ON WATER. 157 



to its outline, instead of the straight line, or 

 uniform curve, which usually characterise it. 

 One bold promontory shooting into the 

 water would divide the length of the head, 

 especially where it is of considerable extent. 

 Great improvement also might, I conceive, 

 be made in the construction of the head, 

 by giving variety to its surface, instead of 

 making it a dead level, as is usually the case. 

 This variety of surface would, moreover, give 

 opportunity for planting near the water, which 

 an not be safely done on the level surface 

 for fear of injuring the puddle bank. 



It is seldom that the head can be con- 

 structed so as to unite easily with the ground 

 beyond it ; for which reason, the drive or 

 walk should not (if it can be avoided) pass over 

 it. Indeed, under no circumstances should 

 you be permitted to walk all round a piece 

 of water, as, its limits being thus betrayed, its 

 extent is ascertained ; whereas, when the 

 walk is so conducted as occasionally to come 

 upon the water, and that at the best points 

 of view, and to be constrained by the inter- 

 vention of planting, &c. again to leave it, 

 not only is the apparent extent, as well as 

 the variety, greatly increased, but the wish to 



