MACKIE — THE GEOLOGY OF THE SEA. 
247 
.silicates or felsjiars must have played, as they still continue to do, in 
the chemistry of the sea. Hence the analyses and study of the per* 
colating waters of the remaining areas of those ancient gneissic and 
bottom-most sedimentary rocks must be one of the chief points in the 
elucidation of our question. Was the sea always salt ? 
Another point of interest still remains, namely the study of the 
present constitnoits of those old primeval rocks. If water dissolves 
out certain substances from them, and the dissolution has been going 
on for ages, it follows there must be a diminished quantity or absence 
of certain soluble materials, and by consequence a proportionate pre- 
dominence of insoluble, in the residue of which their present consti- 
tution consists. 
The yellow corn waves its golden seed-bearing spikes in the 
summer's breeze, and the harvest is reaped and stored into barns ; 
but the farmer in the inclement days of winter spreads his fields with 
manure, and ploughs and furrows the soil, opening and turning it 
up to the rain, the frost, and the air. 
And why ? To replace that which the corn has extracted, and 
that the elements in their chemistry and powers may manipulate 
fresh substances required for another crop. 
If we analysed the soil before the crop was gi'own, would it 
be the same as after the crop was reaped ? Assm'edly not, so if we 
analyse the ancient rocks after ages of loss of certain ox-iginal 
ingi-edients by the incessant dissolving action of percolating springs, 
they would shew, as we have remarked, a poverty of some sub- 
stances and a superabundance of others. Thus it is that while 
we find the potash small in quantity in alkaline and saline waters, we 
find it locked up in superabundant proportion in orthoclase and other 
indissoluble forms in the constituent rocks of the earth's crust, while 
soda, which we find abundant in alkaline and saline springs, is 
observed gradually to diminish in quantity from the oldest granitic 
and gneissic rocks through their regenerated materials in the paleo- 
zoic, secondary, tertiary, and recent eras, becoming less and less as 
their periods of formation approach om' own. 
These conditions and results are readily and mutually explained 
by the soluble substances found in mineral-bearing waters, and by the 
actual constitution of the residue of the rock-mass. So the double- 
