crust would necessarily be thrown into
folds. When an apple dries and shrivels in winter, the surface becomes
covered with ridges. Or again, if we place some sheets of paper between
two weights on a table, and then bring the weights nearer together, the
paper will be crumpled up.

[Illustration: Fig. 17.--Adapted from Ball's paper "On the Formation of
Alpine Valleys and Lakes," _Lond. and Ed. Phil. Mag._ 1863, p. 96.]

In the same way let us take a section of the earth's surface AB (Fig.
17), and suppose that, by the gradual cooling and consequent contraction
of the mass, AB sinks to A'B', then to A''B'', and finally to A'''B'''.
Of course if the cooling of the surface and of the deeper portion were
the same, then the strata between A and B would themselves contract, and
might consequently still form a regular curve between A''' and B'''. As
a matter of fact, however, the strata at the surface of our globe have
long since approached a constant temperature. Under these circumstances
there would be no contraction of the strata between A and B
corresponding to that of those in the interior, and consequently they
could not lie flat between A''' and B''', but must be thrown into folds,
commencing along any line of least resistance. Sometimes indeed the
strata are completely inverted, as in Fig. 19, and in other cases they
have been squeezed for miles out of their original position. This
explanation was first, I believe, suggested by Steno. It has been
recently developed by Ball and Suess, and especially by Heim. In this
manner it is probable that most mountain chains originated.[43]

The structure of mountain districts confirms this theoretical
explanation. It is obvious of course that when strata are thrown into
folds, they will, if strained too much, give way at the summit of the
fold. Before doing so, however, they are stretched and consequently
loosened, while on the other hand the strata at the bottom of the fold
are compressed: the former, therefore, are rendered more susceptible of
disintegration, the latter on the contrary acquire greater powers of
resistance. Hence denudation will act with more effect on the upper
than on the lower portion of the folds, and if continued long enough, so
that, as shown in the above diagram, the dotted portion is removed, we
find the original hill tops replaced by valleys, and the original
valleys forming the hill tops. Every visitor to Switzerland must have
noticed hills where the