Pharmacodynamics of Salts and Drugs 103 



That is, the logarithm 0} the ratio between equivalent precipitating con- 

 centrations of two salts, divided by the difference between the differ- 

 ences of potentials of the ions of the two salts ought to give a cofislant* 



Wc thus have for the first time a formula for application to pro- 

 toplasm which states clearly at the outset that the effect of any salt 

 solution on the protoplasm will depend upon what ions are already 

 in combination with the protoplasm. In other words, if we supplant 

 most of the ions in any cell by sodium, and then apply calcium chloride, 

 the effect will be different from that obtained if calcium chloride is 

 applied before the sodium chloride. Furthermore, the same salt will 

 act differently on different cells, if only those cells have different ions 

 in them. Both of these necessary conclusions of the theory have 

 been established by observation. To make this perfectly clear, the 

 difference in potential between the protoplasm and the salt solution 

 which we started to measure is the difference in potential between the 

 systems ionized colloid — ionized salt. 



However, this formula cannot be applied directly as it stands to 

 protoplasm as a whole, because it only applies to colloidal solutions 

 in which the colloids are all of one sign. In protoplasm, however, it 

 is certain that we have colloids of both signs and very probably 

 amphoter colloids; i.e., colloids which are both positive and nega- 

 tive at different parts of the molecule. t We probably have, in other 



* This formula may, I think, be substituted with advantage for that of the tension coefficient. It 

 is in reality the numerator of the tension coefficient. 



t.\ number of facts speak for the presence of such twin ions in colloidal albumin solutions and in 

 protoplasm. For example, if egg albumin is dialy7,ed nearly free from salts, and then coagulated so as 

 to form a weakly alkaline colloidal solution of albumin, and if this solution is then made acid, it is well 

 known that the albumin becomes predominantly electropositive. That is shown by the colloid migrating 

 slowly to the cathode in an electric field, and also by the combining power of the albumin, since it now 

 combines readily with picric and other acids to form albumin picrate, tannate, and so on. Nevertheless, 

 if the solution is not too acid, it will combine still with the metals in some measure. This is undoubtedly 

 due to the composition and character of the albumin. The alkali albumin first obtained by heating 

 is a salt. When the albumin has add added to it, there is formed, in the first instance, the free acid of 

 the albumin, which is not much dissociated. In addition, the acid is added to the amido-group, and in 

 an excess of acid hydrolytic decomposition being greatly reduced, the dissociation takes place so as to make 

 the albumin predominantly electropositive. However, the ionization of the acid is not entirely prevented, 

 although it is greatly reduced, so that in some places we must have some hydrogen ions being formed, 

 leaving the albumin electronegative at certain places. It is probably this small percentage of hydrogen 

 ions which can still be replaced by the metals. I found that, as a matter of fact, the heavy metals mer- 

 cury and copper, although they would not in themselves cause a precipitate if the solution were sufficiently 

 acid, yet they rendered the albumin far more easily precipitated than it was before. This is to be antici- 

 pated on our \-iew that the colloidal particles are in some places negative and in other places positive. Not 

 only does this appear to be the case for albumin, but in protoplasm there is also reason for believing that 

 both ions of the salt are actually bound by the protoplasm, and especially in Fundulus eggs. It will be re- 

 membered that Du Bois Raymond long ago assumed that such polarized particles might exist in proto- 

 plasm. 



