PERPENDICULAR TOWERS OF SOMERSET. 53 



the frame; and it is manifestly impossible to enclose a spire 

 within the same frame as the tower, which frame in fact is 

 formed by the buttresses. 



Buttresses, then, are essential to a perfect tower, and 

 ought to extend, at least apparently, as high as the 

 comice moulding. Great care should be taken in the 

 arrangement of the windows and the treatment of panel 

 work, ornamental niches, etc., lest the cuntinuity of the 

 wall work be fiittered away ; and the spire ought to grow 

 as it were out of the base of the tower, — that is to say, if 

 the lines of the spire be continued to the ground, the 

 points at which they touch it ought to coincide with the 

 extemal lines of the bases of the buttresses. 



I am not sm'e whether this is exactly the case or not 

 with any spire ; but it will be found that those of the 

 fourteenth Century, at all events, approach nearer to it 

 than those of any other period, while in many of our most 

 admu'ed Perpendicular towers, the principle of spire- 

 growth is altogether abandoned, and those of fi'ame work 

 and lateral continuity very imperfectly carried out. 

 Those early Romanesque towers, which are probably of 

 Anglo-Saxon date, being destitute of buttresses, and 

 having generally each story of rather smaller area than 

 the one below, cannot really be said to have any frame 

 work ; for the pilaster-like strips of stone which we observe 

 at Earls Barton, Sompting, and elsewhere, are in fact a 

 mere matter of construction, performing the same office to 

 the nibble masoniy as the wooden frame, in what in 

 these days is called a brick noggin, does to the brick 

 work set in it ; and have rather the effect of frittering 

 away the lateral continuity, by dividing and subdividing 

 the wall into small compartments, than of conveying any 

 idea of unity in the design of the wliole l)uilding ; while 



