[MORAN] RADIUM STANDARD SOLUTIONS 81 
(d) The position of the rod and charging-key during observa- 
tions affects the electrical capacity and this accounted for a slight 
error in the earlier observations. Later, it was kept in a definite 
position after charging the leaf, by pushing it back until it touched 
the inside of the electroscope in the case of the charging rod. The 
rod to which the gold-leaf was attached was maintained at the same 
standard position on the scale by slightly contracting or releasing 
the earth-wire. 
(e) The leaf was always charged to the same division on the 
scale. Separate tests were made, however, by charging the leaf to 
varying positions on the scale; but it did not seem to make any apprec- 
iable difference. The same discrepancies occurred whether the leaf 
was charged to 25 or 35 on the scale. 
(3) The microscope was in good condition, with a clear scale. 
It was fixed in position at the beginning of the experiment, focussed 
on the leaf and not moved afterwards. 
(4) All errors in calculation were avoided by checking each 
experiment at the time it was done and re-checking a number of them 
at a time at a later date. 
(5) Errors due to heating effects were avoided by having the 
light reflected into the electroscope from the source, the latter being 
further prevented from heating the electroscope by having it screened 
with paper. Heat was also turned off in the coils beneath the electro- 
scope. This was very variable at times and in an earlier stage of the 
work may have introduced an error by causing convection currents 
in the electroscope. : 
(6) The pressure of the air inside the electroscope was always 
maintained at the pressure of the atmosphere obtaining at the time of 
the experiment. A rubber tube on the three-way tap connected to a 
glass tube dipping into water, furnished a means of showing when 
the air-pressure inside the electroscope was atmospheric. 
(7) The stop-watch used in the earlier part of the work was new 
and guaranteed to keep good time. It was tested with the clock 
in the Elementary Laboratory of the Physics Building for ten consec- 
utive instances over half-minute intervals; but it did not show any 
serious error. However, as a reliable jeweller claimed that, owing 
to the mechanism of it, there was likely to be a large error introduced 
at irregular intervals, it was decided to get a new chronometer, which 
by going continuously eliminated the possible source of error in the 
stop-watch. 
The net result of these investigations was to prove that errors 
might be due to— 
(a) Differing positions of the charging-rod and key. 
