400 PELTIER ON THE CAUSES OF THE 
sary, by a greater elevation, to cause it to extend beyond this elec- 
trically homogeneous envelope. There is another cause of electric 
indication, which does not seem to have been sufficiently guarded 
against, viz. the chemical action of vapours on the oxidizable 
wire or rods of which the apparatus consists. An iron bar, like 
that of a lightning conductor, gives a continued electric current ; 
a moist copper wire, twenty to thirty meters long, likewise gives 
one. These currents are always negative from top to bottom, 
and they increase in proportion to the humidity which has been 
deposited. Another cause of error is that which arises from 
the metals which enter into the construction of the building. 
When the walls and the supports are rather damp and they 
have become conductors, the oxidation which these metals un- 
dergo furnish a chemical current to the neighbouring conductor 
which carries it to the ground, and it is then difficult to say 
what the direction of the current will be: it will depend on the 
moisture of the vicinity to which the supports are attached. Of 
the electric apparatus which I possess, there is one which con- — 
stantly gives an electric current, because it is formed of a bar of 
iron and of a zine vane painted in oil colour. In damp weather 
the needle of the rheometer ascends to 80°, without there being 
any electric action of the atmosphere. We must therefore take 
care not to register such results as atmospheric products. For 
more security we must only expose to the air a tuft of platina 
wire, and cover the conducting wire with silk, covered over with 
an oily varnish, and insulate it in the best manner possible. 
With these precautions, the height of our buildings is hardly 
ever sufficient to enable us to obtain an electric current beneath 
a serene sky, a current which in these circumstances is always 
resinous from bottom to top. 
44. When by cooling, these resinous vapours are condensed 
and have formed opake clouds, they preserve the quantity of 
electricity which they possessed ; but this electricity is no longer 
distributed with the primitive uniformity which it had in elas- 
tic vapour not at its point of saturation. The molecules ga- 
thered then into small spheres or vesicles no longer retain their 
entire electric tension ; in this new state they conduct better, 
and the electricity is differently distributed among them ; their 
agglomeration into small masses or cloudy flakes gives them the 
periphery of a body, as well as preserving the individuality of 
the constituent parts. There are then two perfectly distinct 
