Beruflich Dokumente
Kultur Dokumente
Stokes's Law
Author(s): R. A. Millikan
Source: Science, New Series, Vol. 32, No. 822 (Sep. 30, 1910), pp. 436-448
Published by: American Association for the Advancement of Science
Stable URL: http://www.jstor.org/stable/1634813
Accessed: 16-01-2017 00:40 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
American Association for the Advancement of Science is collaborating with JSTOR to digitize,
preserve and extend access to Science
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
436 SCIENCE [N. S. VOL. XXXII. No. 822
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTEMBER 30, 1910] SCIENCE 437
enclosed, and the temperature controlled so the charge e,, which it carries, its speed v,
that the air within the condenser is altogether under gravity, and its speed v, under the in-
stagnant. The droplet, once inside the con- fluence of an electrical field of strength F, are
denser, is illuminated through a small window given by the simple equation
by a beam from an arc light, so that it appears
v- mgg mg
in the field of view of the observing cathe- (1)
tometer telescope like a bright star on a black
VFV---- en - mg F or e- -
background. This star, of course, falls under This equation involves no assumption what-
the action of gravity toward the lower plate, ever save that the speed of the drop is propor-
but before it reaches it, an electrical field of tional to the force acting upon it, an assump-
strength between 3,000 volts and 8,000 volts tion which is fully and accurately tested ex-
per centimeter is thrown on between the perimentally in the following work. Further-
plates, and, if the droplet had received a more, equation (1) is sufficient not only for
charge of the proper sign and strength asthe
it correct determination of the relative values
was blown out through the atomizer, it is of all of the charges which a given drop may
pulled up by this field against gravity, towardhave after the capture of a larger or smaller
the upper plate. Before it strikes this plate number of ions, but it is also sufficient for the
the field is thrown off, the plates short-cir- establishment of all of the assertions made
cuited, and the time required by the drop to
above, except 3, 4 and 6, and for the estab-
fall under gravity the distance corresponding
lishment of 4 no other exact relationship is
to the space between the cross hairs of the ob-
needed. However, for the sake of obtaining a
serving telescope is accurately determined.
provisional estimate of the value of m in equa-
Then the rate at which the droplet moves up
under the influence of the field is measured tion (1), and therefore of making a provi-
sional determination of the absolute values of
by timing it through the same distance when
the field is on. This operation is repeated and the charges carried by the drop, Stokes's law
the speeds checked an indefinite number of will, for the present, be assumed to be correct,
times, or until the droplet catches an ion from but it is to be distinctly borne in mind that
among those which exist normally in air, or the conclusions just now under consideration
which have been produced in the space be- are not at all dependent upon the validity of
tween the plates by any of the usual ionizing this assumption.
agents like radium or X-rays. The fact that This law states that if , is the coefficient of
an ion has been caught, and the exact instant viscosity of a medium, X the force acting upon
at which the event happened are signaled to a spherical drop of radius a in that medium,
the observer by the change in the speed of the and v the velocity with which the drop moves
droplet under the influence of the field. From under the influence of the force, then
the sign and magnitude of this change in
X = 67r/Cav. (2)
speed, taken in connection with the constant
speed under gravity, the sign and the exact The substitution in this equation of the
value of the charge carried by the captured resulting gravitational force acting on a
ion are determined. The error in a single ob- spherical drop of density o in a medium of
servation need not exceed one third of one
density p gives the usual expression for the
per cent. Furthermore, it is from the values
rate of fall, according to Stokes, of a drop
of the speeds observed that all of the conclu-
under gravity, viz.,
sions above mentioned are directly and simply
deduced. 2 ga2
VIv=u !~-
9C (a- p).
(3)
? 3. The Deduction of the Relative Values
of the Charges Carried by a Given Droplet.- The elimination of m from (1) by means of
The relations between the mass m of a drop, (3), and the further relation
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
438 SCIEi VCE [N. S. VOL. XXXII. No. 822
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTEMBER 30, 1910] 439
SCIENCE
TABLE I
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
440 SCIENCE [N. S. VOL. XXXII. No. 822
caught in the first change, i. e., from 71 to that quantity. These observations are the
380, carried exactly the same charge as the justification for assertions 1 and 2 of the in-
negative ion caught in the change from 380 troduction.
to 71; or again, that the negative ion caught Before discussing assertion 4 it is desirable
in the change from 71 to 39, had exactly the to direct attention to three additional con-
clusions which can be drawn from table I.:
same charge as the positive ion caught in
the change from 39 to 71. First, since the time of the drop in the field
varied in these observations from 380 seconds to
Furthermore, the exact value of the charge
6.7 seconds, it will be seen that the resultant
caught in each of the above cases is obtained
moving force acting upon the drop was varied
in terms of mg from the differences in the
in the ratio 1 to 55, without bringing to light
values of e,, given by equation (1), and if it
the slightest indication of a dependence of e1
be assumed that the value of m is approxi-
upon the velocity. Independently of theory,
mately known through Stokes's law, then the
therefore, we can assert that the velocity of this
approximately correct value of the charge on
drop was strictly proportional to the moving
the captured ion is given by the difference force. The certaintywith which this conclusion
between the values of e, obtained through can be drawn may be seen from a consideration
equation (4). The mean value of this dif- of the following numerical data. Although we
ference obtained from all the changes in had the upon our drop all possible multiples of
latter half of table I. (see Differences) isthe unit 4.917 X 10-10 between 4 and 17, save
4.93 X 10-10. only 15, there is not a single value of e, given
Now it will be seen from the first observa- in the table which differs by as much as .5
tion given in the table that the charge which per cent. from the final mean e1. It is true
was originally upon this drop and which wasthat the observational error in a few of the
obtained not from the ions in the air, but fromsmaller times is as much as 1 or 2 per cent.,
the frictional process involved in blowing the but the observational error in the last half of
spray, was 34.47 X 10-10. This number comes the table should nowhere exceed .5 per cent.
within one seventh of one per cent. of being In no case is there here found a divergence
from the final value of el of more than .4 per
exactly seven times the charge on the positive
cent.
or on the negative ion caught in the ob-
Second, since the charge on the drop was
servations under consideration. Mr. Harvey
multiplied more than four times without
Fletcher and myself, who have worked to-
changing at all the value of G, or the value of
gether on these experiments since December,
e1, the observations prove conclusively that in
1909, studied in this way between December the case of drops like this, the drag which the
and May from one to two hundred drops air exerts upon the drop is independent
which had initial charges varying between the of whether the drop is charged or uncharged.
limits 1 and 150, and which were upon as In other words, the apparent viscosity of the
diverse substances as oil, mercury and gly- air is not affected by the charge in the case
cerine, and found in every case the originalof drops of the sort used in these experiments.
charge on the drop an exact multiple of the Third, it will be seen from the table that in
smallest charge which we found that the drop general a drop catches an ion only when the
caught from the air. The total number of field is off. Were this not the case there
changes which we have observed would be be- would be many erratic readings in the col-
tween one and two thousand, and in not one umn under F, while in all of the four and one
single instance has there been any changehalf hours during which these experiments
which did not represent the advent upon thelasted, there is but one such. A moment's
drop of one definite invariable quantity ofconsideration will show why this is. When
electricity, or a very small exact multiple ofthe field is on, the ions are driven with enor-
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTEMBER 30, 1910] SCIENCE 441
mous speed to the plates as soon as they are 16e' 16 X (4.901 X 10-o) 195 X 10-4 ergs.
formed, their velocities in the fields here used
= y 1 .95 X 10-" ergs.
r - .000197
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
442 SCIENCE [N. S. VoL. XXXII. No. 822
changes which correspond to the addition or ions were alike in number and mobility, the
subtraction of 2 or 3 times this amount. The chance that the drop would catch a negative
conclusion, however, that valency is exhibited
ion would be as many times its chance of
in gaseous ionization is not to be so easily catching a positive ion as the distance from
drawn. The arguments for it which are fur- the drop to the negative plate was times the
nished by our experiments will be presented distance from the drop to the positive plate.
fully elsewhere. Space here only permits theSimilarly, if we wished to load the drop posi-
statement that the only strong argument fur- tively it was held by the field close to the
nished by table I. is found near the end of the negative plate. On account of the slightly
table where, when the field was on, the drop greater mobility of the negative ions and also
caught a double negative ion, while I was on account of the somewhat greater numbers
looking at it. in which they occur, we found, in general, a
Some idea of the intensity of ionization slightly greater tendency of the drops to take
used in these experiments may be gained fromup negative than positive charges. In view,
the statement that during the observations therefore, of the greater ease with which nega-
recorded in the first half of the table, a closed tive drops could be held for long intervals
tube of radium, containing 500 mg. of radium without being lost to the plates most of the
bromide of activity 3,000, stood about five feet drops studied have been of negative sign.
away from the testing chamber, so that its ? 5. The Failure of Stokes's Law.-When
y rays could enter this chamber. At the end the values of el were computed, as above, for
of the observations in the group in which different drops, although each individual drop
G -- 23.14, this radium was brought up to showed the same sort of consistency which was
within a few inches of the testing chamber, exhibited by the drop of table I., the values of
and six elementary charges were forced upon el at first came out differently even for drops
the drop in a manner which will presently be showing the same value of the velocity under
explained. The radium was then taken en-
gravity. This last irregularity was practically
tirely out of the room, so that the changes re-
completely eliminated by blowing the drops
corded in the last half of the table are entirely
into air which was strictly dust-free, but even
due to such ionization as exists in air under
then drops of different sizes as determined by
normal atmospheric conditions.
vI always gave consistently different values of
There is but one more comment to be made
upon table I. At a point indicated in the TABLE II
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTEMBER 30, 1910] SCIENCE 443
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
444 SCIENCE [N. S. VOL. XXXII. No. 822
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTEMBE -30, 1910] SCIENCE 445
TABLE X I I I I I
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
446 SCIENCE [N. S. VOL. XXXII. No. 822
studied throughout a period of 47 consecutive and since further a sphere must have a higher
days. The three drops which have been ex- rate of fall than a body of any other form
cluded all yielded values of el from two to fourwhatever having the same mass and density,
per cent. too low to fall upon a smooth elyvthe hypothesis of binary drops to account for
curve like that shown in Fig. 1, which is the an occasional low value of e1 is at least nat-
graph of the results contained in table XII. ural. After eliminating dust we found not
It is probable that these three drops corre- more than one drop in ten which was irregu-
sponded not to single drops, but to two drops lar. The drop shown in table I. is perhaps the
stuck together. Since we have never in all best illustration of the case under considera-
our study observed a drop which gave a value tion which we have observed. It yields a
of eI appreciably above the curve of Fig. 1,value of e, which is four per cent. too low to
or fall on the curve of Fig. 1. This is as large a
TABLE XI departure from this curve as we have thus far
Negative drop No. 32 obtained.
Distance between cross hairs = 1.003 cm. ? 6. The Correction of Stokes's Law.-The
Temperature = 23.2? C. procedure actually adopted for correcting
Stokes's law will be detailed elsewhere. The
end result is this. An equation of the follow-
G (sec.) F (sec.) n I e?X1010 I el101o
F==8.5 - 8.7) ing form is made to replace Stokes's equa-
V= 8577 - 8.3 123 622.40 tion (2):
_- 8.5)
Changed without radium. X = 6riav (1 + A (5)
M 2.44 28.4
- 28.7
28.7 29 g 2(a -p) ,+A (6)
F= 28.70 v a
V- 8573 "
"2.46 28.4 104 524.25 5.040
2.54 29.0
2.46 29.0 in which a is the radius of the drop, I the
2.45 28.8
mean free path of the gas molecule, and A an
" 2.43 28.6
undetermined constant which we obtain from
Change forced with radium.
G -2.462 "2.44 15.7 our observations. It turns out that A is iden-
2.48 15.7
tical to within the limits of observational
F= 15.72 - 15.7 111 558.78 5.034
V- 8568 - 15.7 error (not more than 1 or 2 per cent.) with the
15.8
value deduced by Cunningham8 from the
Change forced with radium.
F= 59.1 kinetic theory considerations, provided the
V= 8565 '"2.50 j591
2.50 59.1 } 100 503.42 5.034
f of his formula7 is made equal to zero. This
F= 60.0
means that the value of A given by our obser-
V- 8563 F 2.5
F 2.45 9} }100 503.22 5.032
60.2II
Change forced with radium. vations is .815. The values of a in tables XII.
F= 81.5 and XIII. are computed from (6), in which
-- 8561 - :81 } 99 498.12 1 5.031
a is now the only unknown.
Change forced with radium.
" 2.44 19.0 ? 7. The Absolute Value of e.-Using now
F= 20.0 " 2.50 20.1 108 543.41 5.032
(6) instead of (4) to combine with (1) and
V- 8555 " 2.42 20.0 J__
denoting by e the absolute value of the ele-
-
".41.} I
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
SEPTrEnBaW 3,0, 1910] SCIENCE 447
Table XIII. contains the values of e obtained the measurement of this constant, After a
from all of the observations recorded in table prolonged and very careful study of all the
XII. except the first four and the last six. data available on the viscosity of air I have
These are omitted not because their introduc- chosen as the most probable value of jp at
tion would change the final value of e, for 15?.0001785.
as For reasons which will be de-
tailed elsewhere it is thought that the error
a matter of fact this is not appreciably altered
by including them, but solely because of in thethis value is less than one half of one per
experimental errors involved in work upon cent.
either exceedingly slow or exceedingly fast It is most interesting that the agreement
drops. When the velocities are exceedingly between Cunningham's rational formula and
slow residual convexion currents introduce our experimental results is so perfect. How
errors, and when they are exceedingly perfect fast it is may be seen graphically from
the time determination becomes uncertain. Fig. 2, in which the curve is computed from
The final mean value of e is 4.9016 X 1010.
7 under the assumption of e -4.9016 and our
The probable error computed from the number- experimentally determined values of e are
of observations shown in the last column and plotted about this curve, every observation
their average divergence should be about, onecontained in Table XII. being shown in the
tenth of one per cent. Since, however, the figure.
co- Nevertheless, it is to be particularly
efficient of viscosity of air is involved in the
emphasized that the correctness of our final
formula the accuracy with which e is known value of the elementary electrical charge is
is limited by that which has been obtained completely
in independent of the correctness of
any theory whatever as to the cause
TABLE XII
of the failure of Stokes's law for small
1 .001315 .0000313 7.384 6.0 other than oil may lead to other values of A,
2 .001673 358 6.864 4.0
3 .001927 386 6.142 2.5
TABLE XIII
4 .006813 755 5.605 1.5
5 .01085 967 5.490 .5
6 .01107 979 5.496 .7 No. Velocity Radius e1Xlloo Prob. eXlO10 Per C
7 .01164 .0001004 5.483 .4 cm./sec. cm. Error. Error.
8 .01176 1006 5.482 .4
9 .01193 1016 5.458 .8 5 .01085 .0000967 5.490 .5 4.892 .20
10 .01339 1084 5.448 .5 6 .01107 979 5.496 .7 4.889- .26
11 .01415 1109 5.448 .4 7 .01164 .0001004 5.483 .4 4.903 .03
12 .01868 1281 5.349 .5 8 .01176 1006 5.483 .4 4.916 .28
13 .02613 1521 5.293 .5 9 .01193 1016 5.458 .8 4,891 .22
14 .03337 1730 5,257 .5 10 .01339 1084 5.448 .5 4.908 .10
15 .04265 1954 5.208 .5 11 .01415 1109 5.448 .4 4.921 .42
16 .05360 2205 5.148 .4 12 .018688 1281 5.349 .5- 4.900 .03
17 .05534 2234 5.145 .5 13 .02613 1521 5.293 .5 4.910 .17
18 .06800 2481 5,143 .7 14 .03337 1730 5.257 .5 4.918 .34
19 .07270 2562 5.139 .5 15 .04265 1954 5.208 .5 4.913 .21
20 .08843 2815 5.102 .3 16 .05360 ' 2205 5.143 .4 4.884 .36
21 .09822 2985 5.107 .4 17 .05534 2234 5.145 .5 4.885 .34
22 .1102 3166 5.065 .4 18 .06800 2481 5.143 .7 4.912 .21
23 .1219 3344 5.042 .5 19 .07270 2562 5.139 .5 4.913 .01
24 .1224 3329 5.096 .5 20 .08843 2815 5.102 .3 4.901 .01
25 .1267 3393 5.061 .5 21 .09822 2985 5.107 .4 4.915 .27
26 .15145 3712 5.027 .5 22 .1102 3166 5.065 .4 4.884 .36
27 .1644 3876 5.050 .3 23 .1219 3344 5.042 .5 4.882 .40
28 .2027 4297 4.989 .7 24 .1224 3329 5.096 .5 4.923 .44
29, .2175 4447 5.046 .4 25 .1267 3393 5.061 .S 4.894 .15
30 .3089 5315 4.980 1.0 26 .15145 3712 5.027 .5 4.880 .44
31 .3969 6047 5.060 1.0 27 .1644 3876 5.050 .3 4.903 .03
32 .4074 6104 5.033 1.0
Final mean e = 4.9016
33 .4735 6581 4.911 1.5
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms
448 SCIENCE [N. S. VOL. XXXII. No. 822
+i 44 4i I I ! ,-i` i, ,~
7 1 fX i~~~~~~~~t-- W~ ~-i ~qlt-~ l l0 gi-14f
FIG. 1
FIG. 1
Hii i 444-1 i 1 i i-
iii
+4- -H i-H-
f44- -H-
76
-.L4-
7't
71.
-4t:
75
. . . . .......
60
4F
6.6
. . . . . . . - - - - -
T T
6.1
It.
H+
+ .. .... + H 4+F-
q.g
qg
7tt
FF
FxlG. FIG. 2
but the value of e should in no way be af- Cunningham's theory, and we.shall soon be in
fected thereby. It is of immense interest toa position to settle this point and to make a
know whether varying the mean free path byfurther communication upon it.
varying the pressure will affect the value of A R. . MILLIKAN
in the way in which it ought according to THE UNIVERSITY OF CHICAGO
This content downloaded from 129.97.58.73 on Mon, 16 Jan 2017 00:40:01 UTC
All use subject to http://about.jstor.org/terms