PART lit. ARCHITECTURE. 99 



as the cross, the circle, the triangle, &c. This we may observe 

 in most cathedral windows. 



It is not asserted, that this was the origin of the pointed 

 style of Grecian architecture ; it is merely mentioned to shew 

 the prevalence of uniformity in its parts : and in general it may 

 be observed, that there are none of these parts, nor any mould- 

 ings or subdivisions of tracery, but may be referred to circular 

 poles or stems of young trees ; and let me add farther, that 

 were there more intricacy or more contrasts in the columns, 

 their character would be materially injured by the consequent 

 harmony in place of simple variety ; the eye would be too 

 much fixed, and the mind prevented from conceiving with suf- 

 ficient facility the essential idea of roundness. From this, and 

 a number of other arguments which might be produced, the 

 theory which I have suggested may serve as a good general 

 guide, whether in design, or in guarding against innovation in 

 this style, under the idea of improvements in the general forms, 

 or in the composition of mouldings. Errors of the latter kind 

 are common ; an example of which must strike a critic in 

 Gothic architecture in the recent alterations of Hereford Cathe- 

 dral. In an altar-skreen in that of York also are introduced 

 Gothic battlements ! which have not the least connexion with 

 the internal finishing of cathedrals, especially when of so large 

 a size as is there used : it destroys all unity of effect, and pro- 



