a 
| Fuly 2, 1885] 
millimetre, no regular change in specific resistance could be 
detected, the actual variations lying within 2°5 per cent. 
The conclusion was thus arrived at that the specific resistance 
of the liquid of which a soap film is formed does not differ from 
that of the same liquid in mass, at all events when the thickness 
is greater than 374. 107° mm., and that comparatively small 
changes in the temperature or hygroscopic state of the air in 
contact with the film are attended with great alterations in the 
specific resistance, which indicate a considerable change in 
composition. 
The method of experiment made it possible to determine the 
amount of this change. Solutions were made up representing 
‘liquide glycérique” which had lost or gained given percentages 
of water, their specific resistances were determined at various 
temperatures, and approximate formulze obtained by which the 
percentage of water present could be calculated if the specific 
resistance and temperature were known. 
*The results of the application of this method of analysis to a 
film are shown in the accompanying figure. The abscissz 
represent time, the ordinates of curve I. represent the average 
thickness of the film. It will be observed that the film con- 
tinued to get thinner during the whole time that it was under 
observation. The electrical observations, however, proved that 
at first the product of the resistance and thickness steadily 
increased, indicating a continuous loss of water. Curve II. 
shows the number of parts of water in 100 of the solution lost 
at the times indicated by the abscissez. After a while a piece of 
blotting paper which had been hung up inside the case was 
moistened with water. While this was being done the obser- 
vations were interrupted. On their renewal it was found that 
although the film thinned as steadily as before, the product of 
the resistance and thickness diminished instead of increasing. 
Curve III. shows the steady absorption of water which followed 
the moistening of the air. These experiments proved that it is 
possible for a film to undergo great changes in composition 
without any indication of the fact being afforded by the colours 
it displays. They show that if the composition of the ‘‘liquide 
glycérique”’ is to be kept constant, all change in the temperature 
and hygrometric state of the air must be as far as possible 
prevented. In later experiments this condition has been secured 
by placing the film box in the centre of a water tank, and by 
keeping an endless band of linen hung up within the case, and 
which dips into the liquid, continually moistened. Observations 
made with this apparatus show that these precautions which are 
certainly necessary are also sufficient. 
The second point to which special attention has hitherto been 
given by Prof. Reinold and myself is the measurement of the 
thickness of very thin films. If the thickness is Jess than a 
certain magnitude, the films appear black, and thus their colour 
gives orly a limit to and not a measure of their thickness. 
Black films display many remarkable properties. In general 
there is a sudden change in thickness at the edge of the black 
NATORE 
211 
indicated by the omission of several colours, or sometimes of one 
or two orders of colours. It is only under rare conditions that a 
gradual change in thickness can be observed from the white to 
the black of the first order. 
To determine the thickness of the black its resistance was 
measured, and the thickness calculated on the assumption that 
the specific resistance was the same as that of the liquid in mass. 
The observations were made in several different ways and 
proved that the thickness of the black portion remains constant 
in any given film, however much its area may alter. Thus, in 
the case of a group of films measured by Wheatstone’s bridge, 
the average resistance of a black ring I mm. in breadth was 
1761 megohms when the total breadth was 2 mm., and 1,761 
megohms when the total breadth lay between Io and 12 mm. 
Again, the resistance of the part of the film between the 
needles used in the electrometer method was practically the 
same when the black had extended over the whole film (40 mm. 
long) as it had been when only the upper 11 mm. were black. 
The final measurement differed from the mean by only o*r per 
cent. Again, in another film the resistance of the black per 
millimetre remained the same to within 2°5 per cent. for an hour 
and a half. 
On the other hand the experiments also proved that the thick- 
ness of the black was different in different films. The values 
found varied between 7°2x107® and 14°2x10° mm. These 
differences are quite outside the possible error of experiment. 
If they were due to changes in the constitution of the liquid of 
which the films were formed, it is very improbable that the 
specific resistance of individual films would not have shown 
progressive changes. As has been stated, none such were 
observed. The mean thickness of the five films made of ‘‘liquide 
glycérique” which were observed was I1’9 x 107 mm., while 
that of thirteen films made of soap solution without any glycerine 
was 11°74 x Io-§ mm. 
The assumption made in these calculations that the specific 
resistance of a film, the thickness of which is ten or twelve 
millionths of a millimetre, is the same as that of the liquid in 
mass, is not justified by the previous experiments, which had 
proved it to hold good only to the much greater thickness of 
370x107 mm. It was therefore desirable to check the results 
by an independent method. For this purpose fifty or sixty plane 
films were formed side by side in a glass tube which was placed 
in the path of one of the interfering beams in a Jamin’s Inter- 
ferential Refractometer. The compensator was adjusted so that 
it had to be moved through a large angle to cause one inter- 
ference band to occupy the position previously held by its neigh- 
bour, z.¢. to alter the difference of the paths of the interfering 
rays by one wave-length. This angle was determined for the 
red light of known wave-length transmitted by glass coloured 
with copper oxide. When the films had thinned to the black 
they were broken by means of a needle which had been included 
in the tube along with them, and which was moved, without 
touching the tube, by a magnet. The rupture of the films 
produced a movement of the interference fringes which was 
measured by the compensator, and from which, in accordance 
with well-known principles, the thickness of the films could be 
deduced. 
The mean thickness given by seven experiments on films made 
of “liquide glycérique” was 10°7x10° mm., that obtained 
from nine experiments on films made of soap solution was 
12°rx10-® mm. The mean of these, or 11°4x10 6 mm., 
differed only by 0°4 x 10-6 mm., from the mean thickness deduced 
from the electrical experiments. 
The last point to which reference is necessary is one which 
lies outside the main line of the enquiries above described, but 
which is nevertheless not without interest. In the course of the 
observations it was noticed that the rate of thinning of a film 
seemed to be affected by the passage of the electric current 
through it. Some experiments made on this point last year 
proved the fact beyond the possibility of doubt. The current 
appears to carry the matter of the film with it, so that it thins 
more rapidly if the current runs down, and less rapidly if the 
current runs up than if no current is passing. This may be 
shown as a lecture experiment. 
A vertical rod which can be moved up and down by rackwork 
is passed through the centre of the cover of a glass film-box. 
To the lower extremity is attached a horizontal platinum wire, 
from which another similar horizontal wire is suspended by two 
silk fibres. A film is formed by lowering the whole into the 
liquid with which the lower part of the vessel is flooded. The 
