4 U. S. BUREAU or FISHERIES. 



membranes in plants and animals approach ideal semipermeability 

 while they are living. Ideal semipermeability with respect to par- 

 ticular dissolved substances has been achieved and is found in living 

 organisms. 



It is to be remembered that in case of semipermeable membranes 

 the solvent will flow from the less con/centrated to the more concen- 

 trated side of the membrane, so that if we wish to extract water w^e 

 need only to make the outside more concentrated than the inside. If 

 we wish to add water, we make the outside less concentrated than the 

 inside ; that is, we use pure water outside, as has sometimes been done 

 unfairly to swell oysters and make them appear " fat." 



It is also to be remembered that the degree of permeability of mem- 

 branes does not necessarily remain unalterable. The permeability 

 of the membrane can very readily be changed, as will be seen later. 

 There is reason for believing, for example, that the permeability of 

 fish to salt increases after death — for stale fish strike through more 

 quickly than fresh fish — and that permeability increases at tempera- 

 tures near the freezing point of water. 



The tissues of fish consist mostly of cells. Each cell is a bag of 

 semiliquid, like the white of egg. The surface of every cell either is 

 or acts like a semipermeable membrane. If we surround the cell with 

 water, the inside will be more concentrated than the outside and 

 water will go in. If we surround the cell with strong salt solution, 

 water will pass out to the salt. Some salt wall also pass into the cell, 

 which fact shows that the cell wall is not ideally semipermeable. 



But what of the protein within the cell ? Why does it not come out 

 while the salt is going in? In order to answer these questions it is 

 necessary to pass from a consideration of the nature of the mem- 

 brane in osmosis to a consideration of the nature of the dissolved 

 substance. 



By a great many experiments it has been found that some dissolved 

 substances never pass through membranes under any circumstances, 

 while others will pass through some membranes. It is found that 

 those which never pass through are also those which on drying out 

 do not crystallize but shrink to a tough mass. They are called col- 

 loids. Examples of them are glue, albumen, gelatin, and soap. The 

 smallest possible particle of these substances is comparatively large, 

 too large, we may imagine, to go through the texture of the mem- 

 brane. They are not only large of molecule but complex in structure. 

 The bulk of animal bodies consists of colloids called proteins, dis- 

 solved in water. The other class of substances, those that may pass 

 through membranes and which on drying out crystallize in regular 

 geometrical shapes, are the crystalloids. Examj^les of this class are 

 salt, sugar, and like substances. It is not to be supposed, however, 

 that all crystalloids will pass with equal facility through any given 

 membrane. Nearly all membranes are in some measure selective of 

 particular crystalloids. The ideal semipermeable membrane permits 

 none to pass, but as membranes degenerate from ideal semipermea- 

 bility to complete permeability they permit more and more of these 

 dissolved things to pass through. 



The phenomena of osmosis having been briefly reviewed, one may 

 readily perceive the imi)orl:ance of applying the principles to the 

 salting of fish. Salt is brought in contact with the exterior of the 



