THICKNESS AND ELECTRICAL RESISTANCE OF THIN LIQUID FILMS. 517 
In fact the observations are perhaps open to criticism of a precisely opposite kind, 
viz., that in view of the irregularities in the value of /3 obtained by our ordinary 
method of measurement, it is strange that on this occasion we should have obtained 
the same result in the case of two consecutive films. Such coincidences were not 
uncommon ; but the objection is, perhaps, best answered indirectly by comparing the 
throws in the case of salted and unsalted solutions, and thus determining whether 
the ratio of the computed apparent thicknesses is about the same when transient and 
continuous currents are used. This experiment we have performed. With a solution 
containing 3 per cent, of KNO 3 the throw was 19 divisions of the scale. With an 
unsalted solution when the black had descended to the same level the throw was 
9'5 divisions. The specific resistances of the two liquids were 36 and 890 ohms per 
cubic centim. respectively. If rq and q are the resistance and calculated thickness of 
the unsalted film, and is the specific resistance of the liquid of which it is formed ; 
and if 7\, t. 2 , have similar meanings with respect to the salted film, we have the 
equation 
h _ P\ '\ 
4 Pa'^’i 
or, remembering that the throws are inversely proportional to the resistances, 
h 
h 
890 9-5 
-V — 19 
36 19 
A 
4 :. 
But 1 . 2 , the thickness of a salted film, is about 12 p-./r. Hence the calculated 
apparent thickness of the unsalted film, = 149 p.p. This number is intermediate 
between the largest and smallest of those obtained by the continuous current method 
and exhibited in Table VII, and does not differ very widely from the mean, which is 
about 160. 
On the whole then the evidence clearly points to the conclusion that the passage of 
the current does not produce the change of specific conductivity under discussion. 
(4.) Passing now to the question as to whether absorption of CO 3 from the air is 
the efficient cause, it must be remembered—and this remark applies also to the 
possibility of the results having been affected by evaporation or change of tempera¬ 
ture—that all our observations are based on a comparison between two solutions 
which differ only in the addition to one of them of not more than 3 per cent, of KNO3, 
or 1 per cent, of Na 2 S 04 . If, therefore, the change in specific conductivity were 
ascribed to any one of these causes, it is not only necessary to assume that the cause 
was competent to produce the change, but that its efficiency was very greatly modified 
by the addition of the salt. 
As it is difficult to argue from the results of experiments conducted on the liquid 
in mass to the behaviour of a film, and as it is probable that, if sufficient time were 
allowed, the amount of CO^ absorbed by a liquid containing a substance such as 
