84 RICHTHOFEN 
NATURAL SYSTEM 
Cretaceous period by thousands of feet. This is not only true for those ranges which, 
like the Andes and Rocky Mountains, have been intensely volcanic, but also for such 
as show a less immediate connection with eruptive activity in the voleanic era, as is 
the case with the Pyrenees, the Alps and the Himalaya. The elevation which sue- 
ceeded the Cretaceous period appears to have everywhere been slow, intermittent, and 
partly retrograde, in the Eocene epoch, while its chief phase is traceable to the Miocene 
and subsequent epochs, by the relative altitude above the level of the sea to which 
sediments of different ages have been raised. 
There can scarcely be any doubt that this acceleration of the changes of level, 
and the extrusion of voleanic rocks, have had an intimate connection. But the mode 
of this connection has by no means been ascertained, and appears to be quite an intri- 
cate subject. It was formerly believed that the eruptive rocks were themselves the 
upheaving agents, their intrusion and ejection having been the mechanical means of 
the elevation of mountain ranges. But geological observation and argumentation have 
conclusively demonstrated that such could not be the case, and made it probable that 
both elevation and eruption are the most conspicuous symptoms of different agencies 
which were dependent on one common cause. When that first theory had to be aban- 
doned, another was put in its place, to the effect that the ejection of rocks must have 
been attended by subsidence ; and, although we shall show in the following pages 
that the rise of the crust greatly predominated in the vicinity of the eruptions, there 
can be no doubt that subsidence was probably in all cases among their intricate effects. 
But the cause generally ascribed to it is not at all capable of explaining it. The 
assumption that eruption must be attended by subsidence is usually made on the ground, 
that the removal of a certain volume of matter by ejection, from a deep-seated place, 
would cause the formation of a vacuity, corresponding in extent to the volume 
removed, and that consequently the overlying portions of the crust would settle down. 
This assumption appeared to be corroborated by the observation that the region sur- 
rounding a volcano subsides during its activity, while it rises in periods of rest. But 
it is evident that the size of a current of lava is not proportionate to the settling of an 
area of hundreds of square miles to the amount of several feet. If, moreover, the 
views here advocated, regarding the cause and mode of eruptive activity, are correct, 
that is, if the ejection of rocky matter is the effect of the increase of volume which it 
undergoes on passing into the state of aqueous fusion, then the eruption is only the 
discharge of the surplus matter which has no room within the space of the fissure ; no 
vacuity could therefore be formed, and the commonly adopted cause of the subsidence 
would not exist. There are, however, two other causes which would produce sub- 
sidence, and have probably been acting in every case of massive and volcanic erup- 
tions. The first of them is the contraction of the liquid mass in the conduits, by cool- 
ing. Its amount must be considerable in proportion to the space of the fissure, but 
small when compared with the volume of a mountain range on the surface. Its effects 
will be local and abrupt. In the case of volcanoes, it is the probable cause of the 
familiar phenomenon of the subsideuce of the bottom of craters, and of those less fre- 
quent cases where whole portions of the cone are suddenly engulfed. As regards 
extensive accumulations of rocks by massive eruptions, there are certain features of 
(122 
