180 



Scientific Proceedings^ Royal Dublin Society. 



This unsteadiness was principally due to fluctuations in the a.c. supply, but 

 also partly to the fact that the apparatus was insufficiently shielded and 

 imperfectly insulated. A large sheet-metal box was procured, and the valve, 

 resistances, condenser, and filament-battery were placed on sheets of ebonite 

 within. The connexions to the a.c. and d.c. supplies and to the galvanometer 

 were, of necessity, still -exposed; but, nevertheless, a considerable improvement 

 in the steadiness of the readings was found. It appeared likely that the 





48 



Fig. 5. 



50 



52 



remaining unsteadiness was chiefly due to the a.c. fluctuations. It must be 

 remembered that, when the greater part of the plate-current is balanced with 

 the potentiometer, the current through the galvanometer is a difference effect. 

 Just as a small change in the capacity C produces a large effect on the 

 galvanometer, so will small variations in any of the other quantities which 

 occur in equation (I). 



