TEST OF AGE OF VOLCANIC ROCKS. 
521 
carbonate of lime, silex, iron, and other mineral ingredients, 
whereby gradual changes in the constitution of the rocks 
may be superinduced. Every geologist is aware how often 
silicified trees occur in volcanic tuhk, the perfect preserva¬ 
tion of their internal structure showing that they have not 
decayed before the petrifying material was supplied. 
The porous and vesicular nature of a large part, both of 
the basaltic and trachytic lavas, affords cavities in which 
silex and carbonate of lime are readily deposited. Minerals 
of the zeolite family, the composition of which has already 
been alluded to, p. 500, occur in amygdaloids and other trap- 
rocks in great abundance, and Daubree’s observations have 
proved that they are not always simple deposits of sub¬ 
stances held in solution by the percolating waters, being oc¬ 
casionally products of the chemical action of that water on 
the rock through which they are filtered, and portions of 
w^hich are decomposed. From these considerations it fol¬ 
lows that the perfect identity of very ancient and very mod¬ 
ern volcanic formations is scarcely possible. 
Tests by Superposition. — If a volcanic rock rest upon an 
aqueous deposit, the volcanic must be the newest of the two; 
but the like rule does not hold good where the aqueous for¬ 
mation rests upon the volcanic, for melted matter, rising 
from below, may penetrate a sedimentary mass without 
reaching the surface, or may be forced in conformably be¬ 
tween two strata, as h below D in the annexed figure (Fig. 
597), after Avhich it may cool down and consolidate. Super¬ 
position, therefore, is not of the same value as a test of age 
in the unstratified volcanic rocks as in fossiliferous forma¬ 
tions. We can only rely implicitly on this test where the 
volcanic rocks are contemporaneous, not where they are in¬ 
trusive. Now, they are said to be contemporaneous if pro¬ 
duced by volcanic action which was going on simultaneously 
with the deposition of the strata with which they are asso¬ 
ciated. Thus in the section at D (Fig. 597), we may perhaps 
ascertain that the trap h flowed over the fossiliferous bed c, 
and that, after its consolidation, a was deposited upon it, a 
and c both belonging to the same geological period. But, 
on the other hand, we must conclude the trap to be intru¬ 
sive, if the stratum a be altered by h at the point of contact. 
