32 
IOWA ACADEMY OF SCIENCES. 
in which a salt is dissolved, evaporation may increase the 
concentration and thus increase the cohesion. Second, in 
the use of water or any pure liquid certain substances may 
be dissolved from the air, thus increasing cohesion. Either 
of these two explanations leads to the suggestion that co- 
hesion in a liquid is a function of the molecular weight of 
the dissolved substance, and that cohesion may be used to 
find the molecular weight of a dissolved substance in the 
same way as varying the freezing point, or the boiling 
point in the surface tension method. I have tested this 
in the case of sodium chloride dissolved in water, and find 
that with solutions containing a half gram molecule, a gram 
molecule, and a gram and a half molecule there is a con- 
stant ratio in the number of grams to separate the disk 
from the solution. Similar tests have been made with 
water solutions of urea and a constant ratio found. 
Tests are also being made with naphthalin and urea dis- 
solved in benzene. But these latter experiments could not 
be completed in time to embody them in this paper, except 
to show their indicated results. Neither could the con- 
stants for the different substances be computed like 
the constants used in the freezing point and the boiling 
point methods. Neither could these latter experiments be 
carried to a degree of completion to enable one to judge 
whether the method of cohesion will be a practical method 
for laboratory purposes. There will, perhaps, be two ob- 
jections to it. First, on account of a considerable quan- 
tity of the liquid solution being required in a test, the 
expense will be greater than in the freezing point and the 
boiling point methods. Second, evaporation will change 
the degree of concentration of the solution in the case of 
volatile liquids. The advantages of the method are the 
inexpensiveness of the apparatus and the ease of ma- 
nipulation. 
