128 EFFECTS of HEAT 
upon a piece of foft butter. The carbonate had fpread very 
much on the infide of the tube, and had rifen round its lip, as 
fome falts rife from their folution in water. In this manner, 
a {mall quantity of the carbonate had reached the outer tube, 
and had adhered to it. The black colour frequently mention- 
ed as accompanying the union of the carbonates with the 
porcelain, is here very remarkable. 
On the 26th of February, I made an experiment, in which 
the carbonate was not weighed, and no foreign fubftance 
was introduced to aflift the compreflion. The temperature 
was 46°. The pyrometer had been affected by the contact 
of a piece of chalk, with which it had united ; and fome of 
the carbonate muft have penetrated the fubftance of the py- 
rometer, fince this laft had vifibly yielded to preflure, as ap- 
peared by a {welling near the contact. I obferved in thefe ex- 
periments, that the carbonate had a powerful action on the 
tubes of Cornifh clay, more than on the pounded filex. Per- 
haps it has a peculiar affinity for argil, and this may lead to 
important confequences. The chalk had vifibly firft fhrunk 
upon itfelf, fo as to be detached from the fides, and had then 
begun to run -by fucceffive portions, fo as ftill to leave a pil- 
lar in the middle, very irregularly worn away ; indicating 
a fucceflive liquefaction, like that of ice, not the nei of 
a mafs foftening all at once. 
Ow the 28th of February, I made an medina with 
oyfter-fhell unweighed, finely ground, and pafled through the 
clofeft fieves. The pyrometer gave 40°. The piece of chalk 
below it had been fo foft, as to fink to the depth of half an 
inch into the mouth of the iron air-tube, taking its impreffion 
completely. A fimall part of this lump was contaminated 
with iron, but the reft was ina fme ftate. The tube had a 
rent in it, through which the carbonate, united with the mat- 
ter of the tube, had flowed in two or three places. The 
fhell 
