STUDIES ON EXOSMOSIS 
that the aggregate amount of material should be the same in the 
different solutions. The strips were protected from evaporation until 
all were cut; each lot was then placed in a test-tube, rinsed with 
distilled water which was allowed to drain off for thirty seconds, and 
the solutions were then poured in. 
In these experiments three solutions were used: sodium chloride 
0.22 M, calcium chloride 0.16 or 0.17 M, cerium chloride 0.050 M. 
In the control experiment distilled water took the place of a salt 
solution. After a period of from fifteen to twenty-five minutes these 
solutions were poured off and the material was rinsed three times with 
distilled water, the second change remaining in contact with the tissue 
two minutes. The last rinsing was allowed to drain off for thirty 
seconds, and then 13.0 cc. of distilled water placed in each test-tube. 
This amount was just sufficient to cover the strips of tissue, which 
were packed loosely in the bottom of the test tube. At the end of 
fifteen minutes the distilled water was poured off into a specially con- 
structed U-tube designed to contain 13 cc. of solution, and its con- 
ductance determined. The solution was then returned to the material, 
and its conductance determined in a similar manner at suitable in- 
tervals. 
Preliminary experiments indicated the possibility that some sub- 
stance gathered on the electrodes, forming there a highly resistant 
layer. It was therefore decided to interpose parchment thimbles 
between the electrodes and the solution. These thimbles were kept 
in distilled water, and were placed (filled with distilled water) in the 
expanded ends of the U-tube just prior to each measurement. Figure 
I will show the arrangement of the U-tube, the bright platinum 
electrodes, and the solutions. It will be seen that the current traversed 
always the same amount of distilled water and the same length of 
column of solution. The distilled water in the parchment thimbles was 
not changed during a single set of three or four readings ; it was possible 
at the end of a set to duplicate so closely the first reading taken, that 
it was evident that the diffusion of electrolytes into the distilled 
water in the thimbles introduced no appreciable error. 
The largest source of error was that introduced by a certain amount 
of distilled water which it was impracticable to remove from the 
outside of the parchment thimbles before introducing them into the 
U-tube. The error thus introduced was not, however, sufficient to 
cause any significant variation in the results of the experiment. The 
