October 26, 1893] 



NATURE 



62:; 



are not merely apparent but real, and this point may now be 

 treated as established. 



We no\r pass on to consider the thicknesses of black films as 

 deduced by the electrical method. The method adopted was 

 in all essentials identical with that previously employed and de- 

 scribed (Phil. Trans. 1883, pt. ii. p. 645, Nature, 1883, 

 lac. cil.). 



The apparent thickness of a black soap film as measured by 

 the electrical method increases as the percentage of added salt 

 diminishes, but in a far larger ratio than would be inferred 

 from the optical method. If the proportion of salt be 

 diminished to zero the thicknesses thus calculated are greater 

 than the greatest thickness at which a film can appear black. 

 Thus with a hard soap solution the apparent thickness rose from 

 lO'6 .a;a to 26*5 M."i as the percentage of KNO3 added was 

 diminished from 3 to 0'5, and became 14S /i/i when the solu- 

 tion contained no salt, this number being the mean value 

 derived from fourteen films, the individual thicknesses of which 

 ranged from 79 to 240 ;ii//. In another set of experiments made 

 with a rather stronger soap solution, the apparent mean thick- 

 ness of the black was 184 ^;t, the extreme values for six films 

 being 84 and 250 /^/i. Similar results were derived from a soft 

 soap solution, the mean apparent thickness obtained from the 

 examination of twenty-three black films being i62ju/x, and the 

 extremes about the same as before, viz. 80 and 252. 



Now a film 148 /n/i thick (to take the smallest of the mean 

 thicknesses given above) could not possibly appear black. 

 According to Newton the beginning of the black occurs when 

 the thickness is 36 iiji, which ij about one-fourth of the 

 smallest mean value obtained from unsalted solutions. We are 

 therefore driven to the conclusion that the close agreement 

 between the results of the optic.il and electrical methods, which 

 has again and again been proved when the solution contains 3 

 percent, of KNOjdoesnot hold in the c.ise of unsalted solutions. 

 The measured thicknesses cmnotbe true thicknesses, and there- 

 fore there must be a difference between the specific conductivity 

 of a film, and that of the liquid from which it is formed. 



Apart from this, however, is the fact that the apparent thick- 

 ness varies considerably from film to film, although all the 

 conditions are maintained as far as possible constant. This is 

 certainly due in some cases to a real variation in thickness. We 

 have frequently seen in the same film two different shades of 

 black separated from each other by a definite sharp line, which 

 is generally very irregu'ar in form. The line which separates 

 the black from the coloured part of a cylindrical film thinning 

 in the normal way is always a horizontal circle. This is rarely 

 the case with the boundary between the two black tint!'. Some- 

 times a patch of the darker black is completely surrounded by 

 the other, sometimes the line of separation is sinuous, or stands 

 higher at one point than at another. It is thus difficult to 

 obtain comparative measures of the thicknesses of the two lints, 

 as the method of experiment employed assumes the thickness of 

 a cylindrical film to be the same at all points on the same 

 horizontal circle. Such measures, however, as have been 

 made indicate that the thickness of the thicker black is about 

 twice as great as that of the thinner. 



The two black tints are not always easy to detect or to dis- 

 tinguish from each other. If only one occurs it is almost im- 

 possible to say whether it is the thinner or thicker variety. 

 Frequently the passage of an electrical current through a film, 

 the black portion of which appears to be homogeneous, dis- 

 closes the existence of the two ditTerent tints by producing 

 or intensifying little white flecks which lie along the 

 boundary between the two. On the suppression of the 

 current the flecks become smaller or disappear, but the atten- 

 tion of the observer having been called to the boundary line, 

 there is no difficulty in distinguishing between the regions on 

 the two sides of it, the thinner appearrng more intensely black 

 than the other. We have never, when experimenting with 

 solutions containing 3 per cent, of KNO3, seen any indication of 

 the two shades of black. If the added salt is reduced to o'5 

 per cent., the phenomenon is seen occasionally ; but with un- 

 salted solutions it is of frequent occurrence. The two varieties 

 of black in a soap film were noticed by .Sir Isaac Newton, who 

 remarks that sunlight is reflected from even the darker spots. 



But to return to the question of the mean apparent thickness 

 of a black film. As has been stated, the optically measured 

 thickness differs little if at all from the true thickness. Ifthe 

 electrical thickness is approximately equal to the optical thick- 

 ness, we may assume that the specific conductivity of the liquid 



is unaltered by the tenuity of the film. If they differ consider- 

 ably the inference is that the specific conductivity has changed. 

 Now in the case of an unsalted solution containing one part of 

 soap dissolved in sixty of water the optical thickness is 277 

 ju/i, while the mean apparent electrical resistance is 160 ,uju. 

 The specific conductivity is therefore greater in the film than in 

 the liquid in mass in the ratio of 5 '8 to I. 



A number of experiments have been carried out for the pur- 

 pose of determining whether the change in the specific con- 

 ductivity is a function of the thickness ol the film, or is peculiar 

 to black films. The result is to show that with an unsalted 

 solution of hard soap the change begins when the film is com- 

 paratively thick. Thus, the ratio of the electrical to the optical 

 thickness (which measures the proportional increase of con- 

 ductivity) is I '66 when the film exhibits the green of the second 

 order (thickness = 641 /i/t) ; it is I '98 at a thickness of 296 

 ii.li., 4'47 at 97 /iM (white of first order) and becomes 58 when 

 the film is black. 



When the solution contains 3 per cent, of KNO3 we know 

 that for the black films the conductivity is the same as for the 

 liquid in bulk. That it remains constant under all circumstanct s 

 is highly probable, though not absolutely certain. 



We have now to inquire into the possible causes of the fact 

 that a black film made of an unsalted soap solution appears to 

 be about six times as great as it really is, or, in other words, 

 that the specific conductivity of the film is six times as great as 

 that of the liquid in mass. This increase might possibly be due 

 to (l) evaporation or absorption of water by the film as it thins, 

 (2) changes of temperature, (3) changes in the chemical consti- 

 tution of the film by the electrolytic action of the current em- 

 ployed, (4) absorption of carbonic acid or of oxygen from the 

 air. In considering these it must be borne in mind that our 

 observations are based on a comparison between two solutions 

 which differ from each other only by the addition to one of 

 them of 3 per cent, of KNO3. If, therefore, the change in con- 

 ductivity were ascribed to any one of these causes it would he 

 necessary to assume not only that the cause was competent to 

 produce the change, but that its efficiency was very greatly 

 modified by the addition of the salt. It is extremely improbable 

 that evaporation or absorption of water, changes of temperature, 

 or absorption of carbonic acid (if occurring in the one liquid), 

 would produce the enormous observed change in the conduc- 

 tivity, while they were inoperative in the case of the other. We 

 have not, however, been satisfied with a priori considerations, 

 but have experimentally examined each of these possible causes. 



With regard to the first, it is sufficient to say that all the pre- 

 cautions which experience has shown to be efficient in securing 

 constancy of composition in the case of liquide glyci:riqtic — a 

 liquid much more susceptible to changes of composition, due to 

 variations of hygrometric state, than plain soap solutions — have 

 been taken. We may be perfectly sure that the change in con- 

 ductivity is not due to the loss or gain of water by the film when 

 thinning. 



Experiments have been made at various temperatures between 

 17" and 27^ C, but there is nothing in the results obtained to 

 indicate that the apparent thickness of the black either increases 

 or diminishes as the temperature changes. Thus, to take four 

 films out of many that might be selected, we have the following 

 results : — 



Temperature 187 ... 21 ... 2II ... 26-3 



Apparent thickness of 



black film in pipe 171 ... 237 ... 201 ... 135 



There is no doubt that the relatively small changes of tempera- 

 ture which occurred in our experiments are not the cause of 

 the large increase in the apparent thickness of a black film. 



But the observed result might be due to change in the com- 

 position of the liquid caused by the passage of an electric current 

 through the film. The current employed to measure the re- 

 sistance of the film is always a feeble one ; but in ordtr to 

 produce a rapid thinning, we have frequently passed a current 

 from a battery of 28 Leclanche cells down the film from the 

 moment of its formation. Such a current, though probably 

 never exceeding 100 microamperes, is passed for a long lime, 

 and might conceivably affect the specific conductivity of the 

 liquid. As a specimen of the kind of results obtained, the 

 following, derived from a soft soap solution, may be given. 

 Each of the values of the thickness was obtained from a dif- 

 ferent film, and the number of cells indicated is that employed 



NO. 1252, VOL. 48] 



