help the reader a little.

The danger is, observe, that the midmost stone at _f_ pushes out the
side ones: then if we can give the side ones such a shape as that, left
to themselves, they would fall heavily forward, they will resist this
push _out_ by their weight, exactly in proportion to their own
particular inclination or desire to tumble _in_. Take one of them
separately, standing up as at _g_; it is just possible it may stand up
as it is, like the Tower of Pisa: but we want it to fall forward.
Suppose we cut away the parts that are shaded at _h_ and leave it as at
_i_, it is very certain it cannot stand alone now, but will fall forward
to our entire satisfaction.

Farther: the midmost stone at _f_ is likely to be troublesome chiefly by
its weight, pushing down between the others; the more we lighten it the
better: so we will cut it into exactly the same shape as the side ones,
chiselling away the shaded parts, as at _h_. We shall then have all the
three stones _k_, _l_, _m_, of the same shape; and now putting them
together, we have, at C, what the reader, I doubt not, will perceive at
once to be a much more satisfactory arrangement than that at _f_.

Sec. III. We have now got three arrangements; in one using only one
piece of stone, in the second two, and in the third three. The first
arrangement has no particular name, except the "horizontal:" but the
single stone (or beam, it may be,) is called a lintel; the second
arrangement is called a "Gable;" the third an "Arch."

We might have used pieces of wood instead of stone in all these
arrangements, with no difference in plan, so long as the beams were kept
loose, like the stones; but as beams can be securely nailed together at
the ends, we need not trouble ourselves so much about their shape or
balance, and therefore the plan at _f_ is a peculiarly wooden
construction (the reader will doubtless recognise in it the profile of
many a farm-house roof): and again, because beams are tough, and light,
and long, as compared with stones, they are admirably adapted for the
constructions at A and B, the plain lintel and gable, while that at C
is, for the most part, left to brick and stone.

Sec. IV. But farther. The constructions, A, B, and C, though very
conveniently to be first considered as composed of one, two, and three
pieces, are by no means necessarily so. When we have once cut the stones
of the arch into a shape like that of _k_, _l_, and _m_, they will hold
tog