of Edinhurgli, Session 1883 - 84 . 
419 
for a minute before a strong fire, and then took three readings. 
The deflections were — 
195, 215, 220, 
giving an average of 210, which is nearly double the previous 
average. Hair is known to be highly hygroscopic ; the heating 
drives off the moisture, and the brush, when brought to the electro- 
meter, is found to be electrified positively. Flannel, after being 
warmed before the fire, is also electrified positively. These facts 
seem to favour the idea that electricity is produced by evaporation. 
It is not advisable to have the brush very dry, for it is then more 
troublesome to take away the charge from it before using it a second 
time. 
In the record of results appended (Table I.) I have entered all the 
average readings obtained. Each entry is the average generally of 
ten deflections, sometimes of fifteen and of five. The number of 
single observations made is upwards of one thousand. The course 
of the experiments breaks up into four series. 
In the case of the first series I experimented with a disc of 
copper and a disc of zinc, each of which had one side highly 
polished. The two averages for zinc obtained on the 26th Novem- 
ber are nearly equal to one another, and to the average of all the 
average readings. It is necessary to choose a standard number for 
one of the metals, and to compare the others with it ; hence copper 
is always taken as 100. The electrometer was duly replenished 
each morning, and all the conditions were preserved constant as far 
as possible ; but various circumstances caused the magnitude of the 
deflections to vary considerably from day to day. The first entry of 
26th November was got by observing the first swing of the electro- 
meter, the second by observing the permanent deflection. In all 
the subsequent observations it was the permanent deflection which 
was noted. One side of the copper disc was not so highly polished 
as the other; the rougher side gave an average of 123; the more 
polished taken subsequently gave an average of 126. I do not con- 
sider this comparison as conclusive ; for I afterwards found that a 
disc of copper gave a decidedly larger deflection than a disc of brass 
copper-plated; and one of the differences between the two discs 
was, that the latter had a much smoother surface. Another expe- 
2 E 
VOL. XII. 
