2. Improvements in Measurements with Quadrant 
Electrometers. 
Part II. Simplified arrangements for accurate and 
continuous work. 
By V. H. Jackson, M.A., and A. T. Muxerser, M.A. 
n a previous paper ' we discussed the difficulties connec- 
ted with the use of sensitive quadrant electrometers in India, 
and described arrangements which we had found adequate at 
Patna to give quite satisfactory results during the dampest 
w pera: The principles underlying these improvements 
were : 
(1) The —— of all wr camerpten insulators. By 
a special key direct to the amber insula- 
tors of one pair of qua a nts we reduced the 
number of insulators required for all the ordinary 
operations in electrometer work to two. a ( (a) 
the insulation of these quadrants; and (5) the in- 
ee ot the wire connecting the electrometer to 
any apparatus outside 
(2) Protection from the Ws of moisture and electro- 
static disturbances in the atmosphere. We enclosed 
the electrometer and attached key in an outer 
earth-connected metal case, and operated the key 
connections from the outside, the air inside the 
case being kept dry by strong sulphuric acid. 
In the actual arrangements then in use, the mere ane 
on e- 
was enclosed in a square zinc case fitte wooden 
ork. The connections with the aga cups o “of the He 
were made in the usual manner, by ot ing down staal 
pieces of brass with amalgamated points by means of silk loops 
attached to cotton threads passing through small holes in the 
case. Drying was effected by soldering up one ee of the 
case after the strong sulphuric acid was introduced, but com- 
munication with the outside air was still possible through the 
four small holes for the threads. These arrangements were 
evidently open to considerable improvement. The operation 
of soldering alone took several minutes, during which the acid 
moisture gradually entered the gase owing to diffusion or tem- 
perature differences. For these reasons, as we pointed out, 
| Journal and Proceedings, A.8.B., Vol. X, No. 6, 1914, pages 227-240. 
