38 



liant colours and varied forms would be 

 fully reflected. 



These are a few of the numberless va- 

 rieties, which it is within the compass of 

 art to imitate : they nevertheless have sel- 

 dom, if ever, been tried in the style, or for 

 the purposes that I have mentioned ; not 

 even those which arise from planting. But 

 as rocks with cascades, have been imitated 

 with success, there can be no diflaculty in 

 placing trunks, or roots of trees, or in 

 imitating many effects of stone, or of 

 rocks, on a smaller scale; especially where 

 there is no motion to disturb them. With 

 regard to the tints of soil, if sand, or any 

 rich-coloured earth, be placed where it will 

 be supported by stones, roots, or ledges 

 of rocks, as it often is in nature, it will 

 probably remain undisturbed ; as there 

 would be no current, or flood to affect it. 



In all I have written on the subject of 

 improvement, one great purpose has been 

 to point out the aflinity between land- 

 scape-painting, and landscape-gardening ; 

 in this case, the affinity is very close in- 



