52 CHEMISTRY: HARKINS AND EWING Proc. N. A. S. 
attraction of the carbon is extremely high, and while the difference be- 
tween the compression of ether and that of water does not vary in such a 
way as to make it possible to give an exact estimate, the pressure would 
probably need to be more than twenty thousand atmospheres higher than 
the normal internal pressure of the liquid. The film, according to Williams, 
has an average thickness of 4 X 10~^ cm., which is of the order of one or 
two molecules in thickness for molecules of the dimensions used in our 
work. The magnitude of this pressure, which is the amount that the 
adhesional pressure betwen the carbon and the liquid exceeds that in 
the liquid itself, is not at all surprising. The deductions from the com^ 
pressibility theory indicate, if the theory is on a sound basis, that the 
average pressure given in our experiments is of the order of twice to three 
times as great as the highest pressures obtained by Bridgman. 
This theory leads to low values for the density of active carbon, or a 
density of about 1.6. Some of the charcoals which we have studied do 
not absorb more ether than they do water, or this compressive effect 
seems wanting. These are just the charcoals which are extremely in- 
efficient in adsorbing gases, that is, cedar wood charcoal, with a density 
of 1.50, and beechwood charcoal with a density of 1.65, which values are of 
the same order of magnitude as that calculated for cocoanut shell charcoal, 
or as the density in water of the rather low-grade cocoanut shell charcoal 
used by Miss Homfray. A charcoal made from anthracite coal alone has a 
density of 1.876 in water, and only 1.906 in ether, while a sample of car- 
bonite (made from anthracite and pitch) of 19 minutes accelerated chlor- 
picrin test had a density of 1.934 in water and 2.075 in ether. 
The data of table 1 suggest still another theory, that the differences in 
the volumes of the liquids absorbed are due to their viscosity, the idea 
being that some of the pores are so fine that the liquids which are more 
viscous are less able to penetrate than those which are more fluid, at least 
in the time allowed for penetration, from 24 to 48 hours. The first re- 
sults which we obtained indicated that the quantity of liquid absorbed 
increases with the time, but this time effect disappeared when we adopted 
the extremely thorough method of outgassing previously described. We 
have heard that Professor Hulett of Princeton University has found such 
a time effect, but presumably on less highly outgassed charcoals. Mr. 
Monroe is now engaged in a study of this effect under the conditions used 
in our tests. The principal evidence against the viscosity theory to be 
found in the data of table 1, is that the viscosity value obtained for propyl 
alcohol is badly out of order, since propyl alcohol is twice as viscous as 
water, yet much more of the propyl alcohol penetrates the pores. This 
accords exactly with the compressibility theory. 
Experiments on the heat of adsorption of liquids gave some interesting 
results. A highly outgassed cocoanut shell charcoal gave a heat of ad- 
sorption equal to 10.5 calories per gram of charcoal when immersed in water 
