404 
CHEMISTRY: RICHARDS AND COOMBS 
THE DETERMINATION OF SURFACE-TENSION 
By T. W. Richards and L. B. Coombs 
WOLCOTT GIBBS MEMORIAL LABORATORY. HARVARD UNIVERSITY 
Received. May 20. 1915 
This investigation is part of a series of investigations having for their 
object the study of the fundamental properties of liquids. It is hoped 
that, when a number of these properties have been determined with 
great accuracy^ the essential relations between them may be discovered 
with greater certainty than is possible at present. 
Among the significant properties of Hqmds surface-tension stands out 
as one of the most interesting. This somewhat unfortunately named 
property affords us a valuable clue concerning the cohesive forces which 
bind the substance together and cause it to become a liquid. Therefore, 
its exact determination is a matter of far-reaching importance to any- 
one who seeks to understand the fundamental nature of the liquid state. 
A glance at the published data concerning surface-tension leads one 
to conclude that much remains to be done. For example, the values 
obtained by experienced men for water at 20° vary all the way from 70.6 
to 78, according to different methods. Even a single method (for ex- 
ample, that of the rise in capillary tubes) has yielded results in the last 
twenty-five years varying from 70.6 to 72.7, and no satisfactory^ evidence 
is forthcoming as to the reasons for the difference. 
It seemed therefore worth while not only to study the surface-tension 
of a variety of new Hquids, but also to discover the reason for the diver- 
gences between different methods, and to obtain results of absolute as 
well as of relative accuracy for liquids already studied. The present 
work, although only preHminary, seems to have been successful in lo- 
cating several of the heretofore not adequately heeded sources of error. 
The method chosen was the well known method of capillary rise in 
carefully measured tubes, because this method seems to be one of the 
most direct and least Hkely to lead to insoluble mathematical compli- 
cations. The method has been used by many experimenters in the past. ' ' 
Several features of the present work deserve emphasis. In the first 
place, the careful selection of the capiUary tubes and the caHbration by 
means of short columns of mercury received especial attention. Cor- 
rection was made for the meniscuses of these columns in determining 
the exact diameter of the tube. Again, great care was taken to de- 
termine exactly the position of the meniscuses, both of the larger and of 
the smaller surface, by means of a finely adjustable black screen behind 
the tubes to be measured. It was found that the exact position was 
