FORMATION OF GARDENS. 117 



character of the soil and sub-soil, and the kind of material at 

 hand to form a surface. Where the ground is naturally wet, or 

 w}}ere there is a liability of the accumulation of water, the soil 

 should be taken out to the depth of at least twenty inches, — 

 the section formed by the excavation forming an obtuse angle 

 towards the centre, or forming the segment of a circle. These 

 excavations should lead into drains, at the lowest points, to carry 

 off the water that percolates through among the stones with 

 which they are filled. They may be filled to within two inches 

 of the intended surface of the walk, — the largest in the bottom, 

 and the smaller toward the surface. This forms a durable and 

 drv walk at all seasons; and, where the soil contains a consid- 

 erable quantity of stones, which have been thrown out in the 

 process of trenching, or the rubbish of building-materials, this 

 affords a good medium of getting them out of the way. 



On dry, gravelly ground, however, these excavations are use- 

 less, so far as drainage is concerned; and, shovelling aside the 

 mere surface-soil, the walk may be laid dow^n on the substratum 

 beneath it. If the walks are on a level, or nearly so, the water 

 generally finds its way off as quickly as it falls, and the cost of 

 excavation is saved. 



The surface of walks may be formed of grass, gravel, or sand. 

 Good gravel is the best, sand the very worst, and grass can only 

 be introduced with propriety in particular places. Sand, or 

 loose gravel, makes a very uncomfortable walk, and, when of 

 great length, is tiresome and disagreeable to walk upon. 



A very common error, among those not acquainted with the 

 proper method of making walks, is, to lay on too much surface- 

 material; and, in many places, we have seen trenches taken 

 out for walks, and filled, to the depth of a foot or more, with 

 gravel, which, if laid on a hard surface to the depth of an inch 

 or less, v/ould have made a good w'alk, but which, at such a 

 depth, all the walking, rolling, and pressing of years could never 

 make it bind. It requires more skill than is generally supposed 

 to make good walks. Among all the operations of the garden- 

 maker there is scarcely one which we are so much disposed to 

 find fault with, as in the making of walks ; and this is precisely 



