18 Journal of the Asiatic Society of Bengal. [NS., XVI, 
the point and its mercury cup is broken. If the amalgamation 
of the point is not satisfactory, or the mercury surface not 
clean, a slight coherer effect is noticeable at the contact, and 
the zero of the electrometer is liable to small shifts even when 
the point is pulled slowly out of the mercury as in the origi- 
nal arrangement. These effects are made all the more pro- 
bable by any vibrations which occur. but with clean mercury 
and properly amalgamated points no trouble of this kind is 
experienced. 
(3) The acid used should be boiled with a small quantity 
of ammonium sulphate, as recommended for Kelvin electro- 
meters. Its specific gravity is thus usually about 1-840 
indicating a proportion of about 95%, of pure sulphuric acid. 
By means of a standardised sliding condenser attached 
to (M) the external connection, the capacities of each part of 
the instrument may be determined with very considerable 
of the electrometers now in use in our Jaboratory. The capa- 
city of the quadrants with the key, using a 6» quartz fibre 
iving sensitiveness 250 mm. per volt when the needle is charged 
to 14 volts, is 28:5 E.S.U.; that of the external connector 
with its mercury cup is 776 E.S.U. The capacity of the 
cylindrical air-condenser as calculated from its dimensions, 
and neglecting end-corrections, was 80°3; but when determined 
by this method its true value, including its connection to the 
mercury cup, is obtained and is 83°8+0°5 E.S.U. 
has proved to be a very difficult one, especially during very 
damp weather. The arrangement used by Gockel’ or Scherring,® 
consisting of a corrugated ebonite insulator with external metal 
protection and sodium drying, was designed specially for use as 
an external insulator under such conditions, but it is quite use- 
less for electrometer measurements of any considerable degree 
of accuracy, except in very dry weather when simple insulators 
work at least equally satisfactorily. Ebonite is so sensitive to 
moisture that its use in connection with electrostatic measure- 
ments when exposed to ordinary air should be avoided, when- 
1 Phys. Zeit., Vol. 5, 1904, pages 294-296. 
Sw bs 6 6G, 1906, page 228. 
. »» 5, 1904, ,, 452. 
>) 9° 
