382 SCIENCE OF GARDENING Part II. 



the mansion; Whatever may be the surface destined for a court or square of buildings, 

 as a stable-yard or farmery, it must be reduced to a plane or planes connected in such a, 

 way as not to interfere with utility or effect. It is not essential that the surface be 

 formed to a perfect level, or to any one slope, but that order and connection should enter 

 into the choice of the slopes, whatever that may be. In kitchen-gardens it sometimes 

 happens that a level, or one general slope, may be adopted ; but much more frequently 

 that different slopes enter into the composition of the enclosed surface. These subordi- 

 nate planes or surfaces are all so connected as to balance and harmonise, and present to 

 the intelligent eye a work, not of chance, but of design and reflection. In a seemingly level 

 garden it often happens that not one of the compartments is level ; but each compartment 

 of itself forms one plane, diverging from the centre, north wall, or some other point of the 

 garden, and terminating on the same level, at the extreme corners of the compartment, or 

 at the lower extremity of the garden. Besides these means, the formation of raised bor- 

 ders, and the furniture of gardens, such as espaliers, bushes, &c. enable the designer 

 to harmonise forms and surfaces seemingly the most incongruous and unsuitable for a 

 scene of culture. 



1 9.55. There are two modes of reducing an irregular surface to one plane. The first is 

 by taking sections of the surface in parallel lines at every ten or twenty feet distance, 

 according as the surface may be more or less irregular ; laying down these sections on 

 paper geometrically, and from the whole finding a mean section. The stakes of all the 

 parallel lines of levels still remaining in the ground, it will be easy to transfer the mean 

 section -by raising these stakes in some places, and lowering them in others, as the scale 

 of the diagram will direct. The second and more general mode is by approximation, or 

 trial and correction, which, in all ordinary cases, is sufficiently correct. Suppose an irre- 

 gular surface, 100 feet square, is to be reduced to a level or plane. The degree of slope 

 is first ascertained (by the American or any other level) from the highest side of the 

 square tp the lower, and it is found, we shall suppose, that the ground will not easily 

 reduce to a horizontal surface. It is, therefore, determined to reduce it to a slope ; and 

 for this purpose a certain height is determined on by the eye for the extremities of the 

 slope ; in fixing on which, the object is to adjust the slope to the earth, so as the former 

 may be completed without exterior aid or superfluity. Supposing the lower side of the 

 plot to be twenty-five inches below the level of the upper side, then the fall is a quarter 

 of an inch in each foot, and a few lines of stakes can be run across the ground in the 

 direction of the slope, with their tops adjusted to this declivity. Or this may be omitted, 

 and the same end attained by borning-pieces used after the ground has been roughly 

 levelled. But this is one, among many parts of the business of a gardener, which can 

 more readily be acquired by practice than verbal instruction. 



1956. Walks are spaces in gardens formed for the purposes of inspecting the garden, 

 recreation, and carrying on the operations of gardening. As one great requisite is, that 

 they should always be dry, the bottom of the walk in most cases forms a drain. There 

 are three descriptions of walks common to gardens, those of gravel, sand, and grass. 

 All walks consists of two parts, their substrata and surface-covering. The substratum 

 is generally placed in an excavation, the section of which is a segment of a circle, or an 

 inverted pointed arch, being deepest in the centre, where, in wet soils and situations, a 

 notch or drain is often formed to carry off the water which oozes from the sides of the 

 bottom, or sinks through the gravel. In all ordinary cases, however, the water will run 

 off" without this notch, provided the general levels of the bottoms of the walks or the drains 

 which cross them, or lead from them, be contrived accordingly. The foundation of the 

 walks is to be filled with stones, the largest at bottom ; or with rubbish of old buildings, 

 flints, or any other similar materials, observing always to place the smallest at top. When 

 this is done, before the covering of gravel, sand, or turf is laid on, the substratum 

 should be well rolled, so as it may never afterwards vary its position, either with the 

 weight of the covering, or any weight which may pass over it. 



1957. The covei-ing of gravel 373 

 (jig. 373. a) need seldom be thicker 

 than six inches, and generally four 

 inches will be sufficient. That this 

 gravel may bind in so thin a stratum, 

 it is requisite that it be free from 

 larger stones than those the size of a 

 pigeon's egg, that the general size be 

 that of large gooseberries or plums, and that there be about a sixth part of rusty sandy 

 matter to promote its binding. The choice of gravel is seldom within the power of the 

 gardener ; but, in general, pit-gravel is to be preferred to river-gravel, as binding better, 

 and having a better color. Gravel abounding in oxide of iron, if laid down where it is 

 finally to remain, when newly taken out of the pit, and well watered and rolled, will often 

 bind into one compact body like what is called pudding-stone. Such gravels, however, 



