516 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



there were any considerable residual charge, be larger than for those 

 with 0.0017 second of charge, and this relation is found in scarcely more 

 than half the cases. On the other hand, there is a continual increase 

 of the throws, due probably to a temperature influence. Of course 

 any effects due to a temperature-coefficient of capacity will be highly 

 magnified in measuring differential throws, as is done here. 



In the last three measurements, recorded above and in all the follow- 

 ing ones, the 50 microfarad condenser was connected across the storage 

 battery. 



TABLE XLVII. 

 " Pure Par. R " vs. Air. 



TABLE XLVIII. 

 " Pure Par. P " vs. Air. 



