442 SCIENCE PROGRESS 



who measured the height of the meniscus in capillary tubes and 

 deduced a finite contact-angle from the fact that the height 

 of the meniscus was always less than the radius of the tube, 

 or that of Sentis, who attempted to determine the contact-angle 

 from the observation of refraction phenomena at the line 

 separating solid from liquid, are not very satisfactory; the most 

 promising method of investigation is that which consists in 

 determining the surface-tension of the same sample of liquid 

 by two methods, one method being independent of, the other 

 dependent on a knowledge of the contact-angle. The contact- 

 angle is then assumed to be zero, and the surface-tension is 

 calculated in each case. If the two results agree within the 

 limits of experimental error, the assumption is justified. If the 

 results are not in agreement — and the criterion of this is that 

 the mean values shall differ by more than the sum of the prob- 

 able errors — the knowledge of the value of the surface-tension 

 given by the first method enables one, from the equation of 

 condition of the second method, to calculate a value for the angle 

 of contact. Experiments of this kind have been successfully 

 carried out by Magie, who compared the values of the surface- 

 tension given by measurements of the partial and total depths 

 of large bubbles of air blown in the liquid and imprisoned 

 beneath a concave lens ; the present writer is now engaged on 

 a comparison of the results given by the capillary-rise method 

 and by Jaeger's method. 



The importance of surface-tension determinations in exhibit- 

 ing the relation between capillary constants and chemical con- 

 stitution warrants a discussion of one or two points of interest 

 concerning, especially, the proper conditions under which results 

 should be compared. Broadly speaking, the results of past 

 researches have shown that surface-tension, considered from the 

 physico-chemical point of view, is almost entirely additive in 

 character. Constitutive influences are, however, to a minor 

 degree exhibited. For example, the surface-tensions of ortho-, 

 meta-, and para-xylene are given by Feustel as 3*21, 3 '08, and 

 3*03 respectively (measured in milligrams-weight per millimetre, 

 at i9°*2 C). But it is clear that, in discussing the relations 

 between surface-tension and chemical composition and constitu- 

 tion, especially with respect to the constitutive influence, the 

 conditions for comparison must be carefully chosen and specified. 

 The surface-tension of a liquid is affected by the amount of 



