the thermometer for high Degrees of Fie at . 399 
commencement of rednefs up to intenfefire, the moft ftriking 
differences of the greateft part of them from the old feemed to 
originate in the lower ftages of heat ; and of thofe which were 
got from the neighbourhood of the old, the variations from it 
in the higher ftages feemed, for the moft part, to be only com- 
feqnences of thofe differences in the lower ones. 
I have mentioned, in the fi r ft Paper, that the original ther- 
mometer-pieces had their bulk enlarged a little on the approach 
of ignition ; but that by the time they became vifibly red-hot 
throughout, they are reduced to their former dimenfions again ; 
and at this moment the thermometric diminution begins. The 
new clays had their bulk enlarged in a much greater propor- 
tion, and the enlargement was of much longer continuance : 
fome of them required a heat of 15 degrees to deftroy the in- 
creafe which ignition had produced in their bulk, and bring 
them back to their original dimenfions : after this period, moft 
of them diminifhed pretty regularly, and uniformly with the 
old, being nearly fo many degrees behind it, in all the fu ex- 
ceeding ftages of heat, as they required to bring, them back 
from the enlarged ftate. 
I. have mentioned alfo, in my former paper, that a quantity 
of air is extricated from the clay, moft rapidly at the period in 
which the augmentation of bulk takes place ; and that the 
augmentation was probably owing to this air forcing the par- 
ticles of the clay a little afunder, previous to the inftant of 
its efcape. It was therefore prefumed, that the greater exten- 
fion of thefe new clays might be owing, either to a greater 
quantity, or ftronger adhefion, of this combined air : and as 
clay, kept moift for a length of time, in certain circum- 
fiances, undergoes a procefs feemingly analogous to fermenta- 
tion^, 
