162 
TRANSLATION. 
rich, in the oxides of iron, especially in certain regions. From this source 
have emanated those great masses of Magnetic oxide which have hurst 
forth in the fashion of the igneous rocks ; and which in Italy, and the 
Ural Mountains, as in Scandinavia, are coimected with the amphibolic 
or pyroxenic rocks. In the upper layer are collected, by preference, the 
lighter or more volatile bodies, such as the alkaline metals, fluorine, 
boron, &c., &c.; and, in fact, it is in the granitic rocks arising from this 
layer, that we find commonly the fluo or boro-silicated minerals, as mica, 
topaz, tourmaline, &c., &c. 
Part II. 
OX THE CHANGES WHICH HAVE TAKEN PLACE IN THE COMPOSITION OE THE 
TWO INCANDESCENT LAYERS, EROM WHICH THE ERUPTIVE ROCKS HAVE 
EMANATED. 
Comparative Chemical Composition of the Principal Types of Erup¬ 
tive Rocks. —It has appeared to me interesting to investigate the changes 
which have taken place, from the earliest epochs of the globe, in the 
nature of the incandescent layers which constitute the seat of the erup¬ 
tions : for this purpose it is necessary to compare the principal types of 
rocks which have burst forth during successive geological epochs. I have 
determined their average composition, and the limits of the variations 
which take place in the proportions of their elements, both from my own 
researches and from the chemical analyses published by various writers, 
among whom I may mention, in particular, MM. Gmelin, Abisch, Du- 
frenoy, Ebelmen, Delesse, and Ch. Deville. The results at which I have 
arrived are contained in the annexed Table. 
Changes in the Composition of the Acid Layer deduced from a Com¬ 
parison of the two great Families of Siliceous Rocks. —The siliceous rocks 
comprise two great families: the Granites and the Trachytes: their sepa¬ 
ration is very definite chronologically, for the former belong to the pri¬ 
mary or secondary periods, and the latter to the tertiary, quaternary, 
and modern periods. If, now, we compare the average composition of 
the two fundamental types of those families, granite and trachyte, and 
consider that they represent the most abundant products of the siliceous 
layer, we shall find that in the long course of ages that divide the pri¬ 
mary and tertiary periods from each other, the following changes took 
place in the composition of the fluid mass which nourished the eruptions: 
viz., there was a diminution of too or t^o in the proportion of silica, and 
of to 1 tin the potash ; but that the proportions of lime and iron oxides were 
almost doubled, and that of soda tripled. If, again, we compare the com¬ 
position of the trachytes of the tertiary period with that of the trachytic 
lavas of the actual epoch (and we may cite as a type lava that of the 
Arso, which was spread over the isle of Ischia in 1301), we shall find 
that the proportion of silica has diminished still further, remaining, 
however, greater than the quantity contained in the various rocks ejected 
from the basic layer; while the soda has increased by more thanAfe. 
