386 



objects that tliey are " thudly, to have a variety 

 in the direction of the parts, but fourthly, ta 

 have those parts not angular, hut melted as it 

 ti'ere into each other, '^ the question is whether 

 this principle (for it is only one) can be ap|)lied in 

 any, and in what degree to the temple of Tivoli ; 

 which, as it is well known, is a circular building, 

 surrounded by columns in the same direction. 

 The forms of temples, as indeed of all buildings, 

 may be divided into two general classes : circular 

 or round ; and square or angular : the second, 

 by far the most numerous, is excluded by the 

 words " not angular." The principle therefore, 

 if applicable at all, must be applied to round 

 buildings ; and if the spirit of what Mr. Burke 

 has said be attended to, I believe it will apply as 

 much as can be expected in such cases : for the 

 lines in all circular objects have a perpetual, though 

 uniform variation ; and as they are constantly and 

 insensibly retiring from the eye, they answer to, 

 the description of " melting as it were into eacK 

 other," much more than the liiws in square, that 

 is, in any other buildings. 



I must here make my reader acquainted with 

 some dextrous manoeuvres of my antagonist. The 

 principle in discussion, as 1 began by remarking, 

 though divided into t^wo parts, is only one ; for it 

 is obvious that if you take the third part singly, 

 without the limitation in the fourth^ you totally 

 pervert !Mr. Burke's manifest intention. This, 

 however, is precisely what Mr. Knight has done ; 



