DIALYSIS. 611 



includes the direction of the current, is due to imbibition, dependent 

 upon capillarity, or capillary attraction. It is plain, however, that 

 to cause the fluid to pass beyond the membrane, some other force must 

 be called into play ; and this involves the second essential condition 

 of osmosis, to wit, diffusion. For "liquids flow out of capillary 

 tubes which are filled with them, only when some other force or cause 

 acts, because capillary attraction cannot produce motion beyond the 

 limits of the solid body which determines the capillary action."* 



The property of diffusion, common to liquids and gases, has been 

 carefully studied by Prof. Graham, and is well illustrated by what he 

 terms "jar or phial diffusion," detailed in the text. 



While it is true that colloids are of low diflfusibility in their natural 

 condition, they may, however, be caused to undergo certain chemical 

 or physical changes, in consequence of which they become capable of 

 osmosis, and thus become practically crystalloids. Thus albumen, 

 which does not ordinarily osmose, when converted into albuminose by 

 the gastric or intestinal juices, becomes capable of absorption and is 

 taken into the bloodvessels. So oil, when converted into an emulsion 

 by the pancreatic fluid, becomes also capable of absorption arid enters 

 the lacteals. 



Crystalloids not only diffuse among themselves, but they will gradu- 

 ally diffuse themselves through colloids. A simple experiment showing 

 this fact may be thus performed : In the centre of a mass of any colloid 

 substance, as boiled starch, let there be placed any highly colored 

 crystalloid, as bichromate of potash ; in a short time, it will diffuse itself 

 throughout the entire mass, giving it a bright-yellow color. If, on 

 the other hand, a colloid, as burnt sugar, or caramel, be placed in 

 the centre of a similar mass, at the end of many days it will be found 

 to have diffused itself scarcely at all. 



In consequence of this property, when crystalloids and colloids are 

 mingled, as in the egg, the former may be caused to diffuse themselves 

 away from the latter. 



This is beautifully shown by a simple arrangement, known as the 

 egg-endosmometer, prepared by removing the shell from one end of an 

 egg in such manner as to leave the membrane intact. In the opposite 

 end, a small opening is made, around which is cemented a glass tube. 

 The egg is then placed in a wine-glass of water, with the end upon 

 which the membrane is intact downward. In a short time, the con- 

 tents of the egg will begin to appear in the tube, showing that water 

 has passed through the animal membrane and caused the displacement 

 of the egg's contents. It will be found, also, that the salts or crystal- 

 loids of the egg have osmosed through the membrane, at the lower end, 

 by adding a solution of nitrate of silver to the water in the wine-glass, 

 when an abundant precipitate of the chlorides will take place. All the 

 tests for albumen may. however, be applied to this water without any 

 response, none of the colloids, or albumen, have passed through the 



* Liebig, Researches on the Chemistry of Food and the Motion of the Juices in 

 the Animal Body. 



