264 THE HUMAN BOD Y. 



portion of such substances which gets through; and if 

 the pressure is slight the water or other solvent may alone 

 pass, leaving all the rest behind on the filter. Under 

 moderate pressure the blood may thus lose by filtration 

 only such bodies as water and salines; while an increase of 

 arterial pressure may lead to the passage of albumen and 

 fibriiiogen. Under healthy conditions, for example, the urine 

 contains no albumen, but anything increasing the capillary 

 pressure in the kidneys will cause it to appear. Dialysis 

 or osmosis has already been considered (p. 42); by it sub- 

 stances pass through the intermolecular pores of a mem- 

 brane independently of the pressure on either side, and for 

 its occurrence two liquids of different chemical constitution 

 are required, one on each side of the membrane. At least 

 if diffusion takes place, as is probable, between two exactly 

 similar solutions, the amount and character of the sub- 

 stances passing opposite ways in a given time are exactly 

 equal, so that no change is produced by the dialysis; which 

 practically amounts to the same thing as if none occurred. 

 When a solution is placed on one side of a membrane allow- 

 ing of dialysis, and pure water on the other, it is found that 

 for every molecule of the dissolved body that passes one 

 way a definite amount of water, called the endosmotic 

 equivalent of that body, passes in the opposite direction. 

 Crystalline bodies as a rule (haemoglobin is an exception) 

 have a low endosmotic equivalent or are readily dialyzable; 

 while colloids such as gum and proteids, have a very high 

 one, so that to get, by dialysis, a small amount of albumen 

 through a membrane, a practically infinite amount of water 

 must pass the other way. Accordingly, if we find such 

 bodies in a secretion we cannot suppose that they have been 

 derived from the blood by osmosis. 



The Chemical Processes of Secretion. As above point- 

 ed out certain secretions, called transudata, seem to be pro- 

 ducts of filtration and dialysis alone, containing only such 

 substances as those which are found in the blood plasma, 

 more or less altered in relative quantity by the ease or diffi- 

 culty with which they severally passed through the layers 

 met with on their way to the surface. But in many cases 



