356 HOW CROPS GROW. 
Water is imbibed by the membrane of bladder much 
more freely than alcohol; on the other hand, a film of 
collodion (nitro-cellulose left from the evaporation of its 
solution in ether,) is penetrated much more easily by aleo- 
hol than by water. If now these liquids be separated by 
bladder, the apparent flow will be towards the alcohol; 
but if a membrane of collodion divide them, the more 
rapid motion will be into the water. 
When a vigorous chemical action is exerted upon the 
-membrane by the liquid or the dissolved matters, osmose 
is greatly heightened. In experiments with a septum of 
porous earthenware (porcelain biscuit,) Graham found 
that in case of neutral organic bodies, as sugar and alco- 
hol, or neutral salts, like the alkali-chlorides and nitrates, 
very little osmose is exhibited, i. e., the diffusion is not 
perceptibly greater than it would be in absence of the 
porous diaphragm. 
The acids,—oxalic, nitric, and chlorhydric,—manifest a 
sensible but still moderate osmose. Sulphuric and phos- 
phoric acids, and salts having a decided alkaline or acid 
reaction, viz., acid oxalate of potash, phosphate of soda, 
and ourbonates of potash and soda, exhibit a still more 
vigorous osmose. For example, a solution of one part of 
carbonate of potash in 1,000 parts of water gains volume 
rapidly, and to one part of the salt that passes into the 
water 500 parts of water enter the solution. 
In all cases where diffusion is greatly modified by a 
membrane, the membrane itself is strongly attacked and 
altered, or dissolved, by the liquids. When animal mem- 
brane is used, it constautly undergoes decomposition and 
its osmotic action is exhaustible. In case earthenware is 
employed as a diaphragm, lime and alumina are always 
found in the solutions upon which it exerts osmose. 
Graham asserts that to induce osmose in bladder, the 
chemical action on the membrane must be different on the 
two sides, and apparently not in degree only, but also in 
