METHODS OF COLLOID RESEARCH 107 



feature consists of a collodion membrane (see Fig. 22) protected by 

 a platinum wire netting. In the case of dextrins which lie on the 

 border line between colloids and crystalloids, W. BILTZ decreased the 

 permeability of the membrane by adding cupric ferrocyanid. He 

 filled the collodion sac with 1 per cent potassium ferrocyanid solution 

 and placed it in 1 per cent copper sulphate solution. After twenty- 

 four hours he washed the sac for twenty-four hours in running water. 

 The method of procedure recommended by FOUARD, impregnation 

 with tannin and gelatin and subsequent tanning with sublimate, has 

 not proven effective, according to BILTZ. Above the collodion mem- 

 brane is a glass cap with a vertical tube. The union of netting and 

 cap is at 6. The fluid is mixed by an electromagnetic stirrer c. The 

 electrodes d permit the measurement of the conductivity. The entire 

 instrument is placed in a thermostat. Readings of the rise in the 

 tube are made with a cathetometer. 



Osmotic Compensation Method 



This method determines whether crystalloids present in a colloidal 

 solution are free or in any way bound, e.g., adsorbed. L. MICHAELIS 

 and P. RoNA* 2 have developed this method in their attempt to solve 

 the question whether the grape sugar, always present in the blood 

 and which, strange to say, does not pass through the kidney, is free 

 or bound in any way. For this purpose we place the fluid to be in- 

 vestigated (in this instance blood) in a fish bladder and suspend it 

 in a glass cylinder containing an isotonic fluid (in this case water 

 with 0.95 per cent NaCl). To the surrounding fluid, in a series of 

 experiments, there is added varying quantities of the crystalloids in 

 question. In this instance sugar is added ; we shall continue to describe 

 the above experiment as an example. If more sugar has been added 

 than is present in the blood it will diffuse into the blood; the sugar 

 content of the surrounding fluid will decrease. If very little or no 

 sugar is added, the sugar will diffuse from the blood into the sur- 

 rounding fluid. This will occur whether the sugars in the blood are 

 free, osmotically active or even if a portion is adsorbed. In the 

 latter case the free sugar will at first diffuse away so that the balance 

 between the adsorbed and the free sugar is disturbed; previously 

 adsorbed sugar may become free and likewise diffuse away. The 

 sugar content of the outer fluid remains constant, only if it accu- 

 rately expresses the free sugar content of the blood. If the total 

 sugar has been determined previously, we may calculate what per- 

 centage is adsorbed or otherwise bound, and what proportion is 

 free or osmotically active. With this method, L. MICHAELIS and 



