during geologic ages so that the shrinkage should accumulate
within, until finally collapse came, giving an era of uplift,
it is obvious that we could account for such cycles. There is
very clear evidence that the outermost layer of the earth's
crust is but a thin shell like the outer shuck or exocarp of a
butternut, so thin that it is not at all possible that it can
sustain itself for more than a hundred miles or so, or for more
than a very few years at the outside. Hayford's[1]
investigations are the latest that show that the continents
project because, on the whole, they are lighter, they float,
that is, above the level of the oceans because there is a mass
of lighter rock below, like an iceberg in the sea. Here the
likeness between nut and earth fails and it would be more like
the earth if the outer shuck were thicker in certain large
areas. If this extra lightness or "isostatic compensation" is
equally distributed, Hayford finds[2] that the most probable
value of the limiting depth is 70 (113 km.) miles, and
practically certain that it is somewhere between 50 (80 km.)
and 100 (150 km.) miles; if, on the other hand, this
compensation is uniformly distributed through a stratum 10 (16
km.) miles thick at the bottom of the crust so that there is a
bulging of the crust down into a heavier layer below to balance
the projection of the mountains above, as I think much more
likely, then the most probable depth for the bottom of the
outer layer is 37 (60 km.) miles. This layer is much thinner
than the outer layer of the figure and is supposed to yield to
weight placed as, though more slowly than, new thin ice bends
beneath the skater.

[1] The figure of the earth and isostasy from measurements in