5^7 
the natural production oj Saltpetre. 
for I have repeatedly observed that after a careful removal of 
it, its reproduction takes place to the same proportional extent 
in the same parts : and this dilference in the degree of the 
efflorescence often takes place not by a gradual transition, but 
as abruptly as if the line of separation had been marked by a 
graver; so that the part of the surface on one side of the line 
shall be almost totall}^ devoid of any efflorescence, while on the 
other side it shall resemble the accumulation of hoar-frost on 
the small branches of a tree. 
The saltpetre formed in summer scarcely appears to con- 
tain a particle of any calcareous salt : that formed in winter 
contains most evident traces of such a salt, though probably 
even in winter the amount of this is not much above one part 
in two hundred of the whole mass. 
A frosty, clear, and dry state of the atmosphere, is parti- 
cularly favourable to the natural production of nitre : but there 
seems to be a limit to its formation, on the same spot, even 
under the most favourable circumstances ; the quantity of the 
nitre not continuing to increase after it has proceeded to a 
certain extent. 
In a moist state of the atmosphere the formation either does 
not take place at all, or goes on slowly : and if that state of 
the atmosphere which is unfavourable to the production of 
nitre continue a sufflcient length of time, the nitre already 
formed gradually disappears. At the commencement of these 
observations I attributed the occasional disappearance of the 
nitre to its mechanical removal from the wall, and supposed 
that it must have been by accident brushed off; but repeated 
observations convinced me this was not the case, its disap- 
pearance proceeding as gradually as its previous appearance; 
MDCCCXIV. 3 X 
