he church is, of course, rudely represented, and
the two upper stories of it reduced to a small scale in order to form a
background to the figures; one of those bold pieces of picture history
which we in our pride of perspective, and a thousand things besides,
never dare attempt. We should have put in a column or two, of the real
or perspective size, and subdued it into a vague background: the old
workman crushed the church together that he might get it all in, up to
the cupolas; and has, therefore, left us some useful notes of its
ancient form, though any one who is familiar with the method of drawing
employed at the period will not push the evidence too far. The two
pulpits are there, however, as they are at this day, and the fringe of
mosaic flowerwork which then encompassed the whole church, but which
modern restorers have destroyed, all but one fragment still left in the
south aisle. There is no attempt to represent the other mosaics on the
roof, the scale being too small to admit of their being represented
with any success; but some at least of those mosaics had been executed
at that period, and their absence in the representation of the entire
church is especially to be observed, in order to show that we must not
trust to any negative evidence in such works. M. Lazari has rashly
concluded that the central archivolt of St. Mark's _must_ be posterior
to the year 1205, because it does not appear in the representation of
the exterior of the church over the northern door;[153] but he justly
observes that this mosaic (which is the other piece of evidence we
possess respecting the ancient form of the building) cannot itself be
earlier than 1205, since it represents the bronze horses which were
brought from Constantinople in that year. And this one fact renders it
very difficult to speak with confidence respecting the date of any part
of the exterior of St. Mark's; for we have above seen that it was
consecrated in the eleventh century, and yet here is one of its most
important exterior decorations assuredly retouched, if not entirely
added, in the thirteenth, although its style would have led us to
suppose it had been an original part of the fabric. However, for all
our purposes, it will be enough for the reader to remember that the
earliest parts of the building belong to the eleventh, twelfth, and
first part of the thirteenth century; the Gothic portions to the
fourteenth; some of the altars and embellishments to the fif