GROUND, ROCK, WATER 147 



However well the rockwork may be done, it will still depend on the Rock Planting 

 planting for its success. It is essential not only to choose plants which 

 naturally live — or at least can live — In rocky places, but to preserve 

 a scale relation between rock and planting which will not dwarf the 

 rock, and a texture-contrast so that each material may accent the 

 effect of the other. (See again Plate 27 and also 35.) The intended 

 planting must be borne in mind when the rocks are arranged, not only 

 so that proper loam-spaces may be left for the plants, but so that when 

 the planting is finally set out it shall not overpower or conceal rock- 

 compositions which were apparently striking enough before the plant- 

 ing was introduced. 



As any rockwork can at best be only suggested on plan, nothing 

 but skillful, patient, practical superintendence will give results worthy 

 of consideration, and no mean ability in visualizing the final result is 

 required of the superintendent who can see his way efficiently and di- 

 rectly through the confusion and structural exigencies of the early stages 

 of the work to a final consistent natural eff^ect. 



In the informal modeling of the ground surface which the land- Minor 

 scape architect is called on to do in his usual practice, particularly in ^ !^^ "■' 

 his smaller work, he has seldom the untrammeled opportunity to ap- Surface 

 proximate to some one landscape character, as in the various examples 

 which we have just discussed. He is engaged rather in providing for 

 all the economic necessities, — roads, drainage, balance of cut and fill, 

 and so on. Nevertheless he suggests such natural character as he can, 

 — at any rate he displays the natural forms of his trees, shrubs, flowers, 

 and turf to best advantage ; and with all this he makes a ground surface 

 which shall be as interesting and as compositionally excellent as pos- 

 sible. Some general considerations relating to this kind of ground- 

 modeling are worth discussion. 



In any open area of considerable size which has only slight modula- 

 tions of surface, the line of view of the observer is almost certain to be 

 nearly parallel to the surface. A very slight elevation will therefore 

 probably conceal what lies immediately behind it, and on account of 

 this fact the landscape architect can greatly change the appearance of 

 the surface without any great change in the relative elevations of its 

 parts. 



