xx PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 
exceedingly hard and compact. The hollows were thus necessarily 
filled with the solution at the time permeating through the rocks; 
and as no spring water is pure water, but on the contrary contains 
something in solution which it has taken up, in most cases, after 
the fall of the water as rain upon the land, we feel assured that 
there is no rock without a solution of some kind working its way 
through it, if it be above the level of the sea, and the water can 
percolate into the atmosphere. It will remaim more stagnant if 
the rock be beneath the sea-level, the movement in the latter case 
being probably slight under ordinary conditions, though no doubt 
differences of temperature would produce the usual effects upon the 
moistened mass. 
The replacement of the matter of fossil shells by other substances, 
as above noticed, is a filling up of cavities upon the same principle as 
the vesicle or hollow in an igneous rock by the siliceous matter con- 
stituting an agate. In agates we have frequently not only proof of 
the successive deposits of siliceous matter, but evidence also im the 
cavities of the interior hollows, such as are noticed by Mr. Hamil- 
ton, of the crystallization of silica in a more definite manner, the par- 
ticles of silica having adjusted themselves under better conditions 
than in the chalcedonic bands forming the agate, so that rock-crystals 
are the result. The mammillated surfaces of the chalcedonie bands 
correspond with the mammillated surfaces of layers of many other 
minerals, being the grouping together of bundles of crystals radiating 
from a multitude of pomts on a previous surface of some kind, fre- 
quently not of the same mineral. The mammillated surface of ma- 
lachite bands filling up cavities in mineral vems may be taken as a 
familiar example of this arrangement, and still better perhaps, as the 
crystals are better seen, such minerals as wavellite. Common cry- 
stallized carbonate of lime, fillmg large fissures of limestone, some- 
times exhibits the succession of crystallized layers, with roughly-mam- 
millated surfaces, in a very illustrative manner. 
Mr. Hamilton remarks that the outer, or rather first-formed coat- 
ings of the large agates, these outer coatmgs sometimes too thin 
to be usefully employed, correspond in mineral structure with the 
whole contents of the mmor cavities. This may perhaps be ex- 
plained, if we consider that the siliceous solutions, the deposits from 
which entered both kinds of cavities, were abundant for a given 
time, after which they either entered less easily into the cavities, the 
first coatmgs of chalcedonic matter presenting a certain amount of 
obstacle to the free entrance of the solutions, or the solutions them- 
selves contained less silica. In the well-known agate-form bodies 
found in the dolomitic limestones and conglomerates of Somerset- 
shire and Gloucestershire, the evidence of changes in the kind of 
solutions which have entered the cavities in which these potato- 
stones, as they are commonly and locally termed, have been formed, 
is often very clear and interesting. Layers of chalcedonic matter, as 
in the agates of Oberstein, are usually first formed, after which the 
silica had better conditions for forming rock-crystals. The siliceous 
