Hy) RICHTHOFEN — NATURAL SYSTEM 
in yoleanic countries, and conspicuous in those numerous instances where elevation is, 
or has been in former times, proceeding more rapidly along the crests of mountain 
ranges than at either foot of them. Adherents of the theory, that all oscillations of 
the earth’s crust are only due to its contraction, consider elevation as real in these 
cases, and have tried to explain it by the assumption, that folds must be formed by the 
subsidence of the ample shell on the contracted nucleus ; that these folds would increase 
in amplitude in consequence of a lateral pressure caused by further subsidence, and 
thus an absolute rise would be effected. It is, indeed, very probable that this process 
is of vast importance in the formation of mountain ranges ; but it cannot explain the 
total amount of the changes of level. If Fourier’s calculation, that, taking the present 
loss of heat by the globe as standard, the radius of the latter should have shortened 
seventeen centimeters in twenty-five centuries, is correct, the diminution would have 
been 1,700 centimeters in 250,000 years, and about six hundred feet in one million 
years. It is manifest how utterly insignificant would be the corresponding diminution 
of the earth’s cireumference, when compared with the vast changes of level which the 
surface must be supposed to have undergone during such a length of time. 
We are bound, for these reasons, to consider, not alone the process of elevation 
of mountain ranges, but also the secular rise of extensive regions, as realities which 
cannot be exclusively explained by the contraction of the globe. But if so, there must 
exist another force, the effects of which oppose those of contraction. The united action 
of both, and the periodical prevalence of either of them, would then be capable of 
explaining the alternation of elevation and subsidence at every single place, and the 
contemporaneous action of both in neighboring regions. This antagonistic force con- 
sists, probably, in an increase of volume attending the slow and perfect crystallization 
of matter, by cooling down, during immense periods, from a viscous state. With a 
great number of bodies, contraction by loss of heat appears to continue to the very 
moment of solidification, and to increase during the latter when no time is allowed for 
crystallization, but to be diminished, and finally reversed, in the same measure as oppor- 
tunity is given for a slow and perfect crystallization. In the special case of rocks 
made up of silicates, this must remain a supposition which is not proved, but is emi- 
nently probable.* 
18 An increase in yolume by crystallization has been found, by experiment, to take place in numerous instances, in fact 
in most cases when a perfect crystallization from a molten state has been obtained, provided that the volume of the substance 
experimented on could be determined immediately before the act of crystallizing. In respect to those silicates which make 
up crystalline rocks (not taking into account the water entering into their composition), experiments could hitherto not 
be made, because it is impossible, with our present means, to allow the viscous mass sufficient time for crystallizing. Yet, 
there are some suggestive facts which have, in some measure, the value of experiments. We mention among them par- 
ticularly one which was observed by Ferd. Zirkel in analyzing, with the aid of the microscope, the texture of the 
minerals participating in the composition of eruptive rocks. He found that “glass cavities” contain ordinarily several 
of those vacuities which are also exhibited by the cavities filled with water, and have been supposed by Sorby 
and others to have originated from the contraction, by cooling, of the enclosed matter. If the substance contained 
in the glass cavities bears signs of an incipient crystallization, by having partly a “lithoid” texture, the vacuities are 
of rarer occurrence, while they are entirely wanting in the so-called stone-cavities, when the lithoid texture pervades the 
whole mass filling the cavity. It is scarcely possible to make an experiment more convincive than this natural occurrence, 
which is more open to subtle observation and measurement than is ordinarily the case with experiments on cognate subjects. 
It needs hardly to be mentioned, in connection with this question, that those experiments which have been made by Bishof 
(90) 
