321 
agatine and other silicious nodules may have been formed, to the 
entire exclusion of the solvent. 
But this assumed process being even allowed, it must be ad- 
mitted, that the following variations of it must be liable to occur. 
The aperture, by which the cavity is supplied, may be closed by the 
crystals at its sides meeting, previously to the filling of the cavity 
by crystallizations : or, instead of the liquid being admitted by one 
aperture, it may percolate through numerous small openings into 
the cavity. It is therefore necessary to see how the process will be 
affected by these variations. 
The aperture may become closed, under two circutnstances : there 
may, or there may not, exist some small fissures, or pores, by which 
the fluid remaining in the cavity, after the supply has failed, may 
drain off. If the former be the case, the hollow, empty, crystalline 
nodule remains ; and if the latter, the water is inclosed, and the 
enhydros, so often mentioned by lithologists, is formed. 
It now remains to inquire, what would be the consequence of 
the fluid infiltrating, through the substance forming the top and 
sides of the cavity, instead of entering by one aperture only. Here, 
as in the other instances, condensation would commence at the 
circumference, or sides of the cavity ; but would go on so slowly, as 
to admit the further infiltration of the fluid, for a considerable time ; 
and the same porosity of the substance forming the bottom of the 
mass, would likewise allow the gradual exit of a superfluous fluid. 
The process which is thus carried on, in this small cavity, will be 
found to agree, almost in every respect, with that, which we shall 
soon see has taken place, in caverns adorned with calcareous 
stalactites. As in those, so the surrounding substance in these 
smaller cavities differs, in closeness of texture, in different parts ; 
opposing more resistance to the infiltration of the fluid, in some 
parts than in others : hence the fluid oozes in a partial and irregular 
manner, according to the resistance it meets with. In some parts, 
VOL. I. 
T T 
