36 



three entirely separate determinations were made, for each 



of which the reading was several times repeated. The results 



were in excellent agreement. It was found that the difference 



in electrical potential between the C.GCSra solution of Cu(NC3)2 



and the same solution containing also a C .05m concentration 



of Ca(KC3)2» was 6mv., the simple solution having the higher 



potential. From this observation it appears, by substitution 



in the formula given above, that 



C 

 Cl = 1-26 , 



or that the concentration of copper ions in the combination 

 solution is i^qp, » °^ C.794 as great as that in the simple 

 one. Now, from Jones' conductivity tables (^.£.page 57 ) 

 it is found by interpolation that the Cu(KC3)o in our simple 

 solution (O.COSm) is about 9C.6 per cent, dissociated at 25°C. 

 It thus appears that the Cu(NC2^2 ^^ ^^® combination solution 

 here considered must be dissociated to an extent e^iual to 

 0.794 X 9C.6, or 71.9 per cent. 



It has been shown earlier in this paper ( page 32 ) 

 that the Ca(NC2)2 i" this particular combination should de- 

 crease the dissociation of the copper salt, on account of 

 the common ion, to 76 percent. There is thus a difference 

 of about 4 between the percentage of dissociation of the 

 Cu(NC3)2 in this mixture, calculated from the concentration 

 of the JIC3 ion, and that derived by the use of the electrical 

 potential. This difference may, of course, be due to the 

 formation of a double salt, following the hypothesis of Clark 

 above mentioned, but it makes no difference in the present 

 discussion whether it be 76 or 72 per cent, of the Cu(NC3) 



