70 
which latter becoming indurated, might take the form of 
chert. The irregular forms in which chert occurs in the 
mountain limestone afford a strong argument in favour of 
this opinion, for it has the very appearance of having been 
deposited where moisture or small streams of water occurred. 
The hydrofluo-silicic acid thus formed would decompose car- 
bonate of lime forming the fluor spar in question, together 
with free carbonic acid gas. Now, fluor spar is soluble in 
free carbonic acid, and these two being the results of the 
above decomposition, we have ample means of conveying it 
into caverns, fissures, &c, where the volatile ingredients 
slowly evaporating, leave the fluor spar to crystallize. The 
carbonic acid, liberated by this decomposition, has the power 
of dissolving carbonate of lime as well as fluor spar, as I 
have before shown. Hence, the water filtering through 
these rocks would become saturated with this compound, and 
accumulating in any locality where the carbonic acid could 
evaporate, the carbonate of lime would crystallize in the form 
of dog tooth or other calcareous spar. 
It is possible that the period in which these changes took 
place was one of long duration, and the formation of these 
crystals exceedingly slow. Indeed, it is necessary to admit 
that such was the case, in order to account for the production 
of such large crystals as these caverns frequently contain. We 
may easily conceive how slow the process would be, when we 
remember that a large mass of igneous matter, like the toad- 
stone, would require ages to cool down, and, during its 
existence in a hot state, would probably be constantly 
furnishing materials for their formation. A time, how- 
ever, would arrive when this process of cooling would be 
completed. Fluoride of silicon would cease to be evolved, 
and carbonic acid to be liberated ; all further production 
of fluor spar and calcareous spar would here be arrested, 
and the putrefactive decomposition of these animal remains 
