ON THE PERPENDICULAR OF SOMERSET. 39 



we find the most perfect of all, a shaft forming a member of 

 the piei' carried up stralght from the ground. This is the 

 more remarkable at Yatton, as its coved roof did not 

 require any roof-shafts at all ; they are clearly added 

 whoUy for the improvement of the general efFect. And I 

 think we may fairly add St. Cuthbert's ; the shafts, of 

 coiirse, cannot rise from the ground, but they somehow 

 look as if the designer would have made them do so, had he 

 planned the church from its foundations. 



In the other variety no shafts rise from the ground ; but 

 a niche is placed between each bay of the clerestory, sup- 

 ported by a shaft corbelled off above the pillars ; the same 

 figure, usually an angel, serves for a finial to the niche, and for 

 a corbel to the roof. This confusion is clearly a mistake in 

 decorative construction,* and, together with a certainwant 

 of simplicity in the whole, must make us considerthis form 

 abstractedly inferior to the other. Nevertheless it is one 

 of the most gorgeous magnificence, and it will be observed 

 that it is very nearly identical with that of the splendid 

 nave of St. Mary's in Oxford, the chief difference being 

 that the latter has no shaft below the niche, a point on 

 which the advantage lies on the side of the Somersetshire 

 examples. Of these the grandest is Martock, but the same 

 plan is also foUowed at Taunton and Bruton, which resemble 

 each other in so many points. 



Of the means of fiUing up spandrils, the most natural is 

 by figures similar to those which are used in the spandrils 

 of doorways, or by other analogous pi-ocesses. Of these 

 there is an early example in the choir at Ely, and they 

 seem so natural a development from the figures often 

 inserted in the same position in Early Gothic buildings, 

 that one wonders they are not more commonly met with in 

 * See Historv of St. David's, 



