588 
CHEMISTRY: HARKINS AND HUMPHERV 
i{r/a). Of these only the results obtained on benzene have been 
plotted in figure 1. This curve is higher than that obtained by us, 
which means that their values for the surface-tension are less than ours. 
This, however, does not mean that the drop-weight results given by Mor- 
gan and McCann are incorrect, but that the capillary-constant for 
benzene which they use, and which they do not themselves determine, 
is lower than that found by us. They use an average constant taken 
from the results of Ramsay and Shields and other workers. Since 
the reading of this paper the supposition that these workers have in 
general obtained too low a capillary-constant has been confirmed by 
Richards and Coombs,^ who considers that the results of Ramsay and 
.55 
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0,8 0.9 1.0 I.I 1.2 1.3 1.4 1.5 1.6 
Values of 
FIG. 1.— (Note) THE CURVE HAS BEEN DETERMINED ACCURATELY FOR VALUES OF - UP 
TO 0.7. FOR VALUES GREATER THAN THIS ONLY THE GENERAL FORM OF THE CURVE HAS 
BEEN DETERMINED. 
Shields, Renard and Guye, and Walden and Swinne, are too low, and 
that the discrepancy is of the order of 3%. 
For values of r/a between 0.9 and 1.2 Morgan and McCann get a 
straight line for the values of the function, which would mean that 
Tate's law holds. Our determinations in this region have not yet been 
made with any great accuracy, since most of the work up to the present 
time has been on values of r/a between 0.3 and 0.7; but at least it can 
be easily seen that the general form of their curve is very different from 
ours. 
