ON WATER. 175 



" when varied in their shape and accom- 

 " paniments, have the same effect as forest 

 " thickets, circular islands that of clumps ; 

 " and the same system which gives rise to 

 " round distinct clumps, of course, produces 

 " islands equally round and unconnected. As 

 " the prevailing idea has been to show a 

 " great uninterrupted extent, whether of grass 

 " or of water, islands on that account have 

 " been but little in fashion : I have, indeed, 

 " seldom seen more than one in any piece of 

 " artificial water, and that apparently made 

 " rather for the sake of water-fowl than of 

 " ornament. When one of these circular 

 " islands is too near the shore, the canal 

 " which separates them is mean, and the 

 " island from most points appears like a pro- 

 " jection from the shore itself ; and when, on 

 " the other hand, it is nearly in the centre 

 " (a position of which I have seen some very 

 " ridiculous instances), it has much the same 

 " unnatural, unmeaning look, as the eye 

 " which painters have placed in the middle 

 " of the Cyclops' forehead ; and that is one 

 6C of the few points on which the judgment 

 " of painters seems to me to be nearly on a 

 " level with that of gardeners : they have 



