Kohlenberg—Osmosis and Osmotic Pressure. 271 
brane formers” to the liquids on each side of the membrane 
to repair such leaks. It is obvious that in any osmotic ex¬ 
periment the composition of the septum is always in a state 
of change, though the extent of this may be slight in some 
cases. Here lies one of the chief difficulties of measuring 
osmotic pressures (which are equilibrium pressures) with 
soft diaphragms to which Eaoult alludes. 
It must be borne in mind that the application of the gas 
laws, either in simple or modified form, to dilute solutions is 
based upon the experiments which Pfeifer made with cop¬ 
per ferrocyanide membranes, and which Morse and Frazer 
have recently sought to verify. But these experiments have 
all been made without stirring and with but one membrane, 
and hence are not final. Furthermore, the osmotic pressures 
of sugar solutions in pyridine, using vulcanized caoutchouc 
as the semi-permeable membrane, show definitely that the gas 
laws do not obtain here at all. In the face of the experi¬ 
mental facts which we now have as showing the nature of 
the osmotic process and the magnitude of the osmotic pres¬ 
sures under different conditions, the general, indiscriminate 
application of the gas laws in their simple or somewhat mod¬ 
ified form to all dilute solutions, and even to some that are 
not dilute, as now in vogue, can not be too greatly deplored. 
To speak of the osmotic pressure of any isolated solution 
without specifying what membrane separates it from what 
other liquid is nonsense, 1 in the light of the facts here pre¬ 
sented. And further, to assume that solutes are polymerized 
or dissociated in dilute solutions because the osmotic pres¬ 
sures developed by the latter in given cases happen to devi¬ 
ate from values computed from the gas laws is evidently 
equally unjustifiable practice. 
SUMMARY. 
In this paper it has been shown that whether osmosis will 
take place in a given case or not depends upon the specific 
nature of the septum and the liquids that bathe it; and if 
1 Compare also views expressed by Van Laar, Cbemisch Weekblad, 2, 
1—16. (1905.) 
