ettled foundation, level and
firm, out of sight. But this has not been done in some of the noblest
buildings in existence. It cannot always be done perfectly, except at
enormous expense; and, in reasoning upon the superstructure, we shall
never suppose it to be done. The mind of the spectator does not
conceive it; and he estimates the merits of the edifice on the
supposition of its being built upon the ground. Even if there be a vast
table land of foundation elevated for the whole of it, accessible by
steps all round, as at Pisa, the surface of this table is always
conceived as capable of yielding somewhat to superincumbent weight, and
generally is so; and we shall base all our arguments on the widest
possible supposition, that is to say, that the building stands on a
surface either of earth, or, at all events, capable of yielding in some
degree to its weight.

[Illustration: Fig. II.]

Sec. VIII. Now, let the reader simply ask himself how, on such a surface,
he would set about building a substantial wall, that should be able to
bear weight and to stand for ages. He would assuredly look about for the
largest stones he had at his disposal, and, rudely levelling the ground,
he would lay these well together over a considerably larger width than
he required the wall to be (suppose as at _a_, Fig. II.), in order to
equalise the pressure of the wall over a large surface, and form its
foot. On the top of these he would perhaps lay a second tier of large
stones, _b_, or even the third, _c_, making the breadth somewhat less
each time, so as to prepare for the pressure of the wall on the centre,
and, naturally or necessarily, using somewhat smaller stones above than
below (since we supposed him to look about for the largest first), and
cutting them more neatly. His third tier, if not his second, will
probably appear a sufficiently secure foundation for finer work; for if
the earth yield at all, it will probably yield pretty equally under the
great mass of masonry now knit together over it. So he will prepare for
the wall itself at once by sloping off the next tier of stones to the
right diameter, as at _d_. If there be any joints in this tier within
the wall, he may perhaps, for further security, lay a binding stone
across them, _e_, and then begin the work of the wall veil itself,
whether in bricks or stones.

Sec. IX. I have supposed the preparation here to be for a large wall,
because such a preparation will give us the bes