A. T. CAMERON AND M. S. HOLLENBERG 



417 



TABLE HI— Continued. 



Experiments with NalOs. 



Experiments with NaNOj. 



other temperatures were made, with approximately the same degree 

 of variation. In this way the effect of temperature has been ascer- 

 tained fairly accurately. These temperature results are shown in 

 Table II. 



The modified Locke solutions employed (Table I) were equimo- 

 lecular. Mixtures in known proportions gave solutions in which defi- 



