3:<2 



NA TURE 



\Augtist 1 6, 1877 



to make the bubble as accurately cylindrical as pos- 

 sible. 



A current could be passed through the film by attaching 

 wires to two binding screws, one of which was affixed to 

 the brass tube, while the other, attached to the ebonite 

 cover, was in conducting communication with the crucible 

 by means of a platinum wire, which dipped into the 

 mercury with which it was surrounded. 



This piece of apparatus was included in the arm of a 

 Wheatstone's bridge, by means of which and a delicate 

 reflecting galvanometer the resistance of the film was 

 measured. As the latter continually altered and generally 

 slowly increased, the known resistance on the bridge was 

 made up to a certain amount, and the moment when the 

 resistance of the cylinder became equal to it noted. The 

 part of the total resistance thus measured, which was 

 due to the coloured portions of the films was calculated 

 on the assumption that Ohm's law held good, a process 

 which made it necessary to determine the shapes of the 

 films, which increased its thickness from above to below. 



This was done by measuring with a cathetometer the 

 breadths of the bands of colour displayed by the film when 

 viewed by light incident at 45°, and noting the time at 

 which each observation was taken. This operation was 

 performed at least twice, and thus the rate of motion of 

 the bands of colour was determined, from which data 

 their positions and the form of the film could be calculated 

 for the moments at which the electrical observations were 

 made. 



The result of these measurements was curious and will 

 perhaps be best understood by means of the figure. The 

 curves there given show the thicknesses of all parts of a 

 particular him at the hours named. The upper parts 

 represent the black portions of the films, the lower corre- 

 spond to the coloured portions, and the sudden change in 

 thickness between the two is well marked. The curves also 

 illustrate very well the phenomena generally obser\ed after 

 the black began to form, namely, that the portion of the film 

 in contact with it became thicker, though never thicker 

 than any of the lower parts of the film, which, on the other 

 hand, continued to become thinner, so that at last the 

 whole assumed one uniform tint, which changed but slowly, 

 and sometimes in such a manner as to indicate an increase 

 in thickness. This latter phenomenon is due to absorption 

 of moisture from the air, but it seems not unlikely that the 

 thickening observed at the lower edge of the black may 

 have been in part due to the fact that the sudden decrease 

 in thickness caused by the formation of the black portion 

 of the film must necessitate the rapid removal of a com- 

 paratively large quantity of liquid from its lower edge. 

 The curves given in the figure, together with all the others 



drawn from the data acquired, prove that the coloured 

 parts of the films observed did not increase regularly in 

 thickness from top to bottom, but that rapid changes 

 occurred somewhat similar to, though far less marked than. 



that observed at the lower edge of the black. Some films 

 thicker than tho«e from which the figure was drawn showed 

 two such shoulders as those curves exhibit below the 

 black, and it was remarked that a rapid change of slo^e 

 in the surface of the film generally began at points where 

 its thickness was such as to cause it to show the yellow of 

 the second order when seen through the telescope of tie 

 cathetometer. The thicknesses of the films are neces- 

 sarily represented on a much larger scale than their 

 lengths, as the latter would cover many yards of pap;r 

 if magnified to the same degree as the former. Hence 

 the figures give only a very exaggerated representa- 

 tion of the forms of the films. For instance, in the case 

 represented in the last figure, the difterence in thickness 

 between the black and the rest of the film, which was the 

 yellow of the second order, was certainly not greater than 

 320-millionths of a millimetre. If, now, this increase in 

 thickness took place in one-hundredth of a millimetre (a 

 distance which would be inappreciable on the scale of the 

 figures), the colours of the first order and those of tl'.e 

 second inside the yellow would probably not have been 

 visible if crowded together in so small a space, yet the 

 transition from the black to the yellow would, in such a 

 case, be made by an easy slope of 3 in 100, instead of by 

 the abrupt change represented. We are thus, in all 

 probability, dealing throughout with very small curvatures, 

 and it is only the extraordinary sensitiveness of the tests 

 of the thickness of a film at our command which enables 

 us to detect the slight changes of slope which occur. 



Having thus determined the shape of the coloured 

 portion of a film it is easy to calculate its resistance, and 

 by subtracting this calculated result from the observed 

 total resistance of the film to deduce that of the black 

 part alone, and thence to calculate the resistance of a ring 

 of the black of the diameter of the crucible, and i mm. 

 broad. If the number so obtained varies in d fferent 

 experiments, the inference is that either the thickaess of 

 the black portions of the films varies, or that the 

 assumption that Ohm's law was applicable to the rest of 

 the film was incorrect ; if it is constant we craw the 

 double conclusion that the thickness of a black film is 

 constant, and that the coloured films obey Ohm's law. 

 Such a method affords, however, only an unsatisfactory 

 test of the latter of these points, as the resistance of the 

 coloured portion of a film is in general so small a fraction 

 of that of the black part, that a considerable apparent 

 divergence from Ohm's law, and consequent error in its 

 calculated value, would not produce much error in the 

 resistance of the black film. 



With regard to the first point, however, viz., the 

 variations to which the thickness of a black film is 

 subject, the experiments furnish some more trustworthy 

 information. Five cylinders were observed on five 

 different days, and in all thirty-six determinations of the 

 resistances of black films were made. The breadth of the 

 black ring varied on different occasions between v^ m.m. 

 and irS; m.m., and the total resistances measured lay 

 between 37 and 22 megohms. The highest and lowest 

 values of the resistance of a ring of the black i mm. 

 broad obtained from individual experiments, differ 'rom 

 one another by about eleven per cent, of the mean value, 

 but the means of each day's results agree to within about 

 five per cent., and the means of four out of the five days' 

 results agree to v/ithin two-and-a-half percent. V/e thus 

 learn that the thickness of the black films formed in the 

 same manner and with the same apparatus, varied very 

 little on the different occasions on which they were 

 observed. Uy treating the experiments in th-s way, we 

 mingle together indiscriminately, all the olservations 

 obtained from any one cylinder, whatever the breadth of 

 the black film may have been. If, however, it diminished 

 in thickness from below to above, we should e.xpect that 

 obscivations taken when tlie black part 01' a film was 

 large and therefore, on the average, thinner, would, on the 



