196 Xant>5cape Brcbitecture 



divide the run into equal parts, another uniformity 

 would be added without removing the former; for 

 regularity always suggests a suspicion of artifice, 

 and artifice detected no longer deceives: our imagin- 

 ation would industriously join the broken parts and 

 the idea of the broken line would be restored. 



" Whatever break be chosen, the position of it must 

 be oblique to the line which is to be broken. A 

 rectangular division produces sameness; there is no 

 contrast between the forms it divides; but, if it be 

 oblique, while it diminishes the part on one side, 

 it enlarges that on the other. Parallel lines are liable 

 to the same objection as those at right angles: though 

 each by itself be the perfect line of beauty, yet, if they 

 correspond, they form a shape between them, whose 

 sides want contrast. On the same principle forms 

 will sometimes be introduced less for their intrinsic 

 than their occasional merit, in contrasting happily 

 with those about them: each sets off the other, and 

 together they are a more agreeable composition than 

 if they had been beautiful, but at the same time more 

 similar. One reason why tame scenes are seldom 

 interesting is, that, although they often admit of 

 many varieties, they allow of few, and those only 

 faint, contrasts. We may be pleased with the 

 number of the former, but we can be struck only by 

 the force of the latter. These ought to abound in the 

 larger and bolder scenes of a garden especially in such 

 as are formed by an assemblage of many distinct and 

 considerable parts thrown together; as when several 



H>M 



