Kohlenberg—Osmosis and Osmotic Pressure. 213 
change of solute. The effect of temperature was considered. 
The degree of permeability of the membranes for various sub¬ 
stances was noted, with the result that membranes as semiper- 
meable as any known hitherto were found, which are yet not 
precipitated membranes. Furthermore it was ascertained why 
the membranes were semipermeable in some cases and not in 
others. It was consequently possible to foretell for which sub¬ 
stances the membranes were permeable and for which sub¬ 
stances not permeable. Further, this work has cast light on the 
process of dialysis. It has been possible to separate different 
dissolved crystalline substances from each other by dialysis, and 
also to separate dissolved crystalline from non-crystalline 
bodies by having the non-crystalline substances pass through 
the septum and the crystalline ones remain behind in solution 
in the dialyzer. The direct measurements made with the semi¬ 
permeable membranes employed, moreover, showed that the gas 
laws do not hold at all in these cases. It was also found that in 
making direct measurements of osmotic pressures it is necessary 
to stir the liquids separated by the membrane, a very important 
fact which has been entirely overlooked in all osmotic pressure 
determinations hitherto made. The necessity of such stirring 
was really accidentally discovered, as will appear from details 
given below, after having worked in vain for over a year try¬ 
ing to get reliable, concordant results without stirring. The 
discovery that agitation of the liquids is essential in osmotic 
pressure measurements is of paramount importance in deciding 
as to the nature of the osmotic process. 
In the presentation of the experimental work which now fol¬ 
lows, the experiments will not be described in the order in 
which they were actually performed for the reason that greater 
clearness will be obtained by detailing the results in the light 
of the theory which was gradually evolved in the course of the 
investigation. 
A TYPICAL CASE OP OSMOSIS. 
If chloroform A, Fig. 1, be placed in a glass tube and a layer 
of water B be poured upon it, and again a layer of ether C be 
carefully poured upon the water, and the whole be allowed to 
