Manchester Memoirs, Vol. Ixv. (1921), No. $ 5 



Generally speaking, one can say that the methods classified under the 

 grouping (B) should in no case be employed unless one has sound in- 

 dependent evidence that the contact angle is either zero, or so small that 

 its neglect introduces no appreciable error into the observed value of the 

 surface tension. 1 



It is to the methods classified under the heading (A) that we must 

 turn to find our ideal method, and of these (7), (4) and (10) promise very 

 well. From my own experience, I incline to say that any of these three 

 methods will give good results, and Jaeger's method is particularly con- 

 venient in practice. The maximum pressure-excess can be very exactly 

 measured, and temperature control is relatively easy. It is not sufficiently 

 clearly realised that the method yields absolute and not comparative values, 

 and a certain suspicion appears to exist that the experimental conditions 

 do not approximate closely enough to a statical state. These doubts I 

 hope to settle by the experiments previously mentioned. 



Method (4) too, is very convenient and exact if the conditions of the 

 theory are properly observed. By bending a thin rod into a circle of large 

 radius an anchor ring is formed, and suspended by three threads from a 

 balance pan so that its plane is horizontal. If the ring be allowed to touch 

 the surface of a vessel containing liquid and be then withdrawn, a distinct 

 maximum pull is observed, and the surface tension of the liquid can be 

 accurately calculated in terms of the maximum pull and the dimensions 

 of the ring. The method might with advantage be much more widely 

 used, as the weighings can be made accurately, and temperature control 

 is not too difficult. 



Of the methods included under the general heading B, it is worth 

 noting that (12^) has recently been recommended as a working practical 

 method. 2 Two vertical plates inclined at a small angle are dipped into the 

 liquid under examination, and the angle between the plates is varied until 

 the surface of the liquid between them coincides with one of a series of 

 confocal hyperbolas drawn on one of the plates. 



A simple and rapid way of roughly measuring the surface tension of 

 very small quantities of liquids is indicated by Kiplinger. 3 A short 

 column of the liquid is introduced into a capillary tube of known bore 

 which is then tilted until the meniscus at the lower end becomes plane. 

 The surface tension readily follows from a knowledge of the angle of tilt. 

 The results are uniformly 3 or 4 per cent, too low. 



Of the various balance methods classified under (B), I have found the 

 measurement of the surface tension pull on a sphere suspended beneath 

 a balance pan very convenient and sensitive. 



If the surface tension pull on such a sphere of radius R is equivalent 

 to a force of mg dynes, then it can be shown that 4 



m = W |R - *- _ -J, 



1 The neglect of a contact angle of 8° introduces a 1 per cent, error into the value 

 ofT. 



2 Grummach and Bein, Zeit. Instrutnentenk, June, 1919, p. 195; Set. Abs. A., 22, 

 500, 1919. 



3 your. Amer. Chem. Soc, 42, 472, ig2o. 



4 Ferguson, Phil. Mag., Nov. 1913, p. 925. 



