Pharmacodynamics of Salts and Drugs 115 



these. Then it is seen that whether the sah dissolves the edestin 

 or converts it into a curdy mass depends mainly on the base. The 

 reason why so slight differences exist in the solvent power of the 

 cations, sodium, potassium, lithium, barium, calcium, and magne- 

 sium, is shown by an examination of their ionic potentials, which 

 are very low and probably about the same in each. The solvent 

 power of manganese and ferrous chlorides is low. 



Cu, Cd, Cr, Co, Fe'", Pb, Hg, Cu, Al, Zn chlorides and nitrates 

 all fail to dissolve. 



The solvent powers of the anions arrange themselves in descending 

 order as follows: CrO^, SO3, 8^03, I, Br, CI, SO4, which is almost 

 certainly the descending order of the ionic potential of these ions. 

 The fact that these bivalent ions of high ionic potential dissolve, in- 

 stead of precipitating, the colloid, and that the valence of the anion 

 is unimportant, is, in my opinion, good evidence that the edestin 

 in such solutions is electronegative and not electropositive; for, as 

 has been shown by Hardy and many others, the valence of the ion 

 of the same sign as the colloid is immaterial, but toward a colloid of 

 opposite sign it is very important. 



The peculiarity of the dissolving action of the heavy metal ace- 

 tates' also receives a possible explanation. In these solutions which 

 dissociate acetic acid the edestin becomes electropositive. Con- 

 sequently it is no longer precipitated by the positive ions, but receives 

 energy from them, and is rendered more positive, and hence more 

 soluble, than it was before. The cause of the failure of the chlorides 

 to dissolve the edestin in acid solution would be that given by 

 Osborne, that in such solutions, where there are many hydrogen 

 ions, it is changed into edestan as an insoluble product. 



I append here also a summary of the results of Pauli^ illustrating 

 the same facts, showing the parallelism between the solvent or pre- 

 cipitating power of the anion upon albumin and its potential energy 

 content. The parallelism is certainly unmistakable. 



Order of precipitation SCN>I>Br>NO.,>Cl>C,H,Oj 



Ionic potential 83( ?) — . 79- 1.27— 1.694 



g) The phenomena 0} stimulation of cells. — I found that the salts 

 ranged themselves very simply in the order of their encrg}' content, 



'Osborne and Harris, loc. cit., p. 165. 



' Pauli, Hojmeister's BeUrdge, 1905, 6, p. 249. 



