Kohlenberg—Osmosis and Osmotic Pressure . 251 
nally the outer layer of muslin firmly holding the disc in the 
proper position. The perforations in the disc were about 1 mm. 
in diameter and the disc itself was about 0.5 mm. thick. By 
observing closely the precautions here laid down, cells were 
practically always obtained without fail which were perfectly 
tight, rigid and capable of withstanding pressures up to the 
point of bursting the glass bell of the thistle tube. Further¬ 
more, on taking such an apparatus apart even after experi¬ 
ments that had continued for weeks, the membrane was found 
to be intact and in perfect condition. Experiments demon¬ 
strating that such membranes do not leak have already been 
given above, ISTos. 1 and 2. We have here then a simple, direct 
and certain method of preparing so-called semipermeable mem¬ 
branes for osmotic pressure measurements, which are yet not 
precipitated membranes: and furthermore we know why the 
membranes are semipermeable for the solutions that come into 
question. 
After the membrane has been put into place as described the 
apparatus is charged with the required amount of mercury and 
then with the solution to be tested, the tube being drawn off at 
B before the small flame of a blowpipe so as to leave no air in 
the apparatus. The latter is then placed in a large vessel con¬ 
taining the pure solvent, and the rise of the liquid in the 
manometer tube is observed from time to time, a pair of sensi¬ 
tive thermometers graduated to 0.1 degree being placed in the 
solvent near the osmometer to indicate the temperature. The 
apparatus was set up in a basement room, whose temperature 
changed but slightly and only very gradually during the course 
of the experiments. It is highly essential that the temperature 
be kept as constant as possible during the experiments and that 
sudden fluctuations be avoided. From what has been said it is 
evident that the experiments were finally set up in much the 
same way in which Pfeifer performed his tests:, only in this 
case additional care was taken to cover up the dish so as to 
minimize evaporation and to protect the surface of the pyri¬ 
dine from the moisture of the air. All measurements were 
made with the open manometer and the height of the mercury 
column was estimated to within 1 mm. in each case. It was 
