624 



NA TV RE 



[October 26, 1893 



Mekong; Mr. W. II. Cozens-IIarcly, on surveys and research 

 in Montenegro ; and Mr. E. J. L. Berkeley, on British East 

 Africa. It is also hoped that the Prince of Monaco, Sir 

 Archibald Geikie, and Mr. J. Y. Buchanan may contribute 

 papers. If Mr. and Mrs. Bent return in time from their pro- | 

 jecled exploring journey in Iladramant, an account of their 

 work will be looked forward to before the close of the session. 



A NUMBER of the Journal oi the Manchester Geographical 

 Society just issued (January to June, 1893) contains a paper on 

 the Yoruba country, Abeokuta, and Lagos, by the Rev. J. T. 

 F. Ilalligey, which gives some vivid descriptions of native life 

 and n anners. 



Dr. Gerhard Schott's physical observations on a 

 voyage in a sailing ship from Hamburg to the China coast and 

 back are published as an Ergiinzungsheft of Peteimanns 

 MitUitungcn. In the discussion of his svork Dr. Schott takes 

 account of previous lesearches on the parts of the ocean he 

 traversed, and his paper is an interesting addition to our know- 

 ledge of oceanography. The memoir is divided into two parts : 

 hydrography, including a discussion of surface temperature .is 

 affected by diurnal range, 'rainfall, and wind, the specific gravity 

 of surface water, surface currents and drifts, and observations on 

 waves ; and meteorology, dealing with the instruments employed, 

 the record of air-temperature, humidity, and cloudiness. The 

 memoir is, of course, well illustrated by mips and diagrams. 



THE THICKNESS AND ELECTRICAL 

 CONDUCTIVITY OF THIN LIQUID FILMS. 



PN August, 1883, an article was published in Nature (vol. 



xxviii. p. 389), signed byiProf. Riicker and myself, giving an 



account up to date of our researches on liquid films. Since that 



lime our work has from time to time as opportunity offered been 



ontinued and further results have been obtained, a brief account 

 cummarise the results to which attention was drawn in 1883. 

 of which I now propose to give. It may he useful first to briefly 

 A c)lindrical soap film when allowed to thin under the action 

 of gravity shows in succession the tints of the various orders of 

 Newton's Colours, and finally becomes black. The thickness 

 of any part of the film may be determined (supposing the re- 

 fractive index to be known) from the colour it exhibits when light 

 is reflected from it at a definite angle. The mean thickness of a 

 horizontal ring of the cylindrical film may also be determined 

 by measuring the electrical resistance of the ring, and by assuming 

 the specific conductivity of the film to be the same as that of the 

 liquid in mass. In the case of a liquid consisting of a mixture 

 ot soap solution and glycerine with a liltle potassium nitrate 

 added to increase the conductivity, we proved by comparing the 

 thickness of a film obtained by the optical method with the thick- 

 ness deduced from its electrical resistance, that down to a thick- 

 ness of 374 /i,u (micromillimetres) — corresponding to colours of 

 the second order of Newton's scale— the specific conductivity of 

 the liquid remains unaltered. When the film becomes thinner 

 than 374M/U, and exhibits the colours of the first order, estimates of 

 its thickness derived fromcolour observations are less trustworthy, 

 and when these colours are replaced by black, we only know 

 from the colour that the thickness of the film has less than a 

 certain maximum value. Assuming, however, the specific re- 

 sistance to be unchanged when the film became black we showed 

 that the thickness of such a black film does not diflTer much from 

 1 2w». 



Experiments were then carried out by the electrical method on a 

 solution of oleate of soda (hard soap) containing 3 per cent, of 

 KNO3 but no glycerine. Black films made of this solution were 

 found to have a mean thickness of i r 7/1^, showing that the thick- 

 ness of the black is practically the same whether the solution 

 does or does not contain glycerine. As this result, however, 

 depends upon the validity of the assumption that the specific re- 

 sistance of a black film is the same as that of a large quantity of 

 the liquid, it was desirable if possible to measure the thickness 

 in question by a method free from the assumption involved in the 

 electrical method. p"or this purpose an optical method depend- 

 ing upon interference phenomena (Phil. Trans. 1883, p. 652) was 

 employed. Two glass tubes about 16 inches long and % inch in 

 diameter were placed horizontally side by side and were tra- 

 versed by two interfering beams of light, the interference bands 

 being produced by thick glass plates. The tubes were filled with 

 plane soap films, each tube containing from 40 to 60 films and 

 having its ends closed by pieces of plate-glass. After an hour or 



NO. 1252, VOL. 48] 



more, when the films had thinned sufficiently to appear black, the 

 position of the central interference band in the field was noted, 

 and its displacement when the films were broken, first in one tube 

 and then in the other, carefully measured. From these measure- 

 ments the average thickness of a black film could be easily de- 

 duced, the only assumption made being that the refractive index 

 of the liquid is unaltered by the tenuity of the film. The average 

 thickness of about 900 films was found to be iZ'ifjL/j.. This result 

 justified the assumption made in the electrical method with 

 regard to the constancy of the specific conductivity of the liquid. 



The results established before the recent work was begun 

 were therefore as follows : — (a) The thickness of a black soap 

 film formed of a solution containing one part of oleate of soda 

 dissolved in 40 of water with 3 per cent, of KNO3 added is about 

 twelve micromillimetres. {/>) It is practically the same when to 

 the soap solution is added two-thirds of its volume of glycerine. 

 (<r) From this it follows that the specific conductivity of such a 

 solution is the same whether the liquid be considered in large 

 quantity or in the form of a minutely thin film. (</) The thick- 

 ness of the black, though often varying from film to film, is 

 always the same in the same film — i.e., is independent of area 

 and age. With regard to these results it may be said at once 

 that they have all been repeatedly and completely confirmed by 

 subsequent investigation. 



We now come to the more recent work. Since in the earlier 

 experiments the solutions were always of the same strengi has re- 

 gards soap, and always contained not less than 3 per cent, of KNO, 

 it was important to determine whether the thickness of a blcck 

 film is or is not dependent upon the proportion of soap or salt 

 in the solution. The optical method was first employed. The 

 strength of the soap solution being kept constant, viz. 2 grammes 

 of hard soap to 100 cc. of water, the proportion of salt was 

 diminished from 3 per cent, to zero. Under these circumstance--, 

 the mean thickness of a black film was found to steadily in- 

 crease from i2^;utoabout 24/i/t. A similar large increase in the 

 thickness was found when the solution contained glycerine, or 

 was made of soft instead of hard soap. When no metallic salt 

 is present, and the strength of the soap solution varies, the 

 thickness of the black increases as the solution becomes more 

 dilute. Thus for a hard soap solution, when the per- 

 centage of soap was 3'3, the thickness was found to be 21 '6 i^/a 

 and rose to 29^3 fi/i as the percentage of soap diminished to 

 I '25. If, on the other hand, the solution contains as much as 

 3 per cent, of KNO3, variation in the^ proportion of soap has 

 liltle or no influence on the thickness of the black. This is 

 shown by the following table : — 



Hard Soap Solution, containing 3 per cent, of KNO3. 

 Percentage of soap in 1 ,, 



the solution I ^ 5 ... 2 o ... i bb 



Mean thickness of the 1 

 black in /t^u S 



2-S 

 '3i 



II-6 



.. 1. 14 

 .. 12 I 



The results above given have been deduced from the optical 

 method of measurement, and the question arises whether the 

 large increase in the thickness of black films formed from an 

 unsalted solution is real, or whether it is due to some incorrect 

 assumption. The only point where error is possible is in the 

 hypothesis that the refractive index is the same as that of the 

 liquid in mass. The thickness of a film varies inversely as ft - I 

 (n being the refractive index), and as the refractive index of the 

 soap solution is i'34, it would have to be reduced to I'lyin 

 order that the calculated thickness might be doubled. It 

 appears therefore a priori extremely improbable that the mere 

 addition of 3 per cent, of KNO3 should so completely change 

 the optical properties of the liquid that whereas if the salt be 

 added the refractive index is practically the same in the thin 

 films and in the liquid in mass, yet without the salt the refrac- 

 tive index should be as much as 13 per cent, less than that of the 

 liquid in mass. It may further be mentioned that Drude ( IVied. 

 Ann. xliii. p. 169, 1891), by an optical method quite difTerent 

 from that employed by us, has compared the refractive indices 

 of black and coloured films, of which the latter may unquestion- 

 ably be taken as nearly if not quite identical with that of the 

 liquid in mass, and has shown that they do not difi'er by more 

 than I part in 140. Such a variation would not affect the ap- 

 parent thickness of the films as measured by the optical method 

 by more than 3 per cent., whereas, as we have seen, the 

 presence or absence of the salt alters the apparent thickness by 

 ICO per cent. On the whole, then, the evidence is very strong 

 that the differences of thickness indicated by the optical method 



