of Edinburgh, Session 1885 - 86 . 
533 
is saturated, and further observations must be made before the actual 
daily variation can be determined. A careful comparison of the obser- 
vations with the ordinary meteorological sheets of the Observatory, 
however, brings out some points of very great interest, especially 
when looked at in the light of the theory advanced by Professor C. 
Michie Smith, in a paper communicated by him to the Society 
on 1st June 1885 (Trans. R.S.E . , xxxii. p. 583). As the result 
of observations on atmospheric electricity, made on the summit 
of Dodabetta, the highest hill in the N eilgherries, Professor 
Smith finds that on the edge of a dissolving mist the potential is 
lower than the normal, while it is higher on the edge of a condensing 
mist. A study of the following table, with the notes appended, 
shows at once that in nearly every case where the top of Ben Nevis 
became clear for a short time, a strong current was found coming 
up the line, while as soon as the summit was again enveloped, the 
current was reversed. The connection between the moisture of the 
air and the earth currents is still further shown by studying the 
rainfall. During a fall of rain or snow the current is almost always 
found to pass down the cable ; and, in the case of a sudden shower, 
the current sometimes drove the mirror of the galvanometer violently 
off the scale. A cessation of rain or snow will be seen in most cases 
to have exactly the opposite effect. 
If we assume that the summit of Ben Nevis takes the potential 
of the masses of vapour covering it, and consider the earth-plate 
at the base as the real “ earth ” or zero of potential, it is at once 
obvious that these results agree with those of Professor Smith, and 
form important additional evidence of the truth of his theory, a 
conclusive proof of which would be of the greatest value in investi- 
gations connected with thunderstorms. 
Following out the above conclusions, we should expect the mean 
daily curve for earth currents to be intimately connected with that 
for relative humidity, but as yet no observations have been taken 
except in saturated air. Further study of the curves shows that the 
currents are probably affected by other causes besides those con- 
nected with cloud. The approach of cyclones seems to disturb the 
electrical conditions to a certain extent, but the observations are 
not sufficiently numerous to allow of any definite inferences being 
drawn. My object in writing this paper has been chiefly to draw 
