ELECTRIC PHA NOMENA OF THE ATMOSPHERE. 399 
space decreases as the square of the resinous inductions which 
are interposed between our instruments and it. 
42. The indication of the electrometers depending on the 
point of their neutralization and on the purity of the atmosphere, 
it would be necessary to fix the place of this point of neutraliza- 
tion, and to decide once for all on the moment which ought to 
be assumed for the maximum of resinous tension of the globe. 
To have the absolute measure of the terrestrial tension a com- 
plete absence of resinous vapour would be required ; that is to 
say, it would be necessary to operate at the pole itself, when all 
the vapours have disappeared, and to measure, at the winter 
solstice, the divergence of a standard electrometer which has 
been raised one decimeter. If we possessed an experiment thus 
made, all other electrometers might be regulated by this stan- 
dard electrometer, as barometers are regulated by a standard 
barometer. This means not being practicable, that must be 
chosen which is the most so for each climate. Thus in Russia, 
at Kasan for example, a frost of —25° C., which has lasted 
twenty days, might be taken; at Berlin a frost of the same dura- 
tion of 15° C. might be taken. At Paris it would be necessary 
to profit by a winter which would give during ten successive 
days a cold of 10° C. with an easterly wind. Severe and long 
_ frosts are too rare at Paris to serve as starting-points; it is 
better to adopt a less degree of cold, but which more frequently 
occurs, unless for the purpose of comparing the electrometers 
with those of Berlin, St. Petersburg or Kasan. It is obvious 
that the lower the temperature is, the less interposed vapour 
there will exist, and the nearer we shall be to the absolute ten- 
sion of the earth. 
43. During all these regular and successive changes, the fixed 
| apparatus indicate nothing; they have always time to put them- 
selves in equilibrium. ‘These instruments are inactive unless 
their length is considerable, unless the electricity of induction, 
radiating from an extended surface of great curvature, as is the 
case with metallic wires, leaves in the instrument only the 
repelled electricity. When the exploring wire is very long, the 
radiation is sufficient to produce a continued current and cause 
the rheoscopes to act. These instruments, like static elec- 
trometers, indicate differences only; the current will also diminish 
in proportion to the resinous vapours which the atmosphere, by 
which the instrument is surrounded, may contain, and it is neces- 
