitation. This hypothesis was suggested by Laplace, and may
conveniently be named after him. It was also tacitly adopted by Young, in
connexion with the still more special hypothesis which Young probably had
in view, namely that the force in each case was constant within a limited
range, the same in all cases, and vanished outside that range.
As an immediate consequence of this hypothesis we have from (3)
K = K0<r2, T=TQ<r'-, (44,45)
where K0, T0 are the same for all bodies.
But the most interesting results are those which Young* deduced relative
to the interfacial tensions of three bodies. By (12), (43),
r^wT.; (46)
so that by (42), (45),
T^^-rtfT. (47)
According to (47), the interfacial tension between any two bodies is
proportional to the square of the difference of their densities. The densities
0"i> 0",, °"s being in descending order of magnitude, we may write
T3l = (<r, - <r2 + <r2 - <rs)2 T0 = TK + Ta + 2 (^ - cr2) (<r, - <r3) T0 ;
so that Tai necessarily exceeds the sum of the other two interfacial tensions.
We are thus led to the important conclusion, so far as I am aware hitherto
unnoticed, that according to this hypothesis Neumann's triangle is necessarily
imaginary, that one of three fluids will always spread upon the interface of
the other two.
Another point of importance may be easily illustrated by this theory,
viz. the dependency of capillarity upon abruptness of transition. "The
reason why the capillary force should disappear when the transition between
two liquids is sufficiently gradual will now be evident. Suppose that the
transition from 0 to a is made in two equal steps, the thickness of the
intermediate layer of density £<r being large compared to the range of the
molecular forces, but small in comparison with the radius of curvature. At
* Works, Vol. i. p. 463.
1890] ON THE THEORY OF SURFACE FORCES. 415
each step the difference of capillary pressure is only one quarter of that due
to the sudden transition from 0 to <r, and thus altogether half the effect is
lost by the interposition of the layer. If there were three equal steps, the
effect would be reduced to one third, and so on. When the number of steps
is infinite, the capillary pressure disappears altogether*."
According to Laplace's hypothesis the whole energy of any number of
contiguous strata of liquids is least when they are arranged in order of density,
so that this is the disposition favoured by the attractive forces. The problem
is to make the sum of the interfacial tensions a minimum, each tension being
proportional to the square of the difference of densities of the two contiguous
liquids in question. If the order of stratification differ from that of densities,
we can show that each step of approximation to this order lowers the sum
of tensions. To this end consider the effect of the abolition of a stratum
a-n+l, contiguous to <rn and trn+2. Before the change we have
(<rn - o-n+iY + On+1 - o-n+2)2,
and afterwards (<rn — <rn+2)2. The second minus the first, or the increase in
the sum of tensions, is thus
2 (<rn - o-n+1) (<rn+1 - <rn+8).
Hence, if <rn+l be intermediate in magnitude between <rn and <rn+», the sum
of tensions is increased by the abolition of the stratum ; but, if o-n+, be not
intermediate, the sum is decreased. We see, then, that the removal of a
stratum from between neighbours where it is out of order and its introduction
between neighbours where it will be in order is doubly favourable to the
reduction of the sum of tensions ; and since by a succession of such steps we
may arrive at the order of magnitude throughout, we conclude that this is
the disposition of minimum tensions and energy.
So far the results of Laplace's hypothesis are in marked accordance with
experiment ; but if we follow it out further, discordances begin to manifest
themselves. According to (47)
vr^v^ + v^, ........................... (48)
a relation not verified by experiment. What is more, (47) shows that
according to the hypothesis TK is necessarily positive; so that, if the
preceding argument be correct, no such thing as mixture of two liquids
could ever take place.
But although this hypothesis is clearly too narrow for the facts, it may
be conveniently employed in illustration of the general theory. In extension
of (25) the potential at any point may be written
V-fffvTIWdBdyd*, ........................ (49)
and the hydrostatical equation of equilibrium is
(50)
Laplace's Theory of Capillarity," Phil. Mag. October 1883, p. 315. [Vol. n. p. 234.]
416 ON THE THEORY OF SURFACE FORCES. [176
By means of the potential we may prove, independently of the idea of
surface tension, that three fluids cannot rest in contact. Along the surface
of contact of any two fluids the potential must be constant. Otherwise,
there would be a tendency to circulation round a
circuit of which the principal parts are close and lgl '
parallel to the surface, but on opposite sides. For in L. t
the limit the variation of potential will be equal and
opposite in the two parts of the circuit, and the
resulting forces at corresponding points, being proportional also to the
densities, will not balance. It is thus necessary to equilibrium that there
be no force at any point; that is, that the potential be constant along the
whole interface.
It follows from this that if three fluids can rest in contact, the potential
must have the same constant value on all the three intersecting interfaces.
But this is clearly impossible, the potential on each being proportional to the
sum of the densities of the two contiguous fluids, as we see by considering
places sufficiently removed from the point of intersection.
According to Laplace's hypothesis, then, three fluids cannot rest in
contact ; but the case is altered if one of the bodies be solid. It is necessary,
however, that the quality of solidity attach to the body of intermediate
density. For suppose, for example (Fig. 9), that Fig 9
the body of greatest density, o-1} is solid, and that
fluids of densities <r2, <r3 touch it and one another.
It is now no longer necessary that the potential
be constant along the interfaces (1, 2), (1, 3);
but only along the interface (3, 2). The potential at a distant point of this
interface may be represented by <r2 + <r3. But at the point of intersection the
potential cannot be so low. as this, being at least equal to ^ + <r3, even if the
angle formed by the two faces of (2) be evanescent. By this and similar
reasoning it follows that the conditions of equilibrium cannot be satisfied,
unless the solid be the body of intermediate density ov,.
One case where equilibrium is possible admits of very simple treatment.
It occurs when o-2 = ^ (o^ + <r3), and the conditions are satisfied by supposing
(Fig. 10) that the fluid interface is plane and per-
pendicular to the solid wall. At a distance from 0 Fig< 10'
the potential is represented by o-j -f o-3 ; and the same
value obtains at a point P, near 0, where the sphere
of influence cuts into (2). For the areas of spherical
surface lost by (1) and (3) are equal, and are replaced
by equal areas of (2) ; so that if the above condition
between the densities holds good, the potential is
constant all the way up to 0. The sub-case, where
<r3 = 0, o-2=i0"i> was given by Clairaut.
1890] OX THE THEORY OF SURFACE FORCES. 417
If the intermediate density differ from the mean of the other two, the
problem is less simple : but the general tendency is easily recognized If, for
example. <r. > ± (<rt + <r,), it is evident that along a perpendicular interface
the potential would increase as 0 is approached. To compensate this the
interface must be inclined, so that, as 0 is approached, er, loses its importance
relatively to <rs. In this case therefore the angle between the two faces of
(1) must be acute.
The general problem was treated by Young by Kg- U-
means of superficial tensions, which must balance
when resolved parallel to the surface of the solid,
though not in the perpendicular direction. In this f — * —
way Young found at once
mt ........................... (51)
or rather, in terms of the more special h vpothesis.
fo - <rjf cos 6 + (a, - <r2)« = (<rs - <r,f. ............... (52)
From this we deduce
in agreement with what we found above for a special case. The equation
may also be written
0-,coG!40 + <r,siiiȣ0 = <rt; ........................ (54)
or if, as we may suppose without real loss of generality, <rt = 0.
a form given by Laplace. In discussing the equation (53) with <r, = 0.
Yonng* remarks: — "Supposing the attractive density of the solid to be
very small, the cosine will approach to — 1, and the angle of the liquid to
two right angles; and on the other hand, when <TJ becomes equal to <r,. the
cosine will be 1, and the angle will be evanescent, the surface of the liquid
coinciding in direction with that of the solid. If the density <ra be still
further increased, the angle cannot undergo any further alteration, and the
excess of force will only tend to spread the liquid more rapidly on the solid,
so that a thin film would always be found upon its surface, unless it were
removed by evaporation, or unless its formation were prevented by some
unknown circumstance which seems to lessen the intimate nature of the
contact of liquids with solids."
The calculation of the angle of contact upon these lines is thus exceed-
ingly simple, but I must admit that I find some difficulty in forming a
definite conception of superficial tension as applied to the interface of a solid
and a fluid. It would seem that interfacial tension can only be employed in
* Workt, YoL L p. 46*. I hare introduced an i magnificent change in the notation.
R. IIL 27
418 ON THE THEORY OF SURFACE FORCES. [176
such cases as the immediate representative of interfacial energy, as conceived
by Gauss. This principle, applied to a hypothetical displacement in which
the point of meeting travels along the wall, leads with rigour to the required
result.
In view of the difficulties which have been felt upon this subject, it seems
desirable to show that the calculation of the angle of contact can be made
without recourse to the principle of interfacial tension or energy. This
indeed was effected by Laplace himself, but his process is very circuitous.
Let 0PM be the surface of fluid (o-j) resting against a solid wall ON of
density <r2. Suppose also that o-3 = 0, and that there is no external pressure
on OM. At a point M at a sufficient distance
from 0 the curvature must be uniform (or the
potential could not be constant), and we will
suppose it to be zero. It would be a mistake,
however, to think that the surface can be straight
throughout up to 0. This we may recognize by
consideration of the potential at a point P just near enough to 0 for the
sphere of influence to cut the solid. As soon as this occurs, the potential
would begin to vary by substitution of <r., for <r1} and equilibrium would fail.
The argument does not apply if 6 = ^TT.
We may attain the object in view by considering the equilibrium of the
fluid MNO, or rather of the forces which tend to move it parallel to ON.
Of pressures we have only to consider that which acts across MN, for on OM
there is no pressure, and that on ON has no component in the direction
considered. Moreover, the solid o-2 below ON exercises no attraction parallel
to ON. Equilibrium therefore demands that the pressure operative across
MN shall balance the horizontal attraction exercised upon OMN by the
fluid o-j which lies to the right of MN. The evaluation of the attraction in
such cases has been already treated. It is represented by MN. o~i2K0, subject
to corrections for the ends at M and N. The correction for M is by (41)
oyTo (2 sec 6 — cos 6), and for N it is <rfTt. On the whole the attraction in
question is therefore
<r* {MN. K0 -
We have next to consider the pressure. In the interior of MN, we have
a-izK0; but the whole pressure MN.<r*Kn is subject to corrections for the
ends. The correction for M we have seen to be Za-j* T0 sec 6. In the
neighbourhood of N the potential, and therefore the pressure, is influenced
by the solid. If cr2 were zero, the deficiency would be 2<r12710. If o-2 were
equal to <rlt there would be no deficiency. Under the actual circumstances
the deficiency is accordingly
1890] ON THE THEORY OF SURFACE FORCES. 419
so that the expression for the total pressure operative across MN is
<r, {MN. ^K, - 2^ T0 sec 0 - 2 (o-, - <ra) T0\.
If we now equate the expressions for the pressure and the resolved attraction,
we find as before
In connexion with edge-angles it may be well here to refer to a problem,
which has been the occasion of much difference of opinion — that of the
superposition of several liquids in a capillary tube. Laplace's investigation
led him to the conclusion that the whole weight of liquid raised depends only
upon the properties of the lowest liquid. Thereupon Young* remarks: _
" This effect may be experimentally illustrated by introducing a minute
quantity of oil on the surface of the water contained in a capillary tube,
the joint elevation, instead of being increased as it ought to be according
to Mr Laplace, is very conspicuously diminished ; and it is obvious that since
the capillary powers are represented by the squares of the density of oil and
of its difference from that of water, their sum must be less than the capillary
power of water, which is proportional to the square of the sum of the separate
quantities."
But the question is not to be dismissed so summarily. That Laplace's
conclusion is sound, upon the supposition that none of the liquids ivete the
walls of the tube, may be .shown without difficulty by the method of energy.
In a hypothetical displacement the work done against gravity will balance
the work of the capillary forces. Now it is evident that the liquids, other
than the lowest, contribute nothing to the latter, since the relation of
each liquid to its neighbours and to the walls of the tube is unaltered by
the displacement. The only effect of the rise is that a length of the tube
before in contact with air is replaced by an equal length in contact with the
lowest liquid. The work of the capillary forces is the same as if the upper
liquids did not exist, and therefore the total weight of the column supported
is independent of these liquids.
The case of Young's experiment, in which oil stands upon water in a glass
tube, is not covered by the foregoing reasoning. The oil must be supposed
to wet the glass, that is to insinuate itself between the glass and air, so that
the upper part of the tube is covered to a great height with a very thin layer
of oil. The displacement here takes place under conditions very different
from before. As the column rises, no new surface of glass is touched by oil,
while below water replaces oil. The properties of the oil are thus brought
into play, and Laplace's theorem does not apply.
* Works, Vol. L p. 463.
27—2
420 ON THE THEORY OF SURFACE FORCES. [176
That theory indicates the almost indefinite rise of a liquid like oil in
contact with a vertical wall of glass is often overlooked, in spite of Young's
explicit statement quoted above. It may be of interest to look into the
question more narrowly on the basis of Laplace's hypothesis.
If we include gravity in our calculations, the hydrostatic equation of
equilibrium is
p = const. + aV — gpz, ........................ (56)
where z is measured upwards, and V denotes as before the potential of the
cohesive forces. Along the free surface of the liquid the pressure is constant,
so that
<rV=oaK0 + gpz> ........................... (57)
z being reckoned from a place where the liquid is deep and the surface
plane.
At a point upon the surface, whose distance from the wall exceeds the
range of the forces,
>.i ........................ (58)
or, if we take the problem in two dimensions,
<rV = K + T/R, .............................. (59)
where R is the radius of curvature, and K, T denote the intrinsic pressure
and tension proper to the liquid and proportional to tr2. Upon this equation
is founded the usual calculation of the form of the surface.
When the point under consideration is nearer to the wall than the range
of the forces, the above expression no longer applies. The variation of V on
the surface of the thin layer which rises above the meniscus is due not to
variations of curvature, for the curvature is here practically evanescent, but
to the inclusion within the sphere of influence of the more dense matter
constituting the wall. If the attraction be a simple function of the distance,
such as those considered above in illustrative examples, the thickness of the
layer diminishes constantly with increasing height. The limit is reached
when the thickness vanishes, and the potential attains the value due simply
to the solid wall. This potential is a K0, the intrinsic pressure within the
wall being a-'*K0 ; so that if we compare the point above where the layer
of fluid disappears with a point below upon the horizontal surface, we find
gpz = or (</ - o-) K0 ............................ (60)
By this equation is given the total head of liquid in contact with the wall ;
and, as was to be expected, it is enormous.
The height of the meniscus itself in a very narrow tube wetted by the
liquid is obtained from (57), (58). If R be the radius of curvature at the
centre of the meniscus,
.............................. (61)
1890] ON THE THEORY OF SURFACE FORCES. 421
and R may be identified with the radius of the tube, for under the circum-
stances supposed the meniscus is very approximately hemispherical.
The calculation of the height by the method of energy requires a little
attention. The simplest displacement is an equal movement upwards of the
whole body of liquid, including the layer above the meniscus. In this case
the work of the cohesive forces depends upon the substitution of liquid for
air in contact with the tube, and therefore not merely upon the interfacial
tension between liquid and air, as (61) might lead us to suppose. The fact
is that in this way of regarding the subject the work which compensates
that of the cohesive forces is not simply the elevation against gravity of the
column (z), but also an equal elevation of the very high, though very thin,
layer situated above it. The complication thus arising may be avoided by
taking the hypothetical displacement so that the thin layer does not accom-
pany the column (z). In this case the work of the cohesive forces depends
upon a reduction of surface between liquid and air simply, without reference
to the properties of the walls, and (61) follows immediately.
Laplace's integral K was, as we have seen, introduced originally to
express the intrinsic pressure, but according to the discovery of Dupre* it
is susceptible of another and very important interpretation. " Le travail de
desagrdgation totale d'un kilogramme d'un corps quelconque e'gale le produit
de 1'attraction au contact par le volume, ou, ce qui £quivaut. le travail de
desagregation totale de 1'unite de volume e'gale 1'attraction au contact."
Attraction au contact here means what we have called intrinsic pressure.
The following reasoning is substantially that of Dupre.
We have seen (2) that l-Knur -ty (z) represents the attraction of a particle
m placed at distance z from the plane surface of an infinite solid whose
density is <r. The work required to carry m from z — 0 to z — oo is therefore
1 TJr (z) dz = ma-K0,
Jo
by (4) ; so that the work necessary to separate a superficial layer of thickness
dz from the rest of the mass and to carry it beyond the range of the
attraction is a^dzK^. The complete disaggregation of unit of volume into
infinitesimal slices demands accordingly an amount of work represented by
a*Kt, or K. The work required further to separate the infinitesimal slices
into component filaments or particles and to remove them beyond the range
of the mutual attraction is negligible in the limit, so that K is the total work
of complete disaggregation.
A second law formulated by Dupre* is more difficult to accept. " Pour un
meine corps prenant des volumes varies, le travail de desagregation restant
* ThSorie Mgcanique de la Chaleur, 1869, p. 152.
Van der Waals gives the same result in his celebrated essay of 1873.— German Translation,
1881, p. 81.
422 ON THE THEORY OF SURFACE FORCES. [176
a accomplir est proportionel a la densite ou en raison inverse du volume."
The argument is that the work remaining to be done upon a given mass at
any stage of the expansion is proportional first to the square of the density,
and secondly to the actual volume, on the whole therefore inversely as the
volume. The criticism that I am inclined to make here is that Dupre"s
theory attempts either too little or too much. If we keep strictly within the
lines of Laplace's theory the question here discussed cannot arise, because the
body is supposed to be incompressible. That bodies are in fact compressible
may be so much the worse for Laplace's theory, but I apprehend that the
defect cannot be remedied without a more extensive modification than Dupre"
attempts. In particular, it would be necessary to take into account the work
of compression. We cannot leave the attractive forces unbalanced ; and the
work of the repulsive forces can only be neglected upon the hypothesis that
the compressibility itself is negligible. Indeed it seems to me, that a large
part of Dupre's work, important and suggestive as it is, is open to a funda-
mental objection. He makes free use of the two laws of thermodynamics, and
at the same time rests upon a molecular theory which is too narrow to hold
them. One is driven to ask what is the real nature of this heat, of which we
hear so much. It seems hopeless to combine thermodynamics with a merely
statical view of the constitution of matter.
On these grounds I find it difficult to attach a meaning to such a theorem
as that enunciated in the following terms*: — "La derivee partielle du travail
mecanique interne prise par rapport au volume egale 1'attraction par metre
carre qu'exercent 1'une sur 1'autre les deux parties du corps situees des
deux cote's d'une section plane," viz. the intrinsic pressure. In the partial
differentiation the volume is supposed to vary and the temperature is sup-
posed to remain constant. The difficulty of the first part of the supposition
has been already touched upon; and how in a fundamental theory can we
suppose temperature to be constant without knowing what it is ? It is
possible, however, that some of these theorems may be capable of an inter-
pretation which shall roughly fit the facts, and it is worthy of consideration
how far they may be regarded as applicable to matter in a state of extreme
cold.
With respect to the value of K, Young's estimate of 23,000 atmospheres
for water has already been referred to. It is not clear upon what basis he
proceeded, but a chance remark suggests that it may have been upon the
assumption that cohesion was of the same order of magnitude in liquids and
solids. Against this, however, it may be objected that the estimate is unduly
high. Even steel is scarcely capable of withstanding a tension of 23,000
atmospheres.
* Loc. cit. p. 47.
1890] ON THE THEORY OF SURFACE FORCES. 423
So far as I am aware, the next estimates of K are those of Dupre. One
of them proceeds upon the assumption that for rough purposes K may be
identified with the mechanical equivalent of the heat rendered latent in the
evaporation of the liquid, that in fact evaporation may be regarded as a
process of disaggregation in which the cohesive forces have to be overcome.
This view appears to be substantially sound. If we take the latent heat of
water as 600°, we find for the work required to disintegrate one gram
of water
600 x 4-2 x 107 C.G.S.
One atmosphere is about 106 C.GJ3. ; so that
#=25,000 atmospheres.
The estimates of his predecessors were apparently unknown to Van der Waals,
who (in 1873) undertook his work mainly with the object of determining
the quantity in question. He finds for water 11,000 atmospheres. The
application of Clausius's equation of virial to gases and liquids is obviously
of great importance ; but, as it lies outside the scope of the present paper,
I must content myself with referring the reader to the original memoir and
to the account of it by Maxwell *.
One of the most remarkable features of Young's treatise is his estimate
of the range a of the attractive force on the basis of the relation T=^aK.
Never once have I seen it alluded to ; and it is, I believe, generally supposed
that the first attempt of the kind is not more than twenty years old.
Estimating K at 23,000 atmospheres, and T at 3 grains per inch, Young
finds f that " the extent of the cohesive force must be limited to about the
250 millionth of an inch " ; and he continues, " nor is it very probable that
any error in the suppositions adopted can possibly have so far invalidated this
result as to have made it very many times greater or less than the truth."
It detracts nothing from the merit of this wonderful speculation that a
more precise calculation does not verify the numerical coefficient in Young's
equation. The point is that the range of the cohesive force is necessarily
of the order T/K.
But this is not all. Young continues : — " Within similar limits of uncer-
tainty, we may obtain something like a conjectural estimate of the mutual
distance of the particles of vapours, and even of the actual magnitude of the
elementary atoms of liquids, as supposed to be nearly in contact with each
other ; for if the distance at which the force of cohesion begins is constant at
the same temperature, and if the particles of steam are condensed when they
approach within this distance, it follows that at 60° of Fahrenheit the distance
of the particles of pure aqueous vapour is about the 250 .millionth of an inch ;
* Nature, Vol. x. p. 477 (1874). See also Vol. XL pp. 357, 374. [Maxwell's Scientific Papen,
Vol. n. pp. 407, 418.]
t Work*, Vol. i. p. 461.
424 ON THE THEORY OF SURFACE FORCES. [176
and since the density of this vapour is about one sixty thousandth of that
of water, the distance of the particles must be about forty times as great ;
consequently the mutual distance of the particles of water must be about
the ten thousand millionth of an inch. It is true that the result of this
calculation will differ considerably according to the temperature of the sub-
stances compared.... This discordance does not, however, wholly invalidate
the general tenour of the conclusion... and on the whole it appears tolerably
safe to conclude that, whatever errors may have affected the determination,
the diameter or distance of the particles of water is between the two
thousand and the ten thousand millionth of an inch." This passage, in spite
of its great interest, has been so completely overlooked that I have ventured
briefly to quote it, although the question of the size of atoms lies outside the
scope of the present paper.
Another matter of ' great importance to capillary theory I will only
venture to touch upon. When oil spreads upon water, the layer formed is
excessively thin, about two millionths of a millimetre. If the layer be at
first thicker, it exhibits instability, becoming perforated with holes. These
gradually enlarge, until at last, after a series of curious transformations, the
superfluous oil is collected in small lenses. It would seem therefore that
the energy is less when the water is covered by a very thin layer of oil,
than when the layer is thicker. Phenomena of this kind present many
complications, for which various causes may be suggested, such as solubility,
volatility, and — perhaps more important still — chemical heterogeneity. It is
at present, I think, premature to draw definite physical conclusions; but
we may at least consider what is implied in the preference for a thin as
compared with a thicker film.
Fig. 13. Fig. 14. Fig. 15.
The passage from the first stage to the second may be accomplished in
the manner indicated in Figs. 13, 14, 15. We begin (Fig. 13) with a thin
layer of oil on water and an independent thick layer of oil. In the second
stage (Fig. 14) the thick layer is split in two, also thick in comparison with
the range of the cohesive forces, and the two parts are separated. In the
third stage one of the component layers is brought down until it coalesces
with the thin layer on water. The last state differs from the first by the
1890] ON THE THEORY OF SURFACE FORCES. 425
substitution of a thick film of oil for a thin one in contact with the water,
and we have to consider the work spent or gained in producing the change.
If, as observation suggests, the last state has more energy than the first, it
follows that more work is spent in splitting the thick layer of oil than is
gained in the approach of a thick layer to the already oiled water. At some
distances therefore, and those not the smallest, oil must be more attracted (or
less repelled) by oil than by water. The reader will not fail to notice the
connexion between this subject and the black of soap-films investigated by
Profe. Reinold and Rticker [PhU. Trans. 172, p. 645, 1884].
[1901. Continuations of the present memoir under the same title will
be found below, reprinted from Phil. Mag. xxxm. pp. 209, 468, 1892.
Reference may be made also to PhU. Mag. XLVHI. p. 331, 1899.]
177.
CLERK-MAXWELL'S PAPERS*
[Nature, XLIII. pp. 26, 27, 1890.]
THE gratitude with which we receive these fine volumes is not unmingled
with complaint. During the eleven years which have elapsed since the
master left us, the disciples have not been idle, but their work has been
deprived, to all appearance unnecessarily, of the assistance which would have
been afforded by this collection of his works. However, it behoves us to look
forward rather than backward ; and no one can doubt that for many years
to come earnest students at home and abroad will derive inspiration from
Maxwell's writings, and will feel thankful to Mr Niven and the committee
of friends and admirers for the convenient and handsome form in which they
are here presented.
Under the modest title of preface, the editor contributes a sketch of
Maxwell's life, which will be valued even by those who are acquainted with
the larger work of Profs. Lewis Campbell and W. Garnett; and while
abstaining from entering at length into a discussion of the relation which
Maxwell's work bears historically to that of his predecessors, or attempting
to estimate the effect which it had upon the scientific thought of the present
day, he points out under the various heads what were the leading advances
made.
In the body of the work the editor's additions reduce themselves to a few
useful footnotes, placed in square brackets. Doubtless there is some difficulty
in knowing where to stop, but the number of these footnotes might, I think,
have been increased. For example, the last term in the differential equation
of a stream-function symmetrical about an axis is allowed to stand with a
wrong sign (Vol. I. p. 591) and on the following page the fifth term in the
* The Scientific Papers of James Clerk-Maxwell. Two Vols. Edited by W. D. Niven.
(London: Cambridge University Press, 1890.)
1890] CLERK-MAXWELL'S PAPERS. 427
expression for the self-induction of a coil should be — £Trcosec20, and not
To a large and enterprising group of physicists, Maxwell's name at once
suggests electricity, and some, familiar with the great treatise, may be
tempted to suppose that this book can contain little that is new to them.
It was De Morgan, I think, who remarked that a great work often over-
shadows too much lesser writings of an author upon the same subject. In
the present case it is true that much of the " Dynamical Theory of the
Electro-magnetic Field " was subsequently embodied in the separate treatise.
Nevertheless, there were important exceptions. Among these may be noticed
the experimental method of determining the self-induction of a coil of wire in
the Wheatstone's balance. By adjustment of resistances, the steady current
through the galvanometer in the bridge is reduced to zero; but at the
moment of making or breaking battery contact, an instantaneous current
passes. From the magnitude of the throw thus observed, in comparison
with the effect of upsetting the resistance-balance to a known extent, the
self-induction can be calculated. The letter to Sir W. Grove, entitled
"Experiment in Magneto-electric Induction" (Vol. u. p. 121), will also be
read with interest by electricians. It gives the complete theory of what
is sometimes called " electric resonance."
There can be little doubt but that posterity will regard as Maxwell's
highest achievement in this field his electro-magnetic theory of light,
whereby optics becomes a department of electrics. The clearest statement
of his views will be found in the note appended to the " Direct comparison
of Electro-static with Electro-magnetic Force" (Vol. n. p. 125). Several
of the points which were then obscure have been cleared up by recent
researches.
Scarcely, if at all, less important than his electrical work was the part
taken by Maxwell in the development of the Dynamical Theory of Gases.
Even now the difficulties which meet us here are not entirely overcome ; but
in the whole range of science there is no more beautiful or telling discovery
than that gaseous viscosity is the same at all densities. Maxwell anticipated
from theory, and afterwards verified experimentally, that the retarding effect
of the air upon a body vibrating in a confined space is the same at atmo-
spheric pressure and in the best vacuum of an ordinary air-pump.
Besides the more formal writings, these volumes include several reviews,
contributed to Nature, as well as various lectures and addresses, all abound-
ing in valuable suggestions, and enlivened by humorous touches. Among
the most noticeable of these are the address to Section A of the British
Association, the lectures on Colour-vision, on Molecules, and on Action at
a Distance, and, one of his last efforts, the Rede Lecture on the Telephone.
Many of the articles from the Encyclopaedia Britannica are also of great
428 CLERK-MAXWELL'S PAPERS. [17 7
importance, and become here for the first time readily accessible to foreigners.
Under " Constitution of Bodies," ideas are put forward respecting the break-
ing up of but feebly stable groups of molecules, which, in the hands of Prof.
Ewing, seem likely to find important application in the theory of magnetism.
A characteristic of much of Maxwell's writing is his dissatisfaction with
purely analytical processes, and the endeavour to find physical interpretations
for his formula?. Sometimes the use of physical ideas is pushed further than
strict logic can approve * ; but those of us who are unable to follow a Sylvester
in his analytical flights will be disposed to regard the error with leniency.
The truth is that the limitation of human faculties often imposes upon us,
as a condition of advance, temporary departure from the standard of strict
method. The work of the discoverer may thus precede that of the systematizer ;
and the division of labour will have its advantage here as well as in other
fields.
The reader of these volumes, not already familiarly acquainted with
Maxwell's work, will be astonished at its variety and importance. Would
that another ten years' teaching had been allowed us ! The premature death
of our great physicist was a loss to science that can never be repaired.
* " With all possible respect for Prof. Maxwell's great ability, I must own that to deduce
purely analytical properties of spherical harmonics, as he has done, from ' Green's theorem ' and
the ' principle of potential energy," seems to me a proceeding at variance with sound method,
and of the same kind and as reasonable as if one should set about to deduce the binomial
theorem from the laws of virtual velocities or make the rule for the extraction of the square root
flow as a consequence from Archimedes' law of floating bodies." Sylvester, Phil. Mag. Vol. n.
p. 306. 1876.
178.
ON PIN-HOLE PHOTOGRAPHY.
[Philosophical Magazine, xxxi. pp. 87—99, 1891.]
IT has long been known that the resolving power of lenses, however
perfect, is limited, and more particularly that the capability of separating
close distant objects, e.g. double stars, is proportional to aperture. The
ground of the limitation lies in the finite magnitude of the wave-length of
light (X), and the consequent diffusion of illumination round the geometrical
image of even an infinitely small radiant point. It is easy to understand
the rationale of this process without entering upon any calculations. At
the focal point itself all the vibrations proceeding from various parts of the
aperture arrive in the same phase. The illumination is therefore here a
maximum. But why is it less at neighbouring points in the focal plane
which are all equally exposed to the vibrations from the aperture ? The
answer can only be that at such points the vibrations are discrepant. This
discrepance can only enter by degrees ; so that there must be a small region
round the focus, at any point of which the phases are practically in agree-
ment and the illumination sensibly equal to the maximum.
These considerations serve also to fix at least the order of magnitude of
the patch of light. The discrepancy of phase is the result of the different
distances of the various parts of the aperture from the eccentric point in
question; and the greatest discrepancy is that between the waves which
come from the nearest and furthest parts of the aperture. A simple calcu-
lation shows that the greatest difference of distance is expressed by 2rar//,
where 2r is the diameter of the aperture, / the focal length, and ar the
linear eccentricity of the point under consideration. The question under
discussion is at what stage does this difference of path introduce an important
discrepancy of phase ? It is easy to recognize that the illumination will not
be greatly reduced until the extreme discrepancy of phase reaches half a
wave-length. In this case
2ar =/X/2r,
430 ON PIN-HOLE PHOTOGRAPHY. [178
which may be considered to give roughly the diameter of the patch of light.
If there are two radiant points, the two representative patches will seriously
overlap, unless the distance of their centres exceed 2#. Supposing it to be
equal to 2x, which corresponds to an angular interval 2x/f, we see that the
double radiant cannot be resolved in the image, unless the angular interval
X/2r.
Experiment* shows that the value thus roughly estimated is very near
the truth for a rectangular aperture of width 2r. If the aperture be of
circular form, the resolving power is somewhat less, in the ratio of about
11 : 1.
It is therefore not going too far to say that there is nothing better
established in optics than the limit to resolving power as proportional to
aperture. On the other hand, the focal length is a matter of indifference,
if the object-glass be perfect.
This is one side of the question before us. We now pass on to another,
in which the focal length becomes of paramount importance.
" The function of a lens in forming an image is to compensate by its
variable thickness the differences in phase which would otherwise exist
between secondary waves arriving at the focal point from various parts of
the aperture. If we suppose the diameter of the lens (2r) to be given, and
its focal length (/) gradually to increase, these differences of phase at
the image of an infinitely distant luminous point diminish without limit.
When / attains a certain value, say /i, the extreme error of phase to be
compensated falls to i\. Now, as I have shown on a previous occasion f,
an extreme error of phase amounting to ^X, or less, produces no appreciable
deterioration in the definition ; so that from this point onwards the lens
is useless, as only improving an image already sensibly as perfect as the
aperture admits of. Throughout the operation of increasing the focal length,
the resolving power of the instrument, which depends only upon the aperture,
remains unchanged; and we thus arrive at the rather startling conclusion
that a telescope of any degree of resolving power might be constructed
without an object-glass, if only there were no limit to the admissible focal
length. This last proviso, however, as we shall see, takes away almost all
practical importance from the proposition.
"To get an idea of the magnitudes of the quantities involved, let us
take the case of an aperture of ^ inch, about that of the pupil of the eye.
The distance flt which the actual focal length must exceed, is given by
* " On the Kesolving Power of Telescopes," Phil. Mag. August 1880. [Vol. i. p. 488.]
f Phil. Mag. November 1879. [Vol. i. p. 415.]
1891] ON PIN-HOLE PHOTOGRAPHY. 431
so that {approximately}
Thus, if \=t^, r = 1V, ^ = 800.
"The image of the sun thrown on a screen at a distance exceeding
66 feet, through a hole £ inch in diameter, is therefore at least as well
defined as that seen direct. In practice it would be better defined, as the
direct image is far from perfect. If the image on the screen be regarded from
a distance /, , it will appear of its natural angular magnitude. Seen from a
distance less than /,, it will appear magnified. Inasmuch as the arrange-
ment affords a view of the sun with full definition (corresponding to aperture)
and with an increased apparent magnitude, the name of a telescope can hardly
be denied to it.
" As the minimum focal length increases with the square of the aperture,
a quite impracticable distance would be required to rival the resolving power
of a modern telescope. Even for an aperture of four inches /, would be five
miles*."
A more practical application of these principles is to be found in landscape
photography, where a high degree of definition is often unnecessary, and
where a feeble illumination can be compensated by length of exposure. In
a recent communication to the British Association t it was pointed out that a
suitable aperture is given by the relation
2r>=/\; ................................. (1)
and a photograph was exhibited in illustration of the advantage to be derived
from an increase of/. The subject was a weather-cock, seen against the
sky, and it was taken with an aperture of ^ inch [inch = 2'54 cm.] and at
a distance of 9 feet. The amount of detail in the photograph is not markedly
short of that observable by direct vision from the actual point of view. The
question of brightness was also considered. As the focal length increases,
the brightness (B) in the image of a properly proportioned pin-hole camera
diminishes. For
£ oc r1//1 oc r'X'/r4 oc X2//-8 * X//. .................. (2)
There will now be no difficulty in understanding why a certain aperture
is more favourable than either a larger or a smaller one, when f and X are
given. If the aperture be very small, the definition is poor even if the aid
of a lens be invoked. If, on the other hand, the aperture be large, the
lens becomes indispensable. The size of the aperture should accordingly be
increased up to the point at which the lens is sensibly missed ; and this, as
we have seen, will occur in the neighbourhood of the value determined by (1).
* " On Images formed without Reflection or Refraction," Phil. Mag. March 1881. [Vol. i.
p. 513.]
t Brit. Assoc. Report, 1889, p. 493.
432 ON PIN-HOLE PHOTOGRAPHY. [178
A more precise calculation can be made only upon the basis of a detailed
knowledge of the distribution of light in the image.
The question of the best size of aperture for a pin-hole camera was first
considered by Petzval*. His theory, though it can hardly be regarded as
sound, brings out the failure of definition when the aperture is either too
large or too small, and, as is very remarkable, gives (1) as the best relation
between r, f, and X. The argument is as follows : — If the hole be very
small, the diameter of the patch of light representative of a luminous point
is given by
D=f\/r,
the measurement being made up to the first blackness in the diffraction-
pattern. " This formula is only an approximate one, applicable when r is
very small ; in the case of a larger aperture, its diameter must be added to
the value above given, that is to say,
D = 2r+f\/r.
From the last formula we can at once deduce the best value for r ; in other
words, the size of the aperture which corresponds to the least possible value
of D, and therefore to the sharpest possible image. In fact, differentiating
the last expression, and setting in the ordinary manner, dD/dr = 0, we find
at once
which corresponds to
D = 2V(2/X)."
The assumption that intermediate cases can be represented by mere addition
of the terms appropriate in the extreme cases of very large and very small
apertures appears to be inadmissible.
The complete determination of the image of a radiant point as given by
a small aperture is a problem in diffraction, solved only within the last years
by Lommelf. In view of the practical application to pin-hole photography,
I have thought that it would be interesting to adapt Lommel's results to
the problem in hand, and to exhibit upon the same diagram curves showing
the distribution of illumination in various cases. For the details of the
investigation reference must be made to Lommel's memoir, or to the account
of it in the Encyclopaedia Britannica, Art. " Wave Theory," p. 444. But it
may be well to state the results somewhat fully. [These results, having been
already given — equations (1) to (19), Vol. in. pp. 135, 136 — are not now
repeated.]
* Wien. Site. Ber. xxvi. p. 33 (1857) ; Phil. Mag. xvn. (1859), p. 1.
t "Die Beugungserscheinungen einer kreisrunden Oeffnung uud eines kreisrunden Schirm-
chens," Aus den Abhandlungen der k. layer. Akademie der Wiss. n. Cl. xv. Bd. n. Abth.
(Miinchen, 1884.)
1891] OX PIN-HOLE PHOTOGRAPHY. 433
At the central point of the image where z = 0, F0 = 1, V^ = 0,
and
r, _ 4 „:„, /_
X 2ab '
4 . firtaa + b\
- sinM— (20)
In general by (10), (11),
if with Lommel we set
"""(Itf+dtf (22)
Also
7> -a**'"' <M>
In these formulae fT,2, Uf, and therefore by (22), (23) J/2 and 72 are known
functions of y and z. The connexion with r and f is given by the relations
In Lommel's memoir are given the values of M* for integral values of e
from 0 to 12 when y has the values TT, 2?r, 3?r, &c. If we regard a, b, X
as given, each of these Tables affords a knowledge of the distribution of
illumination as a function of £ for a certain radius of aperture by means of the
two equations (24). In each case £ is proportional to z ; but in comparing
one case with another we have to bear in mind that the ratio of £ to z varies.
As our object is to compare the distributions of illumination when the
aperture varies, we must treat £, and not z, as the abscissa in our diagrams.
Another question arises as to how the scale of the ordinate 1- should be dealt
with in the various cases. If we take (23) as it stands, we shall have curves
corresponding to the same actual intensity of the radiant point. For some
purposes this might be desirable ; but in the application to photography the
deficiency of illumination when the aperture is much reduced would always
be compensated by increased exposure. It will be more practical to vary
the scale of ordinates from that prescribed in (23), so as to render the
illumination corresponding to an extended source of light, such as the sky,
the same in all cases. We shall effect this by removing from the right-hand
member of (23) a factor proportional to the area of aperture, proportional
that is to i*, or y. Thus for any value of y equal to STT, we shall require to
plot as ordinate, not M2 simply, but sJ/2, and as abscissa, not z simply, but
zj *Js. The following are at once deduced from Lommel's tables III. — VI.
434
ON PIN-HOLE PHOTOGRAPHY.
[178
z
ZIJ1 = Z
M*
0
0
•8106 Max.
1
1
•6286
2
2
•2772
3
3
•0623
4
4
•0269
5
5
•0306
6
6
•0121
i
7
•0018
8
8
•0051
9
9
•0037
10
10
•0004
11
11
•0013
12
12
•0016
3-8317
•0263 Min.
4-7153
•0320 Max.
7-0156
•0018 Min.
8-3060
•0055 Max.
10-1735
•0003 Min.
11-5785
•0019 Max.
z
*/v/2
23f2
0
•ooo
•8106 Max.
1
•707
•6316
2
1-414
•3117
3
2-121
•1560
4
2-829
•1438
5
3-536
•1077
6
4-243
•0426
7
4-950
•0200
8
5-657
•0227
9
6-364
•0125
10
7-071
•0034
11
7-778
•0053
12
8-485
•0046
3-5977
2-544
•1440 Min.
3-8317
2-710
•1442 Max.
7-0156
4-961
•0198 Min.
7-8879
5-578
•0229 Max.
10-1735
7-193
•0032 Min.
11-4135
8-070
•0059 Max.
1891]
OX PIN-HOLE PHOTOGRAPHY.
435
*
2/N/3
SAP
0
•ooo
•2702 Max.
1
•577
•2159
2
1-154
•1631
3
1-732
•2110
4
2-310
•2449
5
2-887
•1734
6
3-464
0916
7
4-041
•0739
8
4-619
•0651
9
5-195
•0335
10
5773
•0156
11
6-350
•0178
12
6-927
•0122
1-9969
1-153 -1631 Min.
3-8317
2-212 -2467 Max.
7-0156
4-050 -0737 Min.
7-0878
4-092 -0739 Max.
10-1735 5-871 O154 Min.
11-0361 6-374 -0178 Max.
z
i
2/2 43P
0 -000
•0000 Min.
1 -500
•0056
2 1-000
•0609
3 1-500
•1594
4 2OOO
•1947
5
2-500
•1515
6
3-000
•1293
7
3-500
•1399
8
4-000
•1148
9
4-500
•0658
10
5-000
•0484
11
5-500
•0458
12 6-000
•0280
3-8317 1-9158
•1961 Max.
5-8978 2-9489
•1291 Min.
7-0156 3-5078
•1399 Max,
10-1735 5-0867
•0483 Min.
10-3861 5-1930
•0483 Max.
28—2
436
ON PIN-HOLE PHOTOGRAPHY.
[178
As it appeared desirable to trace the curve corresponding to a smaller value
of y than any given by Lommel, I have calculated by means of (12), (13) the
value of |J/2, that is of
corresponding to z = 0, 1, 2, 3, 4.
The results are as follows :
9 = 1*-
.
JM-
0
•000
•4748
1
1-414
•3679
2
2-828
•1590
3
4-243
•0272
4
5-657
•0041
The various curves, or rather the halves of them, are plotted in the Figure,
and exhibit to the eye the distribution of light in the images corresponding
to the different apertures. It is at once evident that y = \TT is too small, and
that y = STT is too great. The only question that can arise is between y = TT
and y = 2-Tr. The latter has decidedly the higher resolving power, but the
advantage is to some extent paid for in the greater diffusion of light outside
the image proper. In estimating this we must remember that the amount
of light is represented, not by the areas of the various parts of the diagrams,
but by the volumes of the solids formed by the revolution of the curves round
the axis of /2. In virtue of the method of construction the total volume is
the same in all cases. The best aperture will thus depend in some degree
upon the subject to be represented ; but there is every reason to think that
in general y — 2?r will prove more advantageous than y = TT. It will be
convenient to recall that
2r2 a + b
or, if we write a = <x> , b=f,
ab
(25)
The curve y = TT thus corresponds to (1) ; and we conclude that the aperture
may properly be somewhat enlarged so as to make
r* = \f. ................................. (26)
In the general case when a is finite, T//TT represents four times the number
of wave-lengths by which the extreme ray is retarded relatively to the
central ray; for
V(a2 + r2) + y(62 + r2) - a - b_ r* a + b
X ~~2X ab
approximately.
1891]
OX PIN-HOLE PHOTOGRAPHY.
437
According to (26) the aperture should be enlarged until the retardation
amounts to
In the image of a double star the curves of brightness proper to the two
components are superposed. If the components are equal, resolution will
be just beginning when the distance of the geometrical images asunder is
the double of the value of £ for which /* has about one-half its maximum
value. By inspection of the curve for y = 2-n- we see that there will not be
much appearance of resolution until z/V2 = 1'5. The corresponding angular
interval between the two components is
2f 1-5 x V2 //X\
7=~~^~v(fJ (27)
438 ON PIN-HOLE PHOTOGRAPHY. [178
This may be regarded as defining the maximum separating power as a
function of X and /.
Passing on from the theoretical discussion, I will now describe certain
laboratory observations upon the defining power of various apertures. A
succession of such, of gradually increasing magnitude, were perforated in
a piece of thin sheet zinc, and were measured under the microscope. The
diameters, in fractions of an inch, are as follows : —
(1) (2) (3) (4) (5) (6)
•0210, -0240, -0262, "0290, -0326, "0366.
The objects, whose images were examined, are (1) a grating cut out of sheet
zinc, and (2) a pair of equal round holes a quarter of an inch apart. The
period of the grating is also a quarter inch, and the transparent and opaque
parts are equally wide. Behind the grating, or double hole, was placed a
paraffin lamp and a large condensing lens. The distance a between the
objects and the apertures under test was about 18 feet.
The best -image with a given aperture is obtained by bringing the eye
immediately behind, without the use of a focusing lens. But the image
formed at a sufficient distance beyond, and examined with a focusing glass of
low power, is nearly as good. Thus at a sufficient distance (6) the largest
aperture gives the best image, but at a given distance behind the case is
otherwise. For example, when the image was formed at 8 inches distance,
(2) and (3) were about equal as respects the double hole, while (1) was
decidedly inferior, and that not apparently from want of light. In the case
of the grating (3) had perhaps the advantage over (2).
A photograph of the double hole was now taken under the same circum-
stances with an exposure of 80 minutes. Aperture (2) was here decidedly
better than (3), and (1) was almost as good as (2). The (negative) image
formed by (5) exhibited a pair of white spots near the centre of a patch of
black, corresponding to the state of things indicated in the curve y = 4?r.
The difference between the photograph and the result obtained by eye
observation is readily explained by the smaller effective wave-length in the
former case.
The difference just spoken of is intensified when the light is white. In
one experiment upon cloud-light a = 21 feet, b = 10 inches. In the resulting
photograph, obtained upon an Ilford plate with an exposure of 30 minutes,
the image from (2) was decidedly the best.
We may utilize the last result to calculate the relation between aperture
and focus most suitable for out of door photography. We have
(2r)2 (- + 1} = (-0240)2 fgL + ^~) = 10-5 x 5"99 inches = 10~4 x 1-52 cm.
\fl o/ \£o£ it//
1891] ON PIN-HOLE PHOTOGRAPHY. 439
Thus, if a = x , as may usually be supposed in landscape-photography, the
most favourable diameter of aperture is given by
(2r)2//= 10~5 x 6-0 inches = lO- x To cm.,
the first number being employed if r and f are measured in inches, and
the latter when the measures are in centimetres*. If /= 12 inches,
2r = -027 inch. If /= 7 x 12 = 84 inches, 2r = '071 inch.
The experimental determination of the best value of y is more easily
effected by eye observations. In order to render the wave-length more
definite, an orange-red glass was employed. With a = 18 feet, 6 = 8 inches,
the image formed by aperture (3) was judged to be decidedly the best,
(2) was next, while (1) and (4) were decidedly behind. Thus we may take
as the most favourable aperture 2r = '026 inch.
The mean wave-length of the light employed was found with the aid
of a grating by a comparison with a soda flame : —
Mean X : \D = 239 : 226 ;
so that
\ = ^— x 5-89 x 10~5 = 6-23 x 10~5 centim.
Hence
/7r_2r2/i 1\ _ (-026)2 7 2-54 x 105
\ \a bj 2 54 6'23
agreeing very well with what was expected from the curves.
If we now assume that the best value of y is T8, we can calculate
backwards from the photographic results to find the mean X there effective.
We have
10~4 x 1-52 = (2r)2 (- + 1} = ^ = 2X x T8 ;
' \a bj TT
whence
X = 4-2 x 10"5 centim...
a little less than that belonging to Fraunhofer's line G.
To test the improvement of definition which according to (27) accom-
panies an increase of f, I have used an aperture of '07 inch and a focal length
of 7 feet. The aperture was perforated in sheet zinc, and was placed in the
shutter of a room otherwise completely darkened. The subject was a group
of cedars, and, being somewhat dark in the shadows, required an exposure
of about an hour and a half, even in sunshine. The resulting 12 x 10-inch
photographs fully bear out expectations. To appear in natural magnitude
* The effect of varying the diameter of aperture in photographic landscape work has been
tested by Capt. Abney; but I am not in possession of the data as to focal length necessary
for a comparison with the above.
440 ON PIN-HOLE PHOTOGRAPHY. [178
the pictures would of course need to be held 7 feet distance from the eye ;
but even at 3 or 4 feet the apparent definition is sufficient. I have also
taken panoramic pictures with an aperture of '027 inch and a focal distance
of 12 inches; but in this case there is nothing that could not equally well
be done with an ordinary portable camera.
TERLING PLACE, WITHAM, ESSKX,
Dec. 2, 1890.
179.
SOME APPLICATIONS OF PHOTOGRAPHY.
[Proc. Roy. Inst. xin. pp. 261—272, Feb. 1891 : Nature, XLIV.
pp. 249—254, 1891.]
ONE of the subjects to which I propose to invite your attention this
evening is the application of instantaneous photography to the illustration
of certain mechanical phenomena which pass so quickly as to elude ordinary
means of observation. The expression " instantaneous photography " is not
quite a defensible one, because no photography can be reallv instantaneous —
some time must always be occupied. One of the simplest and most commonly
used methods of obtaining very short exposures is by the use of movable
shutters, for which purpose many ingenious mechanical devices have been
invented. About two years ago we had a lecture from Prof. Muybridge, in
which he showed us the application of this method — and a remarkably
interesting appb'cation it was — to the examination of the various positions
assumed by a horse in his several gaits. Other means, however, may be
employed to the same end, and one of them depends upon the production
of an instantaneous light. It will obviously come to the same thing whether
the light to which we expose the plates be instantaneous, or whether by
a mechanical device we allow the plate to be submitted to a continuous light
for only a very short time. A good deal of use has been made in this way of
what is known as the magnesium flash light. A cloud of magnesium powder
is ignited, and blazes up quickly with a bright light of very short duration.
Now I want to compare that mode of illumination with another, in order to
be able to judge of the relative degree of instantaneity, if I may use such
an expression. We will illumine for a short time a revolving disc, composed
of black and white sectors ; and the result will depend upon how quick the
motion is as compared with the duration of the light. If the light could be
truly instantaneous, it would of necessity show the disc apparently stationary.
442 SOME APPLICATIONS OF PHOTOGRAPHY. [179
I believe that the duration of this light is variously estimated at from
one-tenth to one-fiftieth of a second ; and as the arrangement that I have
here is one of the slowest, we may assume that the time occupied will be
about a tenth of a second. I will say the words one, two, three, and at the
word three Mr Gordon will project the powder into the flame of a spirit lamp,
and the flash will be produced. Please give your attention to the disc, for
the question is whether the present uniform grey will be displaced by a
perception of the individual black and white sectors. [Experiment.] You
see the flash was not instantaneous enough to resolve the grey into its
components.
I wish now to contrast with that mode of illumination one obtained by
means of an electric spark. We have here an arrangement by which we can
charge Leyden jars from a Wimshurst machine. When the charge is sufficient,
a spark will pass inside a lantern, and the light proceeding from it will be
condensed and thrown upon the same revolving disc as before. The test will
be very much more severe ; but severe as it is, I think we shall find that the
electric flash will bear it. The teeth on the outside of the disc are very
numerous, and we will make them revolve as fast as we can, but we shall find
that under the electric light they will appear to be absolutely stationary.
[Experiment.] You will agree that the outlines of the black and white
sectors are seen perfectly sharp.
Now, by means of this arrangement we might investigate a limit to the
duration of the spark, because with a little care we could determine how fast
the teeth are travelling — what space they pass through in a second of time.
For this purpose it would not be safe to calculate from the multiplying gear
on the assumption of no slip. A better way would be to direct a current
of air upon the teeth themselves, and make them give rise to a musical note,
as in the so-called siren. From the appearance of the disc under the spark
we might safely say, I think, that the duration of the light is less than a
tenth of the time occupied by a single tooth in passing. But the spark is
in reality much more instantaneous than can be proved by the means at
present at our command. In order to determine its duration, it would be
necessary to have recourse to that powerful weapon — the revolving mirror;
and I do not, therefore, propose to go further into the matter to-night.
Experiments of this kind were made some twenty years ago by Prof. Rood,
.of New York, both on the duration of the discharge of a Leyden jar, and also
on that of lightning. Prof. Rood found that the result depended somewhat
upon the circumstances of the case; the discharge of a small jar being
generally more instantaneous than that of a larger one. He proved that in
certain cases the duration of the principal part of the light was as low as one
twenty-five-millionth part of a second of time. That is a statement which
probably conveys very little of its real meaning. A million seconds is about
1891] SOME APPLICATIONS OF PHOTOGRAPHY. 443
twelve days and nights. Twenty-five million seconds is nearly a year. So
that the time occupied by the spark in Prof. Hood's experiment is about the
same fraction of one second that one second is of a year. In many other
cases the duration was somewhat greater ; but in all his experiments it was
well under the one-millionth part of a second. In certain cases you may
have multiple sparks. I do not refer to the oscillating discharges of which
Prof. Lodge gave us so interesting an account last year; Prof. Rood's
multiple discharge was not of that character. It consisted of several detached
overflows of his Leyden jar when charged by the Rhumkorff coil. One
number mentioned for the total duration was one six-thousandth part of a
second; but the individual discharges had the degree of instantaneity of
which I have spoken.
It is not a difficult matter to adapt the electrical spark to instantaneous
photography. We will put the lantern into its proper position, excite the
electric sparks within it, causing them to be condensed by the condenser
of the lantern on to the photographic lens. We will then put the object
in front of the lantern-condenser, remove the cap from the lens, expose the
plate to the spark when it comes, and thus obtain an instantaneous view
of whatever may be going on. I propose to go through the operation of
taking such a photograph presently. I will not attempt any of the more
difficult things of which I shall speak, but will take a comparatively easy
subject, — a stream of bubbles of gas passing up through a liquid. In order
that you may see what this looks like when observed in the ordinary way, we
have arranged it here for projection upon the screen. [Experiment.] The
gas issues from the nozzle, and comes up in a stream, but so fast that you
cannot fairly see the bubbles. If, however, we take an instantaneous picture,
we shall find that the stream is decomposed into its constituent parts. We
arrange the trough of liquid in front of the lantern which contains the
spark-making apparatus — [Experiment] — and we will expose a plate, though
I hardly expect a good result in a lecture. A photographer's lamp provides
some yellow light to enable us to see when other light is excluded. There
goes the spark ; the plate is exposed, and the thing is done. We will develop
the plate, and see what it is good for ; and if it turns out fit to show, we will
have it on the screen within the hour.
In the meantime, we will project on the screen some slides taken in the
same way and with the same subject. [Photograph shown.] That is an
instantaneous photograph of a stream of bubbles. You see that the bubbles
form at the nozzle from the very first moment, contrasting in that respect
with the behaviour of jets of water, projected into air. [Fig. 1, Plate I.]
The latter is our next subject. This is the reservoir from which the
water is supplied. It issues from a nozzle of drawn-out glass, and at the
moment of issue it consists of a cylindrical body of water. The cylindrical
444 SOME APPLICATIONS OF PHOTOGRAPHY. [179
form is unstable, however, and the water rapidly breaks up into drops which
succeed one another so rapidly that they can hardly be detected by ordinary
vision. But by means of instantaneous photography the individual drops can
be made evident. I will first project the jet itself on the screen, in order
that you may appreciate the subject which we shall see presently represented
by photography. [Experiment.] Along the first part of its length the jet
of water is continuous. At a certain point it breaks into drops, but you
cannot see them because of their rapidity. If we act on the jet with a
vibrating body, such as a tuning-fork, the breaking into drops occurs still
earlier, the drops are more regular, and assume a curious periodic appearance,
investigated by Savart. I have some photographs of jets of that nature.
Taken as described, they do not differ much in appearance from those
obtained by Chichester Bell, and by Mr Boys. We get what we may regard
as simply shadows of the jet obtained by instantaneous illumination ; so that
these photographs show little more than the outlines of the subject. They
show a little more, on account of the lens-like action of the cylinder and
of the drops. Here we have an instantaneous view of a jet similar to the one
we were looking at just now. [Fig. 2, Plate I.] This is the continuous part ;
it gradually ripples itself as it comes along ; the ripples increase ; then the
contraction becomes a kind of ligament connecting consecutive drops; the
ligament next gives way, and we have the individual drops completely
formed. The small points of light are the result of the lens-like action of the
drops. [Other instantaneous views also shown.]
The pictures can usually be improved by diffusing somewhat the light
of the spark with which they are taken. In front of the ordinary condensing
lens of the magic lantern we slide in a piece of ground glass, slightly oiled,
and we then get better pictures showing more shading. [Photograph shown.]
Here is one done in that way; you would hardly believe it to be water
resolved into drops under the action of a tremor. It looks more like mercury.
You will notice the long ligament trying to break up into drops on its own
account, but not succeeding. [Fig. 3, Plate I.]
There is another, with the ligament extremely prolonged. In this case it
sometimes gathers itself into two drops. [Fig. 4, Plate I.]
[A number of photographs showing slight variations were exhibited.]
The mechanical cause of this breaking into drops is, I need hardly
remind you, the surface tension or capillary force of the liquid surface. The
elongated cylinder is an unstable form, and tends to become alternately
swollen and contracted. In speaking on this subject I have often been
embarrassed for want of an appropriate word to describe the condition in
question. But a few days ago, during a biological discussion, I found that
there is a recognised, if not a very pleasant, word. The cylindrical jet may
PLATE I
Fig. 1.
Fk. 2.
Fig. 3.
Fig. 4.
To face p. 444.
1891] SOME APPLICATTOXS OF PHOTOGRAPHY. 445
be said to become varicose, and the varicosity goes on increasing with time,
until eventually it leads to absolute disruption.
There is another class of unstable jets presenting many points of analogy
with the capillary ones, and vet in many respects quite distinct from them.
I refer to the phenomena of sensitive flames. The flame, however, is not the
essential part of the matter, but rather an indicator of what has happened.
Any jet of fluid playing into a stationary environment is sensitive, and
the most convenient form for our present purpose is a jet of coloured in
nncolonred water. In this case we shall use a solution of permanganate
of potash playing into an atmosphere of other water containing acid and
sulphate of iron., which exercises a decolourising effect on the permanganate,
and so retards the general clouding up of the whole mass bv accumulation
of colour. [Experiment.] Mr Gordon will release the clip, and we shall get
a jet of permanganate playing into the liquid. If everything were perfectly
steady, we might see a line of purple liquid extending to the bottom of the
trough : but in this theatre it is almost impossible to keep anvthing steady.
The instability to which the jet is subject now manifests itself, and we get
a breaking away into clouds something like smoke from chimneys. A heavy
tuning-fork, vibrating at ten to the second, acts upon it with great advantage,
and regularizes the disruption. A little more pressure will increase the
instability, and the jet goes suddenly into confusion, although at first, near
the nozzle, it is pretty regular.
It may now be asked "i What is the jet doing f "" That is just the question
which the instantaneous method enables us to answer. For this purpose the
permanganate which we have used to make the jet visible is not of much
service. It is too transparent to the photographic rays, and so it was
replaced by bichromate of potash. Here the opposite difficulty arises : for
the bichromate is invisible by the yellow light in which the adjustments
have to be made. I was eventually reduced to mining the two materials
together, the one serving to render the jet visible to the eye and the other to
the photographic plate. Here is an instantaneous picture of such a jet as
was before you a moment ago, only [now] under the action of a regular
vibrator. It is sinuous, turning first in one direction and then in the other.
The original cylinder, which is the natural form of the jet as it issues from
the nozzle, curves itself gently as it passes along through the water. It
thus becomes sinuous, and the amount of the sinuosity increases, until in
some cases the consecutive folds come into collision with one another.
[Several photographs of sinuous jets were shown, two of which are reproduced
in Figs. 5, 6, Hate IL]
The comparison of the two classes of jets is of great interest. There is
an analogy as regards the instability, the vibrations caused by disturbance
gradually increasing as the distance from the nozzle increases: but there is
446 SOME APPLICATIONS OF PHOTOGRAPHY. [179
a great difference as to the nature of the deviation from the equilibrium
condition, and as to the kind of force best adapted to bring it about. The
one gives way by becoming varicose; the other by becoming sinuous. The
only forces capable of producing varicosity are symmetrical forces, which act
alike all round. To produce sinuosity, we want exactly the reverse — a force
which acts upon the jet transversely and unsymmetrically.
I will now pass on to another subject for instantaneous photography,
namely, the soap film. Everybody knows that if you blow a soap bubble it
breaks — generally before you wish. The process of breaking is exceedingly
rapid, and difficult to trace by the unaided eye. If we can get a soap film on
this ring, we will project it upon the screen and then break it before our eyes,
so as to enable you to form your own impressions as to the rapidity of the
operation. For some time it has been my ambition to photograph a soap
bubble in the act of breaking. I was prepared for difficulty, believing that
the time occupied was less than the twentieth of a second. But it turns out
to be a great deal less even than that. Accordingly this subject is far more
difficult to deal with than are the jets of water or coloured liquids, which one
can photograph at any moment that the spark happens to come.
There is the film, seen by reflected light. One of the first difficulties we
have to contend with is that it is not easy to break the film exactly when
we wish. We will drop a shot through it. The shot has gone through, as
you see, but it has not broken the film ! and when the film is a thick one,
you may drop a shot through almost any number of times from a moderate
height without producing any effect. You would suppose that the shot in
going through would necessarily make a hole, and end the life of the film.
The shot goes through, however, without making a hole. The operation can
be traced, not very well with a shot, but with a ball of cork stuck on the end
of a pin, and pushed through. A dry shot does not readily break the film ;
and as it was necessary for our purpose to effect the rupture in a well defined
manner, here was a difficulty which we had to overcome. We found, after
a few trials, that we could get over it by wetting the shot with alcohol.
We will try again with dry shot. Three shots have gone through and
nothing has happened. Now we will try one wetted with alcohol, and I
expect it will break the film at once. There ! It has gone !
The apparatus for executing the photography of a breaking soap film will
of necessity be more complicated than before, because we have to time the
spark exactly with the breaking of the film. The device I have used is to
drop two balls simultaneously, so that one should determine the spark and
the other rupture the film. The most obvious plan was to hang iron balls
to two electro-magnets [connected in series], and cause them to drop by
breaking the circuit, so that both were let go at the same moment. The
PLATE II
Fig. 5.
Fig. 6.
To face p. 446.
SOME APPLICATIONS OF PHOTOGRAPHY.
±47
method was not quite a success, however, because there was apt to be a little
hesitation in letting go the balls. So we adopted another plan. The balls
were not held by electro-magnetism but by springs (Fig. 8) pressing laterally,
and these were pulled off by electro-magnets. The proper moment for
putting down the key and so liberating the balls, is indicated by the tap
of the beam of an attracted disc electrometer as it strikes against the upper
stop. One falling ball determines the spark, by filling up most of the
interval between two fixed ones submitted to the necessary electric pressure.
Another ball, or rather shot, wetted with alcohol, is let go at the same
moment, and breaks the film on its passage through. By varying the
distances dropped through, the occurrence of one event may be adjusted
relatively to the other. The spark which passes to the falling ball is,
however, not the one which illuminates the photographic plate. The latter
occurs within the lantern, and forms part of a circuit in connexion with the
outer coatings of the Leyden jars*, the whole arrangement being similar to
DESCMPTIOX OF FIG.
A, B, Electrodes of Wimshorst machine.
C, D, Terminals of interior coatings of Ley-
den jars.
E, F, Balls on inflating supports between
which the discharge is taken.
G, Attracted disc of electrometer.
H, Knife edge.
I, Scale pan.
J, Stops limiting morement of beam.
K. Sparking balls in connexion with exte-
rior coatings of jars. [These exterior
coatings are to be joined by an im-
perfect conductor, such as a tabk.]
L. Lantern condenser.
M. Soap film.
X, Photographic camera.
O, Daniell celL P, Key.
O, Electromagnets. R, Balk.
In practice there were two sets of three jars each.
448 SOME APPLICATIONS OF PHOTOGRAPHY. [179
that adopted by Prof. Lodge in his experiments upon alternative paths of
discharge. Fig. 8 will give a general idea of the disposition of the apparatus.
[Several photographs of breaking films were shown upon the screen ; one
of these is reproduced in Fig. 7, Plate II.*]
This work proved more difficult than I had expected; and the evidence
of our photographs supplies the explanation, namely, that the rupture of the
film is an extraordinarily rapid operation. It was found that the whole
difference between being too early and too late was represented by a displace-
ment of the falling ball through less than a diameter, viz. \ inch nearly.
The drop was about a foot. The speed of the ball would thus be about
100 inches per second ; and therefore the whole difference between being too
soon and too late is represented by -^ second. Success is impossible, unless
the spark can be got to occur within the limits of this short interval.
Prof. Dewar has directed my attention to the fact that Dupre, a good
many years ago, calculated the speed of rupture of a film. We know that
the energy of the film is in proportion to its area. When a film is partially
broken, some of the area is gone, and the corresponding potential energy is
expended in generating the velocity of the thickened edge, which bounds the
still unbroken portion. The speed, then, at which the edge will go depends
upon the thickness of the film. Dupre took a rather extreme case, and
calculated a velocity of 32 metres per second. Here, with a greater thick-
ness, our velocity was, perhaps, 16 yards [say 15 m.] a second, agreeing fairly
well with Dupre's theory.
I now pass on to another subject with which I have lately been engaged,
namely, the connexion between aperture and the definition of optical images.
It has long been known to astronomers and to those who study optics
that the definition of an optical instrument is proportional to the aperture
employed ; but I do not think that the theory is as widely appreciated as it
should be. I do not know whether, in the presence of my colleague, I may
venture to say that I fear the spectroscopists are among the worst sinners
in this respect. They constantly speak of the dispersion of their instruments
as if that by itself could give any idea of the power employed. You may
have a spectroscope of any degree of dispersion, and yet of resolving power
insufficient to separate even the D lines. What is the reason of this ? Why
is it that we cannot get as high a definition as we please with a limited
aperture ? Some people say that the reason why large telescopes are
necessary, is, because it is only by their means that we can get enough light.
That may be in some cases a sufficient reason, but that it is inadequate in
others will be apparent, if we consider the case of the sun. Here we do
* The appearance of the breaking bubble, as seen under instantaneous illumination, was first
described by Marangoni and Stephanelli, Nuovo Cimento, 1873.
SOME APPLICATION'S OF PHOTOGRAPHT. 449
not want more light, but rather are anxious to get rid of a light already
excessive. The principal raison d'etre of large telescopes, is, that without
a large aperture definition is bad. however perfect the lenses may be. In
accordance with the historical development of the science of optics, the
student is told that the lens collects the rays from one point to a focus at
another : but when he has made further advance in the science he finds that
this is not so. The truth is that we are in the habit of regarding this subject
in a distorted manner. The difficulty is not to explain why optical images
are imperfect, no matter how good the lens employed, but rather how it is
that they manage to be as good as they are. In reality the optical image
of even a mathematical point has a considerable extension: light coming
from one point cannot be concentrated into another point by any arrange-
ment. There must be diffusion, and the reason is not hard to see in a
general way. Consider what happens at the mathematical focus, where, if
anywhere, the light should all be concentrated. At that point all the rays
coming from the original radiant point arrive in the same phase. The
different paths of the rays are all rendered optically equal, the greater actual
distance that some of them have to travel being compensated for, in the case
of those which come through the centre, by an optical retardation due to the
substitution of glass for air; so that all the rays arrive at the same time*.
If we take a point not quite at the mathematical focus but near it, it is
obvious that there must be a good deal of light there also. The only reason
for any diminution at the second point lies in the discrepancies of phase
which now occur ; and these can only enter by degrees. Once grant that the
image of a mathematical point is a diffused patch of light, and it follows that
there must be a limit to definition. The images of the components of a close
double point will overlap ; and if the distance between the centres do not
exceed the diameter of the representative patches of light, there can be no
distinct resolution. Now their diameter varies inversely as the aperture:
and thus the resolving power is directly as the aperture.
My object to-night is to show you by actual examples that this is so.
I have prepared a series of photographs of a grating consisting of parallel
copper wires, separated by intervals equal to their own diameter, and such
that the distance from centre to centre is ^ inch [inch = 2"54 cm.]. The
grating was backed by a paraffin lamp and a large condensing lens ; and the
photographs were taken in the usual way? except that the lens employed
was. a telescopic object glass, and was stopped by a screen perforated with
a narrow adjustable slit, parallel to the wires'*-. In each case the exposure
was inversely as the aperture employed. The first [thrown upon the screen],
is a picture done by an aperture of eight hundredths of an inch, and the
* On thk principle wj may readily eakobte the focal lengths of lenses without me of the lav
of sines. See Phil, Mmy. Dee. 1979. [ToL L p. 439.]
t The distance between the grating and the tekaeope lew was 12 ft. 3 in.
450 SOME APPLICATIONS OF PHOTOGRAPHY. [179
definition is tolerably good (Fig. 8). The next (Fig. 9), with- six hundredths,
is rather worse. In the third case (Fig. 10), I think that everyone can see
that the definition is deteriorating; that was done by an aperture of four
hundredths of an inch. The next (Fig. 11) is one done by an aperture of
three hundredths of an inch, and you can see that the lines are getting
washed out. In focusing the plate for this photograph, I saw that the
lines had entirely disappeared, and I was surprised, on developing the plate,
to find them still visible. That was in virtue of the shorter wave-length of
the light operative in photography as compared with vision. In the last
example (Fig. 12), the aperture was only two-and-a-half hundredths of an
inch, and the effect of the contraction has been to wash away the image
altogether, although, so far as ordinary optical imperfections are concerned,
the lens was acting more favourably with the smaller aperture than with the
larger ones*.
This experiment may be easily made with very simple apparatus; and
I have arranged that each one of my audience may be able to repeat it by
means of the piece of gauze and perforated card which have been distributed.
The piece of gauze should be placed against the window so as to be backed
by the sky, or in front of a lamp provided with a ground-glass or opal globe.
You then look at the gauze through the pin-holes. Using the smaller hole,
and gradually drawing back from the gauze, you will find that you lose
definition and ultimately all sight of the wires. That will happen at a
distance of about 4£ feet from the gauze. If, when looking through the
smaller hole, you have just lost the wires, you shift the card so as to bring
the larger hole into operation, you will see the wires again perfectly.
That is one side of the question. However perfect your lens may be, you
cannot get good definition if the aperture is too much restricted. On the
other hand if the aperture is much restricted, then the lens is of no use, and
you will get as good an image without it as with it.
I have not time to deal with this matter as I could wish, but I will
illustrate it by projecting on the screen the image of a piece of gauze as
formed by a narrow aperture parallel to one set of wires. There is no lens
whatever between the gauze and the screen. [Experiment.] There is the
image — if we can dignify it by such a name — of the gauze as formed by
an aperture which is somewhat large. Now, as the aperture is gradually
narrowed, we will trace the effect upon the definition of the wires parallel
to it. The definition is improving; and now.it looks tolerably good. But
I will go on, and you will see that the definition will become bad again.
Now, the aperture has been further narrowed, and the lines are getting
* [1901. The original photographs were exhibited by projection, and are now reproduced for
the first time. In these reproductions the distinction between Figs. (8), (9), (10) is barely
visible. ]
PLATE III
Fig. 8.
Fig.
Fig. 10.
Fig. 11.
Fig. 12.
nn
To face p. 450.
SOME APPLICATIONS OF PHOTOGRAPHY. 451
washed out. Again, a little more, and they are gone. Perhaps you may
think that the explanation lies in the faintness of the light. We cannot
avoid the loss of light which accompanies the contraction of aperture, but to
prove that the result is not so to be explained, I will now put in a lens. This
will bring the other set of wires into view, and prove that there was plenty
of light to enable us to see the first set if the definition had been good
enough. Too small an aperture, then, is as bad as one which is too large;
and if the aperture is sufficiently small, the image is no worse without a lens
than with one.
What, then, is the best size of the aperture ? That is the important
question in dealing with pin-hole photography. It was first considered by
Petzval, of Vienna, and he arrived at the result indicated by the formula,
2r-2=/X, where 2r is the diameter of the aperture, X the wave-length of
light, and / the focal length, or rather simply the distance between the
aperture and the screen upon which the image is formed.
His reasoning, however, though ingenious, is not sound, regarded as an
attempt at an accurate solution of the question. In fact it is only lately
that the mathematical problem of the diffraction of light by a circular hole
has been sufficiently worked out to enable the question to be solved. The
mathematician to whom we owe this achievement is Prof. Lommel. I have
adapted his results to the problem of pin-hole photography. [A series of
curves* were shown, exhibiting to the eye the distribution of illumination
in the images obtainable with various apertures.] The general conclusion
is that the hole may advantageously be enlarged beyond that given by
Petzval's rule. A suitable radius is r = \/(f\).
I will not detain you further than just to show one application of pin-
hole photography on a different scale from usual. The definition improves
as the aperture increases ; but in the absence of a lens the augmented
aperture entails a greatly extended focal length. The limits of an ordinary
portable camera are thus soon passed. The original of the transparency now
to be thrown upon the screen was taken in an ordinary room, carefully
darkened. The aperture (in the shutter) was '07 inch, and the distance of
the 12 x 10 plate from the aperture was 7 feet. The resulting picture of a
group of cedars shows nearly as much detail as could be seen direct from the
place in question.
* Phil. Mag. Feb. 1891. [Vol. in. p. 437.]
180.
ON THE SENSITIVENESS OF THE BRIDGE METHOD IN ITS
APPLICATION TO PERIODIC ELECTRIC CURRENTS.
Fig. 1.
[Proceedings of the Royal Society, XLIX. pp. 203—217, 1891.]
THE most favourable conditions in the ordinary measurement of resistance
have been investigated by Schwendler* and by 0. Heavisidef- It is here
proposed to treat the problem more generally, so as to cover the application
to conductors endowed with self-induction, or combined with condensers. The
receiving instrument may be supposed to be a telephone, which takes the
place of the galvanometer employed in ordinary testing. In the conjugate
"battery" branch a periodic electromotive force of given frequency is the
origin of the currents.
Special attention will be given to the case where the branches are equal
in pairs, e.g., a = c, b = d (Fig. 1). The advan-
tages of this arrangement are important even
in ordinary resistance testing, and in the gene-
ralised application are still more to be insisted
upon. By mere interchange of a and c and
combination of results, the equality of b and d
can be verified independently of the exactitude
of the ratio a : c.
If any element in the combination, for
example a, be a mere resistance, the differ-
ence of potentials at its terminals ( V) is connected with the current, x, by
the relation V= ax. We have, however, to suppose that a is not merely a
* " On the Galvanometric Besistance to be employed in Testing with Wheatstone's Diagram,"
Phil. Mag. Vol. xxxi. p. 364, 1866.
t " On the Best Arrangement of Wheatstone's Bridge for measuring a given Resistance with a
given Galvanometer and Battery," Phil. Mag. Vol. XLV. p. 114, 1873.
ON THE BRIDGE METHOD. 453
resistance or even combination of such. It may include an electromagnet*,
and it may be interrupted by a condenser. So long as the current is strictly
harmonic, proportional to e^, the most general possible relation between V
and x is expressed by V= (a, + ia^)x, where a, and to, are the real and
imaginary parts of a complex coefficient a, and are functions of the frequency
p/2-jr. In the particular case of a simple conductor, endowed with inductance
L, a, represents the resistance, and a, is equal to pL. In general, Oj is positive ;
but CL, may be either positive, as in the above example, or negative. The
latter case arises when a resistance, R, is interrupted by a condenser of
capacity C. Here a^R, a2 = -ljpC. If there be also inductance L,
a, = R, a.2=pL-I{pC.
Since the parts of a* may be either positive or negative, there is nothing
to hinder its evanescence by compensation. In the above combination of an
electromagnet and condenser, compensation occurs when p*LC=\, that is,
when the natural period with terminals connected coincides with the forced
period. The combination is then equivalent to a simple resistance f ; but a
variation of frequency will give rise to a positive or negative a2.
The case of two electromagnets in parallel is treated in my paper on
" Forced Harmonic Oscillations J " ; and other combinations have been dis-
cussed by Mr Heaviside and myself. But the above examples will suffice
to illustrate the principle that the relation of V to x is one of proportionality,
and may be expressed by the single complex symbol a. We fall back at any-
time upon the case of mere resistance by supposing a to be real. In like
manner 6, c, d, e, and f are symbols expressing the electrical properties of the
remaining branches.
In all electrical problems the generalised quantities a, b, &c., combine,
just as they do when they represent simple resistances. Thus, if a, a' be two
complex quantities representing two conductors in series, the corresponding
quantity for the combination is a + a'. Again, if a, a represent two conductors
in parallel, the reciprocal of the resultant is given by addition of the reciprocals
of a, a'. For, if the currents be x and x', corresponding to a difference of
potentials V at the common terminals,
V=ax = a'x,
so that x + x = V(l/a + I/a').
The investigation of the currents in networks of conductors is usually
treated by " KirchhofFs rules," and this procedure may of course be adopted
in the present case to determine the current through the bridge of a Wheat-
stone combination. But it will be more instructive to put the argument in
* An electromagnet here denotes a conductor with sensible inductance. Iron may be present
if the range of magnetisation be small. Phil. Mag. March 1887. [Vol. n. p. 579.]
t Theory of Sound, § 46, Macmillan, 1877.
+ Phil. Mag. May 1886. [Vol. n. p. 475.]
454 ON THE BRIDGE METHOD IN ITS [180
the form applicable to the forced vibrations of all mechanical systems which
oscillate about a configuration of equilibrium.
If p/%7r represent the frequency of the vibration, the coordinates •^•1, i/r2,
i/r3 . . . determining the condition of the system, and the corresponding forces
MS, *P2> ^3-.. are all proportional to eipt, and the coordinates are linear
functions of the forces*. For the present purpose we suppose that all the
forces vanish, except the first and second. Thus •^r1, -\^2 are linear functions of
M*! and W2, and, conversely, ^Pj, W2 may be regarded as linear functions of ^
and i|r2. We may therefore set
(1)
the coefficient of i/r2 in the first equation being identical with that of ^ in the
second by the reciprocal property. The three constants A,B,G are in general
complex quantities, functions of p.
In the application that we have to make of these equations, tylt ^2» ^i. ^2
will represent respectively currents and electromotive forces in the battery
and telephone branches of the combination. The reciprocal property may
then be interpreted as follows :— If ^2 = 0,
and ** = B*-AC*1 (2)
In like manner, if we had supposed ^ = 0, we should have found
-O .T.
showing that the ratio of the current in one branch to an electromotive force
operative in the other is independent of the way in which the parts are
assigned to the two branches.
We have now to determine the constants A, B, C in terms of the electrical
properties of the system. If i/r2 be maintained zero by a suitable force ^a, the
relation between ^ and ^ is ¥, = A^. In our application, A therefore
denotes the (generalised) resistance to an electromotive force in the battery
branch, when the telephone branch is open. This resistance is made up of /,
the resistance in the battery branch, and of that of the conductors a + c, b -f d
combined in parallel. Thus,
In like manner, C = e +
a+b+c+d
* Theory of Sound, Vol. i. § 107.
APPLICATION TO PERIODIC ELECTRIC CURRENTS. 455
To determine B let us consider the force ¥, which must act in e in oider
that the current through it (^r,) may be aero, in spite of the operation of ¥,.
We have V, = B^. The total current ^r, flows partly along the branch a + c
mad partly along 6 + <L The current through a + c is
a+c
and that through 6^-rf is
Hie difference of potentials at the terminals of e, supposed to be interrupted,
is thus
a-f 6-J-c-fd
---- = - i ...............................
a+6+c+d
By (4), (4 ), (7) the relationship of*1,, ¥z to ^r,, ^r, is completely determined.
The problem of the bridge requires the determination of the current ijr*,,
as proportional to ^, when ¥s = 0, that is, when no electromotive force acts
in the bridge itself and the solution is given at once by simple introduction
into (2) of the values A,C,B fiwrn (4), (4'), <7).
If there be an approximate balance, the expression simplifies. For be — ad
its then small, and B* may be neglected relatively to AC in the denominator
of (2). Urns, as a sufficient approximation in this case, we have
ad-bc
}«+^f^j{/<
in agreement with the equation used by Mr Heaviside for simple resistances.
The following interpretation of the process leads very simply to the
approximate form (8), and may be acceptable to readers less familiar with
the general method. Let us first inquire what E.M.F. is necessary in the
telephone branch to stop the current through it If such a force acts, the
conditions are, externally, the same as if the branch were open, and the
current +t in the battery branch due to an E.M.F. equal to ¥, in that branch
is *ijA, where A is written for brevity as representing the right-hand
member of (4). The difference of potential at the terminals of <• , still sup-
posed to be open, is found at once when ^r, is known. It is equal to
456 ON THE BRIDGE METHOD IN ITS [180
where B is defined by (7). In terms of ^ the difference of potentials is thus
BW^/A. If e be now closed, the same fraction expresses the E.M.F. necessary
in e in order to prevent the generation of a current in that branch.
The case that we have to deal with is when ^ acts in f, and there is no
E.M.F. in e. We are at liberty, however, to suppose that two opposite forces,
each of magnitude BWl/A, act in e. One of these, as we have seen, acting in
conjunction with ^ in /, gives no current in e; so that, since electromotive
forces act independently of one another, the actual current in e, closed without
internal E.M.F., is simply that due to the other component. The question is
thus reduced to the determination of the current in e due to a given E.M.F. in
that branch.
So far the argument is rigorous ; but we will now suppose that we have to
deal with an approximate balance. In this case an E.M-F. in e gives rise to
very little current in f, and in calculating the current in e we may suppose f
to be broken. The total resistance to the force in e is then given simply by
C of equation (4'), and the approximate value for -^» is derived by dividing
by C, as we found in (8).
A continued application of the foregoing process gives ^/^i in the form
of an infinite geometric series :
This is the rigorous solution already found ; but the first term of the series
suffices for practical purposes.
The form of (8) enables us at once to compare the effects of increments of
resistance and inductance in disturbing a balance. For let ad — be, and then
change d to d + d' where d' = d\ + id'2. The value of i/^/^i is proportional to
d', and the amplitude of the vibratory current in the bridge is proportional
to Mod d', that is, to ^(d'j* + d'<?). Thus d\, d'2 are equally efficacious when
numerically equal.
The next application that we shall make of (8) is to the generalised form
of Schwendler's problem. When all else is given, how should the telephone,
or other receiving instrument, be wound in order to get the greatest effect ?
If by separation of real and imaginary parts we set
-«+••*.
the factor in the denominator of (6) with which we are concerned becomes
and the square of the modulus is given by
Mod2 = (X + r,)2 + 02 + r2)2 ...................... (10)
1891] APPLICATION TO PERIODIC ELECTRIC CURRENTS. 457
In this equation e, , r, are essentially positive, while «,, r, may be either
positive or negative. If el and ez are both at disposal, the minimum of (10),
corresponding to the maximum current, is found by making
e, = 0, ^ = -rs ......................... (11)
But this is not the practical question. As in the case of simple resist-
ances, what we have to aim at is not to render the current in the bridge
a maximum, but rather the effect of the current. Whether the receiving
instrument be a galvanometer or a telephone, we cannot in practice reduce its
resistance to zero without at the same time nullifying the effect desired. We
must rather regard the space available for the windings as given, and merely
inquire how it may best be utilised. Now the effect required to be exalted
is, cceteris paribus, proportional to the number of windings (w); and, if the
space occupied by insulation be proportional to that occupied by copper, the
resistance varies as m*. So also does the inductance ; and accordingly, if the
instrument be connected to the bridge by leads sensibly devoid of resistance
and inductance,
el + ie. = m*(el + ie2), ..................... .--(12)
where e,, e2 are independent of in. The quantity whose modulus is to be made
a minimum by variation of m is thus
gt + i'e. + r, + t'rg _ rt -f m'
m
and we have
\t ^ _ fo + ro'fi)' + fa + m2
-* + 2 (r,^ + rses) + (e,s
This is a minimum by variation of m when
(13)
We may express this result by saying that to get the best effect the instru-
ment must be so wound that its impedance is equal to that of the compound
conductor ri + tY2. If for any reason the inductances can be omitted from
consideration, then the resistance of the instrument is to be made equal to r, ,
in accordance with Schwendler s rule.
The case of the " batter}7 " branch may often be treated in like manner.
As Mr Heaviside has shown, if a number of cells are available for ordinary
resistance testing, they should be combined, so that their resistance is equal
to that (*,) of the corresponding combination of wires in parallel. Periodic
currents may be conceived to arise from the rotation of a coil in a magnetic
field of given strength. If the space occupied by the windings of the coil be
458 ON THE BRIDGE METHOD IN ITS [180
supposed to be given, their number m will be determined by the condition of
equal impedances. Thus, if
(a_tcX6 + rf,=
a + b + c + d
Mod (/ + if,) = Mod (s, + is2), (15)
in analogy with (13).
The above is the solution of the problem, if the coils of the sending and
receiving instruments represent the whole of their respective branches, and
are limited to occupy given spaces. The inductances and resistances cannot
then be varied independently. But there would often be no difficulty in
escaping from this limitation. The inclusion of additional resistance, external
to the instrument, can only do harm ; but the case is otherwise with induct-
ance, positive or negative. If the inductance of the instrument added to r2.
or to s2, be positive, the total inductance may be reduced to zero by the
insertion of a suitable condenser, and this without material increase of
resistance. If the inductance be already negative, the remedy is not so
easily carried out; but, theoretically, it is possible to add the necessary
inductance without sensible increase of resistance. The greater the frequency
of vibration, the more feasible does this course become. We may, therefore,
without much violence, suppose that the inductances of two branches can be
reduced to zero without additional resistance. Thus,
0; ..................... (16)
and the condition of maximum efficiency of the transmitting and receiving
coils is then given by Schwendler's rule,
* = rlf / = * ......................... (17)
These suppositions form a reasonable basis for further investigation ; but
conclusions founded upon them will be subject to re-examination, especially in
extreme cases. We may also now introduce the promised simplification,
a = c, b = d, ........................... (18)
in accordance with which (8) becomes
_._
^ 46 {e + $(a + 6)} {/+ 2a6/(a -|- 6)} '
Also rI + ir8*=§(a + 6) = 4(a1 + 6I) + ^(as + &,,) ............. (20)
9 (<?L± ~ a*>*).± ?* ± i + a
'(Oi + btf + fa + Hr
. (PI + 6,) (aA + o16a) - (a, + 62) (a^! - a.262)
1891] APPLICATION TO PERIODIC ELECTRIC CURRENTS. 459
It may be well to examine, first, the consequences of (19), in the case of
simple resistances. Here
rs = 0; ................. .(SI)
*» = <>. .................. (23)
In accordance with the plan proposed, we are to make «. = 0, /^ = 0*:
r. fi — *,. Our equation then becomes
Here Oj is still at disposal, and we see that according to (24) it ought to
be diminished without limit. This conclusion does not harmonize with one
obtained by Mr Heavisidet. It must be observed, however, that a, = 0 is
unpractical, involving, as it does, *, = 0, /j = 0. Even according to (24) there
is little to be gained by diminishing a, below, say, ^6,. In this case
ai=i&,, ^r.-fftk, /, = *, = §&, .......... (25)
Such an arrangement as (25) may be recommended for practical use.
When 6, is large, there may be advantage in taking a, relatively smaller
than in the above example. In such cases we approach the limiting condition
of things., and have approximately
*, = »•, = *&„ /, = «k = 2olf ..................... (26)
And the smallness of /j in comparison with 6, may sometimes be a con-
venience.
The next remark that has to be made is that, even when the conductors,
b and d, to be compared are endowed with sensible inductances (positive or
negativeX the problem may still, theoretically, be brought under the above
head. Suppose, for example, that 6, d represent nearly equal elect romagneta.
Their inductances may be compensated by the introduction (in series) of
suitable equal condensers into these branches, so that 6 and d are reduced to
6, and dl. If then we assume a to be a simple resistance (a, = 0), the solution
is as before. Two objections may here be raised. First, on the theoretical
side it has not been proved to be advantageous to assume a. = 0: and,
secondly, the introduction of extraneous condensers*, even with interchange,
* These conditions require no attention in galTanometrie testing with steady currents, being
by F=0, independently of the nature of the i
t L*. cit. p. 130, "In conclusion, if, to ineasi
460 ON THE BRIDGE METHOD IN ITS [180
into the branches to be accurately compared may be a complication unfavour-
able to success.
We will now resume the consideration of (19), supposing that
e = e, + ie2 = i\ - ir.2, /=/ + if2 = s}- is.2, ......... (28)
rl} r2, «„ s2 being given by (20), (21). Thus,
and the question before us is how to make the modulus of the second fraction
on the right a maximum by variation of a. In the denominator of this
fraction 1\ and sl are real, and the modulus of b is \/(V + &22). For the
numerator we have
a b
!
,7T722
Also from the definition of ,9
,*> ,2_
~
,0 that ^ __ W+o.')(V + V) ( 01 _A_12
*,2 + «,f (a, + 6,)2 + (
Thu« Mod
0(1
and this is to be made a minimum by variation of a1} a2-
We shall show presently that (30) can be reduced to zero; but for the
moment we will so far limit the generality of a,, a2 as to suppose that a1 = a-b1,
a2 = #&2> x being real and positive.
(30) then reduces to £V(1 +«) 5 and by (29)
Accordingly, the maximum sensitiveness cannot be attained until x is
reduced to zero, so that al} a» vanish. (31) may be regarded as a generalised
form of (24), free from the limitation that 6» = 0, provided a., be so taken that
We will now suppose in (30) that c^ and 0% are both small, and in the first
instance that bt is finite. We have
1891] APPLICATION TO PERIODIC ELECTRIC CURRENTS. 461
and this reduces ultimately to its first term, depending upon the ratio only of
<*! and 03. The expression vanishes if aa : a., be small enough, so that (30) can
certainly be thus reduced to zero. It is remarkable that the expression for
the sensitiveness shotild be capable of becoming infinite by suitable choice
of a2. If we first suppose that a, is absolutely zero, and afterwards that eij
diminishes without limit, the ultimate value of (32) is ^6, \/(61* + 6./), in place
of zero.
From the practical point of view, these conclusions from our equations
are not particularly satisfactory. We began with certain proposals which,
in ordinary cases, could be carried out; but in the end we are directed to
apply them to an extreme and impossible state of things. We have found,
however, in what direction we must tend in the search for sensitiveness ; and
useful information may be gathered from (32). In practice ctj could not be
reduced below a certain point. The question may then be asked, what is the
best value of a.,, when Oj is given ? From (32) we find at once that
(32) then becoming
*W(«i&i) .................................. (34)
In this case from (29)
<35>
independent of b...
If we suppose in (32) that a,2 = 0, we have
To take a numerical example, let 62 = 0: and suppose o1
according to (33), a, = ± T356j. Also by (20), (21),
The corresponding minimum value of (32), equal to (34), is 6,2/V(10).
But with this value of a^ the gain by allowing a^ to be finite is not great.
If a, = 0,
and the value of (32), equal to (36), is ^6,2.
We see from (36) that when a2 = 0 there is little to be gained by further
reduction of c^. But when a3 is suitably chosen the gain may be worth
462 ON THE BRIDGE METHOD IN ITS [180
having. Thus, in (34), if o1 = T£TF&1, we have -^6,*. Corresponding to this
o2= + T1561 nearly, and
These are not unreasonable proportions, and we see that the use of a?
may be advantageous, even when the subject of measurement is a mere
resistance. It will be remarked too that, except as regards ea, /2, the sign
of a, is immaterial.
When the branches b, d consist of electromagnets, and still more when
they consist of condensers, 6j may be very small. If we suppose it to be zero,
(30) becomes
'* ,07,
Corresponding to this from (20), (21),
* = *«., * = -*(«, + *,), ..................... (38)
._ 2a1622
._ 1a 222 ,
'
From (37) we see that the increase of a, is favourable, especially if the
sign be the same as of 62. Even if o^ = 0, (37) now assuming the form
.(40)
can be reduced to zero by taking aa small enough. But of course (37) ceases
to be applicable unless 6, be small relatively to a,. In correspondence
with (40),
0, = ^, *, = -ik; ..................... (41)
As an example of (37), suppose
o^i&g, a, = 46..
Then (37) = ^nearly.
Also approximately
If 62 represent the stiflEness of a condenser, yj must be a positive
inductance, and its magnitude, relatively to /,, would probably constitute
a difficulty.
1891] APPLICATIOX TO PERIODIC ELECTRIC CURRE3TIS. 463
As an example, with ay equal to TKTO, take
Then (37) = (40) = ^bj nearly,
So far as the general theory is concerned, it is a matter of indifference
whether the indicating instrument be in the branch e, or in f. The latter
corresponds to the connexions in De Sauty's method of testing condensers bv
means of the galvanometer. In practice, more space would probably be
available for the coils of a transmitting instrument than of the receiving
instrument, at least, if the latter be a telephone: and this would tell in
favour of choosing that branch for the transmitter which should have the
larger time constant (L/R\
To get an idea of the relative capacities, resistances, and inductances
involved, we must assume a particular pitch. A frequency suitable for
telephonic experiments is 1000 per second, for which p = 2000?r. Thus, if
the value of a, for a condenser of capacity C, and for an inductance L, and
that of a, for a resistance R, are all numerically equal,
1
£ = 20007^ = 200^^.
If R be 1 ohm, equal to 10» C.G.S., the corresponding capacity is T6 x 10~u
C.G.S., equal to 160 microfarads, and the corresponding inductance is 1'6 x 10s
C.G.S. Again, if C be one microfarad, equal to 10~u C.G.S., R is 160 ohms, and
1, i§ 2-5 x 10T cm.
In the preceding calculations e and f are supposed to be adjusted to the
values most favourable to the effect in the receiving instrument. A question,
which arises quite as often in practice, is how to make the best of given
instruments. The full answer is necessarily somewhat complicated : for there
could be no objection to the insertion of a condenser, for example, if the
sensitiveness could be improved thereby. In what follows, however, the
lniMiiillili^ and receiving branches will be supposed to be fully given, so
that « and f are known complex quantities ; and the only question to be
considered is as to the most suitable value of a, assumed to be equal to c.
For this purpose the modulus of the second fraction on the right in (19) is
to be a maximum, or that of
(43)
is to be a minimum, by variation of a. The problem thus arising of deter-
mining the minimum modulus of a function of a complex quantity may be
treated generally.
464 ON THE BRIDGE METHOD. [180
Let F (2) = F (x + iy) = <f> (x, y) + i^r (x, y),
and let it be required to find when the modulus2 of F (z), viz., <£2 + \Jr'2, is a
minimum by variation of IF, y. We have
And in general
d<j>/dx = d^/dy, d<f>/dy = - d-tyjdx ............. (45 )
In order that (44), (45) may both obtain, we must have either <£2 + -\Jr- — 0,
or else
<ty/<fc» = 0, d<j>jdy = Q, d^/dx = 0. d^rfdy = Q.
The latter conditions are equivalent to
F'00 = 0 .................................. (46)
For example, let
F(z)^(z + a.)(^ + llz}) ........................ (47)
where a, ft are complex constants.
The application of (46) gives
2* = «IP, ................................. (48)
and F(z) = {l + V(«/3)}2 ............................ (49)
We see then that the modulus of (43) will be a minimum, when
and in taking the square root the ambiguity must be so determined as to
make the real part of a positive.
Equation (50) coincides with that obtained by Mr Heaviside for the case
where all the quantities are real.
181.
ON VAN DER WAALS'S TREATMENT OF LAPLACE'S PRESSURE
IN THE VIRIAL EQUATION: LETTERS TO PROF. TAIT.
[Nature, XLIV. pp. 499, 597, 1891.]
Ix Part IT. of jour "Foundations of the Kinetic Theoiy of Gases*,"
you take exception to the manner in which Tan der Waals has introduced
Laplace's intrinsic pressure K into the equation of virial. " I do not profess
to be able folly to comprehend the arguments by which Van der Waals
attempts to justify the mode in which he obtains the above equation. Their
nature is somewhat as follows : — He repeats a good deal of Laplace's capillary
work, in which the existence of a large, but unknown, internal molecular
pressure is established, entirely from a statical point of view. He then gives
reasons (which seem, on the whole, satisfactory from this point of view) for
assuming that the magnitude of this force is as the square of the density
of the aggregate of particles considered. But his justification of the intro-
duction of the term a/v* into an account already closed, as it were, escapes
me. He seems to treat the surface-skin of the group of particles as if it
were an additional bounding-surface, exerting an additional and enormous
pressure on the contents. Even were this justifiable, nothing could justify
the multiplying of this term by (v — /3) instead of by t? alone. But the whole
procedure is erroneous. If one begins with the virial equation, one must
keep strictly to the assumptions made in obtaining it, and consequently
everything connected with molecular force, whether of attraction or of elastic
resistance, must be extracted from the term 2 (Rr)."
With the last sentence all will agree ; but it seemed to me when I first
read Van der Waals's essay that his treatment of Laplace's pressure was
satisfactory, and on reperusal it still appears to me to conform to the require-
ments above laid down. As the point is of importance, it may be well to
* Ed. Trans. VoL xxxn. Part 2, p. 261.
B. in. 30
466 ON VAN DER WAALS'.S TREATMENT OF [181
examine it somewhat closely. The question is as to the effect in the virial
equation of a mutual attraction between the parts of the fluid, whose range
is small compared with the dimensions of bodies, but large in comparison
with molecular distances.
The problem thus presented may be attacked in two ways. The first,
to which I will recur, is that followed by Van der Waals ; but the second is
more immediately connected with that form of the equation which you had
in view in the passage above quoted.
In the notation of Van der Waals (equation 8)
£2™F2= |2//> - ^Rrcos (R, r),
where V denotes the velocity of a particle m, which is situated at a distance r
from the origin, and is acted upon by a force R, while (R, r) denotes the
angle between the directions of R and r. The intermediate term is to be
omitted if R be the total force acting upon m. It represents the effect of such
forces, f, as act mutually between two particles at distances from one another
equal to p. In the summation the force between two particles is to be
reckoned once only, and the forces accounted for in the second term are,
of course, to be excluded in the third term.
In the present application we will suppose all the mutual forces accounted
for in the second term, and that the only external forces operative are due to
the pressure of the containing vessel. No one disputes that the effect of the
external pressure is given by
- %2Rr cos (R, r} = %pv ;
so that
&mV* = §pv + ^p<l>(pl
if with Laplace we represent by <j>(p) the force between two particles at
distance p. The last term is now easily reckoned upon Laplace's principles.
For one particle in the interior we have
and this, as Laplace showed*, is equal to 3K. The second summation over the
volume gives 3Kv, but this must be halved. Otherwise each force would be
reckoned twice. Hence
= f jro + f Kv = $v(p + K),
showing that the effect of such forces as Laplace supposed to operate is
represented by the addition to p, the pressure exerted by the walls of the
vessel, of the intrinsic pressure K. In the above process the particles situated
near the surface are legitimately neglected in comparison with those in the
interior.
* See also Phil. Mag. October 1890, p. 292. [Vol. in. p. 403.]
1891] LAPLACE'S PRESSURE ix THE VIRIAL EQUATION-. 467
Van der Waals's own process starts from the original form of virial
equation —
where R now refers to the whole force operative upon any particle; and it
appears to me equally legitimate. For all particles in the interior of the
fluid R vanishes in virtue of the symmetry, so that the reckoning is limited
to a surface stratum whose thickness is equal to the range of the forces. Upon
this stratum act normally hoth the pressure of the vessel and the attraction
of the interior fluid. The integrated effect of the latter throughout the
stratum is equal to the intrinsic pressure ; and, on account of thinness of the
stratum, it enters into the equations in precisely the same way as the external
pressure exerted by the vessel. The effect of Laplace's forces is thus repre-
sented by adding K to p, in accordance with the assertion of Van der Waals.
I am in hopes that, upon reconsideration, you will be able to admit that
this conclusion is correct Otherwise, I shall wish to hear more fully the
nature of your objection, as the matter is of such importance that it ought
not longer to remain in doubt.
Sept. 7.
I gather from your letter of September 28 (Nature, October 8, p. 546)
that you admit the correctness of Van der Waals's deduction from the virial
equation (i) when the particles are infinitely small, in which case
a representing a cohesive force, whose range is great in comparison with
molecular distances: and (ii) when, in the absence of a cohesive force, the
volume of the particles is small in comparison with the total volume v, in
which case the virial of the repulsive forces at impact gives
j>(«-6)=i2mF'. (2)
For hard spherical masses, the value of 6 is four times the total volume of
the sphere. But you ask, "How can the factor (r — b).'v, which Van der
Waals introduces on the left (in the first case) in consequence of the finite
diameters of the particles, be justifiably applied to the term in K (or or1)
as well as to that in p ? "
In my first letter I desired to avoid the complication entailed by the
consideration of the finite size of the particles: but it appears to me that
the argument there given (after Van der Waals) suffices to answer your
question. For, if the cohesive force be of the character supposed, it exercises
no influence upon any particle in the interior, and is completely accounted for
30—2
468 LAPLACE'S PRESSURE IN THE VIRIAL EQUATION. [181
by the addition to p of afv2. In so far, therefore, as (2) is correct when there
is no cohesive force, the effect of such is properly represented by
(3)
in which b is to be multiplied by a/V, as well as by p.
October 13.
[1901. Prof. Tait maintained his opinion. In a reply to the second of
the above letters, dated Oct. 17, 1891 (Nature, XLIV. p. 628), he writes:—
" On the question of the introduction by Van der Waals of the factor
(v — b)/ v, whether or not it is applied alike to p and to K, I regret to find
that our views must continue to differ. For it appears to me that when once
the various terms of the virial equation have been correctly extracted from
the expression S (Rr), we have no right to modify any of them. There seems
therefore to be no doubt whatever that the procedure in Van der Waals's
sixth chapter is entirely wrong in principle: — except in so far as (in the
German version) he borrows some correct expressions from Lorentz. The
meanings of v and p, in the term pv of the virial equation, are (from the very
beginning of the inquiry) definitely assigned as total volume and external
pressure : — so that this term cannot in any way be altered. No more can the
term Smus/3, Or the ratio of these two terms...."]
182.
OX THE VIRIAL OF A SYSTEM OF HARD COLLIDING BODIES.
[Nature, XLV. pp. 80—82, 1891.]
A RECENT correspondence [see Vol. in. p. 465] has led me to examine the
manner in which various authors have treated the influence of the finite size
of molecules in the virial equation, and I should like to lay a few remarks
upon the subject before the readers of Nature.
To fix the ideas, we may begin by supposing that the molecules are equal
hard elastic spheres, which exert no force upon one another except at the
instant of collision. By calling the molecules hard, it is implied that the
collisions are instantaneous, and it follows that at any moment the potential
energy of the system is negligible in comparison with the kinetic energy.
If the volume of the molecules be very small in comparison with the
space they occupy, the virial of the impulsive forces may be neglected, and
the equation may be written
pv = ^mVa, (1)
where p is the pressure exerted upon the walls of the enclosure, v the volume,
TO the mass, and V the velocity of a molecule.
In his essay of 1873 Van der Waals took approximate account of the finite
size of the molecules, using a peculiar process to which exception has been
taken by Maxwell and other subsequent writers. It must be said, however,
that this method has not been proved to be illegitimate, and that at any rate
it led Van der Waals to the correct conclusion —
p(v-b) = &mV>, (2)
in which 6 denotes four times the total volume of the spheres. In calling
(2) correct, I have regard to its character as an approximation, which was
sufficiently indicated by Van der Waals in the original investigation, though
perhaps a little overlooked in some of the applications.
470 ON THE VIRIAL OF A SYSTEM [182
In his (upon the whole highly appreciative) review of Van der Waals's
essay, Maxwell (Nature, Vol. x. p. 477, 1874 ; Scientific Papers, Vol. 11. p. 407)
comments unfavourably upon the above equation, remarking that in the
virial equation v is the volume of the vessel and is not subject to correction*.
" The effect of the repulsion of the molecules causing them to act like elastic
spheres is therefore to be found by calculating the virial of this repulsion."
As the result of the calculation he gives
(3)
where <r is the density of the molecules, and p the mean density of the
medium, so that /j/<r = &/4v. If we expand the logarithm in (3), we obtain
as the approximate expression, Avhen p/<r is small
pv = ^mV^(l+4>b/v), ........................... (4)
or, as equally approximate,
^(v-46) = iSwF2, ........................... (5)
which does not agree with (2).
The details of the calculation of (3) have not been published, but there
can be no doubt that the equation itself is erroneous. In his paper of 1881
(Wied. Ann. xu. p. 127), Lorentz, adopting Maxwell's suggestion, investigated
afresh the virial of the impulsive forces, and arrived at a conclusion which, to
the order of approximation in question, is identical with (2). A like result
has been obtained by Prof. Tait (Edin. Trans, xxxin. p. 90, 1886).
It appears that, while the method has been improved, no one has
succeeded in carrying the approximation beyond the point already attained
by Van der Waals in 1873. But a suggestion of great importance is con-
tained in Maxwell's equation (3), numerically erroneous though it certainly is.
For, apart from all details, it is there implied that the virial of the impacts
is represented by |S??iF2, multiplied by some function of p/<r, so that, if the
volume be maintained constant, the pressure as a function of V is proportional
to 2wF2. The truth of this proposition is evident, because we may suppose
the velocities of all the spheres altered in any constant ratio, without altering
the motion in any respect except the scale of time, and then the pressure
will necessarily be altered in the square of that ratio.
It will be interesting to inquire how far this conclusion is limited to
the suppositions laid down at the commencement. It is necessary that the
collisions be instantaneous in relation, of course, to the free time. Otherwise,
the similarity of the motion could not be preserved, the duration of a collision,
for example, bearing a variable ratio to the free time. On the same ground,
* In connexion with this it may be worth notice that for motion in one dimension the form
(2) is exact.
1891] OF HARD COLLIDING BODIES. 471
vibrations within a molecule are not admissible. On the other hand, the
limitation to the spherical form is unnecessary, and the theorem remains true
whatever be the shape of the colliding bodies. Again, it is not necessarv
that the shapes and sizes of the bodies be the same, so that application may
be made to mixtures,
In the theory of gases 2 in Vs is proportional to the absolute temperature ;
and whatever doubts may be felt in the general theory can scarcely apply
here, where the potential energy does not come into question. So iar, then,
as a gas may be compared to our colliding bodies, the relation between
pressure, volume, and temperature is
/»=**>(•), ................................. (6)
where $ (t?) is some function of the volume. When r is large, the first
approximation to the form of 0 is
In the case of spheres, the second approximation is
A Ab
where 6 is four times the volume of the spheres,
Thus tar we have supposed that there are no forces between the budies
but the impulses on collision. Many and various phenomena require us to
attribute to actual molecules an attractive force operative to much greater
distances than the forces of collision, and the simplest supposition is a cohesive
force such as was imagined by Young and Laplace to explain capillarity.
We are thus led to examine the effect of forces whose range, though small
in comparison with the dimensions of sensible bodies, is large in comparison
with molecular distances. In the extreme case, the influence of the discon-
tinuous distribution of the attractive centres disappears, and the problem
may be treated by the methods of Laplace. The modification then required
in the virial equation is simply to add to p* a term inversely proportional
to o3, as was proved by Van der Waals; so that (6) becomes
p = T<f>(v)-air*. .............................. (7)
According to (7) the relation between pressure and temperature is linear —
a law verified by comparison with observations by Van der Waals, and more
recently and extensively by Ramsay and Young. It is not probable, however,
that it is more than an approximation. To such cases as the behaviour of
water in the neighbourhood of the freezing-point it is obviously inapplicable.
* It dins appears that, contrary to the assertion of Maxwell, p it subject to correction. It is
pretty clear that he had in view an attraction of much smaller range than that considered bj
Van der Waals.
472 ON THE VIRIAL OF A SYSTEM OF HARD COLLIDING BODIES. [182
In their discussions, Ramsay and Young employ the more general form
p = T<l>(v) + x(v)-> (8)
and the question arises, whether we can specify any generalization of the
theoretical conditions which shall correspond to the substitution of % (v)
for av~\ It would seem that, as long as the only forces in operation are of
the kinds, impulsive and cohesive, above defined, the result is expressed
by (7) ; and that if we attempt to include forces of an intermediate character,
such as may very probably exist in real liquids, and must certainly exist in
solids, we travel beyond the field of (8) as well as of (7). It may be remarked
that the equation suggested by Clausius, as an improvement on that of
Van der Waals, is not included in (8).
Returning to the suppositions upon which (7) was grounded, we see that,
if the bodies be all of one shape, e.g. spherical, the formula contains only two
constants— one determining the size of the bodies, and the second the intensity
of the cohesive force ; for the mean kinetic energy is supposed to represent
the temperature in all cases. From this follows the theorem of Van der
Waals respecting the identity of the equation for various substances, provided
pressure, temperature, and volume be expressed as fractions of the critical
pressure, temperature, and volume respectively. If, however, the shape of the
bodies vary in different cases, no such conclusion can be drawn, except as
a rough approximation applicable to large volumes.
183.
DYNAMICAL PROBLEMS IX ILLUSTRATION OF THE THEORY
OF GASES.
[Phil, Mag. **xn. pp. 424—445, 1891.]
Introduction.
THE investigations, of which a part is here presented, had their origin
in a conviction that the present rather unsatisfactory position of the Theory
of Gases is due in some degree to a want of preparation in the mind of
readers, who are confronted suddenly with ideas and processes of no ordinarv
difficulty. For myself, at any rate, I may confess that I have found great
advantage from a more gradual method of attack, in which effort is concen-
trated upon one obstacle at a time. In order to bring out fundamental
statistical questions, unencumbered with other difficulties, the motion is here
limited to one dimension, and in addition one set of impinging bodies is
supposed to be very small relatively to the other. The simplification thus
obtained in some directions allows interesting extensions to be made in others.
Thus we shall be able to follow the whole process by which the steady state
is attained, when heavy masses originally at rest are subjected to bombard-
ment by projectiles fired upon them indifferently from both sides. The case
of pendulums, or masses moored to fixed points by elastic attachments, is also
considered, and the stationary state attained under a one-sided or a two-sided
bombardment is directly calculated
Collision Formula!.
If M', ^ be the velocities before collision, », v after collision, of- two masses
-P. Q, we have by the equation of energy
p(«*-tf)+Q('!'-o-Q. (i)
and by the equation of momentum,
F) = 0. (2)
474 DYNAMICAL PROBLEMS IN [183
From (1) and (2)
11 -f u = v' + v, i (3)
or, as it may be written,
u' - v =v — u,
signifying that the relative velocity of the two masses is reversed by the
collision. From (2) and (3),
As is evident from (1) and (2), we may in (4), if we please, interchange the
dashed and undashed letters. Thus from the first of (4),
In the application which we are about to make, P will denote a relatively
large mass, and Q will denote the relatively small mass of what for the sake of
distinction we will call a projectile. All the projectiles are equal, and in the
first instance will be supposed to move in the two directions with a given
great velocity. After collision with a P the projectile rebounds and dis-
appears from the field of view. Since in the present problem we have
nothing to do with the velocity of rebound, it will be convenient to devote
the undashed letter v to mean the given initial velocity of a projectile.
Writing also q to denote the small ratio Q : P, we have
If u and v be supposed positive, this represents the case of what Ave may call
a favourable collision, in which the velocity of the heavy mass is increased.
If the impact of the projectile be in the opposite direction, the velocity a",
which becomes u after the collision, is given by
(7)
The symbol v thus denotes the velocity of a projectile without regard to sign,
and (7) represents the result of an unfavourable collision.
Permanent State of Free Masses under Bombardment.
The first problem that we shall attack relates to the ultimate effect upon
a mass P of the bombardment of projectiles striking with velocity v, and
moving indifferently in the two directions. It is evident of course that the
ultimate state of a particular mass is indefinite, and that a definite result can
relate only to probability or statistics. The statistical method of expression
being the more convenient, we will suppose that a very large number of
1891] ILLUSTRATION OF THE THEORY OF GASES. 475
masses are undergoing bombardment independently, and inquire what we are
to expect as the ultimate distribution of velocity among them. If the number
of masses for which the velocity lies between u and tt + du be denoted by
f(u)du, the problem before us is the determination of the form of f(u).
The number of masses, whose velocities lie between u and u + du, which
undergo collision in a given small interval of time, is proportional in the first
place to the number of the masses in question, that is to /"(") du. and in the
second place to the relative velocity of the masses and of the projectiles. In
all the cases which we shall have to consider v is greater than u, so that the
chance of a favourable collision is always proportional to v — u, and that of an
unfavourable collision to v+u. It is assumed that the chances of collision
depend upon u in no other than the above specified ways. The number of
masses whose velocities in a given small interval of time are passing, as the
result of favourable collisions, from below u to above u. is thus proportional to
where «' is defined by (6): and in like manner the number which pass in
the same time from above u to below u, in consequence of unfavourable
collisions, is
........................... (9)
u" being defined by (7). In the steady state as many must pass one way as
the other, and hence the expressions (8) and (9) are to be equated. The
result may be written in the form
(10)
Now, if q be small enough, one collision makes very little impression upon ti :
and the range of integration in (10) is narrow. We may therefore expand
the function / by Taylor s theorem :
so that
(") + cubes of q- "(11)
In the present problem r, = r; bat it will be convenient at this stage to maintain the
476 DYNAMICAL PROBLEMS IN [183
Also fwf(w) dw = f {(w — u) + u] f(w) dw
= \(w - itff(u) + i (w - iiff (u) + ......
+ uf(u) {wf(u) + i(w - u)V» + ...... } ;
so that
/•«"
I wj (w) dw = uf (u) . («" — ur)
J u'
+ {*/(«) + *«/(«)} {(*"- ttf- («'- itf} + ......
= r~ */<«) + (1?^ (/(«) + u/' («)} + cubes of ?• • • -(12>
As far as <f inclusive (10) thus becomes
Or M/(%) {(1 - gr) Wl + (1 + gr) t>} + yf (u) [v,Vz + U' (v, + 2t>)} = 0.
If vl = v, q disappears from the first term as it stands, and will do so in any
case in the limit when it is made infinitely small. Moreover, in the second
term u2 is to be neglected in comparison with v2. We thus obtain
uf(u){l + v/v,}+qv2f'(u) = 0 ..................... (13)
as the differential equation applicable to the determination of f(u) when q is
infinitely small. The integral is
qv- log/(w) + ^ (1 -f v1/v) it? = constant,
or
f(u) = Ae-hu\ ................................. (14)
where
or, if vl = v,
(16)
The ultimate distribution of velocities among the masses is thus a function of
the energy of the projectiles and not otherwise of their common mass and
velocity. The ultimate state is of course also independent of the number of
the projectiles.
, The form of/ is that found by Maxwell. To estimate the mean value
of u* we must divide
[ °V/(w)du by [ Xf(u)du.
J — 00 J — OC
Now
ftfg-uW du = -$qv2 [ue-*1** - je-^i* da},
so that
r + x r + ao
I u2e-u2"^du=^qv* I e~l(1''^ du.
1891] ILLUSTRATION OF THE THEORY OF GASES. 477
The ratio in question is thus ^qtf, showing that the mean kinetic energy of a
mass is one half that of a projectile, deviating from the law of equal energies
first (1845) laid down by Waterston. We must remember, however, that we
have thus far supposed the velocities of the projectiles to be all equal.
The value of A in (14) may be determined as usual. If N be the whole
(very great) number of masses to which the statistics relate,
so that
If we were to suppose that the chances of a favourable or unfavourable
collision were independent of the actual velocity of a mass, there would still be
a stationary state defined by writing v4 = oo in (15). Under these circum-
stances the mean energy would be twice as great as that calculated above.
It is easy to extend our result so as to apply to the case of projectiles
whose velocities are distributed according to any given law F(v), of course
upon the supposition that the projectiles of different velocities do not interfere
with one another. We have merely to multiply by F(v)dv and to integrate
between 0 and x . Thus from (13) we obtain
2u/(«)J vF(v)dv + qf'(u) I »»Jf»<fo = 0 (17)
If .F(w) = <r*'*, we find
/I r
~~ 2A- { ~ J e
so that
Jo v>e-*>'dv = j-fo ve-^dv. (18)
Our equation then becomes
gvng
/(tt) = 4e-*«"« ............................... (19)
The mean energy of the masses is %q/k, and this is now equal to the mean
energy of the projectiles. We see that if the mean energy of the projectiles
is given, their efficiency is greater when the velocity is distributed according
to the Maxwell law than when it is uniform, and that in the former case the
Waterston relation is satisfied, as was to be expected from investigations in
the theory of gases.
It may perhaps be objected that the law e~*^ is inconsistent with our
assumption that v is always great in comparison with u. Certainly there will
478 DYNAMICAL PROBLEMS IX [183
be a few projectiles for which the assumption is violated; but it is pretty
evident that in the limit when q is small enough, the effect of these will
become negligible. Even when the velocity of the projectiles is constant, the
law e~lt~ !&* must not be applied to values of u comparable with v.
The independence of the stationary state of conditions, which at first sight
would seem likely to have an influence, may
be illustrated by supposing that the motion
of the masses is constrained to take place
along a straight line, but that the direction
of motion of the projectiles, striking always
centrically, is inclined to this line at a con-
stant angle 6.
If u' be the velocity of the mass (unity) before impact, and u after impact,
B the impulsive action between the mass and the projectile,
u — u' = B cos 6.
Also, if v, V be the velocities of the projectile (q) before and after impact,
q(v-V} = B\
so that
q (v — V) cos 0 = u — •?/'.
By the equation of energy
w2 - u'2 = q(tf- V*).
From these we find, as before,
\
J .
cos )
This may be regarded as a generalization of (6) ; and we see that it may be
derived from (6) by writing v/cos0 for v, and qcos*0 for q. In applying
equation (10) to determine the stationary state, we must remember that the
velocity of retreat is now no longer w, but w cos 0, so that (10) becomes
* 1 f^ dw=T w cos ef(w) dw-
The entire effect of the obliquity 0 is thus represented by the substitution
of v/cos 0 for v, and of q cos2 0 for q, and since these leave qv2 unaltered, the
stationary state, determined by (15), is the same as if 0 = 0.
The results that we have obtained depend entirely upon the assumption
that the individual projectiles are fired at random, and without distinction
between one direction and the other. The significance of this may be illus-
trated by tracing the effect of a restriction. If we suppose that the projectiles
are despatched in pairs of closely following components, we should expect
that the effect would be the same as of a doubling of the mass. If, again, the
~ /
1891] ILLTSTRATIOX OF THE THEOBT OF GASES. 479
components of a pair were so projected as to strike almost at the same time
upon opposite skies, while ret the direction of the first was at random, we
should expect the whole effect to become evanescent. These anticipations are
confirmed by calculation.
Bv (5) the velocity *,', which on collision becomes •, is
so
.
that the velocity, which after too consecutive collisions upon the same side
becomes «r, is given by
The masses which by single collisions at velocity 0 would ultimately
produce the same effect as these pairs are therefore very approximately 2</.
If the projectiles be distributed in pairs in such a way that the components
of each strike nearly simultaneously and upon opposite sides,
showing that the effect is the same as if the mass were doubled, and the
Telocity reduced from w to qr. Thus, when q is infinitely small, the effect
is negligible in comparison with that obtained when the connexion of the
components of a pair is dissolved, and each individual is projected at
random.
Another Method of Investigation,
The method followed in the formation of equation (10) seems to lead most
simply to the required determination of f(*}i but it is an instructive varia-
tion to consider directly the balance between the numbers of masses which
change their velocities from and to «r.
Tne number of masses whose velocities lie between K and v + du being
f(m)d*, we have as the number whose velocities in a given small interval of
time are expelled from the range rf«,
480 DYNAMICAL PROBLEMS IN [183
This, in the steady state, is equal to the number which enter the range du
from the two sides in consequence of favourable and unfavourable collisions ;
so that
/(u') (v - u') du' + f(u") (v + u") du" - 2 V/(M) du = 0 ....... (20)
By (6), (7), since v is constant,
du' = \±2du, du»
l-q l-q
so that
*^|/(O . (v - u'} + l~qqf(u") • (f + «") - 2v/(M) = 0.
Now
V — U' = r— ^ (W - U), V + U" = ~-J* (V + U),
and thus
fi±|? {(* - 70 /(iO + (v + «) /(*")} - 2»/(«) = 0.
In this
so that
or, when 5 is small enough,
8^{/(W) + W//(W)}+4?V//'(w) = 0 ................ (21)
Accordingly
f(n) + uf(u) + $qv*f"(u) = 0, ..................... (22)
or on integration
It is easy to recognize that the constant C of integration must vanish. On
putting w = 0, its value is seen to be
for /(O) is not infinite. Now f(u) is by its nature an even function of u,
so that /'(O) must vanish. We thus obtain the same equation (14) of the
first order as by the former process.
Progress towards the Stationary State.
Passing from the consideration of the steady state, we will now suppose
that the masses are initially at rest, and examine the manner in which they
acquire velocity under the impact of the projectiles. In the very early stages
of the process the momentum acquired during one collision is practically
1891] ILLUSTRATION OF THE THEORY OF GASES. 481
independent of the existing velocity (u) of a mass, and may be taken to be
±'2qr. Moreover, the chance of a collision is at first sensibly independent
of u. In the present investigation we are concerned not merely, as in
considering the ultimate state, with the mass and velocity of a projectile,
but also with the frequency of impact. We will denote by v the whole
number of projectiles launched in both directions in the unit of time in the
path of each mass. The chance of a collision for a given mass in time dt is
thus represented by vdt. The number of collisions by which masses are
erpelled from the range du in time dt is f(u)du .vdt. The number which
enter the range from the two sides is
{/(« - Zqv) +/(u + 2qv)} du . ±, dt,
so that the excess of the number which enter the range over the number
which leave is
*/(«. + 2,*) -/(»); **. *dt,
df(u t\du
and this is to be equated to L — dL Thus
> ...... (23)
the well-known equation of the conduction of heat. When t = 0, f(u) is to be
zero for all finite values of n. The Fourier solution, applicable under these
conditions, is
where t' is written for ZqriPvt. The total number of masses being 3T, we get
to determine A
so that
........................ (24)
If it be the whole number of collisions (for each mass), n = vt, and we have
4f = 49*i».2ii. .............................. (25)
If the unit of velocity be so chosen that the momentum (2^r) communicated
at each impact is unity, (24) takes the form
(26)
which exhibits the distribution of momentum among the masses after n
impacts. In this form the problem coincides with one formerly treated*
relating to the composition of vibrations of arbitrary phases. It will be seen
• no. M*f. August, 1880, p. 73. [VoL i. pi 491.]
» in 31
482 DYNAMICAL PROBLEMS IN [183
that there is a sharp contrast between the steady state and the early stages
of the variable state. The latter depends upon the momentum of the pro-
jectiles, and upon the number of impacts ; the former involves the energy of
the projectiles, and is independent of the rapidity of the impacts. .
The mean square of velocity after any number (n) of impacts is
N~l f °°u*f(u, n) du = n,
J — 00
or, if we restore 4^2v2,
mean uL = nA<fvl ............................ (27)
It must be distinctly understood that the solution expressed by (24), (25),
(26) applies only to the first stages of the bombardment, beginning with the
masses at rest. If the same state of things continued, the motion of the
masses would increase without limit. But, as time goes on, two causes
intervene to prevent the accumulation of motion. When the velocity of
the masses becomes sensible, the chance of an unfavourable collision increases
at the expense of the favourable collisions, and this consideration alone would
prevent the unlimited accumulation of motion, and lead to the ultimate
establishment of a steady state. But another cause is also at work in the
same direction, and, as may be seen from the argument which leads to (13),
with equal efficiency. The favourable collisions, even when they occur,
produce less effect than the unfavourable ones, as is shown by (6) and (7).
We will now investigate the general equation, applicable not merely to
the initial and final, but to all stages of the acquirement of motion. As
in (20), (23) we have
dt = (i/XO . (v - u') du' + */(*") . („ + u") du" - f(u) .vdu}-
and thus by the same process as for (22)
If we write, as before,
t' = 2q*v2vt, and h = lfqv\ ..................... (29)
we have
Both in the case where the left side was omitted, and also when h vanished,
we found that the solution was of the form
where $ was constant, or a function of t' only. We shall find that the same
form applies also to the more general solution. The factor \/0 is evidently
1891] ILLUSTRATION OF THE THEORY OF GASES. 483
r+»
necessary in order to make I f(u) du independent of the time. By differen-
tiation of (31),
dfjdt' =
/+ df/du
so that (30) is satisfied provided <£ is so chosen as a function of if that
Thus
where, however, the constant must vanish, since <f> = oo corresponds to t' = 0.
Accordingly
which with (31) completes the solution.
If t' is small, (32) gives <f> = l/4,t', in agreement with (24); while if t' be
great, we have <f> = h = 1/qv*, as in (15').
The above solution is adapted to the case where f(u) = 0 for all finite
values of u, when t' = 0. The next step in the process of generalization will
be to obtain a solution applicable to the initial concentration of f(u), no
longer merely at zero, but at any arbitrary value of M; that is, to the case
where initially all the masses are moving with one constant velocity a.
Assume
/=V</>.e-*(u-*)J, .............................. (33)
where <f>, -ty- are functions of t' only. Substituting, as before, in (30), we find
so that the equation is satisfied provided
ig*+ 2^-2/^ = 0, ........................ (34)
and
0 ............................... (35)
The first is the same equation as we found before, and its solution is given by
(32); while (35) gives
' (36)
31—2
484 DYNAMICAL PROBLEMS IN [183
Thus (32), (33), (36) constitute the complete solution of the problem
proposed, and show how the initial concentration at u = a passes gradually
into the steady state when t' = <x> . In the early stages of the process
(37)
to which the factor Nj \/TT may be applied, when it is desired to represent
that the whole number of masses is N. It appears that during the whole
process the law of distribution is in a sense maintained, the only changes
being in the value of u round which the grouping takes place, and in the
degree of concentration about that value.
There will now be no difficulty in framing the expression applicable to an
arbitrary initial distribution of velocity among the masses. For this purpose
we need only multiply (33) by % (a) da., and integrate over the necessary
range. Thus
+a° (38)
<f> being given, as usual, by (32). The limits for a are taken + oo ; but we
must not forget that the restriction upon the magnitude of u requires that
X (u) shall be sensible only for values of u small in comparison with v.
When t' is small, we have from (38),
f(u' ^ = V
ultimately ; so that
Accordingly the required solution expressing the distribution of velocity at t'
in terms of that which obtains when t' = 0, is
<t>(u-ae-My} ....... (39)
We may verify this by supposing that f(u, 0) = e~hu^, representing the steady
state. The integration of (39) then shows that
f(u,t') = e-™,
as of course should be.
An example of more interest is obtained by supposing that initially
f(u, 0) = <r*'M2; .............................. (40)
that is, that the velocities are in the state which would be a steady state
under the action of projectiles moving with an energy different from the
actual energy. In this case we find from (32), (39),
^ ................... <«>
1891] ILLUSTRATION OF THE THEORY OF GASES. 485
We will now introduce the consideration of variable velocity of projectiles
into the problem of the progressive state. In (28) we must regard v as a
function of v. If we use v dv to denote the number of projectiles launched
in unit of time with velocities included between v and v + dv, (28) may be
written
(42)
which is of the same form as before. The only difference is that we now have
in place of (29),
t' = 2ftfvv>dv, ................................. (43)
h = fvdv -=- q fv&dv. ........................... (44)
In applying these results to particular problems, there is an important
distinction to be observed. By definition vdv represents the number of
projectiles which in the unit time pass a given place with velocities included
within the prescribed range. It will therefore not represent the distribution
of velocities in a given space: for the projectiles, passing in unit time, which
move with the higher velocities cover correspondingly greater spaces. If
therefore we wish to investigate the effect of a Maxwellian distribution of
velocities among the projectiles, we are to take, not v = Be~*^, but
(45)
In this case, by (18),
* = %; .................................... (46)
and, as we saw, the mean energy of a mass in the steady state is equal to the
mean energy of the projectiles which at any moment of time occupy a given
space. From (43),
(47)
Pendulums in place of Free Masses.
We will now introduce a new element into the question by supposing
that the masses are no longer free to wander indefinitely, but are moored to
fixed points by similar elastic attachments. And for the moment we will
assume that the stationary state is such that no change would occur in it
were the bombardment at any time suspended. To satisfy this condition it
is requisite that the phases of vibrations of a given amplitude should have a
certain distribution, dependent upon the law of force. For example, in the
simplest case of a force proportional to displacement, where the velocity u is
connected with the amplitude (of velocity) r and with the phase 0 by the
relation t* = r cos 0, the distribution must be uniform with respect to 0, so
that the number of vibrations in phases between 6 and 0 + dd must be dff 2ir
486 DYNAMICAL PROBLEMS IN [183
of the whole number whose amplitude is r. Thus, if r be given, the propor-
tional number with velocities between u and u + du is
du
.(48)
And, in general, if r be some quantity by which the amplitude is measured,
the proportional number will be of the form
<f>(r,u)du, (49)
where <£ is a determinate function of r and u, dependent upon the law of
vibration. If now % (r) dr denote the number of vibrations for which r lies
between r and r + dr, we have altogether for the distribution of velocities u,
If the vibrators were left to themselves, %(r) might be chosen arbitrarily,
and yet the distribution of velocity, denoted by f(u), would be permanent.
But if the vibrators are subject to bombardment, f(u) cannot be permanent,
unless it be of the form already determined. The problem of the permanent
state may thus be considered to be the determination of ^ (r) in (50), so as
to make/(w) equal to e~hu\
We will now limit ourselves to a law of force proportional to displacement,
so that the vibrations are isochronous ; and examine what must be the form
of x (f) in (8) in order that the requirements of the case may be satisfied.
By (15'), if N be the whole number of vibrators,
^he-hu^=r X(r)dr ,ei,
The determination of the form of ^ is analogous to a well-known investigation
in the theory of gases. We assume
X (r) = Ar e~hr\ (52)
where A is a constant to be determined. To integrate the right-hand
member of (51), we write
r2 = w2+7?2. (53)
so that
Thus
A=4Ji.N. (54)
The distribution of the amplitudes (of velocity) is therefore such that the
number of amplitudes between r and r + dr is
N . 4>hr e~hr* dr, (55)
while for each amplitude the phases are uniformly distributed round the
complete cycle.
1891] ILLUSTRATION OF THE THEORY OF GASES. 487
The argument in the preceding paragraphs depends upon the assumption
that a steady state exists, which would not be disturbed by a suspension, or
relaxation, of the bombardment. Now this is a point which demands closer
examination; because it is conceivable that there may be a steady state,
permanent so long as the bombardment itself is steady, but liable to alteration
when the rate of bombardment is increased or diminished. And in this case
we could not argue, as before, that the distribution must be uniform with
respect to 0.
If x denote the displacement of a vibrator at time t,
x = n~lrsm(nt- 6), dx/dt = r cos (nt - 6\
When t = 0,
x — — n~l r sin 6, dx/dt = u = rco80;
and we may regard the amplitude and phase of the vibrator as determined
by u, t) where
u = r cos 6, rj = r sin 0.
Any distribution of amplitudes and phases may thus be expressed by
If we consider the effect of the collisions which may occur at t = 0, we see
that u is altered according to the laws already laid down, while 77 remains
unchanged. The condition that the distribution remains undisturbed by the
collisions is, as before, that, for every constant 77, f(u, rj) should be of the form
e~Ati*, or, as we may write it,
But this condition is not sufficient to secure a stationary state, because, even
in the absence of collisions, a variation would occur, unless f(u,i)) were a
function of r, independent of 6. Both conditions are satisfied, if ^ (77)= A e~hlf,
where A is a constant ; so that
f(u, 77) dudrj = A e~h (M2+»*> dudy = ZirA e~hl*r dr.
Under this law of distribution there is no change either from the progress of
the vibrations themselves, or as the result of collisions.
The principle that the distribution of velocities in the stationary state is
the same as if the masses were free is of great importance, and leads to results
that may at first appear strange. Thus the mean kinetic energies of the
masses is the same in the two cases, although in the one case there is an
accompaniment of potential energy, while in the other there is none. But it
is to be observed that nothing is here said as to the rate of progress towards
the stationary condition when, for instance, the masses start from rest ; and
the fact that the ultimate distribution of velocities should be independent
of the potential energy is perhaps no more difficult to admit than its inde-
pendence of the number of projectiles which strike in a given time. One
488 DYNAMICAL PROBLEMS IN [183
difference may, however, be alluded to in passing. In the case of the
vibrators it is necessary to suppose that the collisions are instantaneous;
while the result for the free masses is independent of such a limitation.
The simplicity of / in the stationary state has its origin in the indepen-
dence of 0. It is not difficult to prove that this law of independence fails
during the development of the vibrations from a state of rest under a vigorous
bombardment. The investigation of this matter is accordingly more com-
plicated than in the case of the free masses, and I do not propose here to
enter upon it.
In a modification of the original problem of some interest even the
stationary distribution is not entirely independent of phase. I refer to the
case where the bombardment is from one side only, or (more generally) is less
vigorous on one side than on the other. It is easy to see that a one-sided
bombardment would of necessity disturb a uniform distribution of phase, even
if it were already established. The permanent state is accordingly one of
unequal phase-distribution, and is not, as for the symmetrical bombardment,
independent of the vigour with which the bombardment is conducted.
But in one important particular case the simplicity of the symmetrical
bombardment is recovered. For if the number of projectiles striking in a
given time be sufficiently reduced, the stationary condition must ultimately
become one of uniform phase-distribution.
Under this limitation it is easy to see what the stationary state must be.
Since the ultimate distribution is uniform with respect to phase, it must be
the same from whichever side the bombardment comes. Under these circum-
stances it could not be altered if the bombardment proceeded indifferently
from both sides, which is the case already investigated. We conclude that,
provided the bombardment be very feeble, there is a definite stationary con-
dition, independent both of the amount of the bombardment and of its
distribution between the two directions. It is of course understood that
from whichever side a projectile be fired, the moment of firing is absolutely
without relation to the phase of the vibrator which it is to strike.
The problem of the one-sided bombardment may also be attacked by a
direct calculation of the distribution of amplitude in the stationary condition.
The first step is to estimate the effect upon the amplitude of a given collision.
From (6), if u' be the velocity before collision, and u after,
u = u' + = — *— (» — u'\
The fraction 2<//(l + q) occurs as a whole, and we might retain it throughout.
But inasmuch as in the final result only one power of q need be retained, it
will conduce to brevity to omit the denominator at once, and take simply
u = u' + 2q(v-u') (56)
1891] ILLUSTRATION OF THE THEORY OF GASES. 489
Thus if p, <f> and r, 0 be the amplitude and phase before and after collision
respectively,
rco80 = />cos$ + 2?(r-pcos$),)
rsm0 = psin£;
so that
f*= p* + 4<7p cos <f> (r - p cos <f>) + 4<f (v - p cos ^)*.
From this we require the approximate value of p in terms of r and ^. The
term in q* cannot be altogether neglected, but it need only be retained when
multiplied by r*. The result is p = r — &r, where
(58)
This equation determines for a given <f> the value of p which the blow
converts into r. Values of p nearer to r will be projected across that value.
The chance of a collision at p, <J> is proportional to (r — p cos ^). Thus if a
number of vibrators in state p, <f> be F(p)dpd$*, the condition for the
stationary state is
r (59)
the integral on the left expressing the whole number (estimated algebraically)
of amplitudes which in a small interval of time pass outwards through the
value r.
By expansion of F(p) in the series
we find
)} +cubes of q.
Again from (58),
I Srd<f> = — qr + <?V/r, I cos <f>Brd<f> = qv,
[^(Sr)1 d<p = 2qV, I ^cos <f> (5r)* d<f> = 0.
The condition for the stationary state is therefore
« (F (r)(- qr + 8V>) - f (r) qV] -rF(r)qv = 0,
F(r) {-2r+ qv*/r\ -F/(r)qv* = 0.
Thus, on integration,
r«-^logr + ^log^(r) = const., .................. (60)
F(r)=Arc-*«'. .............................. (61)
* We hoe assume that the bombardment is feeble.
490 ILLUSTRATIONS OF THE THEORY OF GASES. [183
The mean value of r2, expressed by
is qv2', that is, the mean value of the maximum kinetic energy attained
during the vibration is equal to the kinetic energy of a projectile. The
mean of all the actual kinetic energies of the vibrators is the half of this ;
but would rise to equality with the mean energy of the projectiles, if the
velocities of the latter, instead of being uniform, as above supposed, were
distributed according to the Maxwellian law.
If we are content to assume the law of distribution, p e~hp~, leaving only
the constant h to be determined, the investigation may be much simplified.
Thus from (57) the gain of energy from the collision is
£r2 - ip2 = 2qp cos </> (v - p cos <f>) + 2q*v>.
The chance of the collision in question is proportional to the relative velocity
(v — p cos <£) ; and in the stationary state the whole gain of energy is zero.
Hence
{2qp cos (j>(v-p cos <£)2 + 2gV} - 0.
In the integration with respect to <f> the odd powers of cos </> vanish. Hence
/"oo
2qv p dp e-W (qv* - p2) = 0 ;
Jo
so that
as in (61).
184.
EXPERIMENTS IN AERODYNAMICS*
[Nature, XLV. pp. 108, 109, 1891.]
THE subject of this memoir is of especial interest at the present time,
when the skill of a distinguished inventor is understood to be engaged in
attacking the many practical difficulties which lie in the way of artificial
flight upon a large scale. For a long time the resistance of fluids formed an
unsatisfactory chapter in our treatises on hydrodynamics. According to the
early suggestions of Newton, the resistances are (1) proportional to the
surfaces of the solid bodies acted upon, to the densities of the fluids, and
to the squares of the velocities: while (2) "the direct impulse of a fluid on
a plane surface is to its absolute oblique impulse on the same surface as
the square of the radius to the square of the sine of the angle of incidence."
The author of the workf from which these words are quoted, in comparing
the above statements with the experimental results available in his time
(1882), remarks : — " (1) It is very consonant to experiment that the resistances
are proportional to the squares of the velocities.... (2) It appears from a
comparison of all the experiments, that the impulses and resistances are very
nearly in the proportion of the surfaces.... (3) The resistances do by no
means vary in the duplicate ratio of the sines of the angle of incidence."
And he subsequently states that for small angles the resistances are more
nearly proportional to the sines of incidence than to their squares.
It is probable that the law of velocity tended to support in men's minds
the law of the square of the sine. For, if both be admitted, it follows
that the resistance, normal to the surface, experienced by a plane when
immersed in a stream of fluid, depends only upon the component of the
velocity perpendicular to the surface. That the effect should be independent
• Experiment* im Aerodynamic*. By S. P. Langler. S*if AMWM Cort ritaffem to Kmowledye
(Washington, 1891).
t System of Mechanical Phitotophy. By John Bobison, YoL n., 18».
492 EXPERIMENTS IN AERODYNAMICS. [184
of the component parallel to the plane seems plausible, inasmuch as this
component, if it existed alone, would exercise no pressure : but that such
a view is entirely erroneous has been long recognized by practical men,
especially by those concerned in navigation.
From the law of the simple sine, enunciated by Robison, it follows at
once that the pressure upon a lamina exposed perpendicularly to a stream
may be increased to any extent by imparting to the lamina a sufficiently
high velocity in its own plane. The immense importance of this principle
was clearly recognized by Mr Wenham in his valuable paper upon flight *, and
a few years later the whole subject was discussed by the greatest authority
upon such matters, the late Mr W. Froude, with characteristic insight and
lucidity f.
The theoretical problem of determining the resistance from the first
principles of hydrodynamics is not free from difficulty, even in the case of two
dimensions, where a long rectangular lamina is exposed obliquely to a stream
whose direction is perpendicular to the longer sides. The formula + resulting
from the theory of Kirchhoff, viz.
Trainee
4 + TT sm a r
where p is the density of the fluid, and V is the total velocity of the stream
flowing at the angle a with the plane of the lamina, shows that when a is
small the resistance is nearly proportional to sin a. Moreover, (1) agrees with
the experiments of Vince§.
It will be seen that the laws of resistance were fairly well established
many years ago, at least in their main outlines. Nevertheless, there was
ample room for the systematic and highly elaborate experiments recorded
in the memoir whose title stands at the head of this article. The work
appears to have been executed with the skill and thoroughness which
would naturally be expected of the author, and will doubtless prove of
great service to those engaged upon these matters. The scanty reference
to previous knowledge, which Prof. Langley holds out some promise of
extending in subsequent publications, makes it rather difficult to pick out
the points of greatest novelty. The main problem is, of course, the law of
obliquity, and this is attacked with two distinct forms of apparatus. The
general character of the results, exhibited graphically on p. 62, will be made
apparent from the accompanying reproduction, in which are added a curve D,
corresponding to (1) and E, representing the law of sin2 a. In each case the
* Report of Aeronautical Society for 1866.
f Proc. Inst. Civ. Eng. 1871 (discussion upon a paper by Sir F. Knowles). [See Vol. i. p. 290.]
J See Phil. Mag. December, 1876 [Vol. i. p. 291]. Also Basset's Hydrodynamics, Vol. i.
p. 181.
§ Phil. Trans. 1798.
1891]
EXPERIMENTS IX AERODYNAMICS.
493
abscissa is the angle a and the ordinate is the normal pressure, expressed as
a percentage of that experienced when a =90°. Of Prof. Langley's curves,
Fig. L
A relates to a square plane 12 inches x 12 inches, B to a rectangle 6 x 24 inches,
and C to a rectangle 30 x 4'8 inches, the leading edge (perpendicular to the
stream) being in each case specified first, so that the theoretical curve D
corresponds most nearly to C. It will be seen at a glance that at small
angles the pressure is enormously greater than according to the law of sin1 a.
The differences between A, B, C, anticipated in a general manner by
Wenham and Froude, are of great interest. They demonstrate that in
proportion to area a long narrow wing is more efficient as a support than
a short wide one, and that in a very marked degree.
Up to a certain point there is no difficulty in giving a theoretical account
of these features. When a rectangular lamina is exposed perpendicularly,
there is one point, VJR. the centre, at which the velocity of the stream is
annulled. At this point the pressure attains the full amount, \pV*, due to
the velocity of the stream, while at every other point the pressure is less, and
falls to zero at the boundary. If the lamina is sloped to the stream as in B
and C, there is still a median plane of symmetry ; and at one point in this
plane, but now in advance of the centre, the full pressure is experienced. In
strictness, there is only one point of maximum pressure, whatever may be the
proportions of the lamina. But if the rectangle be very elongated, there
is practically a great difference in this respect according to the manner of
presentation, although the small angle a be preserved unchanged. For when
the long edges are perpendicular to the stream (C), the motion is nearly in
494 EXPERIMENTS IN AERODYNAMICS. [184
two dimensions, and the region of nearly maximum pressure extends over
most of the length. But the case is obviously quite different when it is
the short dimension that is perpendicular to the stream, for then along
the greater part of the length there is rapid flow, and consequently small
pressure.
It will naturally be asked whether any explanation can be offered of the
divergence of C from the theoretical curve D. This is a point well worthy
of further experiment. It seems probable that the cause lies in the suction
operative, as the result of friction, at the back of the lamina. That the
suction is a reality may be proved without much difficulty by using a hollow
lamina, AB (Fig. 2), whose interior is connected with a manometer. If
there be a small perforation at any point C, the manometer indicates the
pressure, positive or negative, exercised at this point, when the apparatus is
exposed to a blast of air.
Fig. 2.
When once the law ot obliquities is known, the problem of aerial main-
tenance presents no further theoretical difficulty. It was successfully treated
many years ago by Penaud*, and somewhat later by Froude, whose interest-
ing letters, written shortly before his death, have recently been published f.
In perhaps the simplest form of the question the level is supposed to be
maintained with the aid, e.g., of screw propulsion, the necessary maintenance
being secured by an aeroplane slightly tilted (a) upwards in front. The work
required to be expended in order to maintain a given weight depends upon
the area of plane, the inclination, and the speed. Penaud's results show that,
if skin friction could be neglected, the necessary work might be diminished
indefinitely, even with a given area of wing. For this purpose, it would only
be necessary to increase the speed and correspondingly to diminish a. But
when skin friction is taken into account, the work can only be reduced to
a minimum, and to do this with a given area of wing requires a definite
(large) velocity, and a definite (small) inclination. The accurate determi-
* See Report of Aeronautical Society for 1876.
t Edinburgh Proceedings. B. E. Froude, 1891.
1891] EXPERIMENTS IX AERODYNAMICS. 405
nation of the tangential, as well as of the normal, force experienced by an
inclined plane is thus of essential importance in the question of flight
The work of Penand seems to be so little known that it has been thought
desirable to recapitulate some of his theoretical conclusions. But we owe to
Penaud not merely sound theory, but the actual construction of a successful
flying machine, in which horizontal flight is maintained by a screw propeller.
In these models the energy is stored by means of stretched india-rubber,
a method available only upon a small scale. It is probable that the principle
of the rocket might be employed with advantage : and even upon a large
scale, the abolition of all machinery would allow of considerable extravagance
in the use of explosive material. This method is especially adapted to the
very high speeds which on other grounds are most suitable.
In the chapter on "The Plane Dropper" some striking experiments are
described, illustrating the effect of a forward movement in retarding the fall
of a horizontal plane. Prof. Langley seems hardly to recognize that there is
nothing really distinctive in this arrangement when he savs: —
" It is, of course, an entirely familiar observation that we can support an
inclined plane by moving it laterally, deriving our support in this case from
the upward component of pressure derived from the wind of advance ; but,
so far as I am aware, this problem of the velocity of fall of a horizontal plane
moving horizontally in the air has never been worked out theoretically or
determined experimentally, and I believe that the experimental investigation
whose results I am now to present is new."
But, apart from the complications which attend the establishment of a
uniform regime, there is no essential difference between the two cases.
The hydrodynamical forces depend only upon the magnitude of the relative
velocity and upon the inclination of this relative velocity to the plane. All
else is a question merely of ordinary elementary mechanics.
It is interesting to note that Prof. Langley's experience has led him to
take a favourable view of the practicability of flight upon a large scale.
Such was also the opinion of Penaud, who (in 1876) expresses his conviction
"that, in the future more or less distant, science will construct a light
motor that will enable us to solve the problem of aviation." But sufficient
maintaining power is not the only requisite : and it is probable that difficulties
connected with stability, and with safe alighting at the termination of the
adventure, will exercise to the utmost the skill of our inventors.
[1901. Some of the problems here referred to are further discussed in
the Wilde Lecture on the Mechanical Principles of Flight (Manchester
Proceedings, VoL xuv. Part n. pp. 1-26, 1899.]
185.
ON REFLEXION FROM LIQUID SURFACES IN THE NEIGH-
BOURHOOD OF THE POLARIZING ANGLE*.
[Philosophical Magazine, xxxm. pp. 1 — 19, Jan. 1892.]
BY the experiments of Jamin and others it has been abundantly proved
that in the neighbourhood of the polarizing angle the reflexion of light from
ordinary transparent liquids and solids deviates sensibly from the laws of
Fresnel, according to which the reflexion of light polarized perpendicularly to
the plane of incidence should vanish when the incidence takes place at the
Brewsterian angle. It is found, on the contrary, that in most cases the
residual light is sensible at all angles, and that the change of phase by
180°, which, according to Fresnel's formula, should occur suddenly, in reality
enters by degrees, so that in general plane-polarized light acquires after
reflexion a certain amount of ellipticity. Although Jamin describes the non-
evanescence at the polarizing angle and the ellipticity in its neighbourhood
as " deux ordres de phenomenes de nature differente," it is clear that they
are really inseparable parts of one phenomenon. If we suppose the incident
light polarized perpendicularly to the plane of incidence to be given, the
vibration which determines the reflected light at various angles may be
represented in amplitude and phase by the situation of points relatively to
an origin and coordinate axes. Thus, according to Fresnel's formula, the
locus of these points is the axis of abscissae XX' itself, the point 0 corre-
sponding to the polarizing angle, at which the reflexion vanishes, and in
passing which there is a sudden change of phase of 180°. If the reflexion
remains finite at all angles, the curve in question meets the axis YY' at
some point P, not coincident with 0, and the corresponding phase differs by
a quarter-period from the phases met with at a distance from this angle. So
* [1901. A preliminary account of the experiments here detailed was given before the British
Association (see Report, Aug. 21, 1891, p. 563). The principal observation had been recorded
still earlier (Phil. Mag. Nov. 1890; Vol. in. p. 396).]
1892] ON REFLEXION FROM LIQUID SURFACES. -k"7
far as experiment can yet show, this curve may be a straight line parallel to
XX', and at a short distance from it. If it lie above XX', the reflexion is
what Jamin characterizes as positive ; if below, the reflexion is negative.
To this order of approximation the behaviour of a transparent body
reflecting light of given wave-length is determined by two constants, (1) the
refractive index p, and (2) the intensity of reflexion at the angle tan"1/*
when the light is polarized perpendicularly to the plane of incidence. The
most convenient form of the second constant for experimental purposes is the
ratio of reflected amplitudes for the two principal planes when the light,
incident at the angle tan"1/*, is polarized at 45° to these planes. It may be
called the ellipticity, and, after Jamin, will be denoted by k. According to
Fresnel k = Q; but Jamin found for water k=— -00577, and for absolute
alcohol k = + "00208. Contrasting liquids with solids, he remarks*, " on vient
de voir que leur polarisation est elliptique, et qu'il est impossible d'en trouver
la cause dans une constitution moleculaire anormale." And, again: — "II
est jusqu'a present impossible de constater une relation simple entre la valeur
du coefficient k et 1'indice de refraction ; tout porte a croire, an contraire, que
ces deux constantes sont independantes, 1'une de 1'autre. Mais, a defaut
de loi precise, on peut remarquer une tendance du coefficient k a diminuer
avec 1'indice, et quand celui-ci est approximativement egal a T45, k est nul :
1'indice continuant a decroitre, le coefficient k reprend des valeurs sensibles et
croissantes."
Since the time of Jamin many valuable observations upon reflexion have
been made by Quincke and others, which it is not necessary for my purpose
further to refer to. In 1889f Drude made the very important observation
that the ellipticity of a freshly split surface of rock-salt is very small, but
that it rapidly increases on standing. And he concludes generally that solid
bodies reflect from natural cleavage surfaces according to the laws of Fresnel.
* "Memoire ear la reflexion de la Lamiere a la surface des Liqaides." Jim. Ckim. mi.
p. 165 (1851).
t Wied, AIM. MXVI. p. 532.
R. ill. 32
498 ON REFLEXION FROM LIQUID SURFACES [185
It is remarkable that fluids are excluded from this statement. Indeed Drude
expressly remarks that in fluids the natural surface-sheet, which forms the
transition from one medium to another, exercises an essential influence upon
the reflexion.
Nearly contemporaneously with Drude's observation I made a first
attempt to examine whether the different behaviour which Plateau found
for water and for alcohol in respect of superficial viscosity, and which I was
disposed to attribute to surface contamination, corresponded to anything in
the phenomena of reflexion near the polarizing angle ; but at that time
1 was misled by a faulty nicol, and failed to observe anything distinctive.
Subsequently, having proved that the superficial viscosity of water was due
to a greasy contamination, whose thickness might be much less than one
millionth of a millimetre, I too hastily concluded that films of such extra-
ordinary tenuity were unlikely to be of optical importance, until, prompted
by a remark of Sir G. Stokes, I made an actual estimate of the effect to
be expected. The thickness required to stop camphor movements, viz.
2 x 10~7centim., is about B^ of \D. This will be one factor in the expression
for the amplitude of reflexion due to grease. There will be another somewhat
small factor expressive of the difference of optical quality between grease and
water; so that the intensity of reflexion at the polarizing angle might on
this view be of the order 10~6, or 10~7, a quantity quite appreciable, when
the incident light is from the sun. Thus encouraged, I returned to the
attack, and on October 2nd, 1890, examined the image of the sun as reflected
from water at the polarizing angle. The arrangements were very simple.
Sunlight, reflected horizontally into the dark room from a heliostat outside,
was again reflected downwards by an adjustable mirror. The water was
contained in a large porcelain dish to the depth of about an inch, and at the
bottom of the dish was placed a piece of darkly-coloured glass, not quite
horizontal, with the view of annulling the diffuse reflexion. The reflected
image was observed with a nicol, from which the glass covering disks had
been removed, simply held in the hand. The appearance of the dark spot,
brought to the centre of the sun's disk, was at once recognized to be
dependent upon the condition of the water-surface. When the surface was
clean, the spot was dark and its width (in the plane of incidence) was about
^ or 1 of the solar diameter. Moreover there was a strong appearance of
colour, brown above and blue below, evidently due to the dependence of the
polarizing angle upon the wave-length of the light. But when the surface
was greasy, even although camphor fragments still rotated briskly, the band
lost its darkness, became diffuse, and showed but little colour. When the
greasy film was about sufficient to stop the camphor movements, the contrast
with the effect of clean water was very marked*.
* Phil. Mag. November 1890, p. 400. [Vol. in. p. 396.]
1892] IX THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 409
The surface was cleansed bj the aid of the expansible hoop employed
in my former experiments. This is made of very thin sheet brass, about
2 inches wide. It is placed npon the water, already reasonably clean, in
its contracted condition, so that the area enclosed is but small. When it is
opened out, say to a circle of about 10 inches diameter, the internal surface
of the water is rendered more clean, and the external less clean, than before.
To get the best result it is desirable to go through the operation of
expanding two or three times, probably because the cleaned water-surface
acquires grease from the internal surface of the brass hoop. It will be
evident that the action depends npon the hoop not being completely wetted*.
Otherwise the grease could repass from the outside back into the interior.
For this reason the hoop cannot be expected to succeed with liquids like
alcohol
By taking advantage of the apparent motion of the sun in altitude, these
observations may be repeated in summer without any heliostat, or reflexion,
other than that from the water itself. Thus on June 26 the dish was placed
on a table below the window of an nndarkened room, and the passage of the
spot across the sun's disk was watched. The spot was central at about 4k 0",
and the instant of centrality could be determined to within 10*, and probablv
to within 5*. On August 15, when the sun's motion in altitude was slower,
centrality occurred at about 3k 10", and the precise instant was less well
determined.
To aee the band at its best requires an unusually good nicol. Whether
on account of residual defects in the nicols, or in the lenses of my eyes, vision
was improved by the use of a horizontal slit, about T'? inch wide, cut out
of black paper, and attached to the cork mounting of the nicol on the side
next the eve. Under these conditions the band seen from clean water
looks black and well defined, and of width amounting to | or £ of the solar
diameter. A still further improvement sometimes attends the use of a
second nicol, held parallel to the first, through which the light passes before
reflexion from the water. With these arrangements the band is visibly
deteriorated by quantities of grease far less than is required to check the
camphor movements.
It has been mentioned that the dark band from clean water was fairly
narrow ; and it will be of interest to inquire what is to be expected upon the
assumption that Fresnel's formulae really express the facts of the case. We
will write
„ sin (0-0,) ^ tan (0-0.). m
* Since imperfect wetting must be attributed to residual grease, it would appear that the
operation of the hoop is incomplete at best. Nererthekas, it » a very useful and convenient
32—2
500 ON REFLEXION FROM LIQUID SURFACES [185
so that the ratio of amplitudes of the two polarized components, corre-
sponding to a primitive polarization at 45°, is
r 008(0 + 0,) „.
S cos (6 - 0a) '
vanishing when 0+0i = ^7r, that is when 0 = tan~1//,. We will suppose
that the angle of incidence has approximately this value, and write 0 + 80,
0i + 80i for 0, 0i respectively. Thus in the neighbourhood of the polarizing
angle the ratio is
cos (0-00
Now
sin 0 = /*sin 0!, cos 080 = /j, cos 01801,
so that
$/J cos 080 sin 0i 80 80
Oc/i — - 2T = — — '• — A~ == — o •
//, cos 0j p, sin 0 fi2
Hence
£ /i2cos2(0- 0,)'
For water
^=1-3336, 0 = tan- V = 53° 8', 0, = 36°52';
and
T/^f = 1-627 80, .............................. (4)
T2/S2 = 2-649 (80)2 ............................ (5)
Let us calculate the ratio corresponding to the upper or lower limb of the
sun when the spot is central ; that is, let 80 be the angular radius of the sun,
whose value in minutes is 16. Thus
~ rt 16-7T
~ 10800 '
and corresponding thereto from (5)
T*/S* = 5-74 x 10~5 (6)
The width of the band actually observed had been estimated at about
£ of the solar diameter, so that at its limits
T*/S* = 2 x 10-".
The band was thus about as narrow as Fresnel's formulae would lead one
to expect, and its deterioration by a film of grease might be anticipated as
at least probable, from the rough estimate above given of the effect of such
a film.
The results so far obtained were already sufficient to show that Jamin's
value of k, viz. - '00577, must be (numerically) much in excess of the truth.
For according to it, since If = 3'33 x 10~5, the minimum illumination at the
centre of the spot should be half as great as Fresnel's formulas make it at
the limb of the sun, so that the whole diameter of the sun would be almost
1892] IN THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 501
equally obscured. The observed narrowness of the band, even in the absence
of all precise measures, thus constitutes a proof that Jamin's k is several
times too great, and suffices to render it almost certain that the water-surface
with which he worked was highly contaminated.
It has already been mentioned that a well-formed band was attended with
a marked appearance of colour. The account of this rendered by Fresnel's
formulae is quite satisfactory. Let us calculate the illumination at the centre
of the band corresponding to /4, due to a change from /i to /4 + 8/4, comparing
it, as usual, with 8*. In the differentiation 0 is to be treated as constant,
and the change in 0,, viz. 80,, is due to 8/4. From (2),
T= cos (£TT + 80Q -80,
S ~ cos (0 - 0, - 80,) " cos (0 - 0,) '
and the relation between 80, and 8/4 is
cos 0, 80, = - sin 0 8/4//4=,
or 80, = - 8/4 //42,
since cos 0, = sin 0. Thus
T 8/4
S~~/4sCOS(0-0,)'
iter,
1 mt
From these data,
•(7)
In the case of water,
1-341, /4B = 1-331, 8/4 = -010.
T*/S* = 3-46 x 10-* ;
showing that if the spot is central for Fraunhofer's line B, the illumination
at the centre for G is more than half as great as is found (6) for B at the
upper and lower limbs of the sun. A considerable development of colour is
thus to be expected, when the band is well formed.
The band may be achromatized with the aid of a suitable prism, held
between the nicol and the eye, but of course at the expense of introducing
colour at the upper and lower limbs of the sun. I had at my disposal a
glass prism of 10°. This diminished, but could not annul, the colour when
held nearly in the position of minimum deviation ; but by sufficient sloping
the band was practically achromatized. When more dispersive materials,
e.g. benzole or bisulphide of carbon, were substituted for water, the develop-
ment of colour is very great, and in the case of the latter made it impossible
to judge of the perfection of the band. The above-mentioned glass prism
was of course quite insufficient for compensation.
The magnitude of these chromatic effects is given at once by Brewster's
law, which we may write in the form
tan (0 + 80) = /4 -I- 8/4.
Thus
8/4
i , Ti. (8>
502 ON REFLEXION FROM LIQUID SURFACES [185
which gives the angular displacement of the centre of the dark band, due
to the change from //, to p, + S/z. Let us inquire what small angle (i) must
be given to a prism of the same material in order that it may be capable
of compensating the colour. The deviation D is equal to (^ — 1) i, so that
$D = SfA. i. Hence, if 8D = S0,
* = r+b> (9)
The necessary angle is thus independent of the dispersive power, and does
not vary rapidly with the refractive power, of the substance. For water,
i = 9/25 in circular measure, or about 22°. For glass (/^ = 1'5) we should
have i= 18°.
An attempt was made to achromatize the band from bisulphide of carbon
with a 15° prism of that material. So far as could be judged the colour was
compensated, but the observation was imperfect on account of the insufficient
angular magnitude of the solar disk.
These experiments on the achromatization of the band had been made in
the hope of thereby reducing its apparent width, seeing that according to (8)
the difference of position for the lines B and G amounts, in the case of water,
to 13', much more than the apparent width of the band. But the width of
the achromatized band could not be set at much less than ^ of the sun's
diameter*. It seems that in estimating the dimensions of the uncorrected
band the eye instinctively allows for the influence of colour.
In experimenting upon water various kinds were tried. Usually the
tap-water (from an open cistern) behaved after expansion as well as did
distilled water. The brass hoop, judiciously applied, appears to be capable
of removing ordinary surface-contamination ; but the appearance of the band
is liable to be deteriorated by suspended matter, which detracts from the
central darkness. So far as could be judged by this method of observation,
the best bands were sensibly perfect. There was no evidence of any
departure from the law of Fresnel.
Similar results were obtained from other liquids, e.g. strong alcohol,
sulphuric acid, benzole. Special interest attached to an observation upon
a saturated solution of camphor, of which the superficial tension is much
lower (-72) than that of pure water. The band was sensibly perfect.
Oleate of soda (?^) was troublesome on account of the difficulty of
avoiding scum. A pretty good band could be obtained, certainly inferior
to the best, possibly owing to residual scum, but much better than from
water greased with olive-oil to the point at which the camphor motions
are just stopped.
* A coloured glass is still less effective than the compensating prism. A reduction in the
intensity of the light necessarily broadens the band. A similar effect occurs if the sun is not
quite clear.
1892] IN THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 503
The results last recorded prove that the optical effect is not determined
by surface-tension, for the tension of the oleate solution is much less than
that of any merely greased surface. A similar conclusion was suggested by
the observed difference of behaviour of various parts of the same surface.
A surface, originally clean, and then greased with olive-oil carried upon
a previously ignited platinum wire, frequently showed streakiness, when the
eye, observing through the nicol, as usual, was focused upon the surface.
Except perhaps in the case of oleate, none of these experiments, many
times repeated, gave any evidence of a real departure of properly skimmed
surfaces from the laws of Fresnel ; and it looked very much as if all the
results enunciated for liquids by Jamin were vitiated by the presence of
greasy films. That a film of extreme tenuity would suffice was certain. The
band from water was very obviously deteriorated by a film of olive-oil, which
needed to be condensed four or five times in order to stop the camphor
movements.
But it was impossible to rest here. It was necessary actually to measure,
or, if that were not possible, to find limits for. the ellipticity of the various
surfaces. And for this purpose a much more elaborate apparatus had to be
installed.
Sunlight, reflected horizontally from the heliostat, passed through a dia-
phragm in the shutter of about £ inch diameter, and thence to a collimating
lens of 23 inches focus. It was next reflected in the required oblique
direction by an adjustable mirror, and caused to traverse the polarizing nicol,
mounted in a circle that allowed the orientation of the nicol to be read
to a minute of angle. After reflexion from the surface under examination
the light traversed in succession a quarter- wave-plate and the analysing nicol,
and was then received into the eye, either directly, or with the intervention
of a small telescope magnifying about twice. In either case the eye was
focused upon the diaphragm, which was provided with cross wires; so that
the rays which fell upon any part of the retina constituted a parallel pencil,
not only at the surface of the liquid, but also in their passage through the
nicols and quarter-wave-plate. The latter was of mica, and both it and the
analysing nicol were mounted so as to be capable of rotation about the
direction of the reflected ray.
The adjustments were made as follows. The analyser and quarter- wave-
plate being removed, the mirror and polarizer were adjusted until the dark
spot was central in relation to the cross wires. A rotation of the mirror,
altering the angle of incidence, moves the spot vertically, while a rotation
of the polarizer moves it horizontally. The zero for the eye-nicol could have
been found by rotating the polarizer and then recovering the dark spot;
but in order to avoid risks of displacement, which might be fatal in such
a delicate inquiry, I preferred to leave the first nicol untouched, and to
504 ON REFLEXION FROM LIQUID SURFACES [185
depolarize the light by the introduction of a parallel plate of quartz. With'
the aid of this the analysing nicol could be set, and then the mica. If, with
the quartz plate in action, the spot is dark and central, all is well adjusted.
On removal of the quartz, the band is now seen in full perfection.
One of the difficulties in these experiments lay in the extreme sensitive-
ness of the liquid surfaces to tremor, a sensitiveness aggravated by the perfect
cleanliness required. It had been thought that it would suffice to mount
the apparatus upon a shelf attached to the walls of the building, and isolated
from the floor. But it appeared that the slightest touch upon the tangent-
screw of the divided circle, such as it is necessary to make at the moment
of observation, entailed a most distracting tremor. A remedy was found in
suspending the dish containing the liquid under examination independently
from the roof.
The work has been greatly retarded by want of sunshine. In order to be
more independent, I tried to work at the Royal Institution by the electric
light. But it appeared impossible to make any observations of value on
account of the tremor by which London is pervaded. Moreover the arc-light
is veiy inferior to sunshine for such a purpose.
The theory of the experiment is as follows. According to Fresnel's
formulas the ratios of the reflected to the incident vibrations are, for the two
planes of polarization, T and 8; in which the reality of T and 8 indicates
that there is no change of phase in reflexion (other than 180°). The
ellipticity is represented by the addition to T of iM, where M is small and
» = V( — !)• Thus if the incident light be polarized in the plane making
an angle a with the principal planes, the reflected vibrations may be
represented by
(T + iM)cosa, 8 sin a.
By the action of the mica, or other compensator, a relative change of
phase 7 is introduced. This is represented by writing for 8 sin a,
8 sin a (cos y + i sin 7).
Thus the vibration transmitted by the analyser, set at angle /9, is
cos a cos £ (T + iM) + 8 sin a sin /3 (cos y + i sin 7) ;
and the intensity of this is
(T cos a cos /3 + 8 sin a sin @ cos 7)"
+ ( M cos a cos /3 -I- 8 sin a sin ft sin 7)* (10)
In order that the light may vanish we must have both
T+ # tan a tan # cos 7 = 0, ....(11)
0 (12)
1892] IX THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 505
In the neighbourhood of the polarizing angle, if, S vary slowly, but T
varies rapidly. Hence, if 7 be given, we may regard (12) as determining
tan a tan /9, while (11) gives T, and thence the precise angle of reflexion for
the dark spot. If there be no ellipticity, 3/ = 0; whence tan a tan £ = 0,
T=0, indicating, as was to be expected, that the dark spot occurs at the
Brewsterian angle.
But this law is not universal. For if there be no compensator, 7 = 0,
and we have as the expression for the intensity,
(Tcos a cos ft + S sin a sin ft? + M* cos1 a cos1 ft.
Hence, if a and ft are small, the second term cannot be made to vanish,
and the brightness is a minimum when
T=-Stanatan£.
The position of the nearly dark spot is thus dependent upon o, ft, and
assumes the Brewsterian position only when either a. or ft vanishes.
In the case of a quarter-wave-plate, 7 = ± \ TT, and the equations become
T = 0, Jt = tanotan£ = + M,S. .................. (13)
The dark spot thus occurs at the Brewsterian angle, while tan a tan ft gives
the value of M ;S, viz. the k of Jamin. Accordingly if ft be set to any
convenient angle*, and a be then adjusted so as to bring the dark spot to the
central position, the product of the tangents of a and ft, each measured from
the zeros obtained in the preliminary adjustments, gives k.
But the following procedure not only affords greater delicacy, but makes
us comparatively independent of the positions of the zeros. Set ft, e.g.,
to +30J, and find o; then reset ft to — 30C. The new value of o would
coincide with the old one if there were no ellipticity : and the difference
of values measures o upon a doubled scale. If o' be the second value, so
that the difference is a —a, then
k = tan 30° tan £ (a' - a),
or as would suffice for all the purposes of the present investigation
-o) ........................ ...(14)
In practice several readings for a would be taken as quickly as possible,
ft being reversed between each. In this way there is the best chance of
distinguishing casual errors of observation from the results of progressive
changes in the condition of the surface under examination. For greater
security against error due to maladjustments, readings were often taken in all
four positions, differing from one another by 90°, of the quarter-wave mica.
The observed differences of a should be reversed in adjacent positions of the
* In my apparatus it was convenient to throw the fine adjustment upon a.
506 ON REFLEXION FROM LIQUID SURFACES [185
mica, and should be identical in the opposite positions, [i.e. those] obtained
from one another by rotation through 180°.
In the above reasoning 7 has been regarded as independent of X, but this
is of course only roughly true. If we neglect the dispersion of the mica, we
may take 7 = 70 + ^7, where 70 relates to the mean ray \, while
S7/7o = _gX/X0 ............................ (15)
If the mica be suitably chosen, %=±^7r.
On this principle of the variability of 7 may be explained an effect which
was puzzling when first observed. When the water-surface was rather highly
contaminated, it was found that the appearance of the spot varied according
to the choice of positions for the mica. In one position and its opposite
the spot was nearly free from colour*, while in the other two positions,
differing from the former by 90°, the coloration was intense. It was evident
that some cause was at work, in one case compensating, and in the other
doubling, the usual Brewsterian coloration.
If M0 be the mean value of M, the setting of the nicols will give, as
before,
tanatan/3 = + M0/S; ........................ (16)
while from (11),
T=±M0cosj ............................... (17)
The angle of reflexion corresponding to darkness is determined by (17), and
both sides of the equation are functions of X. For the mean ray y=±%7r,
and at the correct angle T = 0. For a neighbouring ray at the same angle
of reflexion we have for T,
dT
dx;sx'
and for cosy,
cos ^7r(l + 87/70) = -
Hence the condition of achromatism is
or by (15),
Thus if M0 be of the right magnitude, the colour will be compensated when
7 = ^7r, and doubled when 7 = — ITT, or vice versa.
When the colours were but little dispersed in the plane of incidence,
there could usually be observed on sufficiently contaminated surfaces a
dispersion laterally, indicating a variation of M with X. It was to be
* Attention is here fixed upon the central plane of incidence, colour on the right and left
of the spot being disregarded.
1892] IX THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 507
expected that M should be proportional to X~*. Not much more could be
done experimentally than to verify the direction and order of magnitude of
the effect. Thus it appeared that on a greasy surface the difference of
readings corresponding to £=±30° was greater when the settings were
made for the brown than for the blue side of the spot Of these the former,
due to the absence of blue, represents the setting proper to blue light.
The angles ± 30s were found suitable for ft. It was at first supposed that
advantage would accompany a smaller ft ; but in this case the spot was too
diffused in a horizontal direction to suit the dimensions of the bright field
employed. The adjustment of the spot to centrality (right and left) by
variation of a was then less certain. On the other hand, a too great increase
of ft throws excessive stress upon the readings of a.
The delicacy of the apparatus may be measured by the smallest error of a
visible on simple inspection. When the light was bright and the reflecting
surface steady, a setting for /8 = + 30° was visibly wrong on going over to
/?=-303, when the change afterwards found necessary in the setting of a
exceeded about 2*. Less than this could hardly be recognized on simple
inspection ; but the error of a single setting, arrived at by trials backwards
and forwards, appeared to be less than 1'. Thus the same readings, taken
to the nearest minute, were often recovered many times in succession : but
on other occasions larger differences were met with, and it was often difficult
to judge whether they were due to imperfect observation or to real changes
in the condition of the reflecting surface. In any case it will be a modest
estimate to suppose that a difference of one minute in a can be detected
on repetition. From this we should get, by (14), as the least observable
value of k,
i = ^ tan 30° x tan I7 = -00009.
Jamin's arrangements do not appear to have allowed of the determination
of values of k less than -001.
The first systematic experiments upon cleaned water-surfaces showed that
the ellipticity, if real, was pretty close to the limit of observation. At this
stage I expected to find the marked ellipticity of greasy water gradually
diminishing to zero as the purifying process was carried further, but
remaining always of the same sign, so long as it could be observed at all.
This anticipation was not completely verified- The larger differences of o,
found with ordinary water upon which camphor fragments were fully active,
amounting say to 407, rapidly diminished under the skimming process, so
that the final difference on the purest surfaces seemed just to escape direct
observation. It frequently happened that no displacement of the dark spot
relatively to the cross wires could be detected on the reversal of ft. But
when, in order to the highest accuracy, many sets of alternate readings
were taken, the difference would come out sometimes in one direction and
508
ON REFLEXION FROM LIQUID SURFACES
[185
sometimes in the other. A small difference of 2', or more, on the side of the
contaminated water was easily accepted as due to incomplete cleaning, but
I was for a time sceptical as to the significance of similar small differences in
the opposite direction. That these differences were not errors of observation
was soon apparent ; but I thought that they might be of instrumental origin,
due perhaps to some maladjustment.
The outstanding question was so small that it might perhaps have been
dismissed, but I was unwilling to stop without a determined attempt to get
to the bottom of it. The minute reversal of ellipticity stood its ground
in spite of repeated remountings of the apparatus; but I feared that it
might possibly be due to some, though I was unable to conjecture what,
defect in the optical parts themselves. When, however, the nicols at first
used were replaced by beautiful prisms made by Steeg and Renter and the
effect still remained, it had to be accepted as genuine, and the conclusion
was forced upon me that with some water-surfaces, and those presumably the
cleanest, there is a minute ellipticity in the opposite direction to that of
ordinary water, and such that the difference of settings for a amounted to
about 2'. This corresponds to k = + '0002. It will be understood that this
is a very minute quantity, but it is not without interest from a theoretical
point of view. The fact that k can be positive as well as negative implies
of course its possible evanescence. It is, I think, safe to say that some
samples of water-surfaces polarize light to perfection.
It will now be desirable to give a specimen of actual observations. The
one selected, principally on account of its completeness in respect to the
positions of the quarter- wave mica, is dated April 1, 1891, and relates to
a surface of tap-water, freshly drawn, and skimmed with the aid of the brass
hoop already described. The operation of skimming was repeated after each
readjustment of the mica. In the first column the direction of the arrow
Mica
Analyser,
Polarizer.
Separate Readings of a
(minutes)
Means
Difference
1
+ 30°
-30
31, 32, 32, 31
28, 29, 29, 29
183° 314'
183 28|
+ 2f
+ 30
-30
28, 31, 29, 30
32, 31, 32, 32
183 294
183 31|
-ij
t
+30
33, 33, 33, 33
183 33
1
-30
30, 32, 32, 32
183 31£
+4
-
+ 30
-30
28, 29, 29, 29
31, 32, 31, 29
183 28|
183 30|
-2
1892] IN THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 509
indicates the position of the mica. The second gives the readings of ft, the
third the individual readings of a, the minutes only being entered*. It will
be understood that the readings for ft = ± 30° were taken alternately ; the
first reading under j being 31', the second 28', the third 32', and so on. The
fourth column gives the means, and the fifth the difference of these means,
which represents ellipticity. The second and fourth differences, corresponding
to positions of the mica differing by a right angle from those of the first and
third, must have their signs reversed before combination for a final mean
difference. We get
After the last set the hoop was lifted, so as to allow the return of the
contamination. The readings then became
I +30° I 37, 44, 43 I
I -30° I 17, 14, 12 |.
It would appear that the first of these were premature, insufficient
time having been allowed for the contamination to spread. The difference,
reckoned as before, may now be taken to be about — 30', and is in the
opposite direction to the small effect of the clean surface. For the contami-
nated surface k=- '0026, and for the clean k = + '0001 8 f.
Although the above results, and others of a similar nature, obtained both
with tap-water and with distilled water, render it practically certain that
A: is positive for pure water, I do not regard with the same confidence the
numerical value above recorded. It is difficult to feel sure that the cleansing
was sufficient. A theoretical objection to the hoop method has already
been alluded to; and the more perfect methods depending upon the use of
convection currents^ are scarcely applicable here. Attempts were indeed
made to work with a surface cleaned by an ascending column of fluid, the
column being expanded by heat communicated to it from an immersed
platinum spiral, itself warmed by an electric current. But the readings were
not accordant ; and it appeared that the observations were prejudiced by the
deformations of the surface which are the necessary accompaniment of such
a flow. Doubts as to the perfection of the cleanliness actually attained lead
me to think it possible that the true value of k for an ideal water-surface
may be even twice as great as that actually found.
* In almost all the observations the settings were made by myself, and the readings of a
at the vernier by Mr Gordon. Without two observers the difficulties would have been much
increased.
t The observations so far did not of themselves determine which of the two surfaces has
the positive k according to Jamin's convention. It was evident, however, that it must be the
contaminated and not the clean surface which corresponds to Jamin's determination of a nega-
tive *. Subsequent observations upon reflexion from glass verified this assumption.
J See Roy. Soc. Proc. " On the Superficial Viscosity of Water," Vol. XLVIU. p. 133. [Vol. m.
p. 363.]
510 ON REFLEXION FROM LIQUID SURFACES [185
Opportunities for useful work upon clean surfaces have been very few, for
it is hopeless to attempt observations without a prospect of at least an hour's
almost uninterrupted sunshine. But shorter and more uncertain periods may
be utilized for observations upon contaminated surfaces, as these do not
demand the same care or amount of repetition. As an example of such
I will record the readings of June 6, from a water-surface slightly greased
with oil of cassia.
I +30° I 180° 2' I
-30= 180° 54' Band achromatu.
+30°
-30°
180° 47'
178° 59'
Band strongly coloured,
red above.
The difference of readings is here about - 50', giving k = - '0043. On
trial it was found that camphor fragments would just move. The above is
an example of the effect of the position of the mica upon the coloration
of the band, a subject already discussed.
Experiments were made with the object of comparing different kinds of
oil as to their relative effects, optically and upon camphor. It was found,
as had been expected, that cassia was more powerful optically than olive-oil.
Thus when camphor was nearly dead the difference of readings for olive-oil
was about — 30' and for cassia about — 48'.
Interest was felt in the behaviour of a saturated solution of camphor,
whose surface-tension is much lower than that of clean water. Observations
upon this liquid proved especially difficult, for the dark spot frequently
shifted laterally while under inspection, indicating temporary changes in the
ellipticity of the particular part of the surface in use. There is little doubt
that this complication is due to local evaporation under the influence of light
currents of air. As the camphor evaporates from any part of the surface, the
tension is momentarily raised, and the surface contracts. If the camphor only
were in question, there would probably be no attending optical disturbance,
but the local expansions and contractions of the surface lead to attenuation
and concentration of the greasy matter present. Under favourable circum-
stances the difference in the readings of a might be as low (numerically)
as — 6', and was perhaps due after all to residual greasy matter, other than
camphor. In any case the optical effect of the camphor is much less than
that of an oily film giving the same surface-tension.
With a strong solution of oleate of soda the difference of a could not be
reduced below - 25'. It is difficult to suppose that this can be due to a film
of foreign matter removable by skimming. But the amount of the ellipticity
is very low in relation to the surface-tension, which is only about one-third
of that of clean water. The value of k corresponding to the above readings
is — '0021, only about double of the smallest quantity appreciated by Jamin.
1892] IN THE NEIGHBOURHOOD OF THE POLARIZING ANGLE. 511
That die surface-tension has no definite relation to the ellipticitv is
abundantly evident. For example, camphor was quite active upon a surface
which gave a difference of readings of — 80*, corresponding to k = — -007.
On this occasion the surface had stood for some time without much pro-
tection, and it is possible that the effect may have been partly due to dust.
The last example that I will mention of aqueous solutions is a strong
brine. This gave a somewhat variable difference of about — 5', corresponding
to t = — -00042. In this case there seemed to be unusual difficulty in getting
the surface clean, so that the difference between the brine and pure water is
not improbably due to some secondary cause.
Moot of the available time was spent upon water in its various states,
not only on account of its intrinsic importance, but also because of the
presumably greater simplicity of a clean water-surface. The observations
are made in an atmosphere which contains no very small proportion of
aqueous vapour. When the liquid under examination has an affinity for
water, e.g. alcohol, it is difficult to form a precise idea as to what may be
the condition of the surface. Besides, the arrangements for skimming are
less easilv applied. On the other hand, the liquids of lower tension are leas
likely to acquire a film of grease. For alcohol, and also for petroleum, the
value of * is about + "0010.
The general conclusion to be drawn from these investigations is that the
ellipticitv of the liquids examined is very much less than was supposed by
Jamin, whose results for water and aqueous solutions were almost certainly
vitiated by the presence of greasy contamination. Thus the intensity of
reflexion from clean water is not much more than y^- part of that given
by Jamin Moreover, the value of t is positive, and not negative. It is
even possible that there would be no sensible ellipticitv for the surface of
a chemically pure body in contact only with its own vapour. But the
surfaces at bodies are the field of very powerful forces of whose action we
know but tittle ; and even if there be nothing that could be called chemical
rhaiijjT, file mere want of abruptness in the transition would of itself entail
a complication. There is thus no experimental evidence against the rigorous
applicability of FresneFs formula? to the ideal case of an abrupt transition
between two uniform transparent media.
POSTSCRIPT (October !!>
Solutions of saponine and gelatine, substances which confer the foaming
property, have been examined. With very small quantities the difference
of readings may amount to a degree, mot to be dimimuked by repeated
The value of * is thus - -005.
512 ON REFLEXION FROM LIQUID SURFACES. [185
The suspicion above suggested that the true value of k for clean water
may be numerically higher than is indicated by the results obtained with the
aid of the brass hoop has been verified by some observations upon surfaces
cleansed by heat. The water, as clean as possible, was contained in a large
shallow tin tray. By the application of gentle heat to the part of the tray
under optical examination any residual grease is driven off, in consequence
of the smaller tension of the warmer surface. If the whole surface is fairly
clean to begin with, a very moderate difference of temperature suffices to
keep the grease at bay. The difficulties of the experiment have so far
prevented a complete series of readings; but the following, obtained on
October 2, seem sufficient to establish the fact : —
+30°
-30°
43, 42, 42, 42
47, 47, 48, 47
+30° I 52, 51, 51, 50
I -30° I 47, 46, 47, 45, 48
At the conclusion of the second set the contamination was evidently
returning. It would seem that on the cleanest surfaces the difference of
readings may amount to 5', the necessity of readjustment on passing between
/3 = + 30° being obvious on simple inspection. Corresponding to this
k = + -00042.
186.
ON THE THEORY OF SURFACE FORCES. II.
COMPRESSIBLE FLUIDS.
[Philosophical Magazine, xxxin. pp. 209 — 220, 1892.]
Ix the first part of the paper published under the above title (Phil. May.
Oct. and Dec. 1890 [Vol. HI. p. 397]) the theory of Young and Laplace was
considered, and further developed in certain directions. The two leading
assumptions of this theory are (1) that the range of the cohesive forces,
though very small in comparison with the dimensions of ordinary bodies,
is nevertheless large in comparison with molecular distances, so that matter
may be treated as continuous ; and (2) that the fluids considered are incom-
pressible. So far as I am aware, there is at present no reason to suppose that
the applicability of the results to actual matter is greatly prejudiced by
imperfect fulfilment of (1); but, on the other hand, the assumption of
incompressibility is a somewhat violent one, even in the cases of liquids,
and altogether precludes the application of the theory to gases and vapours.
In the present communication an attempt is made to extend the theory to
compressible fluids, and especially to the case of a liquid in contact with its
own vapour, retaining the first assumption of continuity, or rather of ultimate
homogeneity. There will not be two opinions as to the advantage of the
extension to compressible fluids ; but some may perhaps be inclined to ask
whether it is worth while to spend labour upon a theory which ignores the
accumulated evidence before us in favour of molecular structure. To this
the answer is that molecular theories are extremely difficult, and that the
phenomenon of a change of state from vapour to liquid is of such extreme
importance as to be worthy of all the light that can be thrown upon it. We
shall see, I think, that a sufficient account can be given without introducing
the consideration of molecules, which on this view belongs to another stage
of the theory.
u in 33
514 ON THE THEORY OF SURFACE FORCES. [186
If p denote the ordinary hydrostatical pressure at any point in the interior
of a self-attracting fluid, p the density, and V the potential, the equation of
equilibrium is
dp = pdV. (1)
If, as we shall here suppose, the matter be arranged in plane strata, the
expression for the potential at any point is
(2)
£ + ...... , ........................ (3)
where p' is the density at a distance z from the point in question. Expanding
in series, we may write
z~P +
so that
where
K = 27T f "V (z} dz, L = TT \X z- -dr (z) dz. . . . . .(4)
JO JO
The integrals involving odd powers of z disappear in virtue of the relation
We may use (3) to form an expression for the pressure, applicable to
regions of uniform density (and potential). Thus, integrating (1) from a
place where p = pi to one where /> = />2> we have
= 2K (p22 - p,2) - fdp [2KP + 2Ldn-p/dz- + ...}
= K (p* - p*) - fdp {2Ld*p/dza- + ...}.
In the latter integral each term vanishes. For example,
and at the limits all the differential coefficients of p vanish by supposition.
Thus, in the application to regions of uniform density — uniform, that is,
through a space exceeding the range of the attractive forces —
p^; ........................... (5)
or, as we may also write it,
p = v + Kp\ ................................. (6)
where -or is a constant, denoting what the value of p would be in a region where
p = 0. We may regard -BT as the external pressure operative upon the fluid.
Equation (5) may also be obtained, less analytically, by the argument
OW THE THEORY OF SURFACE FORCES. 515
employed upon a former occasion*, and still more simply perhaps by
consideration of the forces operative upon the entire mass of fluid included
between the two strata in question regarded as a rigid body. It is very
important to remember that it ceases to apply at places where p « varying,
and that unless the strata are plane it requires correction even in its
application to region* of uniform density.
In the case of a uniform medium, (6) gives the relation between the
external pressure v, measured in experiments, and die total internal
l««i«ure pT found by adding to the former the intrinsic pressure Kp*. By
the constitution of the medium, independently of the self-attracting propertv.
there is a relation between p and p, and thence, by (6), between v and p. If
we suppxse that the medium, freed from self-attraction, would obey Boyle's
law. = l-, and
According to (7), when p is very small, «r varies as p. As p increases, w
increases with it, until p = t/ZKT when «• reaches a maximum. Beyond this
point v diminishes as p increases, and this without limit. The curve which
represents the relationship of v and p is a
parabola; and it is evident that all beyond Fi*r. i.
the vertex represents unstable conditions.
For at any point on this portion the pressure
diminishes as p increases. If, therefore, the
original uniformity were slightly disturbed,
without change of total volume, one part of
the fluid becoming denser and the other rarer
than before, the latter would tend still further
to expand and the former to contract. And
according to our equations the collapse would have no limit.
Points on the parabola between O and the vertex represent conditions
which are stable so far as the interior of the fluid is concerned but it may be
necessary to consider the action of the walls upon the fluid situated in their
neighbourhood. The simplest case is when the containing vessel, which may
be a cylinder and piston, exercises no attraction upon the fluid. The fluid
may then be compressed up to the vertex of the parabola without losing its
uniformity or becoming unstable. If, however, there be sufficient attraction
between the walls of the vessel and the fluid, instability leading to total
collapse will set in before the vertex is reached.
It wfll be seen that condensation to a denser state is easily explained,
without any reference to molecules, as a direct consequence of self-attraction
in a medium otherwise obeying Boyle's law. The objection that may be
* "On Laplace's Theotr of Capflluitj^ PKL Mmy. Oe*. 1883. [Tot n. p. ML]
33—2
516
ON THE THEORY OF SURFACE FORCES.
[186
raised at this point is rather that the explanation is too good, inasmuch as
it points to indefinite collapse, instead of to a high, but finite, contraction in
the condensed part.
A simple and well-known modification provides an escape from a con-
clusion which follows inevitably from a rigorous application of Boyle's law.
A provision is required to prevent extreme collapse, and this we may find in
the assumption that a constant must be subtracted from the volume in order
to obtain the quantity to which the pressure is proportional. In this case it
is usual and convenient to express the relation by the volume v of the unit
mass, rather than by the density. We have
p (v — b) = constant,
v — />) = constant,
.(8)
the well-known equation of Van der Waals. Here 6 is the smallest volume
to which the fluid can be compressed ; and under this law the collapse of the
fluid is arrested at a certain stage, equilibrium being attained when the
values of «r are again equal for the condensed and uncondensed parts of the
fluid.
According to (8), there are three values of v corresponding to a given nt.
Below the critical temperature the three values are real, and the isothermal
Fig. 2.
curve assumes the form ABCDEFGH (Fig. 2)
suggested by Prof. James Thomson. The part
DF is unrealizable for a fluid in mass, being
essentially unstable; but the parts AD, FH
represent stable conditions, so far as the
interior of the homogeneous fluid is con-
cerned. The line CG represents the (external)
pressure at which the vapour can exist in
contact with the liquid in mass, and the
isothermal found by experiment is usually
said to be HGECBA. This statement can
hardly be defended. If a vapour be com-
pressed from H through G, it can only travel along the straight line from
G towards E under very peculiar conditions. Apart from the action of the
walls of the containing vessel, and of suspended nuclei, the path from G to F
must be followed. The path from G to E implies that the vapour at G is in
contact with the liquid in mass. This is by supposition not the case ; and the
passage in question could only be the result of foreign matter whose properties
happened to coincide with those of the liquid. If the walls attract the vapour
less than the vapour attracts itself, they cannot promote condensation, and
the path HGF must be pursued. In the contrary case condensation must
begin before G is reached, although it may be to only a limited extent.
1892] OX THE THEORY OF SURFACE FORCES. 517
Probably the latter is the state of things usually met with in practice. So soon
as the walls are covered with a certain thickness of liquid, the path coincides
with a portion of GEC. and the angle at G is only slightly rounded off.
Similar considerations apply at the other end of the straight course. If
the liquid be expanded through C, it will not, in general, pass along CE, but
will continue to pursue the curve CD, and will even attain the limit D, if the
attraction of the walls upon the liquid be not less than that of the liquid
upon itself. In the contrary case separation will suddenly occur at a point
upon the wall, a bubble of vapour will be formed, and a point on the straight
line CE will be attained. It is thus scarcely conceivable that a fluid should
follow the broken course ABCEGH without some rounding of the corners, or
else of overshooting the points C, G. with subsequent precipitation upon the
line CEG.
A very important question is the position of the line CG. Maxwell*
showed that inasmuch as the area of the curve represents work performed
at a constant temperature, it must be the same for the complete course as
for the broken one. The line CG is therefore so situated as to cut off equal
areas above and below.
This discussion is of course quite independent of the precise form of the
relation between p and v. All that is necessary is such a modification of
Boyle's law at great densities as will secure the fluid against indefinite collapse
under the influence of its self-attraction.
We will now pass to the question of the transition from liquid to vapour.
still supposing the strata to be plane. This is a problem considered by
Maxwell in his article upon "Capillary Action" in the Encyclopedia Britan-
fiuxrh Dut his solution appears to me to be vitiated by more than one
oversight. By differentiation of (6) he obtains (with A written for K)
dp = ZApdp,
and thence, by (1),
so that
V = 2Ap + constant,
In the subsequent argument the identity of A with K is overlooked: and
the whole process is vitiated by the illegitimate differentiation of (6), which
is only applicable at places where p is not varying. The final result, which
appears to be arrived at without any assumption as to the physical connexion
between p and p, is thus devoid of significance.
* Sature. Vol. D. p. 358, 1875; SciaUijie Paptn, VoL n. p. 418.
t Scientific Papcrt, Vol. n. p. 560.
518 ON THE THEORY OF SURFACE FORCES. [186
Let us integrate (1) from a place in the vapour round which the density
has the uniform value pl to a place in the liquid where the uniform density
is pz. Thus
rd^=V2-V1 = 2K(p.2-pl}, ..................... (9)
J (i) p
by (3). The external pressure is uniform throughout, and may be denoted by
vf' ; and by (6),
Tx'=pl-Kp? = p,-Kp.? ...................... (10)
At places where p is varying, that is in the transitional layer, •&, as given by
(6), does not represent the external pressure ; but we will still regard it
as defined analytically by (6). Thus
P
By comparison of (9) and (11),
(»p dp
or on integration by parts,
-](2)+f(2)^ = o.
Pj(i> J O)/3
The values of or at the limits are the same, and have been denoted by ta-'.
Hence
I — ~~-^dp = 0 (13)
J (i) />"
Since dp/p"cc dv, this equation, obtained by purely hydrostatical methods
applied to the liquid and vapour and the layer of transition between them,
has precisely the same significance as Maxwell's theorem upon the position of
the line CG in J. Thomson's diagram. In that theorem OT represents the
external pressure that would be exerted by the fluid in various states of
uniform density, some of which are not realizable. In the subject of the
present investigation all the densities intermediate between those of the
vapour and liquid actually occur ; but, except at the extremities, OT no longer
represents external pressure.
The explanation of the stable existence in the transitional layer of certain
densities, which would be unstable in mass, depends of course upon the fact
that in the transitional layer the complete self-attraction due to the density
is not developed in consequence of the rapid variation of density in the
neighbourhood.
The distribution of density in the transitional layer, and the tension of
the surface, can only be calculated upon the basis of a knowledge of the
physical constitution of the fluid as expressed by the relation between p and
1892]
OX THE THEORY OF SURFACE FORCES.
519
p, and by the law of self-attraction. Poisson's contention that the surface-
tension cannot be found upon the supposition of an abrupt transition from
the liquid to its vapour is evidently justified ; and since the thickness of the
layer of transition is necessarily of the order of the range of the attraction,
it follows that the correction for gradual transition is not likely to be small.
A complete calculation of a particular case would be of interest, even on
rather forced suppositions : but the mathematical difficulties are considerable.
An approximate investigation might be conducted as follows : —
From (1) and (3),
If we neglect the terms in d*p dz\ &C-, this becomes
..(14)
where f(p) = fdp p is a function of p given by the constitution of the
medium.
Equation (14) may now be integrated by quadratures-
and
It is possible that a graphical process would be found suitable. Equation < 14i
determines the curvature at any point of the curve, representing the relation
between p and z, in terms of the coordinates and the slope.
When the relation between p and z is known, the calculation of the
surface-tension is a matter of quadratures. Probably the simplest way of
considering the question is to regard the five surface as spherical (liquid
within and vapour without), and to calculate the difference of pressures.
We have from (1),
, ...... (16)
z being measured outwards along the
radius. The question is thus reduced
to the determination of V at the various
points of the layer of transition, for all of
which z = R approximately. Let P (Fig. 3)
be a point at which 7 is to be estimated,
so that OP = z, and let AQB be a spherical
shell of radius z - f , of thickness rf£, and
of density p'. We have first to estimate
the potential dV of this shell at P.
520 ON THE THEORY OF SURFACE FORCES. [186
The element of mass at Q is
If, as before, </> (/) express the ultimate law of attraction, and
we have to multiply the above element of mass by II (/). Now
so that
The element of the potential is therefore
In the integration the limits of/* are AP and BP. The former is denoted
by f, and the latter may be identified with oc , since z or R is supposed to be
a very large multiple of the range of the forces. Accordingly for the potential
at P of the whole shell, we have
where, as usual,
(18)
To find the whole potential at P, (17) must be integrated with respect to f
from — oc to + oc , p' being treated as a function of £. As we need only
consider P near the layer of transition, z in (17) may be identified with R.
If the transition is continuous, we may expand p' in the series
dp ? d*p
p ==p~^dz+r2d^---'
and then at the point P,
'
1892] OX THE THEORY OF SURFACE FORCES. 521
where (as in Maxwell's " Capillary Action ")
When (19) is multiplied by dp dz and integrated across the whole layer of
transition, we get for the part independent of R,
simply, all the other terms in L, M, ... vanishing. Hence by (16), with
integration by parts.
The first term upon the right in (21) is the same as when the strata art-
plane. The second gives the capillary tension (T), and we conclude that when
the transition is continuous
From these results we see that "the existence of a capillary force is connected
with suddenness of transition from one medium to another, and that it may
disappear altogether when the transition is sufficiently gradual*.5'
The series (22) would probably suffice for the calculation of surface-tension
between liquid and vapour when once the law connecting p and z is known.
It is possible, however, that its convergence would be inadequate, and in this
respect it must certainly fail to give the result for an abrupt transition. In
the latter case, where the whole variation of density occurs at one place. (16)
becomes
p,-p1 = 2K(p.f-pl^-(pi-p1)V, .................. (28)
V relating to the place in question. And by (17)
, -f Pl) J"
Thus
, ..................... (24)
if T=-jrW(S)d?.(pt-PlY, ..................... (25)
* "On Laplace's Theory of Capillarity," Phil. Mag. October 1883. [Vol. n. p. 234]
522 ON THE THEORY OF SURFACE FORCES. [186
where (25) agrees with the value of the tension found for this case by
Laplace.
In the application to a sphere of liquid surrounded by an atmosphere of
vapour, equations (9), (11), (12) remain unchanged, in spite of the curvature
of the surface. If «r" denote the external pressure acting upon the vapour,
(26)
(27)
The symbol OT is still regarded as defined algebraically by (6), so that
TSV = TZ", vr.2 = <n" + 2T/R ...................... (28)
Integrating (12) by parts, we find
Pa Pi -M1)P"
or by (28),
In this equation w is a known function of p. If we compare it with (13),
where CT' represents the external pressure of the vapour in contact with a
plane surface of liquid, we shall be able to estimate the effect of the
curvature. It is to be observed that the limits of integration are not
the same in the two cases. If we retain p,, p2 for the plane surface, and
for the curved surface write pl + Sp1, p2 + fy>2, we have from (29)
or., — •or" OT! — •or" ~ f^ iff — •nr"
Oft -- - - a - OP, + --- « -
p22 p,2 JPl P2
or by (28),
.
The limits of integration are now the same as in (13), so that by
subtraction
( ' "\(l l
(-57 -Iff ) --- =~
V>, p
or
This is the value for the excess of vapour-pressure in equilibrium with a
convex surface that is given in Maxwell's " Heat " as a deduction from
Sir W. Thomson's principle.
The application of this principle may be extended in another direction.
When liquid rises in a capillary tube open above, the more attenuated vapour
at the upper level is in equilibrium with the concave surface, and the more
dense vapour below is in equilibrium with the plane surface of the liquid.
1892] ON THE THEORY OF SURFACE FORCES. 523
But, as was pointed out in the former paper, the rise of liquid is not limited
to the height of the meniscus. Above that point the walls of the tube are
coated with a layer of fluid, of gradually diminishing thickness, h^p than the
range of forces, and extending to an immense height. At every point the
layer of fluid must be in equilibrium teith the eapour to be found at the same
level. The data scarcely exist for anything like a precise estimate of the effect
to be expectedr but the argument suffices to show that a solid body brought
into contact with vapour at a density which may be much below the so-called
point of saturation will cover itself with a layer of fluid, and that this laver
may be retained in some degree even in what passes for a good vacuum. The
fluid composing the layer, though denser than the surrounding atmosphere of
vapour, cannot properly be described as either liquid or gaseous.
In our atmosphere fresh surfaces, ejj. of split mica or of mercury, attract
to themselves at once a coating of moisture. In a few hours this is replaced,
or supplemented, by a layer of grease, which gives rise to a Large variety of
curious phenomena. In the case of mica the fresh surface conducts electricity,
while an ok! surface, in which presumably the moisture has been replaced by
grease, insulates well
187.
ON THE RELATIVE DENSITIES OF HYDROGEN AND
OXYGEN. II.
[Proceedings of the Royal Society, L. pp. 448—463, 1892.]
IN a preliminary notice upon this subject*, I explained the procedure by
which I found as the ratio of densities 15'884. The hydrogen was prepared
from zinc and sulphuric, or from zinc and hydrochloric, acid, and was liberated
upon a platinum plate, the generator being in fact a Smee cell, enclosed in
a vessel capable of sustaining a vacuum, and set in action by closing the
electric circuit at an external contact. The hydrogen thus prepared was
purified by corrosive sublimate and potash, and desiccated by passage
through a long tube packed with phosphoric anhydride. The oxygen was
from chlorate of potash, or from mixed chlorates of potash and soda.
In a subsequent paper on the " Composition of Water -f-," I attacked the
problem by a direct synthesis of water from weighed quantities of the two
component gases. The ratio of atomic weights thus obtained was 15'89.
At the time when these researches were commenced, the latest work
bearing upon the subject dated from 1845, and the number then accepted
was 15'96. There was, however, nothing to show that the true ratio really
deviated from the 16 : 1 of Prout's law, and the main object of my work was
to ascertain whether or not such deviation existed. About the year 1888,
however, a revival of interest in this question manifested itself, especially in
the United States, and several results of importance have been published.
Thus, Professor Cooke and Mr T. W. Richards found a number which, when
corrected for an error of weighing that had at first been overlooked, became
15-869.
The substantial agreement of this number with those obtained by myself
seemed at first to settle the question, but almost immediately afterwards
* Roy. Soc. Proc. Vol. XLIII. p. 356, February, 1888. [Vol. in. p. 37.]
t Roy. Soc. Proc. Vol. XLV. p. 425, February, 1889. [Vol. in. p. 233.]
1892] OX THE RELATIVE DENSITIES OF HYDROGEN* AND OXYGEN. 525
there appeared an account of a research by Mr Reiser, who used a method
presenting some excellent features, and whose result was as high as 1.V94H.
The discrepancy has not been fully explained, but subsequent numbers agree
more nearly with the lower value. Thus. Noyes obtains 15*896, and Dittmar
and Henderson give 15'866.
I had intended further to elaborate and extend my observations on the
synthesis of water from weighed quantities of oxygen and hydrogen, but
the publication of Professor E. W. Morley's masterly researches upon the
" Volumetric Composition of Water* " led me to the conclusion that the best
contribution that I could now make to the subject would be by the further
determination of the relative densities of the two gases. The combination
of this with the .number 2'0002t, obtained by Morley as the mean of
astonishingly concordant individual experiments, would give a better result
for the atomic weights than any I could hope to obtain directly.
In all work of this sort, the errors to be contended with may be classed as
either systematic or casual. The latter are eliminated by repetition, and are
usually of no importance in the final mean. It is systematic errors that are
most to be dreaded. But although directly of but little account, casual errors
greatly embarrass a research by rendering difficult and tedious the detection
of systematic errors. Thus, in the present case, almost the only source of
error that can prejudice the final result is impurity in the gases, especially in
the hydrogen. The better the hydrogen, the lighter it will prove : but the
discrimination is blunted by the inevitable errors of weighing. After perhaps
a week's work it may become clear that the hydrogen is a little at fault, as
happened in one case from penetration of nitrogen between the sealed-in
platinum electrodes and the glass of the generator.
Another difficulty, which affects the presentation of results, turns upon
the one-sided character of the errors most to be feared. As has been said,
impure hydrogen can only be too heavy, and another important source of
error, depending upon imperfect establishment of equilibrium of pressure
between the contents of the globe and the external atmosphere, also works
one-sidedly in the same direction. The latter source of error is most to be
feared immediately after a re-greasing of the tap of the globe. The super-
fluous grease finds its way into the perforation of the plug, and partially
blocks the passage, so that the six minutes usually allowed for the escape of
the initial excess of pressure in the globe may become inadequate. Partly
* Amtr. Journ. Sci. March, 1891.
t It should not be overlooked that this number is difficult to reconcile with views generally
held as to the applicability of Avogadro's law to very rare gases. From what «e know of the
behaviour of oxygen and hydrogen gases under compression, it seems improbable that volumes
which are as 2-0002 : 1 under atmospheric conditions would remain as 2:1 upon indefinite
expansion. According to the formula of Van der Waals, a greater change than this in the ratio
of volumes is to be expected. [1901. In later experiments Morley obtained 2-0027.]
526 ON THE RELATIVE DENSITIES [187
from this cause and partly from incomplete washing out of nitrogen from the
generator, the first filling of a set was so often found abnormally heavy that
it became a rule in all cases to reject it. From these and other causes, such
as accidental leakages not discovered at the time, it was difficult to secure a
set of determinations in which the mean really represented the most probable
value. At the same time, any arbitrary rejection of individual results must
be avoided as far as possible.
In the present work two objects have been especially kept in view. The
first is simplicity upon the chemical side, and the second the use of materials
in such a form that the elimination of impurities goes forward in the
normal working of the process. When, as in the former determinations, the
hydrogen is made from zinc, any impurity which that material may contain
and communicate to the gas cannot be eliminated from the generator; for
each experiment brings into play a fresh quantity of zinc, with its accom-
panying contamination. Moreover, the supply of acid that can be included
in one charge of the generator is inadequate, and good results are only
obtained as the charge is becoming exhausted. These difficulties are avoided
when zinc is discarded. The only material consumed during the experiments
is then the water, of which a large quantity can be included from the
first. On the other hand, the hydrogen liberated is necessarily contaminated
with oxygen, and this must be removed by copper contained in a red-hot
tube. In the experiments to be described the generator was charged with
potash*, and the gases were liberated at platinum electrodes. In the case
of a hydrogen filling the oxygen blew off on one side from a mercury seal, and
on the other the hydrogen was conveyed through hot tubes containing copper.
The bulk of the aqueous vapour was deposited in a small rlask containing
strong solution of potash, and the gas then passed over solid potash to a long
tube packed with phosphoric anhydride. Of this only a very short length
showed signs of being affected at the close of all operations.
With respect to impurities, other than oxygen and oxides of hydrogen,
which may contaminate the gas, we have the following alternatives. Either
the impurity is evolved much more rapidly than in proportion to the
consumption of water in the generator, or it is not. If the rate of evolution
of the impurity, reckoned as a fraction of the quantity originally present,
is not much more rapid than the correspondingly reckoned consumption of
water, the presence of the impurity will be of little importance. If on the
other hand, as is probable, the rate of evolution is much more rapid than
the consumption of water, the impurity is soon eliminated from the residue,
and the gas subsequently generated becomes practically pure. A similar
argument holds good if the source of the impurity be in the copper, or even
* At the suggestion of Professor Morley, the solution was freed from carbonate or nearly so,
by the use of baryta, of which it contained a slight excess.
1892]
OF HYDROGEN AND OXYGEN.
527
in the phosphoric anhydride; and it applies with increased force when at
the close of one set of operations the generator is replenished by the mere
addition of water. It is, however, here assumed that the apparatus itself is
perfectly tight.
Except for the reversal of the electric current, the action of the apparatus
is almost the same whether oxygen or hydrogen is to be collected. In the
latter case the copper in the hot tubes is in the reduced, and in the former
case in the oxidised, state. For the sake of distinctness we will suppose that
the globe is to be filled with hydrogen.
The generator itself (Fig. 1) is of the U-form, with unusually long
branches, and it is supplied from Grove cells with about 3 amperes of
electric current. Since on one side the oxygen
blows off into the air, the pressure in the generator
is always nearly atmospheric. Some trouble has
been caused by leakage between the platinum
electrodes and the glass. In the later experiments
to be here recorded these joints were drowned
with mercury. On leaving the generator the
hydrogen traverses a red-hot tube of hard glass
charged with copper*, then a flask containing a
strong solution of potash, and afterwards a second
similar hot tube. The additional tube was intro-
duced with the idea that the action of the hot
copper in promoting the union of the hydrogen
with its oxygen contamination might be more
complete after removal of the greater part of the
oxygen, whether in the combined or in the uncom-
bined state. From this point onward the gas
was nearly dry. In the earlier experiments the
junctions of the hard furnace tubes with the soft glass of the remainder of the
apparatus were effected by fusion. One of these joints remained in use, but
the others were replaced by india-rubber connexions drowned in mercury.
It is believed that no leakage occurred at these joints ; but as an additional
security a tap was provided between the generator and the furnace, and was
kept closed whenever there was no forward current of hydrogen. In this way
the liquid in the generator would be protected from any possible infiltration
of nitrogen. Any that might find its way into the furnace tubes could easily
be removed before the commencement of a filling.
Almost immediately upon leaving the furnace tubes the gas arrives at
a tap which for distinctness may be called the regulator. In the generator
* The copper must be free from sulphur; otherwise the contamination with sulphuretted
hydrogen is somewhat persistent.
528 ON THE RELATIVE DENSITIES [187
and in the furnace tubes the pressure must be nearly atmospheric, but in the
globe there is (at the commencement) a vacuum. The transition from the
one pressure to the other takes place at the regulator, which must be so
adjusted that the flow through it is approximately equal to the production
of gas. At first the manipulation of the regulator was a source of trouble
and required almost constant attention, but a very simple addition gave the
desired control. This was merely a long wooden arm, attached to the plug,
which served both as a lever and as an indicator. Underneath the pointed
extremity was a small table to which its motions could be referred. During
the first two-thirds of a filling very little readjustment was needed, and the
apparatus could be left for half an hour with but little fear of displacing too
much the liquid in the generator. Towards the close, as the motive force fell
off, the tap required to be opened more widely. Sometimes the recovery of
level could be more conveniently effected by insertion of resistance into the
electric circuit, or by interrupting it altogether for a few minutes. Into
details of this kind it is hardly necessary to go further.
From the regulator the gas passed to the desiccating tubes. The first
of these was charged with fragments of solid potash, and the second with
a long length of phosphoric anhydride. Finally, a tube stuffed with glass
wool intercepted any suspended matter that might have been carried
forward.
The connexion of the globe with the generator, with the Toppler, and
with the blow-off, is shown in the accompanying Fig. 2. On the morning
of a projected filling the vacuous globe would be connected with the free end
of the stout-walled india-rubber tube, and secured by binding wire. The
generator being cut off, a high vacuum would be made up to the tap of the
globe. After a couple of hours' standing the leakage through the india-rubber
and at the joints could be measured. The amount of the leakage found
in the first two hours was usually negligible, considered as an addition to
a globeful of hydrogen, and the leakage that would occur in the hours
following would (in the absence of accidents) be still smaller. If the test
were satisfactory, the filling would proceed as follows: —
The electric current through the generator being established and the
furnace being heated, any oxygen that might have percolated into the drying
tubes had first to be washed out. In order to do this more effectively, a
moderate vacuum (of pressure equal to about 1 inch of mercury) was
maintained in the tubes and up to the regulator by the action of the
pump. In this way the current of gas is made very rapid, and the half-hour
allowed must have been more than sufficient for the purpose. The generator
was then temporarily cut off, and a high vacuum produced in the globe
connexion and in the blow-off tube, which, being out of the main current
of gas, might be supposed to harbour impurities. After this the pump would
OF HTDBOGEX AXD OXYGKX.
=o -.
be cot off, the connexion with the generator re-established, and, finally toe
tap of the globe cautiously opened.
Kg •;.
TO GENERATOR.
-c =.v=.
The operation of filling usually occupied from two to three hours. When
the gas began to blow off under an excess of pressure represented bv about
half an inch of mercury, the blow-off cistern was lowered so as to leave the
extremity of the tube free. For two minutes the current of gas from the
generator was allowed to flow through, after which the generator was cut
off, and the globe left in simple communication with the atmosphere, until it
was supposed that equilibrium of pressure had been sufficiently established.
Doubts have at various times been felt as to the interval required for this
purpose. If too little time is allowed, there will remain an excess of pressure
in the globe, and the calculated weight of the filling will come out too high.
On the other hand, an undue prolongation of the time might lead to a
diffusion of air back into the globe. In a special experiment no abnormal
weight was detected after half-an-hours communication, so that the danger
on this side appeared to be small. When the passages through the taps
were free from ^UJMHJ, one or two minutes sufficed for the establishment
of equilibrium, but there was always a possibility of a partial obstruction.
In the results to be presently given four minutes were allowed after the
separation from the generator. It may be remarked that a part of any
minute error that may arise from this source will be eliminated in the
comparison with oxygen, which was collected under like conditions.
R ,,T 34
530
ON THE RELATIVE DENSITIES
[187
The reading of the barometers and thermometers at the moment when
the tap of the globe was turned off took place as described in the former
paper. The arrangements for the weighings were also the same.
In the evacuations the process was always continued until, as tested
by the gauge of the Toppler after at least a quarter of an hour's standing,
the residue could be neglected. Here, again, any minute error would be
eliminated in the comparison of the two gases.
In the case of oxygen, the errors due to contamination (even with
hydrogen) are very much diminished, and similar errors of weighing tell
very much less upon the proportional agreement of the final numbers.
A comparison of the actual results with the two kinds of gas does not,
however, show so great an advantage on the side of the oxygen as might have
been expected. The inference appears to be that the individual results are
somewhat largely affected by temperature errors Two thermometers were,
indeed, used (on opposite sides) within the wooden box by which the globe
is surrounded, and they could easily be read to within ^° C. But in other
respects, the circumstances were unfavourable in consequence of the presence
in the same room of the furnace necessary to heat the copper. An error
of ±0'1°C. in the temperature leads to a discrepancy of 1 part in 1500 in
the final numbers. Some further elaboration of the screening arrangements
actually employed would have been an improvement, but inasmuch as the
circumstances were precisely the same for the two gases, no systematic
error can here arise. The thermometers were, of course, the same in the
two cases.
The experiments are grouped in five sets, two for oxygen and three for
hydrogen. In each set the work was usually continued until the tap of the
globe required re-greasing, or until, owing to a breakage or to some other
accident, operations had to be suspended.
Oxygen.
1891
Weight
Bar.
temp., F.
Globe
temp., C.
June 29 ...
grams
2-5182
70
2Q-85
July 2
2-5173
69
20-60
July 4
2-5172
67i
1975
July 6
2-5193
70*
21-40
July 9 .
2-5174
64
17 -60
July 10
2-5177
65 1
19-05
Mean
2-51785
68
20°
1892]
OF HYDROGEN AND OXYGEN.
531
The six fillings were all independent, except that of July 6, when the
bulk of the oxygen remaining from the previous filling was not removed.
It so happens that this case shows the greatest discrepancy, but there seems
to be no sufficient reason for rejecting it.
Hydrogen.
1891
Weight
Bar.
temp., F.
Globe
temp., C.
July 31
gram
0-15807
fiOi
August 4
August 6
0-15816
0-15811
0-15803
ou$
65
66^
65
18-00
19-20
18"15
August 11
0-15801
66
19-15
August 13
0-15809
68A
20-10
Mean
0-15808
65
18°
Hydrogen.
Hydrogen.
1891
Weight
Bar.
temp., F.
Globe
temp., C.
September 22
September 24
gram
0-15800
0-15820
58
6H
0
14-5
16-3
September 28
0-15792
62
17-6
September 30
October 2
0-15788
0-15783
63*
62
18-1
17-3
Mean
0-15797
61
17°
1891
Weight
Bar.
temp., F.
Globe
temp., C.
October 26
October 28
October 31
gram
0-15807
0-15801
0-15817
55
56
50
o
13-30
14-00
10-95
November 3
0-15790
53*
12-10
November 5
0-15810
55
12-00
0-15798
50
10-70
November 10
November 13
0-15802
0-15807
48
55£
9-30
12-70
Mean
0-15804
53
12°
34—2
532
ON THE RELATIVE DENSITIES
[187
Oxygen.
1891
Weight
Bar.
temp., F.
Globe
temp., C.
November 30
December 3
grams
2-5183
2-5168
53
56
0
12-15
13-55
December 5
2-5172
56i
14'15
December 7
2-5181
58A
14-70
December 8
2-5156
51
11-15
Mean
2-5172
55
13°
In almost every case the weight of the globe full is compared with the
mean of the immediately preceding and following weights empty. The
numbers recorded in the second column are derived from the readings of the
balance by the introduction of corrections —
(1) For the errors of the weights themselves, found by a systematic
comparison, only relative values uncorrected for buoyancy being required.
(2) For the deviation of the mean* barometric reading at the time of
filling from 30 inches (as read upon the vernier).
(3) For the deviation of the temperature of the barometers (Column 3)
from 60° F.
(4) For the deviation of the temperature of the gas (as read upon the
thermometers) from 12° C.
As an example, I will take in detail the calculation for the hydrogen
filling of October 26. After the evacuation of October 24, the working
globe (14) with its compensating volume piece and 0'4778 gm. stood on the
left of the balance with globe (11) on the right. The position of equilibrium
of the pointer, as determined after four different releasements, each observed
in the usual manner, was 19'02 scale divisions. In like manner, after the
evacuation of October 27, with the same weights in use, the equilibrium
position of the pointer was 18'46. After the filling of October 26, the
weights associated with (14) were 0'3220 gm. instead of 0'4778 ; and the
pointer reading was 20'08. So far as the weights are concerned, the value
of the hydrogen would be 0'4778 — 0'3220, or 01558 gm. ; but to this we
must add a correction corresponding to T34 scale divisions, being the
difference between 20'08 and £ (19'02 + 18'46). At the time in question,
the value of a scale division was 0'00020 gm., so that we obtain —
0-1558 + 0-00027 = 0-15607.
There were two barometers.
1892]
OF HYDROGEN AND OXYGEN.
The particular weights in use on this occasion were such that no
correction is necessary in order to allow for their errors.
The mean barometer reading at the time of filling was 29742, so that
the factor required on this account is 30 : 29'742. The correction for
temperature of gas is from 13'3 to 12°.
Log 0-15607 =1-19332
For barometer 0'00375
For temperature 0*00198
Log 0-15814 =1-19905
To this a correction for the temperature of the barometer has still to be
applied. For 1" F. the correcting factor is (1 - 0'000089), or for 5° F.
(1 - 0-000445). From 0*15814 we are thus to subtract 0*00007, giving the
tabular number 0'15807.
A further minute correction to the mean of each set may be made for
the temperature of the glass. A warm globe is larger than a cold one, and
consequently holds more gas. If we suppose that the volume expansion of
the glass per degree C. is 0'000025, we find, corrected to 12 C.—
Hydrogen.
1891
Weight
Bar.
temp., F.
Globe
temp., C.
Corrected
to 12-
Julv
gram
0-15808
0
65
18
gram
0-158056
September
October
0-15797
0-15804
61
53
12
0-157950
0-158040
•
Mean
60
16
0-158015
H
,
Oxygen.
1891
Weight
Bar.
temp., F.
Globe
temp., C. ,
Corrected
to 12s
June
grams
2-51785
68
20
grams
2-51735
November
2-51720
55
13
i -
2-51713
Mean
61*
164
2-57124
The means here exhibited give the weights of the two gases as they
would be found with the globe at 12s C., and the barometers at 60° F. and
at 30 inches. The close agreement of the mean temperatures for the two
534 ON THE RELATIVE DENSITIES [187
gases shows how little room there is for systematic error dependent upon
imperfections in the barometers and thermometers. But the results still
require modification before they can be compared with the view of deducing
the relative densities of the gases.
In the first place, there is a systematic, though minute, difference in the
pressures hitherto considered as corresponding. The terminal of the blow-off
tube is 33 inches below the centre of the globe at the time of filling. In the
one case this is occupied by hydrogen, and in the other by oxygen. If we
treat the latter as the standard, we must regard the hydrogen fillings as
taking place under an excess of pressure equal to |f of the weight of a
column of oxygen 33 inches high ; and this must be compared with 30 inches
of mercury. Hence, if we take the sp. gr. of oxygen under atmospheric
conditions at 0-0014, and that of mercury at 13*6, the excess of pressure
under which the hydrogen was collected is as a fraction of the whole
pressure
33 15 0-0014
and 0-000106 x 0'158 = 0'000017. This, then, is what we must subtract from
the weight of the hydrogen on account of the difference of pressures due to
the gas in the blow-off tube. Thus
H = 0-157998, 0 = 2-51724.
[These numbers are not quite comparable with those given in the former
communication, inasmuch as the standard temperature then used for the
barometers was 55° F. Reduced so as to correspond to 60°, the former
numbers become
H = 0-15797, O = 2-5174.
The agreement is satisfactory, especially when it is remembered that both
gases were prepared by different methods in the two sets of experiments. —
Feb. 17.}
But there is still another and a more important correction to be introduced.
In my former paper it was shown that when the weighings are conducted in
air the true weight of the gas contained in the globe is not given by merely
subtracting the weight of the globe when empty from the weight when full.
When the globe is empty, its external volume is less than when full, and
thus, in order to obtain the true weight, the apparent weight of the gas must
be increased by the weight of air whose volume is equal to the change of
volume of the globe.
In order to determine the amount of this change of volume, the globe
is filled to the neck with recently boiled distilled water, and the effect is
observed upon the level in the stem due to a suction of, say, 20 inches of
mercury. It is not advisable to carry the exhaustion much further for fear
" -" OF HTDBOGEX JLSTD OXTGEf.
of approaching too nearly the point at winch bobbles of vapour
formed internally. La the earlier experiments, described in the
note, the upper surfece of the liquid was in the stem of the globe itself (below
the tap), and the- only difficulty lay in the accurate estimation of a
of volume oocnning in a wide and somewhat irregular tube. Ihe
employed was to produce, by introduction of a weighed quantity of
a rise of level equal to that caused by the suction.
The advantage df this procedure lay in the avoidance of joints and of the
tap itself, but, for the reasons given, the readings were not quite so accurate
as might be desired. I wished, therefore, to supplement, if possible, the
former determination by one in which the change of volume occurred in a
tube narrower and of better shape. With this object in view, the stem of the
globe was prolonged by a graduated tubular pipette attached with the aid
of india-rubber. The tubes themselves were treated with gutta-percha
cement., and brought almost into contact. It had hardly been expected
that the joint would prove unyielding under the applied suction, but it was
considered that the amount of the yielding could he estimated and alk>w«d
for by operations conducted wiitk top domed. The event, however, proved
that the yielding at the joint was scarcely, if at all, perceptible.
The pipette, of bore such that 16 cm. corresponded to 1 CJCL, was graduated
to CKMU and was read by estimation to ODOlcuc. In order the better t<c»
eliminate the changes due to temperature, readings under atniKi^phrrk-
pressnre, and under a suction of 20 inches of mercury, were alternate*!. On
January 28, 1892, a first set gave 0648-0300 = 0348, a s«xa>d gave
0*645-0316=03485, and a third gave 0ti75 - 0326 = 0349. Similar
operations with tap closed* gave no visible movement.
The result of the day s experiments was thus 034&5 for 20 inches, or
0528 for 30 inches,, suction. Similar experiments on January 28, at a
different part of the graduation, gave 0526. On this day the yielding
with tap closed was just visible, and was estimated at CNM»L As a mean
result, we may adopt 0524cjc. The graduation of the pipette was subse-
quently verified by weighing a thread of mercury that occupied a measured
length.
A part of the above-measured volume is due to the expansion of the
water when the pressure is relieved. We may take this at 0000047 of the
volume per atmosphere. The volume itself may be derived with sufficient
accuracy lor the present purpose from the weight of its oxygen contents It
k 2-517//000137, or 1837 CA. The expansion of the water per itmnaphpir
is thus 0000047 x 1837, or 0087 CXL This is to be subtracted from 0524.
and leaves 0437 CJR. This number applies strictly to the
- Foren*!
536 ON THE RELATIVE DENSITIES [187
within the glass, but the change in the external volume of the globe will be
almost the same*.
The correction now under consideration is thus the weight of 0'437 c.c.
of air at the average temperature of the balance room. The density of
this air may be estimated at 0'00122 ; so that the weight of 0'437 c.c. is
O000533 gm. This is the quantity which must be added to the apparent
weights of the gases. The former estimate was 0'00056 gm. The finally
corrected weights are thus
H = 0158531, 0 = 2-51777;
and for the ratio of densities we have
15-882.
This corresponds to a mean atmospheric condition of pressure and tempe-
rature.
If we combine the above ratio of densities with Professor Morley's ratio
of volumes, viz. 2'0002 : 1, we get, as the ratio of atomic weights, 15'880.
If we refer to the table, we see that the agreement of the first and third
series of hydrogen weighings is very good, but that the mean from the second
series is decidedly lighter. This may have been in part fortuitous, but it is
scarcely probable that it was so altogether. Under the circumstances we
can hardly reckon the accuracy of the final results as closer than ^V?r
A word should perhaps be said upon a possible source of systematic
error, viz. mercury vapour. There is no doubt that hydrogen passed over
mercury takes up enough to cause a slow and superficial, but quite distinct,
discoloration of sulphur over which it subsequently flows. In the experi-
ments here recorded, the gas did not, indeed, flow over mercury in mass, but,
inasmuch as mercury was used to secure the tightness of some of the joints,
it is difficult to feel sure of its absence. Again, in evacuations conducted
with a mercury pump can the vacuum be regarded as free from mercury
vapour, which, it must be remembered, would not show itself upon the
gauge of the Toppler? If both the hydrogen and the "vacuum" were
saturated with mercury vapour, the result of the weighings would, according
to Dalton's law, be free from its influence. The same may be said of any
volatile impurity arising from the grease f upon the stopcocks. As the
* For a spherical shell of glass of uniform thickness and with elastic constants following
Poisson's law, the ratio of the difference of the internal and external expansion to either of them
is 4t/3a, where t is the thickness of the shell, and a the mean radius. In the present application
the value of ajt, deduced from the measured circumference and from the weight of glass, is
about 110.
{Perhaps an arrangement in which the external volume is directly measured would have been
preferable. No allowance for expansion of water would then be needed. — Feb. 17.}
t Composed of vaseline and beeswax.
1892] OF HYDROGEN JLSD OXYGEK. 537
matter stands, the results must, I think, be regarded as affected with a
possible error amounting to a fraction of the weight of mercury vapour at
the temperatures employed. But this is probably a very small quantity.
According to Hertz*, the vapour-pressure of mercury at 15° C. would be
about O001 mm. If this be correct, the weight of mercury vapour in an
atmosphere of hydrogen would be as a fraction of the latterf
OOOl ___ 1
It appears that in an investigation of hydrogen aiming at an accuracy
of 1/10,000 the question of mercury vapour requires very careful con-
sideration.
The accompanying table of results found by various experimenters may
be useful for comparison: —
I
t_
Date
A; L.::
•= -. :-•':.>
1 •- : -.:.--
Dumas. ..
1S42
15-96
Regnault
1845
15-96
Barfeish
_
15-S34
Cooke and
Reiser. .
Richards... J
: ss«
15-869
15-949
Rarleieh
»
1889
15-89
Xoyes
Dittmar
1890
1890
15-896
15-866
-
Morfey .
1891
15-879
_
Ledoc
1891
15-905
Rarleish
1893
15-882
In conclusion, I must express my obligations to Mr Gordon, who has
me throughout. The work has been unusually tedious, partly from
its inherent nature, requiring as it does a certainty of 0-1 milligram in the
weighings, and still more from the constant liability to accidents, which may
render nugatory a large amount of preparatory work.
[1901. For further investigations respecting the density of gases see
Proc. Boy. Soc. un. p. 481, 1892; LY. p. 340, 1894; LVIL p. 265, 1895.
These papers will be included in VoL IT.]
* Wifd. Amm. VoL xrn. p. 199.
t [1901. IB the anginal the number* «« 200 and 3800. I o« the eooteOa. to
Pnrf. Japp.]
188.
SUPERHEATED STEAM.
[Nature, XLV. pp. 375, 376, 438, 512, 1892.]
I HAVE noticed a curious misapprehension, even on the part of high
authorities, with respect to the application of Carnot's law to an engine
in which the steam is superheated after leaving the boiler. Thus, in his
generally excellent work on the steam-engine *, Prof. Cotterill, after explaining
that in the ordinary engine the superior temperature is that of the boiler,
and the inferior temperature that of the condenser, proceeds (p. 141): " When
a superheater is used, the superior temperature will of course be that of the
superheater, which will not then correspond to the boiler pressure."
This statement appears to me to involve two errors, one of great
importance. When the question is raised, it must surely be evident that,
in consideration of the high latent heat of water, by far the greater part
of the heat is received at the temperature of the boiler, and not at that of the
superheater, and that, of the relatively small part received in the latter stage,
the effective temperature is not that of the superheater, but rather the mean
between this temperature and that of the boiler. An estimate of the possible
efficiency founded upon the temperature of the superheater is thus immensely
too favourable. Superheating does not seem to meet with much favour in
practice ; and I suppose that the advantages which might attend its judicious
use would be connected rather with the prevention of cylinder condensation
than with an extension of the range of temperature contemplated in Carnot's
rule.
If we wish effectively to raise the superior limit of temperature in a
vapour-engine, we must make the boiler hotter. In a steam-engine this
means pressures that would soon become excessive. The only escape lies in
* Second edition (Spon: London, 1890). [1900. Prof. Cotterill explained subsequently
(Nature, Vol. XLV. p. 414, 1892) that I had misunderstood the passage in question.]
1892] SUPERHEATED STEAM. 539
the substitution for water of another and less volatile fluid. But, of liquids
capable of distillation without change, it is not easy to find one suitable
for the purpose. There is, however, another direction in which we may
look. The volatility of water may be restrained by the addition of saline
matters, such as chloride of calcium or acetate of soda. In this way the
boiling temperature may be raised without encountering excessive pressures,
and the possible efficiency, according to Carnot, may be increased.
The complete elaboration of this method would involve the condensation
of the steam at a high temperature by reunion with the desiccating agent,
and the communication of the heat evolved to pure water boiling at nearly
the same temperature, but at a much higher pressure. But it is possible
that, even without a duplication of this kind, advantage might arise from the
use of a restraining agent. The steam, superheated in a regular manner,
would be less liable to premature condensation in the cylinder, and the
possibility of obtaining a good vacuum at a higher temperature than usual
might be of service where the supply of water is short, or where it is desired
to effect the condensation by air.
[1900. The indications given above were, it would seem, too concise.
See a further correspondence (Nature, Vol. XLV. pp. 413 — 414, 438, 486, 510).
The proposal was to condense the pure steam reversibly by reunion with
the desiccating agent. To this end it would be necessary to have a supply
of solution of the same strength as in the boiler, but at the temperature
and pressure of condensation. Theoretically, the amount of this supply
would need to be an indefinite multiple of the water evaporated and to be
condensed, but in practice a ratio of 4 or 5 times would suffice. To effect,
without serious dissipation of energy, the necessary changes of pressure and
temperature as the solution circulates between the boiler and the condenser,
pumps and regenerators would be needed. The former would do work in
restoring the solution (with the condensed water) to the boiler, but this
would be largely compensated by the solution flowing the other way from
the boiler to the condenser. In the regenerator the heat would be passed
from the outgoing to the returning solution with but a small difference of
temperature and, therefore, with a small dissipation of energy. A similar
use of the principle of regeneration is made in the air liquefying machines
of Linde and Hampson.]
Heat Engines and Saline Solutions.
Mr Macfarlane Gray (p. 414) appears to call in question my assertion
that in a vapour-engine a saline solution may take the place of a simple
liquid when it is desired to replace water by a substance of less volatility,
and that the advantage which Carnot proved to attend a high temperature
540 SUPERHEATED STEAM. [188
can thus be attained without encountering an unduly high pressure. He
contends that "the saline mixture is not the working substance. Carnot's
law refers to the working substance only, and not to anything left in the
boiler."
Perhaps the simplest way of meeting this objection is to point out that
Maxwell's exposition of Carnot's engine (Theory of Heat, chapter viii.) applies
without the change of a single word, whether the substance in the cylinder be
water, mercury, or an aqueous solution of chloride of calcium. In each case
there is a definite relation between pressure and temperature; and (so far
as the substance is concerned) all that is necessary for the reversible operation
of the engine is that the various parts of the working substance should be in
equilibrium with one another throughout.
Let us compare the behaviour of water in Carnot's engine before and
after the addition of chloride of calcium, supposing that the maximum and
minimum pressures are the same in the two cases. The only effect of the
addition is to raise both the superior and the inferior temperatures. The
heat rejected at the inferior temperature may still be available for the
convenient operation of an engine working with pure water. At the upper
limit, all the heat is received at the highest point of temperature — a state
of things strongly contrasted with that which obtains when vapour rising
from pure water is afterwards superheated.
Heat Engines and Saline Solutions.
I have nothing to modify in what I have written under the above
heading (p. 438); but to deal completely with all the questions raised by
Mr Macfarlane Gray (p. 486) would require half a treatise. I will limit
myself to a few brief remarks.
(1) In Carnot's engine there is no (separate) boiler or condenser.
(2) When I spoke of the various parts of the working substance being
in equilibrium with one another, I referred to complete equilibrium, thermal
as well as mechanical. If the temperature varies from one part to another
there is no equilibrium.
(3) On the above understanding the pressure of vapour in equilibrium
with a saline solution of a given strength is a definite function of the
temperature.
(4) Let me suggest that the origin of the difficulty may lie in the phrase
" superheated vapour " which has not so definite a meaning as Mr Gray seems
to ascribe to it. Whether vapour be superheated or no, depends, not only
1892] SUPERHEATED STEAM. 541
upon the condition of the vapour itself, but also upon the bodies with which
it is in contact. Vapour which is merely saturated in contact with a saline
solution must be regarded as superheated when contact with the solution is
cut off. In the first situation it would condense upon compression, and in
the second situation it would not.
In conclusion, I will hazard the prediction that, if the heat-engines of
the distant future are at all analogous to our present steam-engines, either
the water (as the substance first heated) will be replaced by a fluid of less
inherent volatility, or else the volatility of the water will be restrained by
the addition to it of some body held in solution.
189.
ABERRATION*.
[Nature, XLV. pp. 499—502, 1892.]
UNDER this head may conveniently be considered not only the apparent
displacement of the stars discovered by Bradley, but other kindred phenomena
dependent upon the velocity of light bearing but a finite ratio to that of the
earth in its orbit round the sun, and to other astronomical velocities.
The explanation of stellar aberration, as usually given, proceeds rather
upon the basis of the corpuscular than of the wave theory. In order to
adapt it to the principles of the latter theory, Fresnel found it necessary to
follow Young in assuming that the aether in any vacuous space connected
with the earth (and therefore practically in the atmosphere) is undisturbed
by the earth's motion of 19 miles per second. Consider for simplicity the
case in which the direction of the star is at right angles to that of the earth's
motion, and replace the telescope, which would be used in practice, by a pair
of perforated screens, on which the light falls perpendicularly. We may
further imagine the luminous disturbance to consist of a single plane pulse.
When this reaches the anterior screen, so much of it as coincides with the
momentary position of the aperture is transmitted, and the remainder is
stopped. The part transmitted proceeds upon its course through the aBther
independently of the motion of the screens. In order, therefore, that the
pulse may be transmitted by the aperture in the posterior screen, it is evident
that the line joining the centres of the apertures must not be perpendicular
to the screens and to the wave front, as would have been necessary in the case
of rest. For in consequence of the motion of the posterior screen in its own
* This paper was written in 1887, when I was occupied with my article upon "Wave Theory"
for the Encyclopedia Britannica, and at a time when a more extensive treatment was contem-
plated than was afterwards found practicable. Friends upon whom I can rely are of opinion that
its publication may be useful ; and, as I am not able to give it a complete revision, I prefer to let
it stand under its original date, merely warning the reader that very important work has since
been published by Michelson. — January, 1892.
1892] ABERRATION. 543
plane the aperture will be carried forward during the time of passage of the
light. By the amount of this motion the second aperture must be drawn
backward, in order that it may be in the place required when the light reaches
it. If the velocity of light be F, and that of the earth be v, the line of
apertures giving the apparent direction of the star must be directed forwards
through an angle equal to v/V. More generally, if the angle between the
star and the point of the heavens towards which the earth is moving be a,
there will be an apparent displacement towards the latter point, expressed by
sin or . vj V, and independent of the position upon the earth's surface where the
observation is made. The ratio v/V is about
The aperture in the anterior screen corresponds to the object-glass of the
telescope with which the observation would actually be made, and which is
necessary in order to produce agreement of phase of the various elementarv
waves at a moderately distant focal point. The introduction of a refracting
medium would complicate the problem, and is not really necessary for our
present purpose. As has been shown (Phil. Mag. March 1881 [Vol. i. p. 513],
" On Images formed without Reflection or Refraction "), the only use of an
object-glass is to shorten the focal length. Our imaginary screens mav be as
far apart as we please, and if the distance is sufficient, the definition, and
consequently the accuracy of alignment, is as great as could be attained with
the most perfect telescope whose aperture is equal to that in the anterior
screen.
It appears, then, that stellar aberration in itself need present no particular
difficulty on the wave theory, unless the hypothesis of a quiescent aether at
the earth's surface be regarded as such. But there are a variety of allied
phenomena, mostly of a negative kind, which require consideration before
any judgment can be formed as to the degree of success with which the wave
theory meets the demands made upon it. In the first place, the question
arises whether terrestrial optical phenomena could remain unaffected by the
supposed immense relative motion of instruments and of the aether ; whether
reflection, diffraction, and refraction, as ordinarily observed by us, could be
independent of the direction of the rays relatively to the earth's motion. It
may be stated at once that no such influence has been detected, even in
experiments carefully designed with this object in view.
Another class of experiments, with the results of which theory must be
harmonized, are those of Fizeau and Michelson upon the velocity of light in
ponderable refracting media which have a rapid motion (relatively to the
instruments and other surrounding bodies) in the direction of propagation,
or in the opposite direction. These very important researches have proved
that in the case of water the velocity of the ponderable medium is not
without effect; but that the increment or decrement of the velocity of
propagation is very decidedly less than the velocity of the water. On the
544 ABERRATION. [189
other hand, the motion of air, even at high velocities, has no perceptible
influence upon the propagation of light through it.
Again, it has been found by Airy*, as the result of an experiment
originally suggested by Boscovitch, that the constant of stellar aberration
is the same, whether determined by means of a telescope of the ordinary
kind, or by one of which the tube is filled with water. It is clear that,
according to Fresnel's views of the condition of the aether at the earth's
surface, this agreement must involve some particular supposition as to the
propagation of light in moving refracting media.
The theory of these phenomena must evidently turn upon the question
whether the aether at the earth's surface is at rest, absolutely or relatively to
the earthf; and this fundamental question has not yet received a certain
answer. The independence of terrestrial optical phenomena of the earth's
motion in its orbit is, of course, more easily explained upon the latter
alternative; or rather no explanation is required. But in that case the
difficulty is thrown upon stellar aberration, which follows a more simple
law than we should expect to apply in the case of an aether disturbed by
the passage of a body in its neighbourhood. Prof. Stokes has, indeed,
attempted a theory on these lines J, by supposing the aetherial motion to
be what is called in hydrodynamics irrotational. In strictness there is,
however, no such motion possible, subject to the condition of vanishing
absolutely at a great distance, and relatively at the earth's surface ; and it
does not appear that the objection thus arising can be satisfactorily met.
If we start from the experimental facts which have the most direct
bearing upon the question under discussion, we are led to regard Fresnel's
views (doubtless in some generalized form) as the more plausible. From the
results of Fizeau and Michelson relative to air, we may conclude with tolerable
confidence that a small mass of ponderable matter, of very low refracting
power, moving in space, would not appreciably carry the aether with it. The
extension of the argument to a body as large as the earth is not unnatural,
though it involves certainly an element of hypothesis. In like manner, if the
globe were of water, we should expect the aether to be carried forward, but
not to the full amount. The simplest supposition open to us is that, in any
kind of ponderable matter, forming part of a complex mass, the aether is
carried forward with a velocity dependent upon the local refracting power,
but independent of the refracting power and velocity of other parts of the
mass. In the earth's atmosphere, where the refracting power is negligible,
the aether would be sensibly undisturbed.
* Proc. Roy. Soc. xx. 1872, p. 35; xxi. 1873, p. 121.
t An accusation of crudeness might fairly be brought against this phraseology ; but an attempt
to express the argument in more general language would probably fail, and would in any case be
+ Phil. Mag. xxvni. 1846, p. 76 ; xxix. 1846, p. 6.
1892] ABERRATION. 545
If we agree to adopt this point of view provisionally, we have next to
consider the relation between the velocity of luminous propagation in moving
ponderable matter and the refractive index. The character of this relation
was discovered by Fresnel, whose argument may be thrown into the following
form.
Consider the behaviour of the aether when a plate of ponderable matter
(index = /*) is carried forward through vacuum with velocity v in a direction
perpendicular to its plane. If D be the density of the aether in vacuum, and
A the density in the refracting medium, then, according to Fresnel's views as
to the cause of refraction, Dj = p?D. The aether is thus condensed as the plate
reaches it ; and if we assume that the whole quantity of aether is invariable ,
this consideration leads to the law giving the velocity (xv) with which the
denser aether within the plate must be supposed to be carried forward. For
conceive two ideal planes, one in the plate and one in the anterior vacuous
region, to move forward with velocity v. The whole amount of aether between
the planes must remain unchanged. Now, the quantity entering (per unit
area and time) is Dv, and the quantity leaving is DI (v — xv). Hence
x = 1 - ,*-»,
so that the velocity with which the aether in the plate is carried forward is
v ( 1 — fJt~^), tending to vanish as fi approaches unity. If V be the velocity of
light in vacuum and V/p the velocity in the medium at rest, then the absolute
velocity of light in the moving medium is
Vlf*,±v(I -/*-*) (1)
Whatever may be thought of the means by which it is obtained, it is not a
little remarkable that this formula, and no other, is consistent with the facts
of terrestrial refraction, if we once admit that the aether in the atmosphere is
at absolute rest. It is not probable that the aether, in moving refracting
bodies, can properly be regarded as itself in motion ; but if we knew more
about the matter we might come to see that the objection is verbal rather
than real. Perhaps the following illustration may assist the imagination.
Compare the aether in vacuum to a stretched string, the transverse vibrations
of which represent light. If the string is loaded, the velocity of propagation
of waves is diminished. This represents the passage of light through
stationary refracting matter. If now the loads be imagined to run along
the string with a velocity not insensible in comparison with that of the
waves, the velocity of the latter is modified. The substitution of a membrane
for a string will allow of a still closer parallel. It appears that the suggested
model would lead to a somewhat different law of velocity from that of Fresnel;
but in bringing it forward the object is merely to show that we need not
interpret Fresnel's language too literally.
We will now consider a few examples of the application of the law of
velocity in a moving medium ; and first to the experiment of Boscovitch, in
35
546 ABERRATION. [189
which stellar aberration is observed with a telescope filled with water. We
have only to suppose the space between the two screens of our former
explanation to be occupied by water, which is at rest relatively to the screens.
In consequence of the movement of the water, the wave, after traversing
the first aperture, is carried laterally with the velocity v (1 — /*~2), and this
is to be subtracted from the actual velocity v of the aperture of the posterior
screen. The difference is przv. The ratio of this to the velocity of light in
water (V//J,) gives the angular displacement of the second aperture necessary
to compensate for the motion. We thus obtain prl vj V. This angle, being
measured in water, corresponds to v/V in air; so that the result of the
motion is to make the star appear as if it were in advance of its real place by
the angle v/V, precisely as would have happened had the telescope contained
air or vacuum instead of water.
We will now calculate the effect of the motion of a plate perpendicular
to its own plane upon the retardation of luminous waves moving in the same
(or in the opposite) direction. The velocity of the plate is v, its index is //,,
and its thickness is d. Denoting, as before, the velocity of the aether within
the plate by xv, and supposing, in the first place, that the signs of v and V
are the same, we have, for the absolute velocity of the wave in the plate,
V/fJL + XV.
We have now to express the time (t) occupied by the wave in traversing
the plate. This is not to be found by simply dividing d by the above written
velocity ; for during the time t the anterior face of the plate (which the wave
reaches last) is carried forward through the distance vt. Thus, to determine t
we have
whence
Vt_ p,
d l+x-lvV'
The time, t0, which would have been occupied in traversing the same distance
(d + vt), had the plate been away is given by
so that
Thus
Substituting in this Fresnel's value of as, viz. (1 - ^~2), and neglecting as
insensible the square of v/V, we find
v/V) ...................... (5)
ABERRATION. 547
If we suppose that part of the original wave traverses the plate, and that part
passes alongside, (5) gives the relative retardation— that is, the distance
between the wave-fronts which were originally in one plane. It would
appear at first sight that this result would give us the means of rendering
v evident. For the retardation, depending upon the sign of t^F, will be
altered when the direction of the light is reversed, and this we have it in
our power to bring about by simply turning our apparatus through 180°.
A more careful examination will, however, lead us to a different conclusion.
The most obvious way of examining the retardation would be to use
homogeneous light, and, by producing regular interference of the two portions,
to observe the positions of the fringes, and any displacement that might
result from a shift of the apparatus relatively to the direction of the earth's
motion- But if we employ for this purpose a terrestrial flame, e.g. that of a
Bunsen's burner containing sodium, we have to take into account the feet that
the source is itself in motion. For it is evident that the waves which pass in
a given time through any point towards which the source is moving are more
numerous than had the source been at rest, and that the wave-lengths are
correspondingly shortened If v be the velocity of the source, the wave-length
is changed from X to X (1 — r/ F). At a point behind, from which the source is
retreating, the wave-length is X(l +»/F). We shall have occasion to refer
again to this principle, named after Doppler, as applied by Huggins and
others to the investigation of the motion of the heavenly bodies in the
line of sight.
Referring now to (5), we see that, although the absolute retardation is
affected by r, yet that the retardation as measured in wave-lengths remains
unaffected. If, then, there be, in the absence of v, an agreement of phase
between the two interfering beams, the introduction of r will cause no
disturbance. Consequently no shifting of the interference bands is to be
expected when the apparatus is turned so that the direction of propagation
makes in succession all possible angles with that of the earth's motion.
The experiment has been modified by Hoek*? who so arranged matters as
to eliminate the part of the retardation independent of r. As before, of two
parallel beams A and B, one, A, passes through a plate of refracting medium :
the other, B, through air. The beams are then collected by a lens, at the
principal focus of which is placed a mirror. After reflection by this mirror,
the beams exchange paths, B returning through the plate, and A through
air. Apart, therefore, from a possible effect of the motion, there would be
complete compensation and no final difference of path. As to the effect
of the motion, it would appear at first sight that it ought to be sensible.
During the first passage, A is (on account of r) accelerated : on the return,
B is retarded; and thus we might expect, upon the whole, a relative
* Archive, X&rM«M». t m. p. 180 (1968), t. m p. 443 (1869).
35—2
548 ABERRATION. [189
acceleration of A equal to (/u, - l)d. 2v/V. But here, again, we have to
consider the fact that another part of the apparatus, viz. the mirror, partakes
of the motion. In the act of reflection the original retardation of A is
increased by twice the distance through which the mirror retreats in the
interval between the arrival of the two waves. This distance is (with
sufficient approximation) (/A — l)d.v/V; so that the influence of the move-
ment of the mirror just compensates the acceleration of A which would have
resulted in the case of a fixed mirror. On the whole, then, and so long as the
square of v/Fmay be neglected, no displacement of fringes is to be expected
when the apparatus is turned. The fact that no displacement was observed
by Hoek, nor in an analogous experiment by Mascart*, proves that if the
stationary condition of the asther in terrestrial vacuous spaces be admitted,
we are driven to accept Fresnel's law of the rate of propagation in moving
refracting media.
What is virtually another form of the same experiment was tried by
Max well |, with like negative results. In this case, prisms were used instead
of plates ; and the effect, if existent, would have shown itself by the displace-
ment of the image of a spider-line when the instrument was turned into
various azimuths.
On the basis of Fresnel's views it may, in fact, be proved generally that,
so far as the first power of v/V is concerned, the earth's motion would
not reveal itself in any phenomenon of terrestrial refraction, diffraction, or
ordinary refraction. The more important special cases were examined by
Fresnel himself, and the demonstration has been completed by Stokes |.
Space will not allow of the reproduction of these investigations here, and
this is the less necessary, as the experiment of Hoek, already examined,
seems to raise the principal question at issue in the most direct manner.
Another point remains to be touched upon. We have hitherto neglected
dispersion, treating p as constant. In stationary dispersing media, p. may
be regarded indifferently as a function of the wave-length or of the periodic
time. When, however, the medium is in motion, the distinction acquires
significance ; and the question arises, What value of p, are we to understand
in the principal term V//JL of (1) ? Mascart points out that the entirely
negative results of such experiments as those above described indicate that,
in spite of the difference of wave-length due to the motion, we must take the
same value of p. as if the medium and the source had been at rest, or that /*
is to be regarded as a function of the period.
Mascart has experimented also upon the influence of the earth's motion
upon double refraction with results which are entirely negative. The theo-
* Ann. de VEcole Normale, t. in. (1874).
t Phil. Trans. 1863, p. 532.
t Phil. Mag. xxvni. p. 76 (1846). See also Mascart, Ann. de VEcole Normale, t. i. (1872),
t. in. (1874) ; and Verdet, (Euvres, t. iv., deuxieme partie.
1892] ABERRATION. 549
retical interpretation must remain somewhat ambiguous, so long as we remain
in ignorance of the mechanical cause of double refraction.
Reference has already been made to the important experiments of Fizeau
and Michelson upon the velocity of light in moving media. The method, in
its main features, is due to the former*, and is very ingeniously contrived for
its purpose. Light issuing from a slit is rendered parallel by a collimating
lens, and is then divided into two portions, which traverse tubes containing
running water. After passing the tubes, the light falls upon a focussing
lens and mirror (as in Hoek's experiment), the effect of which is to inter-
change the paths. Both rays traverse both tubes ; and, consequently, when
ultimately brought together, they are in a condition to produce interference
bands. If now the water is allowed to flow through the tubes in opposite
directions, one ray propagates itself throughout with the motion of the water,
and the other against the motion of the water ; and thus, if the motion has
any effect upon the velocity of light, a shift of the bands is to be expected.
This shift may be doubled by reversing the flow of water in the tubes.
Fizeau's investigation has recently been repeated in an improved form
by Michelson f.
" Light from a source at a falls on a half-silvered surface, 6, where it
divides : one part following the path bcdefbg, and the other the path bfedcbg.
This arrangement has the following advantages : (1) it permits the use of an
extended source of light, as a gas flame ; (2) it allows any distance between
the tubes which may be desired ; (3) it was tried by a preliminary experi-
ment, by placing an inclined plate of glass at h. The only effect was either
to alter the width of the fringes, or to alter their inclination; but in no case
* Ann. de Chimie, ra. Ivii. (1859).
t American Journal, Vol. MXI. p. 377 (1886).
550 ABERRATION. [189
was the centre of the white fringe affected. Even holding a lighted match in
the path had no effect on this point.
" The tubes containing the fluid were of brass, 28 mm. internal diameter ;
and in the first series of experiments, a little over 3 metres in length, and in
the second series a little more than 6 metres."
Even with the longer tubes and the full velocity (about 8 metres per
second) the displacement on reversal amounted to less than the width of
a fringe. Nevertheless, fairly concordant results were arrived at; and they
showed that the fraction (x) of the velocity of the water (v) by which the
velocity of light is altered is '434, with a possible error of + '02. The
numerical value of the theoretical expression is
ar = 1 - /i-2 = "437,
in very close accordance.
"The experiment was also tried with air moving with a velocity of
25 metres per second. The displacement was about '01 of a fringe; a
quantity smaller than the probable error of observation. The value calcu-
lated from (1 - /*-2) would be "0036."
We have seen that, so far as the first power of v/V is concerned, Fresnel's
theory agrees with all the facts of the case. The question whether it is
possible to contrive an experiment in which v2/ V2 shall be sensible, has been
considered by Michelson*, who, having arrived at an affirmative conclusion,
proceeded to attack this very difficult experimental problem. In Michelson's
apparatus interference is brought about between two rays, coming of course
originally from the same source, one of which has traversed, to and fro, a
distance D parallel to the earth's motion, and the other a like distance in
the perpendicular direction. The phase of the latter ray is considered by
Michelson to be unaffected by the earth's motion. As to the former, it is
retarded by the amount
D D W 2D v2
or, reckoned in distance at velocity V,
2Dv*/V* ..................................... (6)
"Considering only the velocity of the earth in its orbit, the ratio
v/V=lO~i approximately, and v*/V* = IO~8. If D= 1200 mm., or in wave-
lengths of yellow light, 2,000,000, then in terms of the same unit,
"If, therefore, an apparatus is so constructed as to permit two pencils
of light, which have travelled over paths at right angles to each other, to
* American Journal, xxn. p. 120 (1881).
1892] ABERRATION. 551
interfere, the pencil which has travelled in the direction of the earth's motion
will in reality travel '04 of a wave-length further than it would have done
were the earth at rest. The other pencil, being at right angles to the motion,
would not be affected.
" If now the apparatus be revolved through 90°, so that the second pencil
is brought into the direction of the earth's motion, its path will be lengthened
•04 wave-length'. The total change in the position of the interference bands
would be '08 of the distance between the bands, a quantity easily measur-
able."
In the actual experiment, the earth's velocity was not available to the full
extent, and the displacement to be expected on this account was reduced to
•048 : but Michelson considers that some addition to it should be made on
account of the motion of the solar system as a whole. The displacement
actually found was '022 : and when the apparatus was employed in such
azimuths that the rotation should have had no effect in any case, '034. These
results are very small, and Michelson gives reasons for regarding them as
partially systematic errors of experiment. He concludes that there is no
real displacement of the bands, and that the hypothesis of a stationary aether
is thus shown to be inconsistent with fact.
It has, however, been recently pointed out by Lorentz* that Michelson
has overestimated the effect to be expected according to Fresnel's views.
The ray which travels perpendicularly to the earth's motion is not unaffected
thereby, but is retarded to an amount represented by D y2/ V*. The outstanding
relative retardation is thus only D&/V*, instead of the double of this quantity.
Accepting this correction, we have to expect, according to Fresnel's views,
a shift of only '024 of a band in Michelson's experiment.
Under these circumstances Michelson's results can hardly be regarded as
weighing heavily in the scale. It is much to be wished that the experiment
should be repeated with such improvements as experience suggests. In
observations spread over a year, the effects, if any, due to the earth's motion
in its orbit, and to that of the solar system through space, would be separated.
On the whole, Fresnel's hypothesis of a stationary aether appears to be at
the present time the more probable ; but the question must be considered to
be an open one. Further evidence would be most important : but it is difficult
to see from what quarter anything essentially new can be expected. It might
be worth while for astronomers to inquire whether it is really true, as is
generally assumed, that stellar aberration is independent of the position upon
the earth's surface from which the observation is made. Another question
that might, perhaps, be submitted with advantage to an experimental
* "Over den invloed dien de beweging der aarde of de licht vewhijnnselen uitoefent."
(Amsterdam, 1886.)
552 ABERRATION. [189
examination is whether the propagation of light in air is affected by the
rapid motion of heavy masses parallel to, and in the immediate neighbour-
hood of, the ray.
If we once admit the principle that, whatever the explanation may be,
no ordinary* terrestrial observation is affected by the earth's motion, it is
easy to give an account of what must happen when the light comes from an
external source which may have a motion in the line of sight. Imagine,
for example, a spectroscopic examination of a soda flame situated on a star
and vibrating in identical periods with those of terrestrial soda flames. In
accordance with Db'ppler's principle, the wave-lengths are altered by a
relative motion in the line of sight, and the fact may be rendered evident
by a comparison between the spectra of the star and of the terrestrial flame,
held so as to be seen in the same direction. The simplest case is when the
flame is entirely external to the apparatus, so that both lights are treated in
precisely the same way. It is evident that, under these circumstances, the
difference between the two cannot fail to become apparent : and this way
of regarding the matter shows also that the apparent displacement of the
bright lines in the stellar spectrum is dependent upon the relative, and not
further upon the absolute, motions of the star and of the earth. The mean
of observations, equally distributed over the year, would thus give data for
determining the relative motion in the line of sight of the star and of the
solar system.
If the external source be the sun itself, it might be thought that the
spectra must agree almost perfectly, the eccentricity of the earth's orbit
being so very small. But the sun is a revolving body, and consequently
a distinction must be made according to the part of the sun from which the
light proceeds. It is found, in fact, that a very sensible shift takes place in
the positions of the dark lines according as the light under observation comes
from the advancing or from the retreating limb. This circumstance has been
successfully employed by Thollon and Cornu to distinguish between lines
having a solar and a terrestrial origin. In the latter case it is a matter of
indifference from which part of the sun the light proceeds.
In general optical theory the fmiteness of the velocity of light is usually
disregarded. Velocities at least ten times greater than that of the earth in
its orbit are, however, known to astronomers : and such must begin to exercise
a sensible influence upon radiation. Moreover, in so wide a generalization as
the theory of exchanges, the neglect of even a small quantity is unsatisfactory.
Prof. Balfour Stewart has discussed the influence of the motion of a plate
exercising selective absorption upon the equilibrium of radiation within an
enclosure. He argues that a disturbance will ensue, involving a violation of
* This qualification is inserted in order to exclude such an experiment as that of Michelson,
just described, in which an attempt is made to render sensible an effect depending on t>2/F2.
1892] ABERRATION. 553
the second law of thermodynamics, unless compensated by some other effect
not hitherto recognised. It appears, however, more probable that the whole
radiation coming from and through a plate would not be altered by its
motion. Whatever effect (in accordance with Doppler's law) the motion has
upon the radiation from the plate, a similar effect would be produced upon
the absorbing power. On this view the only result of the motion would be
to change the wave-length' of the rays most powerfully emitted and absorbed,
but without disturbing the balance required by the theory of exchanges.
The moving plate would in fact be equivalent to a stationary one of slightly
different quality.
190.
REMARKS ON MAXWELL'S INVESTIGATION RESPECTING
BOLTZMANN'S THEOREM.
[Philosophical Magazine, xxxm. pp. 356—359, 1892.]
THE investigation in question, which was published by Maxwell in the
12th volume of the Cambridge Philosophical Transactions* only a short time
before his death, has been the subject of some adverse criticism at the
hands of Sir W. Thomson f and of Mr Bryan J. The question is indeed a
very difficult one; and I do not pretend to feel complete confidence in
the correctness of the view now to be put forward. Nevertheless, it seems
desirable that at the present stage of the discussion some reply to the
above-mentioned criticisms should be hazarded, if only in order to keep the
question open.
The argument to which most exception has been taken is that by which
Maxwell (Scientific Papers, II. p. 722) seeks to prove that the mean kinetic
energy corresponding to every variable is the same. In the course of it,
the expression (T) for the kinetic energy is supposed to be reduced to a
sum of squares of the component momenta, an assumption which Mr Bryan
characterizes as fallacious. But here it seems to be overlooked that Maxwell
is limiting his attention to systems in a given configuration, and that no
dynamics are founded upon the reduced expression for T. The reduction can
be effected in an infinite number of ways. We may imagine the configu-
ration in question rendered one of stable equilibrium by the introduction
of suitable forces proportional to displacements. The principal modes of
isochronous vibration thus resulting will serve the required purpose. I do
not see the applicability to this argument of the warning quoted from
Routh's Rigid Dynamics. Perhaps the objection is felt that the conclusion
* Scientific Papers, Vol. n. p. 713.
t Proc. Roy. Soc. June, 1891.
J " On the Present State of our Enowledge of Thermodynamics," Brit. Assoc. Report, 1891.
I am indebted to the author for an advance copy of this valuable report.
cannot be true in the absence of
This is a point that may require
argument seems to imply that the
invariantic character.
The nature of the question may be Olnstiated by
mately realized in the billiard-table, viz. the path of an
moving in a plane without loss of energy «?rf Imiitqd within an
The fundamental assumption is that, apart from exceptional eases, the
particle, starting from a given point, will sooner or later tirane that point
in erery direction : and the conclusion founded upon this assumption is that
in the long run all directions through the pjint are equally fa wared. I do
not see that there is here anything to be specially surprised at. If the*
premises are admitted, the conclusion seems natural enough.
In another part of his investigation Maxwell puts forward under the
same reserves the more general hypothesis that not merely does the system
pass through a given configuration with every possible svstem of velocities
consistent with the energy condition, but also through every wnfiguratioa
which can be reached without violation of the same condition. In the
billiard-table example this means that every part of the table is reached
sooner or later; and, as we have seen, every part that is reached is traversed
as much in one direction as in another. In this case, where there is no
potential energy, we may indeed go further*. MaxweiFs equation 1 41 ) shows
that any part of the table is occupied in the long run as much as anv other :
so that all points, as well as all directions, are equally probable.
To my mind the difficulty of Maxwell's investigation lies more in the
premises than in the deductions^. It is easy to propose particular eases tor
which the hypothesis is manifestly untrue. For example, if the table be
circular, a particle projected otherwise than along a diameter will leave a,
central circular area uninvaded, and in the outer zone will not pass through
a given point in every direction, even when the projection is such that the
path is not re-entrant. The question is how &r the considerations advanced
by Maxwell justify us in putting aside these cases as too exceptional to
interfere with the general proposition, which, at any rate in its
to physics, IB essentially one of probability.
Having found Maxwell's demonstration of the fundamental theorem
556
REMARKS ON MAXWELLS INVESTIGATION
[190
difficult to follow, I have sought to simplify it by an arrangement such that
the initial and final times t' and t may be considered as absolutely fixed
throughout the discussion. The following, dependent upon the substitution
for the " action " A of Hamilton's " principal function " 8, seems to meet the
requirements of the case. By definition,
and, as in Thomson and Tait's Natural Philosophy, § 319,
so that
&8 = {2m (xSx + . . .)} - [2m
or in generalized coordinates
...)].
(1)
In this equation all the motions contemplated are unconstrained, and occupy
the fixed time t — t'. The total energy E is variable from one motion to
another, and S is to be regarded as a function of the q's and ^''s.
The initial and final momenta are thus expressed by means of S in
the form
dS dS ,.
so that
dpr' = d*S
dqs dqr'dqg
dpg
dqrf
,
Thus, using 8 with t - t' constant, instead of (as in Maxwell's investigation)
A with E constant, we get
dqj. . . dqn' dpS. . . dpn' = dqj. . . dqn' dq^.. dqn
dqn
dqn
.(4)
* As an example the motion of a particle in two dimensions about a centre of force may be
considered. qr, q, are then the rectangular coordinates of the particle at a fixed time t ; qr', q,'
the coordinates at the fixed time t', while pri pa and pr', pa' are the component velocities at the
same moments.
In equation (3) r and s may be identical.
1892] RESPEcnxG BOLTZMAXX'S THEOREM. 557
On the left side the motion is denned by the initial q's and p's at time f ;
on the right by the initial and final q's and by t - f (not E, which is a
dependent variable).
In like manner
dp^ dp.
••
..(5)
dfr dp*
I dqm' dqn'
By the relation (3) proved above the two determinants in (4) and (5) are
equal, and thus
dp. (5)
the required conclusion.
[1901. For a further discussion of this subject the reader is referred to
a paper on the " Law of Partition of Kinetic Energy" (Phil. May. xux. p. 98,
1900).]
191.
ON THE PHYSICS OF MEDIA THAT ARE COMPOSED OF FREE
AND PERFECTLY ELASTIC MOLECULES IN A STATE OF
MOTION*.
[Phil. Trans. 183 A, pp. 1—5, 1892.]
THE publication of this paper after nearly half a century demands a word
of explanation; and the opportunity may be taken to point out in what
respects the received theory of gases had been anticipated by Waterston, and
to offer some suggestions as to the origin of certain errors and deficiencies in
his views.
So far as I am aware, the paper, though always accessible in the Archives
of the Royal Society, has remained absolutely unnoticed. Most unfortunately
the abstract printed at the time (Roy. Soc. Proc. 1846, Vol. v. p. 604 ; here
reprinted as Appendix I.) gave no adequate idea of the scope of the memoir,
and still less of the nature of the results arrived at. The deficiency was
in some degree supplied by a short account in the Report of the British
Association for 1851 (here reprinted as Appendix II.), where is distinctly
stated the law, which was afterwards to become so famous, of the equality of
the kinetic energies of different molecules at the same temperature.
My own attention was attracted in the first instance to Waterston's
work upon the connection between molecular forces and the latent heat
of evaporation, and thence to a paper in the Philosophical Magazine for
1858, " On the Theory of Sound." He there alludes to the theory of gases
under consideration as having been started by Herapath in 1821, and he
proceeds : —
" Mr Herapath unfortunately assumed heat or temperature to be repre-
sented by the simple ratio of the velocity instead of the square of the
velocity— being in this apparently led astray by the definition of motion
* [From an Introduction to a Memoir, entitled as above, by J. J. Waterston, received Dec. 11,
1845, read March 5, 1846.]
1892] INTRODUCTION TO WATERSTON'S MEMOIR. 559
generally received — and thus was baffled in his attempts to reconcile his
theory with observation. If we make this change in Mr Herapath's defini-
tion of heat or temperature, viz., that it is proportional to the vis viva,
or square velocity of the moving particle, not to the momentum, or simple
ratio of the velocity, we can without much difficulty deduce, not only the
primary laws of elastic fluids, but also the other physical properties of gases
enumerated above in the third objection to Newton's hypothesis. In the
Archives of the Royal Society for 1845 — 1846, there is paper 'On the
Physics of Media that consists of perfectly Elastic Molecules in a State of
Motion,' which contains the synthetical reasoning upon which the demon-
stration of these matters rests. The velocity of sound is therein deduced to
be equal to the velocity acquired in falling through three-fourths of a uniform
atmosphere. This theory does not take account of the size of the molecules.
It assumes that no time is lost at the impact, and that if the impacts produce
rotatory motion, the vis viva thus invested bears a constant ratio to the
rectilineal vis viva, so as not to require separate consideration. It also does
not take account of the probable internal motion of composite molecules ; yet
the results so closely accord with observation in every part of the subject as to
leave no doubt that Mr Herapath's idea of the physical constitution of gases
approximates closely to the truth. M. Kronig appears to have entered upon
the subject in an independent manner, and arrives at the same result ;
M. Clausius, too, as we learn from his paper ' On the Nature of the Motion
we call Heat ' (Phil. Mag. Vol. xiv. 1857, p. 108)."
Impressed with the above passage and with the general ingenuity and
soundness of Waterston's views, I took the first opportunity of consulting the
Archives, and saw at once that the memoir justified the large claims made
for it, and that it marks an immense advance in the direction of the now
generally received theory. The omission to publish it at the time was a
misfortune, which probably retarded the development of the subject by ten
or fifteen years. It is singular that Waterston appears to have advanced no
claim for subsequent publication, whether in the Transactions of the Society,
or through some other channel. At any time since 1860 reference would
naturally have been made to Maxwell, and it cannot be doubted that he
would have at once recommended that everything possible should be done to
atone for the original failure of appreciation.
It is difficult to put oneself in imagination into the position of the reader
of 1845, and one can understand that the substance of the memoir should
have appeared speculative and that its mathematical style should have failed
to attract. But it is startling to find a referee expressing the opinion
that " the paper is nothing but nonsense, unfit even for reading before the
Society." Another remarks "that the whole investigation is confessedly
founded on a principle entirely hypothetical, from which it is the object to
deduce a mathematical representation of the phenomena of elastic media.
560 INTRODUCTION TO WATERSTON's MEMOIR. [191
It exhibits much skill and many remarkable accordances with the general
facts, as well as numerical values furnished by observation The original
principle itself involves an assumption which seems to me very difficult to
admit, and by no means a satisfactory basis for a mathematical theory, viz.,
that the elasticity of a medium is to be measured by supposing its molecules
in vertical motion, and making a succession of impacts against an elastic
gravitating plane." These remarks are not here quoted with the idea of
reflecting upon the judgment of the referee, who was one of the best qualified
authorities of the day, and evidently devoted to a most difficult task his
careful attention; but rather with the view of throwing light upon the
attitude then assumed by men of science in regard to this question, and in
order to point a moral. The history of this paper suggests that highly
speculative investigations, especially by an unknown author, are best brought
before the world through some other channel than a scientific society, which
naturally hesitates to admit into its printed records matter of uncertain
value. Perhaps one may go further and say that a young author who
believes himself capable of great things would usually do well to secure the
favourable recognition of the scientific world by work whose scope is limited,
and whose value is easily judged, before embarking upon higher flights.
One circumstance which may have told unfavourably upon the reception
of Waterston's paper is that he mentions no predecessors. Had he put
forward his investigation as a development of the theory of D. Bernoulli,
a referee might have hesitated to call it nonsense. It is probable, however,
that Waterston was unacquainted with Bernoulli's work, and doubtful whether
at that time he knew that Herapath had to some extent foreshadowed
similar views.
At the present time the interest of Waterston's paper can, of course, be
little more than historical. What strikes one most is the marvellous courage
with which he attacked questions, some of which even now present serious
difficulties. To say that he was not always successful is only to deny his
claim to rank among the very foremost theorists of all ages. The character
of the advance to be dated from this paper will be at once understood when
it is realised that Waterston was the first to introduce into the theory the
conception that heat and temperature are to be measured by vis viva. This
enabled him at a stroke to complete Bernoulli's explanation of pressure
by showing the accordance of the hypothetical medium with the law of
Dalton and Gay-Lussac. In the second section the great feature is the
statement (VII.), that " in mixed media the mean square molecular velocity
is inversely proportional to the specific weight of the molecules." The proof
which Waterston gave is doubtless not satisfactory; but the same may be
said of that advanced by Maxwell fifteen years later. The law of Avogadro
follows at once, as well as that of Graham relative to diffusion. Since the
law of equal energies was actually published in 1851, there can be no
1892] INTRODUCTION TO WATERSTON's MEMOIR. 561
hesitation, I think, in attaching Watereton's name to it. The attainment of
correct results in the third section, dealing with adiabatic expansion, was
only prevented by a slip of calculation.
In a few important respects Waterston stopped short. There is no
indication, so far as I can see, that he recognised any other form of motion,
or energy, than the translator)- motion, though this is sometimes spoken of
as vibratory. In this matter the priority in a wider view rests with Clausius.
According to Waterston the ratio of specific heats should be (as for mercury
vapour) 1'67 in all cases. Again, although he was well aware that the
molecular velocity cannot be constant, there is no anticipation of the law of
distribution of velocities established by Maxwell.
A large part of the paper deals with chemistry, and shows that his
views upon that subject also were much in advance of those generally held
at the time
With the exception of some corrections relating merely to stops and
spelling the paper is here reproduced exactly as it stands in the author's
manuscript. — Dec. 1891.
[1901. It may be added that Waterston's memoir contains the first
calculation of the molecular velocity, and further that it points out the
relation of this velocity to the velocity of sound. The earliest actual
publication of such a calculation is that of Joule, who gives for the velocity
of hydrogen molecules at 0° C. 6055 feet per second (Manchester Memoirs,
VoL EX. p. 107, Oct. 1848; Phil. Mag. Ser. 4, VoL xrv. p. 211: Joule's
Scientific Papers, VoL L p. 295), thus anticipating by eight or nine years
the first paper of Clausius (Pogg. Ann. 1857), to whom priority is often
erroneously ascribed.]
192.
EXPERIMENTS UPON SURFACE-FILMS.
[Philosophical Magazine, xxxni. pp. 363—373, 1892.]
THE experiments here described are rather miscellaneous in character,
but seem of sufficient interest to be worthy of record. The greater
number of them have been exhibited in the course of lectures at the Royal
Institution.
The Behaviour of Clean Mercury.
According to Marangoni's rule, water, which has the lower surface-
tension, should spread upon the surface of mercury; whereas the universal
experience^ of the laboratory is that drops of water standing upon mercury
retain their compact form without the least tendency to spread. To Quincke
belongs the credit of dissipating the apparent exception. He found that
mercury specially prepared behaves quite differently from ordinary mercury,
and that a drop of water deposited thereon spreads over the whole surface.
The ordinary behaviour is evidently the result of a film of grease, which
adheres with great obstinacy.
The process described by Quincke is somewhat elaborate ; but my
experience with water suggested that success might not be so difficult, if only
the mistake were avoided of pouring the liquid to be tried from an ordinary
bottle. In the early experiments upon the camphor movements difficulty
seems to have been experienced in securing sufficiently clean water surfaces.
The explanation is probably to be found in the desire to use distilled water,
and to the fact that the liquid would usually be simply poured from a stock
bottle into the experimental vessel. No worse procedure could be devised;
for the free surface in the bottle is almost sure to be dirty, and is transferred
in great part to the vessel. In my experience water from the dirtiest cistern
will exhibit the camphor movements, provided that it be drawn in the usual
manner from a tap, and that the precaution be taken to give the vessel a
preliminary rinsing.
!892] EXPERIMENTS UPON SURFACE-FILMS. 563
In order to carry out the idea of drawing the liquid from underneath,
an arrangement was provided like an ordinary wash-bottle, and was filled
with tolerably clean mercury. As experimental vessels watch-glasses are
convenient. They may be dipped into strong sulphuric acid, rinsed in
distilled water, and dried over a Bunsen flame. When the glasses are cool
they may be charged with mercury, of which the first portion is rejected
Operating in this way there was no difficulty in obtaining surfaces upon
which a drop of water would spread, although, from causes that could not
always be traced, a certain proportion of failures was met with.
Exposure of the glasses to the atmosphere soon tells upon the success of
the experiment, although on one occasion spreading occurred after a glass
had stood (with protection from dust) for 20 hours. Even so short an
exposure as 10 minutes was found to prejudice the condition of the mercury
surface. Although something here may have depended upon the special
character of the sample of mercury, it will be advisable in repeating the
experiment to pour the mercury at the last moment.
As might be expected, the grease which produces these effects is largely
volatile. In many cases a very moderate preliminary wanning of the watch-
glass makes all the difference in the behaviour of the drop.
So far as I have observed, the spreading of the drop takes place always
in a leisurely fashion. If a little powder of recently ignited magnesia be
dusted over the mercury, there is no violent repulsion of the dust before the
advancing water. But if a small drop of oil be substituted for the water, the
powder is flashed away so quickly that the eye cannot follow the operation.
The difference between the two cases appears to depend upon the atmo-
spheric moisture. As soon as the mercury is poured, it coats itself with an
aqueous film, and the subsequent spreading of the drop takes place upon
a surface whose affinity for water is already largely satisfied. A drop of water
that has spread and then partially gathered up again (as usually happens
after a short interval) shows an interesting behaviour when breathed upon.
The disk contracts somewhat, and then as the breath, which need hardly
be visible, passes off, expands again: and thus a number of times. The
temporary character of the effect indicates that it is due rather to the
moisture of the breath than to any greasy contamination — a view confirmed
by subsequent experiments, in which the breath was replaced by a current
of pure air which had passed through warm water.
In the experiment with a powdered surface, the dust may be driven from
the neighbourhood of a drop of petroleum by the action of vapour, without
actual contact of the liquids.
Drops of Bisulphide of Carbon upon Water.
The behaviour of a drop of CS, placed upon clean water is also at first
sight an exception to Marangoni's rule. So far from spreading over the
36—2
564 EXPERIMENTS UPON SURFACE-FILMS. [192
surface, as according to its lower tension it ought to do, it remains suspended
in the form of a lens. And dust which may be lying upon the surface is not
driven away to the edge upon the deposit of the drop, as would happen in
the case of oil. A simple modification of the experiment suffices, however,
to clear up the difficulty. If after the deposit of the drop a little lycopodium
be scattered over the surface, it is seen that a circular space surrounding
the drop, of perhaps the size of a shilling, remains bare, and this however
often the dusting be repeated, as long as any of the CS2 remains. The
interpretation can hardly be doubtful. The bisulphide is really spreading all
the while, but on account of its volatility is unable to reach any considerable
distance. Immediately surrounding the drop there is a film moving outwards
at a high speed, and this carries away almost instantaneously any dust that
may fall upon it. The phenomenon above described requires that the water
surface be clean. If a very little grease be present, there is no outward flow
and dust remains undisturbed in the immediate neighbourhood of the drop.
With the aid of the vertical lantern, and a shallow dish whose bottom is
formed of plate glass, these experiments are easily shown to an audience.
Movements of Dust.
When dust of sulphur or lycopodium is scattered upon the surface of
water contained in a partially filled vessel, it is found that after a few
seconds the dust leaves the edge and that a clear ring is formed of perhaps
a centimetre in width. Two explanations suggest themselves. The action
may be due to grease communicated to the surface from the edge of the
vessel ; or, secondly, it may be the effect of gravity upon those particles of
the dust which lie within the limits of the capillary meniscus. The first
explanation is rendered improbable by the non-progressive, or at least but
very slowly progressive, character of the effect; and it is negatived by a
repetition of the experiment in a varied form. It is found that if the vessel,
whether of glass or metal, be filled over the brim, so that the capillary
meniscus is convex, then, although as before a bare margin is formed, the
effect is due to a motion of the dust outwards (instead of inwards, as in the
former case), and is therefore not to be attributed to grease.
A similar movement of dust was to be observed in the experiment above
recorded, where magnesia was scattered upon a pool of mercury, and is
undoubtedly due to gravity; but the full explanation is not so simple as
might appear at first sight.
Even in the interior parts of the surface at a distance from the edge the
sulphur particles do not retain their initial positions, but form aggregates
into which continually increasing numbers are attracted. This is also due to
gravity, neighbours tending, as it were, to fall into the depression by which
every particle is surrounded.
1892] EXPERIMENTS UPON SURFACE-FILMS. 565
Camphor Movements a Test of Surface-Tension.
The theory of these movements, due to Van der Mensbrugghe, implies that
they will take place with greater or less vigour so long as the tension of the
surface, which may be in some degree contaminated, is greater than that of a
saturated solution of camphor. If, however, the contamination be so great
that the tension falls below this point, the solution of camphor can no longer
spread upon the surface, and the movements cease. Thus, according to
this theory and to observations* upon a saturated solution of camphor, the
movements are an indication that the actual tension does not fall below 71
of that of pure water.
Although there appeared to be no reason for distrusting this view, it
was thought desirable to examine specially whether the cessation of the
movements was really a question of surface-tension only, without regard
to the character of the contamination. The readiest method of ensuring
the equality of the tensions of two surfaces contaminated with different
materials is to make the two surfaces parts of one surface, for two parts
of the same surface cannot be at rest unless they have the same tension.
The method of experiment was therefore to divide a surface of clean water
contained in a large dish into two parts by a line of dust, and to communicate
different kinds of grease to the surfaces on the two sides of the indicating
line. If, for example, a small chip of wood, slightly greased with olive-oil,
be allowed to touch one part of the surface, the line of dust is repelled by
the expansion of that part, but the effect may be compensated by a slight
greasing of the other part with oil of cassia. By careful alternate additions
the line of dust may be kept central, while the two halves become increas-
ingly greased with the two kinds of oil. At every stage of this process, so
long as the surface is at rest, the tension of all parts is necessarily the same.
A large number of substances have thus been tried in pairs, of which
may be mentioned oils of olive, cassia, turpentine, lavender, cinnamon, anise,
petroleum, pseudocumene. In no case could any difference be detected in
the behaviour of camphor fragments on the two sides. Whenever possible,
the quantities of oil were adjusted to the point at which the movements
were just ceasing. In case of overshooting the mark, the excess of oil could
be easily removed by strips of paper, partially immersed and then withdrawn,
the action being equivalent to an expansion of the surface. In several cases
the volatility of the substance with which the surface was contaminated led
to a subsequent retraction of the line of dust. Thus freshly distilled oil
of turpentine, even at first barely capable of arresting the movements, soon
off.
* Phil. Mag. November 1890. [Vol. ra. p. 394.]
566 EXPERIMENTS UPON SURFACE-FILMS. [192
As was shown by Tomlinson, oil of anise is incapable of arresting the
camphor movements. In the experiment with a partition of dust, olive-oil
will drive oil of anise into a very small space, whose area is doubtless dependent
upon the amount of other impurities present. In this case, as in all others,
the behaviour of camphor is the same on the various parts of the surface.
It may thus be taken as established that the relation of a contaminated
surface to the camphor movements is one of surface-tension only.
A similar method of experimenting may be applied to a rough deter-
mination of the degree of purity of cleansed surfaces. The whole of the
surface under test is lightly dusted over, and olive-oil is applied at several
places close to the circumference until camphor movements are nearly
arrested. After each addition of oil the dusted area contracts, and at the
close of operations it gives a measure of the extent to which the original
contamination must be concentrated in order to stop camphor.
A few numbers may be given as examples, although in all probability
the result is influenced by a variety of circumstances. A circular area of
10 inches diameter, occupied by tap-water, and cleansed by the flexible hoop
described in former papers, was tested on July 28, 1891. The application
of oil, just sufficient to stop the camphor movements, drove the dust into
a central circular patch of 2^ inches diameter. When the surface was in
its natural condition, unpurified by the action of the hoop, the central patch
was of about 5 inches diameter. These numbers, approximately verified
on repetition, show that the natural surface was about 4 times, and the
purified about 16 times better than according to the camphor standard.
The difference between the two cases is less than was expected, and would
perhaps have been greater had distilled water been employed. It must be
remembered also that contact with dust (sulphur) is unfavourable to the
purity of a water surface. In a very good light a special dusting might
probably be dispensed with, the motion of the surface being evidenced by
inevitable motes.
If the dust be applied in the first instance to a small central patch, which
is then touched internally with a very small quantity of oil, the expansion
of the dust in the form of a ring is followed by a slight but unmistakable
rebound. The effect appears to take place when the surface is very clean to
begin with, and is then somewhat difficult of explanation. I am disposed to
think that it must be attributed in all cases to initial contamination. This
is concentrated in front of the rapidly advancing ring, and has not time to
diffuse itself equally over the whole external area. Under the influence of
inertia the expansion of the central area may then proceed so far that its
tension becomes greater than that of the parts immediately surrounding.
1892]
EXPERIMENTS UPON SURFACE-FILMS.
567
Influence of Heat.
For a lecture experiment the effect of heat is best shown by holding a
hot body near the surfece of water contained in a shallow vessel with a glass
bottom. The hot body may be the end of a glass rod heated by a flame, or
more conveniently a small spiral of platinum wire, rendered incandescent at
will by an electric current. The immediate effect of the heat is to lower the
tension of the part of the surface affected; but the visible result depends
entirely upon whether the surface be clean or otherwise. In the former case
the heated surface expands, and an outward current is generated. This
is rendered evident by the clearing away of dust. But if the original
contamination exceed a very small quantity, a moderate expansion of the
heated area brings the tension again up to equality with that of the
surrounding surface, and there is no further action. In this case there is
no visible clearing away of dust under the hot body.
Under favourable circumstances a very slight elevation of temperature
suffices. On July 28 a shallow tin vessel 8x5 inches, the lid of a biscuit-box,
was levelled and filled with tap-water from a rubber hose, after a thorough
preliminary rinsing in situ. A little dust (sulphur) was then scattered over,
and the finger was brought underneath into contact with the bottom of the
dish. After about 20 seconds the dust opened out, and a bare spot was
formed over the finger of about 1| inches diameter. A spirit-flame, applied
for a few seconds under one end of the dish, cleared away the dust from the
larger part of the area. If when quiet was nearly restored, a little fresh
dust was applied, and the experiment with the finger repeated, the effect was
more pronounced than before, and the bared space much larger, showing that
the treatment with the spirit-flame had driven away most of the residual
contamination.
The best effects were obtained with a dish somewhat larger than that
above mentioned ; and in subsequent experiments the difference of tempe-
rature between different parts was more readily maintained by the use of a
Fig. i.
vessel in which the main portions were connected by a comparatively narrow
channel. In this way the tensions of surfaces, contaminated in different
degrees, may be equalized, the wanner purer surface in one compartment
568 EXPERIMENTS UPON SURFACE-FILMS. [192
balancing the colder but greasier surface in the other. And the actual
temperature-difference necessary for equilibrium gives a measure of the small
difference of tensions to be compensated*.
When the surface of the liquid in the tin vessel is but very slightly
greased, a spot can no longer be cleared by the warmth of the finger held
underneath. Indeed the spirit-flame itself soon becomes ineffective. And yet
the greasing may be so slight that camphor fragments move with apparently
unabated vigour.
It is of interest to compare the behaviour of saturated solution of
camphor with that of greasy water. The former can scarcely be brought to
rest, unless covered up. This is doubtless due to evaporation of camphor,
aided by local draughts. A spirit-flame drives away dust in a manner
impossible in the case of a merely greasy surface, whose tension may
nevertheless be decidedly higher than that of the camphorated water.
It may here be mentioned that the lowering of tension by camphor
follows a different law from the lowering caused by soap. In the latter case
the fall of tension requires time, and at the first moment of its formation
a free surface has almost the tension of pure water. Similar experiments
to those formerly recorded f with soapy water have shown that the ratio of
tensions for pure water and for solution of camphor are the same at the
first moment of the formation of a free surface as when the measures are
conducted statically.
Saponine and Soap.
A strong infusion of horse-chestnuts allowed excellent bubbles to be
blown, up to 4 inches [10 cm.] or more in diameter. When the interiors of
equal bubbles of soap and of saponine were brought into communication, the
latter contracted and the former expanded, showing that the tension of the
saponine film was the greater. In order to obtain equilibrium, the diameter
of the saponine bubble required to be about half as great again as that of the
soap bubble. These saponine bubbles exhibited the characteristic wrinkling,
when caused suddenly to contract by withdrawal of part of the contained air.
The foaming of Highland waterfalls is doubtless attributable to dissolved
vegetable matter. In the autumn of 1890 I had an excellent opportunity of
observing these effects in the case of the river Creed at Stornoway. By the
coalescence of smaller ones, hemispherical bubbles of remarkable size, up to
a foot or more in diameter, were frequently formed, and endured for a few
seconds ; and yet not the smallest bubble could be blown from a tobacco-pipe.
However, by collecting some of the foam and allowing it to subside, which
took a good while, I obtained liquid from which bubbles could be blown with
* The lowering of tension per degree Cent, is said to be -0018 of the total value.
t Proc. Roy, Soc. March 1890. [Vol. in. p. 341.]
1892] EXPERIMENTS UPON SURFACE-FILMS. 569
a pipe up to 4 inches diameter. But these bubbles behaved like soap, and
not as had been rather expected, like saponine, remaining perfectly tight
and smooth when the included air was rapidly withdrawn.
Separation of Motes.
In the course of some experiments last year, in illustration of Sir G.
Stokes's theory of ternary mixtures, I had prepared an association* of water,
alcohol, and ether, in which the quantity of alcohol was so adjusted that the
tendency to divide into two parts was almost lost. As it was, division took
place after shaking into two nearly equal parts, and these parts were of
almost identical composition. On placing the bottle containing the liquids
in the concentrated light from an arc lamp, I was struck with the contrast
between the appearance of the two parts. The lower, more aqueous, layer
was charged with motes, while the upper, more ethereal, layer was almost
perfectly free from them. Some years ago I had attempted the elimination
of motes by repeated distillation of liquid in vacuum, conducted without
actual ebullition, but I had never witnessed as the result of this process
anything so clear as the ethereal mixture above described.
The observation with the ternary association, which happened to be the
first examined, is interesting, because the approximate equality of the liquids
suggests that the explanation has nothing directly to do with gravitation.
But the presence of alcohol is not necessary. Ether and water alone shaken
together exhibit the same phenomenon. It would appear that when the two
liquids are mixed together in a finely divided condition, the motes attach
themselves by preference to the more aqueous one, and thus when separation
into two distinct layers follows, the motes are all to be found below f.
An obvious explanation, which, however, stands in need of confirmation,
is that under the play of the capillary forces the energy is least when the
motes, which may be presumed to be denser than either liquid, are in contact
with the denser rather than with the rarer of the two. The density here
referred to is that which occurs in Laplace's theory of capillarity, and may
need to be distinguished from ordinary mechanical density.
I have lately endeavoured to obtain some confirmation of the views above
expressed by the use of other liquids. It would evidently be satisfactory to
exhibit the selection of motes by the upper, instead of by the lower, layer.
Experiments with bisulphide of carbon and water, and also associations of
these two bodies with alcohol, which acts as a solvent to both, gave no
* Association is here employed as a general term denoting the juxtaposition of two or more
fluids. Whether the result is a mixture depends upon circumstances.
t [1901. The clearness of the upper layer, after a mixture of ether and alcohol has been
shaken up with dnst, had already been observed and explained much as above by Bams (Am.
Journ. xxxvn. p. 122, 1889).]
570 EXPERIMENTS UPON SURF ACE-FILMS. [192
definite result, perhaps in consequence of a tendency to the formation of a
solid pellicle at the common surfaces. But with chloroform and water, and
with associations of chloroform, water, and acetic acid (acting as a common
solvent), the experiment succeeded. The motes were always collected in the
upper, more aqueous, layer, even when the composition of the two layers into
which the liquid separated was so nearly the same that a few additional drops
of acetic acid sufficed to prevent separation altogether.
In this and similar cases a marked tendency to foaming may be observed
when the composition is such that separation just fails to take place.
The Lowering of Tension by the Condensation of Ether Vapour.
The suspension of water in an inverted tube of small bore is familiar to
all. The limit of diameter was investigated some years ago by Fi 2
Duprez*. A glass tube, such as that shown in Fig. 2, is ground
true at the lower end, and at the upper end is connected to an
india-rubber tube provided with a pinch-cock. Water is sucked
up from a vessel of moderate size, the rubber is nipped, and by
a quick motion the tube and the vessel are separated, preferably
by a downward movement of the latter. In this way of working
Duprez found that the liquid might remain suspended in tubes
of diameter up to 16 millim., and with the aid of a sliding plate
up to 19'85 millim. The theory is given in Maxwell's article
in the Encyclopaedia Britannica (" On Capillary Action "). For
lecture purposes it is well not to attempt too much. The tube
employed by me had an internal diameter of 14£ millim., and
there was no difficulty in obtaining suspension. The experiment
on the effect of ether- vapour was then as follows : — The inverted
tube, with its suspended water, being held in a clamp, a beaker
containing a few drops of ether was brought up from below until
the free surface of the water was in contact with ether vapour. The lowering
of tension, which follows the condensation of vapour, is then strikingly shown
by the sudden precipitation of the water.
Breath Figures and their Projection.
These figures are perhaps most readily prepared upon the plan described
in Riess's Electricity. The carefully cleaned glass plate upon which the
image is to be received is placed upon a flat metallic slab, and upon it again
rests the coin to be copied, for example, a shilling. The two conductors form
the coatings of a Leyden jar, and are connected by wires to the discharging
* " Sur un cas particulier de 1'equilibre des liquides," Bruxelles Acad. Sci. Mem. xxvi. 1851 ;
xxvm. 1854.
1892J EXPERIMENTS UPON SURFACE-FILMS. 571
terminals of a large Wimshurst machine, the latter being set so as to give
sparks about | inch long. In my experiments about 20 turns of the handle
were found sufficient to impress the latent image.
The projection of the figures, developed upon the glass by breathing,
requires a special arrangement, which it is the principal object of this note
to describe. For this purpose the light simply transmitted by the undimmed
parts of the plate must be intercepted, leaving the image to be formed by the
light diverted from its path by the condensed breath. The arrangement was
as follows: —
The ordinary condenser B (Fig. 3) of the electric lantern was stopped
down to an aperture of £ inch, and provided a somewhat divergent beam
Fig. 3.
SCALED
of light of corresponding diameter. At a distance of 15 J inches from the
condenser was placed the slide C upon which a figure had been impressed.
The focusing lens D was of plate-glass, 6 inches in diameter and 25 inches
focus, and was of course distant from the breath figure by an amount slightly
exceeding its own focal length. Any light that might pass outside was
intercepted by a suitable mounting. So far there was nothing peculiar,
except in respect to the dimensions of the focusing lens. But now between
the latter and the screen was inserted a disk E of black card 2 inches in
diameter, at such a distance (40 inches) from the lens as to receive a well-
defined image of the hot carbons A. By this disk all regularly refracted light
would be stopped, so that the screen would appear dark. If, however, any
part of the prepared glass be dimmed by the breath, light is there diverted
from its path, and thus escaping the stop proceeds to form an image of the
part in question upon the screen. The dewed parts of the breath-figure are
accordingly seen bright upon a dark ground ; and with the arrangement
described, in which the large diameter of the focusing lens is a leading
feature, the projected images are very beautiful. A similar method would
probably be adequate to the projection of smoke-jets.
In conclusion I may mention that the latent images can be developed in
a more durable manner by a deposit of silver, the arrangements being such as
are adopted for the silvering of mirrors, except that the action is stopped at
an earlier stage. The washed and dried deposit may then be protected from
mechanical injury by a coat of varnish.
Fig. 1.
193.
ON THE THEORY OF SURFACE FORCES. III.— EFFECT
OF SLIGHT CONTAMINATIONS.
[Philosophical Magazine, xxxin. pp. 468 — 471, 1892.]
OBSERVATION* having suggested that the lowering of surface-tension of
water due to a film of oil falls off more rapidly when the film is attenuated
than the thickness of the film itself can be
supposed to do, I was led to examine the
question theoretically; and the result shows
that, according to the principles of Young and
Laplace, the lowering of tension due to a very
thin film should be in proportion, not to the
thickness, but to the square of the thickness of
the film. In the calculations which follow the
fluids are supposed to be incompressible, a layer
of density p and thickness a being interposed
between fluids of densities p2 and pt (Fig. 1). The thickness a, as well as the
range of the forces, is supposed to be negligible in comparison with the radius
of curvature R of the surfaces of separation.
By II. (16)f we have for the difference of pressures in the inner and
outer liquids,
Vdp
pfi-(p-pl).V(p,pl)-(pt-p).V(pt,p), ......... (1)
where V(p, p^, V(pz, p) denote the potentials at the surfaces of separation.
* See, for example, Miss Pockels on " Surface Tension," Nature, Vol. XLIII. p. 437 (1891).
t Phil. Mag. Feb. 1892. [Vol. in. p. 519.]
ON THE THEORY OF SURFACE FORCES. 573
Again, by II. (17),
pj
and
Consider now
,p)l ............... (2)
and collect separately the part independent of R, and that proportional to
R-1. For the first we have
(3)
the same as if a = 0.
For the second, omitting the factor — ZTT/R, we get
(P ~
o
Now #=27r/3
Jo
so that p,-pl = K(p.2*-p*) + 2T/R, ...................... (5)
where T= vfa- &}*+(&£*£- ^(p - Pl) (&-?)[ ^(QSd£ ...(6)
The tension of the composite surface is thus given by (6).
If a = 0, we fall back upon the case of a simple sudden transition from /
to pi, and we get as before
r-wfc-fty£V<D« ....................... (7)
Again, if a = x ,
(8)
574 ON THE THEORY OF SURFACE FORCES. [193
This corresponds to the formation of two independently acting tensions
between the two pairs of liquids.
To pass from these verifications to circumstances of novelty, let us now
suppose that a is small compared with the range of the forces. When £ is
small, >/r(£) may be identified with •vJr(O), and we have
ST=-7r(p-Pl)(p2-p). + (Q).a?, ..................... (9)
showing that in the limit 8T is proportional to the square of the thickness a.
According to Young's supposition I. (19)* of a constant attraction within
the range a,
^•(0- **(**- n-i^-n
so that -^(0) = £a3; and more generally whether a be great or small,
(10)
The general formula (6) may be applied also to the case of a thin lamina
by supposing that p.2 = p1 = pa. Thus
(11)
gives the tension of a lamina of density p and thickness a surrounded by
fluid of density /30f. Here again, if a be very small, the integral reduces to
Ja2-«/r(0), so that the tension varies as the square of a.
It must be understood that the lamina is here supposed to be of uniform
constitution, and that thus the result is probably inapplicable to soap-films.
The law that the effect of a film when very thin is proportional to the
square of the thickness is of considerable interest. It is here deduced upon
the basis of Laplace's more special hypothesis that the specific differences
between various bodies in relation to capillary properties may be represented
by a simple coefficient in the expression for the law of force. But it is not
difficult to see that the conclusion is really independent of this restriction,
and that it holds good even though the law of force for every pair of bodies is
entirely arbitrary.
[1901. Subsequent experiments by Miss Pockels and by the author
(Phil. Mag. XLVin. p. 331, 1899) show that the actual falling off is more
sudden than that above calculated, probably in consequence of the dis-
continuity which enters when the layer is only one or two molecules thick.]
* Phil. Mag. Oct. 1890. [Vol. in. p. 404.]
t In Maxwell's solution of this problem, Art. "Capillary Action," Enc. Brit., the tension of
the lamina is given at double the above value.
194.
ON THE QUESTION OF THE STABILITY OF THE FLOW
OF FLUIDS.
[Philosophical Magazine, xxxiv. pp. 59— 70T 1892.]
IT is well known that while Sir G. Stokes's theory of viscous flow gives
a completely satisfactory account of what is observed in the case of capillary
tubes, no theory at present exists to explain the complete change in the laws
of flow which supervenes when the tnbes are of larger diameter and the
velocities not very small. Pro£ Osborne Reynolds* has applied the theory
of dynamical similarity to this question, and has shown both by theory and
experiment that the change in the law of resistance occurs when epic /* has
a certain value, where c is a linear parameter such as the diameter of the
tube, w is the velocity, /* the coefficient of friction, and p the density.
The conclusion is perhaps most easily reached by applying the method of
dimensions to the expression for the ratio (P) of the difference of pressures
at two points along the length of the tube to the distance between the points.
The dimensions of this ratio are those of a force divided by a volume ; and if
we assume that it may be expressed in terms of *t (equal to /» p\ cr p, and ic
in the form
we have the three relations
so that
and
Since n is here indeterminate, all we can infer from dynamical similarity
is that
P = *><r*pf(cwi*), ........................... (2)
where / is an arbitrary function.
* PkO. Trmm,. ctmr. p. 935 (1883).
tOf which the dinMUBonsaie Sin «PMC and -1 in time.
576 ON THE QUESTION OF THE [194
For capillary tubes and moderate velocities P varies as the first power
of iv, so that in (1) n = 1. In this case
P = Avc-*pw, (3)
A being an arbitrary constant. When, on the other hand, cw/v is great,
experiment shows that n = 2 nearly. If this law be exact, (1) gives
P = Bc~lpw\ (4)
independent of v. The second power of the velocity and independence of
viscosity are thus inseparably connected.
In the above theory no account is taken of any variation in the walls of
the tubes. Either they must be perfectly smooth or else the irregularities
mmt be in proportion to the diameters. Under this limitation (2) would
appear to hold good, at least if there be no finite slip at the walls.
The proportionality to p, expressed in (4), has probably not been tested
experimentally. Neither is there any complete theoretical deduction of (4).
But a comparison with Torricelli's law of efflux is significant. The resistance
is the same as if it were necessary to renew continually the velocity of the
liquid at intervals which are proportional to the diameters of the pipes.
The connexion between the alteration in the law of resistance and the
transition from regularly stratified to eddying motion has been successfully
traced by Reynolds. The question is, Why do eddies arise and take
possession ? From the description and drawings given by Reynolds it is
natural to suppose that in the absence of viscosity the stratified motion would
be unstable, and that it is stable in small tubes and at low velocities only in
consequence of the steadying effect of viscosity then acting at an advantage.
It was with this idea that (at an earlier date*) I attempted an investigation
of the stability of stratified flow in two dimensions, fully expecting to find it
unstable. The result, however, was to show that in the absence of viscosity
the stratified flow between two parallel walls was not unstable, provided that
the law of flow were such that the curve representing the velocities in the
various strata was of one curvature throughout, a condition satisfied in the
case in question. To be more precise, it was proved that if the deviation from
the regularly stratified motion were, as a function of the time, proportional to
eint, then n could have no imaginary part.
On the other hand, if the condition as to the curvature of the velocity
curve be violated, n may acquire an imaginary part, and the resulting
disturbance of the steady motion is exponentially unstable, as was shown by
several examples in the paper referred to, and in a later onef in which the
subject was further pursued.
* Proc. Math. Soc. February 12, 1880. [Vol. i. p. 474.]
t Ibid. November 1887. [Vol. in. p. 17.]
1892] STABILITY OF THE FLOW OF FLUIDS. 577
We are thus confronted with a difficulty. For if the investigation in
question can be applied to a fluid of infinitely small viscosity, how are we to
explain the observed instability which occurs with moderate viscosities ? It
seems very unlikely that the first effect of increasing viscosity should be to
introduce an instability not previously existent, while, as observation shows,
a large viscosity makes for stability.
Several suggestions towards an explanation of the discrepancy present
themselves. In the first place, irregularities in the walls, not included in the
theoretical investigation, may play an essential part Again, according to the
view of Lord Kelvin, the theoretical stability for infinitely small disturbances
at all viscosities may not extend beyond very narrow limits: so that in
practice and under finite disturbances the motion would be unstable, unless
the viscosity exceeded a certain value. Two other suggestions which occurred
to me at the time of writing my first paper as perhaps pointing to an
explanation may now be mentioned. It is possible that there may be an
essential difference between the motion in two dimensions to which the
calculations related, and that in a tube of circular section on which observa-
tions are made. And, secondly, it is possible that, after all, the investigation
in which viscosity is altogether ignored is inapplicable to the limiting case
of a viscous fluid when the viscosity is supposed infinitely small. There
is more to be said in favour of this view than would at first be supposed.
In the calculated motion there is a finite slip at the walls, and this is
inconsistent with even the smallest viscosity. And, further, there are kindred
problems relating to the behaviour of a viscous fluid in contact with fixed
walls for which it can actually be proved* that certain features of the motion
which could not enter into the solutions, were the viscosity ignored from the
first, are nevertheless independent of the magnitude of the viscosity, and
therefore not to be eliminated by supposing the viscosity to be infinitely
small Another case that may be instanced is that of a large stream of
viscous fluid flowing past a spherical obstacle. As Sir G. Stokes has shown,
the steady motion is the same whatever be the degree of viscosity ; and yet
it is entirely different from the flow of an inviscid fluid in which no rotation
can be generated. Considerations such as this raise doubts as to the inter-
pretation of much that has been written on the subject of the motion of
inviscid fluids in the neighbourhood of solid obstacles.
The principal object of the present communication is to test the first of
the two latter suggestions. It will appear that, as in the case of motion
between parallel plane walls, so also for the case of a tube of circular section,
no disturbance of the steady motion is exponentially unstable, provided
viscosity be altogether ignored.
» " On the Circulation of Air in Kuudfs Tubes," Pkil. Tmu. November 1885- [VoL n- p. *»-]
B. III. 37
578 ON THE QUESTION OF THE [194
Referring the motion to cylindrical coordinates z, r, 0, parallel to which
the component velocities are w, u, v, we have*
du v2 _ dQ dv uv _ldQ dw _ dQ
di~ r~~dr' di + T~~r~d0' ~dt = ~dz '
d d d v d d
5i=jT+ttj- + -j3 + «'-r,
dt dt dr rdO dz
where — Q= V + p/p, [and V is the potential of the impressed forces].
These are the general equations. In order to apply them to the present
problem of small disturbances from a steady motion represented by
u = 0, v = 0, w=W,
where W is a function of r only, we will regard the complete motion as
expressed by u, v, W + w, and neglect the squares of the small quantities
u, v, w, which express the disturbance.
Thus,
ti A_ _i TIT" ^"H> /Q\
. "dr "3* + "" ^~ ;&' '
which, with the "equation of continuity,"
(pi) + vw^
dr dQ dz
determine the motion.
The next step is to introduce the supposition that as functions of t, z, 6,
the variables u, v, w, and Q are proportional to ei(nf+kz+s0\
We get
i(n + kW)u = , 9, (n + kW)v = -Q, (5)
dW
u~ + i(n + kW)w=ikQ, (6)
-T- (ru) -f isv + i krw = 0 (7)
From these equations three of the variables may be eliminated, so as to
obtain an equation in which the fourth is isolated. The simplest result is
that in which Q is retained. It is
i^_n^,,^ 2* dWdQ_
7~7Tw ^JIT TTT. ~ u W
* Basset's Hydrodynamics, § 470.
1892] STABILITY OF THE FLOW OF FLUIDS. 579
But the equation in u lends itself more readily to the imposition of
boundary conditions. If s = 0, that is in the case of symmetrical disturbances,
the equation in u is obtained at once by differentiation of (8), and substitution
of u from (5). After reduction it becomes
If the undisturbed motion be that of a highly viscous fluid in a circular
tube, W is of the form A + Br*, and the second part of (9) disappears. There
can then be admitted no values of n, except such as make n + k W = 0 for
some value of r included within the tube. For the equation
... ...
dr* r dr r2
being that of the Bessel's function of the first order with a purely imaginary
argument, admits of no solution consistent with the conditions that u = 0
when r vanishes, and also when r has the finite value appropriate to the wall
of the tube. But any value assumed by — k W is an admissible solution
for n. At the place where n -f kW = 0, (10) need not be satisfied, and under
this exemption the required solution may be obtained consistently with
the boundary conditions. It is included in the above statement that no
admissible value of n can include an imaginary part.
If s be not zero, we have in transforming to u to include also terms
arising from the differentiation in (8) of - Qs*/r2, that is
for the second of which we substitute from (5), and for the first from (8)
itself. The result is
Idu
From (11) we may fall back on the case of two dimensions by supposing r
to be infinite. But, in order not to lose generality, we must at the same time
allow s to be infinite, so that, for example, s = k'r. Thus, writing x for r,
and y for r6, we find for the differential equation applicable to the solution in
which all the quantities are proportional to ^0*+**+*'*>,
(12)
agreeing with that formerly discussed except for a slight difference of
notation.
37—2
580 ON THE QUESTION OF THE [194
We will now consider (11) in the abbreviated form,
where a is a positive number not less than unity ; or, again,
d ( du\ . j.
j-(ra j-J -f 6ra-2w = - pj^ ...................... (13)
dr\ drj n + kW
The question proposed for consideration is whether (13) admits of a
solution with a complex value of n, subject to the conditions that for two
values of r, say rj and r2, u shall vanish. This represents the flow of fluid
through a channel bounded by two coaxal cylinders.
Suppose, then, that n is of the form p + iq, and u of the form a + ij3,
where p, q, a, /3 are real. Separating the real and imaginary parts in (13),
we get
and thence
0 A (r« ^} _ a *L (7M W\ = k^WtV + W.q
Pdr( dr) dr\ dr) + kW* + * '
We now integrate this equation with respect to r over the space between
the walls, viz. from r-i to ?v The integral of the left-hand member is
<">
and this vanishes at both limits, /3 and a being there zero. The integral of
the right-hand member of (16) is accordingly zero, from which it follows that
if TFj be of one sign throughout, q must vanish — that is to say, no complex
value of n is admissible.
The general value of Wlf viz.
d?W IdWfrv*-*
dr1 r dr jfe»r» + s'' "
reduces in the case of two dimensions to cPW/dr*, or, as we may then write
it, d2W/da?. Instability, at any rate of the full-blown exponential sort, is
thus excluded, provided d2W/dx2 is of one sign throughout the entire region
of flow limited by the two parallel plane walls.
Commenting upon this argument, Lord Kelvin* remarks that the
disturbing infinity, which arises in (13) when n has a value such that
n + kW vanishes at some point in the field of motion, "vitiates the seeming
* Phil. Mag. Aug. 1887, p. 275.
STABILITY OF THE FLOW OF FLUIDS. 581
proof of stability." Perhaps I went too far in asserting that the motion was
thoroughly stable; but it is to be observed that if n be complex, there is
no " disturbing infinity." The argument, therefore, does not fail, regarded as
one for excluding complex values of n. What happens when n has a real
value such that n + kW vanishes at an interior point, is a subject for further
examination.
The condition for two dimensions that d*W/da* be of one sign throughout
is satisfied for a law of flow such as that of a viscous fluid, and we shall
see that the corresponding condition for (17) in the more general problem
is also satisfied in the case of the steady flow of a viscous fluid between
cylindrical walls at r, and r8. The most general form of W for steady motion
symmetrical about the axis is*
W = Ar* + B\ogr+C, ........................ (19)
in which the constants A, B, C are related by the conditions
+ C,
From the last two equations we derive
A^-rfi + Blogrt/r^O, ..................... (20)
so that A and B have opposite signs. Introducing the value of W from (18),
we obtain as the special form here applicable
which is thus of one sign throughout the range. A small disturbance from
the steady motion expressed by (19) is therefore not exponentially unstable.
The result now obtained is applicable however small may be the inner
radius r, of the annular channel. But the extension to the case of the
ordinary pipe of unobstructed circular section may be thought precarious,
when it is remembered that provision must be made for a possible finite
value of u when r = 0. But although a and /3 may be finite at the lower
limit, the annulment of (17) is secured by the factor r*; so that complex
values of n are still excluded, provided Wl be of unchangeable sign. In the
present case the B of (19) vanishes, and we have
d*W/dr* =
so that (18) gives
W
Wl
satisfying the prescribed condition.
* Basuet's Bydrodynamiet, § 514.
582 ON THE QUESTION OF THE [194
The difficulty in reconciling calculation and experiment is accordingly
not to be explained by any peculiarity of the two-dimensional motion to
which calculation was first applied. It may indeed be argued that the
instabilities excluded are only those of the exponential type, and that there
may remain others on the borderland of the form tcost, &c. But if the
above calculations are really applicable to the limiting case of a viscous fluid
when the viscosity is infinitely small, we should naturally expect to find that
the smallest sensible viscosity would convert the feebly unstable disturbance
into one distinctly stable, and if so the difficulty remains. Speculations
on such a subject in advance of definite arguments are not worth much ; but
the impression upon my mind is that the motions calculated above for an
absolutely inviscid liquid may be found inapplicable to a viscid liquid of
vanishing viscosity, and that a more complete treatment might even yet
indicate instability, perhaps of a local character, in the immediate neighbour-
hood of the walls, when the viscosity is very small.
It is on the basis of such a complete treatment, in which the terms
representing viscosity in the general equations are retained, that Lord Kelvin*
arrives at the conclusion that the flow of viscous fluid between two parallel
walls is fully stable for infinitesimal disturbances, however small the amount
of the viscosity may be. Naturally, it is with diffidence that I hesitate to
follow so great an authority, but I must confess that the argument does not
appear to me demonstrative. No attempt is made to determine whether in
free disturbances of the type eint (in his notation e"°') the imaginary part of n
is finite, and if so whether it is positive or negative. If I rightly understand
it, the process consists in an investigation of forced vibrations of arbitrary
(real) frequency, and the conclusion depends upon a tacit assumption that if
these forced vibrations can be expressed in a periodic form, the steady motion
from which they are deviations cannot be unstable. A very simple case
suffices to prove that such a principle cannot be admitted. The equation
to the motion of the bob of a pendulum situated near the highest point
of its orbit is
d-x/dt2-mia; = X, (21)
where X is an impressed force. If X = cospt, the corresponding part of # is
« — - 255E5.,; . ...(22)
p1 + m2
but this gives no indication of the inherent instability of the situation
expressed by the free " vibrations,"
oc^Ae'M + Be-1** (23)
As a preliminary to a more complete investigation, it may be worth while
to indicate the solution of the problem for the two-dimensional motion ol
viscous liquid between two parallel planes, in the relatively very simple case
* Phil. Mag. Aug. and Sept. 1887.
1892] STABILITY OF THE FLOW OF FLUIDS. 583
where there is no foundation of steady motion. The equation, given in Lord
Kelvin's paper, for the motion of type e?{nt+k* is
(24)
The boundary conditions, say at r = ± a, are that u, (v), and w shall there
vanish, or by (7) that
u = 0, dujdx = 0.
The following would then be the proof from the differential equation that for
all the admissible values of n, p is zero and q is positive.
Writing as before, u = a + i/3, and separating the real and imaginary parts,
we find
Multiply (25), (26) by a, fi respectively, add, and integrate with respect to
a: over the range of the motion. The coefficient of q is
and this is equal to zero in virtue of the conditions at the limits. In like
manner the coefficient of /* is zero, as appears on successive integrations by
parts. The coefficient of p is
so that p = 0.
Again, multiply (25) by £, (26) by a, and subtract. On integration as
before the coefficient of q is
and that of ft, is
Hence q has the same sign as /x, that is to say, q is positive. That n
in f1* is a pure positive imaginary is no more than might have been inferred
from general principles, seeing that the problem is one of the small motions
about equilibrium of a system devoid of potential energy.
Since (24) is an equation with constant coefficients, the normal functions
in this case are readily expressed. Writing it in the form
584 ON THE STABILITY OF THE FLOW OF FLUIDS. [194
we see that the four types of solution are
ekx, e~k\ eik'*, e~ik'*,
where
-k'2 = k* + m/tJL; (28)
or, if we take advantage of what has just been proved,
w-qlr-ie, (29)
where q and //, are positive. It will be seen that the odd and even parts of
the solution may be treated separately. Thus, for the first,
u = Asinhkx + £sink'a;, (30)
and the conditions to be satisfied at x = ± a give
0 = A sinh ka + B sin k'a, 0 = kA cosh ka + k'B cos k'a ; (31 )
so that the equation for k' is
tan k'a tanh ka ,„„*
~W^ ~ka~~
Again, for the solution involving the even functions,
u = Ccoshkx + Dcosk'x, (33)
where
cot k'a coth ka , 0 . N
HF^ = kcT (34)
Equations (32), (34) give an infinite number of real values for k', and
when these are known q and n follow from (29).
The most persistent motion (for which q is smallest) corresponds to a small
value of k, and to the even functions of (33). In this case from (34)
k'a = TT, 2?r, 3-7T, &c.,
the first of which gives as the smallest value of q
q = H-n*/a2 (35)
The corresponding form for u is
M = 6'-fa-«t{l + cos (war/a)} (36)
This type of motion is represented by the arrows in the following diagram : —
On the other hand the smallest value of q under the head of the odd
functions is
g = yuTT2 (l-4303)2/a2, (37)
and the motion is of the type
"T" ~7~
195.
OX THE INSTABILITY OF A CYLINDER OF VISCOUS
LIQUID UNDER CAPILLARY FORCE.
[Philosophical Magazine, xxxiv. pp. 145 — 154, 1892.]
THE main outline of the theory of the instability of a long cylinder of
liquid is due to Plateau, who showed that if the equilibrium surface r = a be
slightly deformed so as to become
r = a + a cos kz, (1)
in which z is measured parallel to the axis, the deformation is stable or
unstable according as ka is greater or less than unity ; that is, according as
the wave-length X of the varicosity is less or greater than 2ira, the circum-
ference of the cylinder. The solution of the merely statical question is,
however, insufficient for the application to the important problem of the
disintegration of a jet of liquid. A deformation of any wave-length
exceeding 2?ra increases exponentially with the time (e9*)', and what we
require to know is the relation between q and X. A value of X, if any, for
which q is a maximum, determines the mode of maximum instability ; and
this in general tends more and more to be the actual mode of disintegration
as the initial disturbances are reduced without limit.
As resisting the capillary force, Plateau seems to have had in view only
viscosity. " Par suite des viscosites interieure et superficielle, le rapport
entre la longueur normale des divisions et le diametre du cylindre surpasse
toujours la limite de la stabilite. II se reduirait sans doute a cette limite
meme si le liquide etait exempt de toute viscosite*." In the case of jets
composed of liquids such as water, the influence of viscosity might be
expected to be small, but it would be erroneous to conclude that the limit
lea = l would then correspond to the mode of maximum instability. Inertia
* Statique experimental* et theorique det liquidei toumit aux teulet forett moltculairet, 1873
Vol. n. p. 231.
37—5
586 ON THE INSTABILITY OF A CYLINDER [195
would necessarily play a leading part, and from a complete investigation* it
appears that the mode of maximum instability corresponds with A, = 4*51 x 2a,
exceeding very considerably the circumference of the cylinder.
The importance attached by Plateau to viscosity suggested an investi-
gation in which this property should be included ; and the results would at
any rate find application to threads of materials like glass and sealing-wax, in
which viscosity would predominate over inertia. Having in my mind some old
experiments upon the behaviour of fine threads of treacle deposited upon
paper, which slowly resolve themselves into drops having a very similar
appearance to those obtained from a jet of water, I rather expected to find
that under the influence of viscosity alone the mode of resolution would be
nearly the same as under the influence of inertia alone. This anticipation
proved to be wide of the mark, the result showing that under viscosity alone
the value of X for maximum instability would be very great. And a little
consideration shows that the retarding forces exerted by the paper support
may well be of quite a different character from those due to mere fluid
viscosity. In the latter case the gathering together from considerable
distances is but little resisted, the motion not differing greatly from that
of a solid body, whereas such a mode of resolution would be greatly impeded
by the contact with paper. In order better to represent such contact forces,
I have considered the problem in the form which it assumes when the resist-
ances are proportional to the absolute velocities of the parts. This admits
of easy solution, and the result illustrates the behaviour of the thread of
treacle in contact with paper, and shows that there is a marked difference
between this case and that of a thread whose disintegration is resisted by
true fluid viscosity.
The introduction of resistances proportional to absolute velocities does
not interfere with the irrotational character of the motion of otherwise
frictionless fluidf. The radial and axial velocities u, w may thus, as usual,
be regarded as derived from a velocity-potential according to the equation
u = d<f>/dr, w = d<f>/dz ......................... (2)
If the resistance is p times the velocity, the general equation of pressure, viz.
p/p = R-d<f>/dt-$U2, ......................... (3)
becomes for the present purpose, where U2 may be neglected,
...(4)
The quantities defining the motion are as functions of z proportional to eikz,
and as functions of t proportional to eint, where k is real, but n may be
* Proc. Math. Soc. November 1878. [Vol. i. p. 361.] See also below.
t Theory of Sound, Vol. n. § 239 (1878).
1892] OF VISCOUS LIQUID UNDER CAPILLARY FORCE. 587
complex. The general equation for the velocity-potential of an incompressible
fluid, viz. Vȣ = 0, thus becomes
of which the solution, subject to the condition to be imposed when r = 0, is
<i>
or rather
(5)
At the same time p is given by
We have now to consider the boundary condition, applicable when r = a.
The displacement £ at the surface is connected with <f> by the equation
%dt = h¥r "(7)
The variable part of the pressure is due to the tension T, which is supposed
to be constant, as is practically the case in the absence of surface-contamination.
The curvature in the plane of the axis is — d*% dz1, or AJ£. The curvature in
the perpendicular direction is (a + £)~l, or I/a — £, a*. Thus
and the boundar condition is
or by (5),
a quadratic equation by which n is determined.
as found in the former paper. In this expression ika Jt' Jt is a real positive
quantity for all (real) values of ka ; so that the displacement is exponentially
unstable if ka< 1, and periodic if ka > 1, as was to be expected. In the
former case the values of in are numerically greatest when ka = ir/4'5.
In the other extreme case where inertia may be neglected in comparison
with viscosity, we have
so that the instability is greatest when ka has the same value as in the first
case.
588 ON THE INSTABILITY OF A CYLINDER [195
The general form of the quadratic is
(m)2 + w. ////> + # (&2 a2 -1) = 0, .................... (12)
where H is positive.
If ka < 1, both values of in are real, one being positive and the other
negative. The displacement is accordingly unstable, and the greatest
instability occurs with the above-defined value of ka. If, on the other hand,
ka > 1, the values of in may be either real or imaginary. In the former case
both values are negative, and in the latter the real parts are negative, so that
the deformations are stable.
The investigation applicable to a real viscous liquid of viscosity /*, or pv,
is much more complicated than the foregoing, mainly in consequence of the
non-existence of a velocity-potential. But inasmuch as the motion is still
supposed to be symmetrical about the axis, the equation of continuity gives
where -»/r is Stokes's current function. For small motions ^r satisfies the
equation *
In the present question i/r as a function of z and t is proportional to ei(nt+kz),
and it may be separated into two parts, fa and fa, of which fa satisfies
and fa satisfies
where
_ _
r dr\r dr
(17)
At the surface we have to consider the normal stress P, and the tangential
stresses. Of the latter one vanishes in virtue of the symmetry, and the other
is to be made to vanish in conformity with the condition that there is to be
no impressed tangential force -f. Thus
or in terms of ty by (13)
* Camb. Tram. 1850. See also Basset's Hydrodynamics, Vol. n. p. 259.
f It is here assumed that there is no " superficial viscosity."
1892] OF VISCOUS LIQUID UXDER CAPILLARY FORCE. 589
Introducing ^,, ^, and having regard to (15), (16), we may express this
condition in the form
afrfk-HJp+J^-o, ........................ (20)
which is to be satisfied when r = a.
Again, for the normal stress,
Herein
For
r dr V dr r dr / r dr
for ^2,
r dr \ dr r dr / r dr ~ r dr '
so that
- ...... (22)
dr dr r J AT dr
The variable part of the capillary pressure is, as we have already seen,
in which g
Thus, the condition to be satisfied when r = a is
+ »- ...... (23)
na r vka dr dr r Jra dr
The forms of ^, ^r, are to be determined by the equations (15), (16), and
by the conditions to be satisfied when r = 0. It will be observed that ^r,
satisfies the condition appropriate to the stream-function when there is a
velocity-potential This would be of the form
4> = <#*J.(ikr\ .............................. (24)
so that
* = J(ru) *-£;£- r ««•/.' (ikr\
Thus
* = ^rJ.'(a-r) .............................. (25)
is the most general form admissible, as may be verified by differentiation. In
this Jt(ikr) satisfies the equation
~ . 0 .................. (26)
590 ON THE INSTABILITY OF A CYLINDER [195
Since (16) differ from (15) only by the substitution of k' for k, the general
form for -^r.2 is
^ = BrJ0'(ik'r) (27)
By use of these values the first boundary condition (20) becomes
2k* A J0' (ika) + (k'2 + k2) BJti' (ik'a) = 0 (28)
We have next to introduce the same values into the second boundary
condition (23). In this
- AikaJ0(ika)
j
by (26). In like manner,
d^fdr = — Bik'aJ0 (ik'a).
Thus
, . ,,,.. . DT,,J v.
— ' — [AJ0'(ika) + BJ0 (ik a)]
pa*
J0(ik'a) + 2k* A JQ" (ika) + 2kk'BJ0"(ik'a)
Bik'aJ0(ik'a)] (29)
rt/U>
Between (28) and (29) we now eliminate the ratio A/B, and thus obtain as
the equation by which [in conjunction with (17)] the value of n is to be
determined
k'(k"-k*)J0'(ika)
(80)
We shall now apply this result to the particular case where the viscosity is
very great in comparison with the inertia. The third part of (30) may then
be omitted, and we have to seek the limiting form of the remainder when k'
is nearly equal to k, as we see must happen by (17). In the first part,
In the second,
T"(i
J° I
and
k'(k'*-k*) J0'(ika) J0Sk
J> (lk a) = ~ ~ •
1892] OF VISCOUS UQOD UXDKR CAPIL.LABT FORCE. 591
Thus the limiting form is
pa.
= _ ..
n J? f- "** --fifc-'
in which, however, we may effect farther simplifications bv means of the
properties of J,. We find by use of (26)
so that, finallv.
_-
In (31) the argument of Jt, «/.' is Ou, or z as we will call it for brevity. And
by a known property J"/ = — «/,. Now
so that if x =
These functions have been tabulated by Prof. A. Lodge * under the notation
I.(x), I,(xl where
(32,
^i+¥:^-6-, ....... (33)
In this notation
^7^(x)/A«(JrX ...... (34)
and we have to consider the march of (34) as a function of jr.
When x is very small,
so that
(34) = - 3 + terms in *• ;
and then from (31)
T
tu — _
Spa
Brit. AM. Report, 1889, p. 3B.
592 ON THE INSTABILITY OF A CYLINDER [195
We shall see that this corresponds to the maximum instability, and it occurs
when the wave-length of the varicosity is very large in comparison with the
diameter of the cylinder. The following table gives the values of (34) for
specified values of x: —
X
(34)
x (34)
o-o
- 3-0000
ro
-3-0188
0-2
- 3-0000
2-0
-3-2160
0-4
- 3-0004
4-0 - 4-458
0-6 - 3-0023
6-0
- 6-247
It will be seen that the numerical value of (34) is least when x •= 0, which is
also the value of x for which the numerator of (31) is greatest. On both
accounts, therefore, in is greatest when x or ka = 0. But over the whole
range of the instability from ka = 0 to ka = 1, (34) differs but little from — 3,
so that we may take as approximately applicable
•<*»
The result of the investigation is to show that when viscosity is paramount
long threads do not tend to divide themselves into drops at mutual distances
comparable with the diameter of the cylinder, but rather to give way by
attenuation at few and distant places. This is, I think, in agreement with
the observed behaviour of highly viscous threads of glass or treacle when
supported only at the terminals. A separation into numerous drops, or a
varicosity pointing to such a resolution, may thus be taken as evidence
that the fluidity has been sufficient to bring inertia into play.
The application of (31) to the case of stability (ka>l) is of less interest,
but it may be worth while to refer to the extreme case where the wave-length
of the varicosity is very small in comparison with the diameter. We then fall
upon the particular case of a plane surface disturbed by waves of length X.
The result, applicable when the viscosity is so great that inertia may be left
out of account, is the limit of (31) when a, or x, is infinite, while k remains
constant, or
Tic
•»«!?•+ Lim x \J*(ix)IJ?(ix) + 1 }.
By means of the expressions appropriate when the argument is large, the
limit in question may be proved to be — 1 ; so that
<37>
If gravity be supposed operative in aid of the restoration of equilibrium, we
should have to include in the boundary condition relative to pressure a term
1892] OF VISCOUS LIQUID UNDER CAPILLARY FORCE. 593
qp% in addition to Jl^: so that the more general result is obtainable by
adding gp If to T. Thus
giving the rate of subsidence of waves upon the surface of a highly viscous
material It could of course be more readily obtained directly.
When gravity operates alone,
in--2£—-±- (39)
2/Jt~ 2**' '
which agrees with a conclusion of Prof. Darwin*. A like result may be
obtained from equations given by Mr Basset f.
* Pkii. Zraxs. 1879, p. 10. In equation (12) write i/«=i, and nuke i= x .
t Hydrvdymamict, ToL n. § 590, equations (21), (27). See also Tmit, EXimb. Pne. 1890, p. 110.
196.
ON THE INSTABILITY OF CYLINDRICAL FLUID SURFACES.
[Philosophical Magazine, xxxiv. pp. 177—180, 1892.]
IN former papers* I have investigated the character of the equilibrium of
a cylindrical fluid column under the action of capillary force. If the column
become varicose with wave-length X, the equilibrium is unstable, provided X
exceed the circumference (27m) of the cylinder ; and the degree of instability,
as indicated by the value of q in the exponential eft1 to which the motion
is proportional, depends upon the value of X, reaching a maximum when
X = 4'51 x 2a. In these investigations the external pressure is supposed to
be constant; and this is tantamount to neglecting the inertia of the sur-
rounding fluid.
When a column of liquid is surrounded by air, the neglect of the inertia
of the latter will be of small importance ; but there are cases where the
situation is reversed, and where it is the inertia of the fluid outside rather
than of the fluid inside the cylinder which is important. The phenomenon
of the disruption of»a jet of air delivered under water, easily illustrated by
instantaneous photography [Vol. III. p. 443], suggests the consideration of the
case where the inside inertia maybe neglected; and to this the present paper
is specially directed. For the sake of comparison the results of the former
problem are also exhibited.
Since the fluid is supposed to be in viscid, there is a velocity-potential,
proportional to eikz as well as to eqt, and satisfying the usual equation
*4 + lf*-*,_0 (!)
dr2 r dr
If the fluid under consideration is inside the cylinder, the appropriate
solution of (1) is
0 = «70(;&r) = /„(&•); (2)
and the final result for q2 is
T (1 - tort) ika . J0' (ika) = T (1 - fro') kaI0' (ka)
^ pa3 J0(ika) pa3 I0(ka)
* (1) " On the Instability of Jets," Math. Soc. Proc. November 1878. [Vol. i. p. 361.]
(2) " On the Capilkry Phenomena of Jets," Proc. Roy. Soc. May 1879. [Vol. i. p. 377.]
(3) "On the Instability of a Cylinder of Viscous Liquid under Capillary Force," supra,
p. 145. [Vol. in. p. 585.]
1892] OS THE INSTABILITY OF CYLINDRICAL FLUID SURFACES. 595
in which T represents the capillary tension, p the density, and, as usual,
But if the fluid be outside the cylinder, we hare to use that solution of
(1) for which the motion remains finite when r = x . This may be expressed
in two ways*. When r is great we have the semi-convergent form
-
and for all values of r the fully convergent series
in which 7 is Euler's constant, equal to '5772..., and
S. = l+| + J+...-H>. ........................ (8)
In this case the solution of the problem becomes
T L'a'-in'v)
<f> being defined by (7). In (9) p represents the inertia of the external fluid,
that of the internal fluid being neglected, while in the corresponding formula
(3) p is the inertia of the internal fluid, that of the external fluid being
neglected- There would be no difficulty in writing down the anal vtical
solution applicable to the more general case where both densities are
regarded as finite,
The accompanying Table gives the values of
[&) - J
to which q in (3) is proportional, and of
<»>
corresponding in a similar manner to (9). In the second case we have
On account of the fector (1 - *») both (10) and (11) vanish when x = 0 and
when x = 1. Beyond x = 1, (10), (11) become imaginary, indicating stability.
It will be seen that when the fluid is internal the instability is a maximum
between or= -6 and x = 7 ; and when the fluid is external, between x = -4 and
x = '5. That the maximum instability would correspond to a longer wave-
length in the case of the external fluid might have been expected, in view of
the greater room available for the flow. The same consideration also explains
the higher maximum attained by (11) than by (10>
* See the writings of Sir G. Stokes; or Theory o/Somrf, $ Ml.
596
ON THE INSTABILITY OF CYLINDRICAL FLUID SURFACES.
[196
X
W
xl^x)
(10)
-+(*)
x+'(z)
(11)
o-o
1-0000
•0000
•0000
oo
1-0000
•oooo
o-i
1-0025
•0050
•0703
2-4270
•9854
6339
0-2
1-0100
•0201
•1382
1-7527
•9551
•7233
0-3
1-0226
•0455
•2012
1-3724
•9169
•7795
0-4
1-0404
•0816
•2567
1-1146
•8738
•8113
0-5
1-0635
•1289
•3015
•9244
•8283
•8198
0-6
1-0920
•1882
•3321
•7774
•7817
•8022
0-7
1-1264
•2603
•3433
•6607
•7353
•7535
0-8
1-1665
•3463
•3269
•5654
•6894
•6625
0-9
1-2130
•4474
•2647
•4869
•6449
•5017
1-0
•oooo
•oooo
In order the better to study the region of the maximum, the following
additional values have been calculated by the usual bisection formula
X
(10)
X
(11)
•65
•3406
•45
•8186
•70
•3433
•50
•8198
•75
•3397
•55
•8147
The value of x for which (10) is a maximum may now be found from
Lagrange's interpolation formula. It is
x = -696,
corresponding to X = 2a x ir\x = 4'51 x 2a, (14)
and agreeing with the value formerly obtained by a different procedure.
In like manner we get for the value of x giving maximum instability in
the case of the external fluid,
x = '485,
and X = 6'48x2a (15)
Some numerical examples applicable to the case of water were given in a
former paper. It appeared that for a diameter of one millimetre the dis-
turbance of maximum instability is multiplied one-thousand-fold in about
one-fortieth of a second of time. This is for the case of internal fluid. If
the fluid were external, the amplification in the same time would be more than
one-million-fold.
END OF VOL. III.
CAMBRIDGE : PRINTED BY J. AND C. F. CLAY, AT THE UNIVERSITY PRESS.
University of California Library
Los Angeles
This book is DUE on the last date stamped below.
APR <•<"'.. 1999
WAY.
FE
JUL
OCTI
MAI
OCT
MAR:
HP
OC3R2HIJ
L OO6 589 1 88 9
001013951 7
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11