58 PHYSICOCHEMICAL BASIS OP PHYSIOLOGICAL PROCESSES 



to those already described (page 4) can be employed. Most of the 

 recent work has been done either with collodion sacs, or with unglazed 

 clay cnps impregnated with some gel, such as silica or gelatin. When 

 such an osmometer, filled with some colloidal solution (like a solution of 

 pure albumin) and provided with a vertical glass tube, is placed in an 

 outer vessel containing water, the fluid will be seen to rise in the ver- 

 tical tube, the height to which it rises being proportional to the osmotic 

 pressure. 



But the observed pressure does not necessarily give us the osmotic 

 pressure of the pure colloid, for to this, even when highly purified, there 

 is almost certain to be attached a considerable amount of inorganic 

 salt, which may be responsible for the osmosis. It has indeed been 

 maintained by some observers that electrolytes form an integral part 

 of certain colloids, being bound to them perhaps by adsorption (see 

 page 65), and that they are essential to the maintenance of the colloidal 

 state. In any case, since electrolytes are always present, the osmotic 

 pressure of the pure colloid can be measured only when means are 

 taken to discount their influence. Several devices have been used, of 

 which the following may be mentioned: 



1. Addition to the fluid outside the osmometer of a percentage of 

 salt equal to that found by chemical analysis to be present in the col- 

 loid. (This method is untrustworthy.) 



2. The use of a limited quantity of fluid on the outside of the osmom- 

 eter so that equality of saline content soon becomes established, by 

 diffusion, in the fluids on the two sides of the membrane. 



3. The use of a membrane which is permeable to electrolytes but 

 not to colloids. 



Even when the greatest care is taken in its measurement, the osmotic 

 pressure of a given colloid has been found to vary considerably not 

 only according to the method used in its preparation, but also accord- 

 ing to the amount of mechanical agitation (shaking, stirring, etc.) to 

 which the colloid solution has been subjected. Regarding the influ- 

 ence of the method of preparation, it was found in one series of experi- 

 ments that albumin that had been repeatedly washed (but still con- 

 tained considerable ash) gave no osmotic pressure, whereas another 

 preparation that had been purified by crystallization twice (and con- 

 tained much less ash) had a pressure of 3.38 mm. Hg. According to 

 these results the ash content of the colloid is not fundamentally re- 

 sponsible for its osmotic pressure. As to the influence of mechanical 

 agitation, the osmotic pressure of a gelatin solution is increased by 

 shaking, while that of a solution of egg albumin is decreased. 



The property upon which the osmotic pressure depends is undoubtedly 



