COMPARISON WITH LANDSCAPE PAINTING. 219 



of improvement. It is true tliat in these seeming disad- 

 vantages there is also some real compensation. If his 

 broad terrene canvas is not so pliant and submissive to 

 his will as that of a painter^ — if he has not so clear a 

 field and so absolute a command over his objects^ — 

 yet the very difficulties and peculiarities of the locality 

 are often suggestive of his most admirable results. 

 Sir Uvedale Price^ in one of the notes to his well- 

 known work^ records the practice of Zucchi, a celebrated 

 Roman painter of castellated scenery. It was the method 

 of that artistj first to dash off a large rock^ the more 

 diversified and picturesque the better^, and he then pro- 

 ceded to build his towers on this crag and on that_, ac- 

 cording as he could find foundations for them; and it 

 was generally observed that he was successful in his 

 castles in proportion as he had been happy in construct- 

 ing their rocky bases. Here it is easy to perceive the 

 process by which pictorial effects were suggested to the 

 mind of the painter. More subtLl_, and therefore more 

 untraceable, may be the operation of suggestion, in the 

 practice of Landscape Gardening, but not less certain is 

 its influence. We beheve that many of its happiest 

 effects have flowed from this som-ce ; and on the other 

 hand it would not be difficult to recount some lamentable 

 instances of failure and error arising from neglect of this 

 principle. The practical conclusion is, that the artist 

 should never venture to design, much less execute, till he 

 has made himself thoroughly acquainted with the natural 

 and acquired characters of the groimd. 



Again, the Landscape Painter has to deal with no 

 more than one view at a time ; and he can not only 

 place the spectator at the point most favourable for con- 

 templating it, but by certain modes of treatment he can 



L 2 



