ON THE PERPENDICULAR OF SOMERSET. 59 
fail to remark a very strong family likeness among them. 
There is a sort of character by which it is not hard to re- 
cognize them; there is a great similarity in proportion, and 
there are also several points of detail wkich most of them 
have in common. Thus there is in all a great tendeney to 
panelling in the form of windows, those portions which are 
requisite being pierced for light and sound ; so common is 
this that, in speaking of windows in a Somersetshire tower, 
one must generally be understood to mean panelled designs 
of this kind, partly blank, partly pierced. Panelling not 
thus grouped into window-patterns, such as we see at 
Cirencester, Wolverhampton, or St. Margaret’s, Leicester, 
—all noble towers, but not resembling any Somersetshire 
model—is by no means common. There is however a 
great deal of surface ornament in the way of decorative 
canopies and pinnacles, a mode of enrichment used lavishly 
at least as early as the tower at Redcliffe; and it has often 
struck me that, in the method of its treatment, the skill of 
the Somersetshire architects is admirably displayed. No 
one who has admired at a distance the magnificent outline 
of the great tower of Gloucester Cathedral can have failed 
to be disappointed on a nearer examination at the frippery 
appearance produced by the excess of ornament of this 
nature; the decorations look as if they were nailed against 
a plain wall, and had nothing further to do withit. Now 
somehow or other the Somersetshire architeets have con- 
trived to avoid this fault in the use of the very same kind 
of decoration; perhaps partly by always keeping it in 
subordination to panelled spaces; whereas at Gloucester 
there are no such spaces except the windows them- 
selves, which, being of small size and deeply recessed from 
the surface, look like insignificant apertures in the 
canopy-work. Another peculiarity is the frequent use of 
H3 
