PASSAGE OF LIQUIDS THROUGH MEMBRANES. 477 



verted into the albumen of the blood. Mialhe, in com- 

 paring the endosrnotic properties of albumen and albumi- 

 nose, has given an explanation of these remarkable phe- 

 nomena. 



Recognizing the fact, which was, indeed, pointed out 

 clearly by Dutrochet, that albumen is capable of inducing a 

 more powerful endosmotic current than almost any other 

 liquid, he has shown that it never itself passes through mem- 

 branes in the exosrnotic current ; but that albuminoids, after 

 transformation by digestion into albuminose, or albumen 

 mixed with gastric juice, pass through animal membranes 

 with great facility. The experiments by which these facts 

 are demonstrated are very conclusive, and are of the highest 

 physiological importance. On removing part of the shell of 

 an egg, so as to expose its membranes, and immersing it in 

 pure water, the passage of water into the egg was rendered 

 evident by the projection of the distended membranes ; but 

 although the surrounding liquid had become alkaline, and 

 the appropriate tests revealed the presence of some of the in- 

 organic constituents of the egg, the presence of albumen 

 could never be detected. 1 When the contents of the egg 

 were replaced by the serum of the blood, the same result fol- 

 lowed. " After six or eight hours of immersion, the serum 

 had yielded to the water in the vessel all its saline elements, 

 chlorides, sulphates, phosphates, which were easily recog- 

 nized by their peculiar reactions, but not an atom of al- 

 bumen." 2 



1 MIALHE, Chimie appliquee d la Physiologic et d la Therapcutique, Paris, 

 1856, p. 141. 



2 Loc. tit., p. 142. 



The above experiment possesses additional interest, as it illustrates a new 

 principle recently announced by Graham, called dialysis. This is dependent simply 

 upon the fact that various membranes and some porous septa, when used in an 

 apparatus like an endosmometer, will allow only " crystalloid " bodies to pass 

 through ; while the organic substances, called by Graham " colloids," are retained. 

 For example, if a fluid containing an inorganic substance, such as a salt of arsenic, 

 mixed with organic matter, be placed in the interior of such an apparatus, and 



