of Edinhurgli, Session 1883-84. 
423 
ing for tin was smaller than any of ten subsequent, and such was 
the case also with the brass ; they were noted, but not included in 
the average. Two series were taken for lead, the former for the 
side which had been roughened by polishing, the latter for the side 
which had not been interfered with : in both cases the first whisk 
gave a nearly null deflection, but the deflection afterwards became 
pretty steady about an average. 
To further elucidate the cause of this initial phenomenon, I 
rubbed each disc several times with a piece of flannel stretched over 
my forefinger. The electricity was not discharged after each rub, 
but was allowed to accumulate : — 
Disc. 
Electricity. 
Lead (rougher side), . 
positive. 
Copper, 
negative. 
Zinc, 
• • T> 
Tin, 
' V) 
Iron, 
" J? 
Brass, . negative, with trace of positive at first. 
The lead disc was tried again, using the same side, and rubbing 
with what was noted to be a specially clean part of the flannel. 
It became negative at first, then positive after several rubs with 
considerable pressure. The same operation was repeated, and with 
the same result. The more polished side was then tried ; it was 
negative at first, and greater pressure in rubbing was required to 
produce positive. Hence, the anomalous electricity produced on 
the lead is undoubtedly due to the fact that the rubbing with the 
flannel cloth abraded the lead (or the oxide of lead), as was indeed 
evident from an inspection of the cloth after the operation. And 
these trials also show why lead and bismuth and graphite wmre so 
prone to become positive under the friction to which they were sub- 
jected by De la Eive and the other experimenters. The ease with 
which these substances make a mark on paper shows that they can 
be easily abraded ; and when the rubbing is so violent as to cause 
abrasion, anomaly may very well be expected. 
The camel’s-hair brush changes the electricity not from nega- 
tive to positive, but from positive to negative. It certainly does 
not abrade the surface. Does the explanation of the anomaly con- 
