s at intervals in the wall, without building columns. The
north arcade at Billingham in Durham, and the thirteenth century arcades
at Tytherington in Gloucestershire consist of arches with large masses
of the earlier wall left between them. Such a method was economical, as
much less dressed stone was required; and we find it employed at Copford
in Essex, where good building stone was hard to get. Nevertheless, it
prevented the free circulation of light from the windows of the aisles,
and practically shut off the aisles from the church.

[Illustration: Fig. 9. Gretton, Northants: arcade of nave showing
blocked window head.]

Sec. 44. There is one obvious consequence of the setting out of aisles on
either side of an existing building which, although an imperfection in
itself, contributes greatly to the variety of the parish church plan.
The builders cannot see both their aisles at one and the same time: the
older church comes in between. In fact, until the nave and aisles are
actually joined, at the close of the work, by the breaking down of the
walls beneath the arches, there can be no opportunity of appreciating
the full effect of the work. There is a famous instance at Beverley
minster of the mistakes to which the presence of the older building may
lead. The aisles of the nave were set out in the fourteenth century on
either side of an older and shorter nave. The south aisle was set out
first, the width of the eastern bay being measured from a new buttress
in the angle of nave and transept. On the north side there was a
thirteenth century buttress in this position: the builders, in setting
out their north aisle, overlooked the fact that this buttress was of
less projection than the newly built one on the other side, with the
result that their buttress measurements throughout varied on both sides,
while the standard of width between the buttresses, which had been
employed on the south side, was retained. Consequently, as the columns,
in a vaulted church, have to be built in line with the buttresses of the
corresponding aisle walls, the columns were not opposite one another,
and the discrepancy increased as the church advanced westward. When the
builders got clear of the intervening building, in the western bays of
the nave, they were able to rectify their mistake slightly; but the
effect is unpleasantly noticeable in the obliquity of the transverse
arches of the vaulting.

Sec. 45. If errors like this could take place