io6 



A. P. Mathews 



that the original ions in combination with the colloids are K 

 and CI.* 



TABLE 5. 



Salts Compared 



CuCU -MnCl, 



HgCl, -MnCU 



NiCh -MnCh 

 AgNOj-HCl.. 



CuCU -CdCl,. 



NiCl, -CdCl, 



CdCU -MgCl, 



HgCh -MgCh 



CuCl, -MgCl, 



ZnCl, -MnCl, 



ZnCl, -NiCU. 



HgCl, -CoCh 



HgCU -CuCh 



CoCh -MnCU 



e' + e' -e!'-e" 



(1.403) 



(1.816) - 



(0.848) - 



(0.8859)' 



(0.7567)" 



(0.201) 



(1-073) 



(2.241) 



(1.828) ■ 



(0.302) 



(0.546) 



(0.972) 



(0.4121)" 



(0.843 )■ 



log 





3 1249 



3.6812 



2.163 



2 . 0044 



1.765s 



0.8035 



2.1SS3 



4.477 



3 ■ 9208 



1.0430 



1. 1 201 



2.4771 



0.5563 



I .2041 



K 



2.23 

 2.03 



2.5s 

 2. 26 



2 33 

 3-997 

 2.010 

 1.998 

 2.I4S 



3 454 

 2.051 

 2.548 

 I 35 

 1.428 



Mean value of .K 2 . 23 



The values of K (Table 5) are on the whole fairly constant for the 

 great majority of the salts compared. The variations from the mean 

 of 2 . 2 are due almost entirely to the fact that mercury is not so poison- 

 ous as it should be, and that cobalt is a good deal less toxic than the 

 theory demands, while nickel is a little more toxic. The explanation 

 of these variations is no doubt to be found in part in the dissociation. 

 I have assumed throughout that the dissociation is complete. This 

 has been done for the sake of simpHcity. It is, however, certain that 

 the dissociation of mercury chloride even in these dilutions is far from 



* Some modification or explanation is necessary of the conclusion of a former paper that oppositely 

 charged ions must have of necessity an opposite action. This is in one sense true. That is, the {wsitive 

 ion in combination with the proteid, if the latter is electronegative, must constantly be neutralizing the 

 negative charge and producing undissociated albumin. It may be stated in this sense that the positive 

 ion always tends to precipitate an electronegative albumin. It happens, however, that the power of 

 neutralizing the charges of the colloid — that is, of reducing ionization — varies greatly in different cations, 

 being greatest in those of high ionic potential, and least in those of low. If, therefore, we have, as we 

 do have in a protoplasmic system, colloids in a state of equilibrium with ions already present, the particular 

 direction of the change in state of that equilibrium produced by the substitution of new positive or neg- 

 ative ions for those already present will depend on the relative potentials of the ions present and those 

 introduced in their places. The actual effect observed, therefore, of replacing an ion of high potential 

 with that of a low, will be the direct opposite of that produced by replacing the ion with an ion of still higher 

 potential, and in this case there will appear to be an antitoxic or antagonisUc action between two ions 

 of the same character of charge. In an earUer discussion of this matter I neglected to take into account 

 the great importance of the ions present in protoplasm. For example, suppose the ions in the protoplasmic 

 system to be mainly sodium; and let us suppose that potassium has a lower potential than sodium, while 

 calcium has a higher potential. If one substitute calcium for the sodium, the result will be to precip- 

 itate in part the electronegative colloids in the protoplasm. If, however, pwtassium be substituted for 

 the sodium, the result will be to dissolve still further the colloids. In this case potassium and calcium 

 will appear to exert an antagonistic action toward each other. If, however, the ions already in the proto- 

 plasm are of higher potential than calcium, then both potassium and calcium will produce the same kind 

 of an action on the protoplasmic colloids. The results obtained by Loeb, Loeb and Giess, Miss Moore 

 and myself on toxic and antitoxic action of salts thus have a very simple explanation. 



