486 PROCEEDINGS OF THE AMERICAN ACADEMY. 



there would be required a correction (as explained in the section on Per- 

 meability) for change of volume of the Stentors used, which was, from 

 the nature of the case, not required in the experiment with distilled 

 water. 



That the destructive action of the excess of a single salt was not 

 osmotic in its nature, is evident from the absence of injury under the in- 

 fluence of milk sugar in a concentration exerting equal or greater osmotic 

 pressure. Mechanical pressure and movement of water, characteristic 

 osmotic processes, resulted in no harm to the cells. Simple osmotic 

 action, that is redistribution of water and salts, as it would occur in a 

 physical experiment, is not sufficient to account for the resiilt, even 

 though the curves may be subdivided into osmotic regions. In fact, 

 the experiments upon Stentor, taken as a whole, permit the conclusion 

 that osmosis, though a factor necessarily present, is secondary in im- 

 portance to the specific activity of physiological salts. The modification 

 of membranes included in that class of phenomena denominated by 

 Ostwald ''mechanical affinity" (Scliafer, '98, p. 275), and the distinc- 

 tively chemical processes included in metabolism, exert a controlling in- 

 fluence superior to that of simple osmotic distribution of salts "and water. 

 An important problem suggests itself here, namely, the determination 

 of the specific activities of the salts with I'eference both to permeability 

 and to their individual metabolic functions. 



It might be supposed that the physiological salts, not being active sub- 

 stances chemically, owe most of their physiological effects to their osmotic 

 action. It is highly probable that the maintenance of a certain mechan- 

 ical pressure in the medium surrounding a cell is at least one of the im- 

 portant functions of physiological salts. But the magnitude of this 

 factor as compared to the chemical, metabolic activity of these salts is 

 strikingly shown by the secondary curves on page 473. As explained on 

 page 466, these are to be so drawn as to connect those concentrations of 

 the different salts which represent approximately equal osmotic pressures. 

 If the salts had produced mean results in proportion to their osmotic 

 pressures these secondary curves would have been parallel to the axis of 

 abscissas. But at equal pressures the mean results have such values as 

 to produce curves, making nearly the largest angle possible with the axis 

 of abscissas. These facts, as represented by the curves, justify the con- 

 clusion that the physiological salts have a distinct chemical, that is, specific 

 activity in addition to their osmotic function. 



We have now attempted a separate estimation of the relative shares, 

 for Stentor, of the osmotic factor and of the qualitative chemical factors, 



