daily use.
Thus the Renaissance manner of building is a convenient style for
dwelling-houses, but the natural sense of all religious men causes them
to turn from it with pain when it has been used in churches; and this
has given rise to the popular idea that the Roman style is good for
houses and the Gothic for churches. This is not so; the Roman style is
essentially base, and we can bear with it only so long as it gives us
convenient windows and spacious rooms; the moment the question of
convenience is set aside, and the expression or beauty of the style is
tried by its being used in a church, we find it fail. But because the
Gothic and Byzantine styles are fit for churches they are not therefore
less fit for dwellings. They are in the highest sense fit and good for
both, nor were they ever brought to perfection except where they were
used for both.

Sec. LVII. But there is one character of Byzantine work which, according
to the time at which it was employed, may be considered as either fitting
or unfitting it for distinctly ecclesiastical purposes; I mean the
essentially pictorial character of its decoration. We have already seen
what large surfaces it leaves void of bold architectural features, to be
rendered interesting merely by surface ornament or sculpture. In this
respect Byzantine work differs essentially from pure Gothic styles,
which are capable of filling every vacant space by features purely
architectural, and may be rendered, if we please, altogether independent
of pictorial aid. A Gothic church may be rendered impressive by mere
successions of arches, accumulations of niches, and entanglements of
tracery. But a Byzantine church requires expression and interesting
decoration over vast plane surfaces,--decoration which becomes noble
only by becoming pictorial; that is to say, by representing natural
objects,--men, animals, or flowers. And, therefore, the question whether
the Byzantine style be fit for church service in modern days, becomes
involved in the inquiry, what effect upon religion has been or may yet
be produced by pictorial art, and especially by the art of the
mosaicist?

Sec. LVIII. The more I have examined the subject the more dangerous I
have found it to dogmatize respecting the character of the art which is
likely, at a given period, to be most useful to the cause of religion.
One great fact first meets me. I cannot answer for the experience of
others, but I never yet met with a Christian