CONSTANT FOR ELECTROLYTES IIEBI5. 



411 



each case to break up into three ions, as, according to Prof. 

 MacGregor's Diagram of Freezing-point depressions, it seems to 

 do. 



In the case of MgS0 4 'each gramme-molecule contains two 

 gramme-equivalents, hence fc Jm ; and since each molecule 

 breaks up into two ions we get l = i. 



In the case of H 3 P0 4 , if each gramme-molecule contains 

 three gramme-equivalents, we have k = \m ; also, if each molecule 

 breaks up into two ions, as Loomis's and Jones's results seem to 

 imply, we have l = % i. 



As the constants, m and i, depend so simply on the constants, 

 k and I, the accuracy of their determination will depend on the 

 accuracy with which we can determine k and I. Dr. MacGregor 

 has shewn that the values of I can be determined with a much 

 greater degree of accuracy than can k. 



Thus if AB or AB' be the true curve representing the rela- 

 tion between 6 and a the curves for different electrolytes bend 

 in different ways then the equation 6=k (1 a)-{-la will repre- 

 sent the straight line AE, i. e., the tangent to the curve at 

 infinite dilution if AC represents unity. 



a / 



A A f 



