THE ORIGIN OF IGNEOUS ROCKS. 
149 
magma, the outflows alternating with one another without 
a rigid order of sequence. The descriptions so far given 
seem to indicate that the groups of rocks erupted at these 
successive periods had the same general range of composi¬ 
tion—that is, the magma from which each group originated 
had the same or very similar chemical compositions. 
But when it is remembered that most of the ancient lavas 
are considerably decomposed or metamorphosed, it is hardly 
to be expected that their original chemical composition has 
remained unaltered; so that it may always remain a matter 
of uncertainty. When this alteration is taken into account, 
and also the fact that the full sequence of events may not be 
represented in a given locality, either because of incomplete 
exposure of the rocks or of their partial erosion, when sub¬ 
aerial deposits, it seems to the writer that the general law of 
sequence as stated in this paper will be found to hold good 
for the volcanic rocks of all epochs. 
The Devonian eruptive rocks of the region of Christiania, 
as briefly described by Brogger, are not entirely in conflict 
with this law of succession, although they are said to have 
commenced with extremely basic varieties. It is stated that 
there are alternations of more and of less silicious varieties, 
the general change being toward highly silicious ones, until 
the final eruptions, which were basic. 
From the mode of occurrence, then, of the rocks of a com¬ 
plete group, it is evident that as the eruptive action in the 
district proceeded the rocks became more highly differen¬ 
tiated. This is more quickly appreciated by studying the 
chemical composition of the various members of a rock 
group. 
CHEMICAL EVIDENCE. 
Before taking up the study of the chemical composition 
of the rocks it is necessary to call attention to the fact that 
the tables of chemical analyses and diagrams of molecular 
proportions are arranged according to the silica percentages 
of all of the rocks of each group, irrespective of the order in 
