148 LANDSCAPE DESIGN 



If his object is to unify the whole surface, so that its total extent 

 and its main form are at once to be seen, the slight elevations and 

 depressions which he may make for variety in the surface will be re- 

 lated sequentially to each other : they will be plainly all undulations 

 and modulations of the same surface. The designer may even take pains 

 in this case that no part of the surface is concealed from the eye : he 

 may depend for interest and variety on the apparent differences in tex- 

 ture and color caused by the slight differences in the relation of the 

 angle of slope to the view of the observer. If the landscape be seen 

 in the low light and long shadows of morning or evening, such slight 

 variations in surface will give great differences in appearance and may 

 quite sufficiently distinguish part from part without interrupting the 

 whole. If, however, the designer wishes to have a more striking and 

 definite difference between certain parts of the surface, he may do so 

 by making a nearer rise conceal a farther hollow. If the line of sight 

 over the top of the rise misses the ground beyond by only a few inches, 

 still it may be enough to bring out the silhouette of the nearer ground 

 against a background considerably more remote. If then either por- 

 tion of the composition is in light when the other is in shadow, the 

 difference may be very marked, so marked perhaps as to change the 

 composition from one of sequential modeling of surface to one of har- 

 mony of masses and contrast of light and shade. 

 Banks There are many cases where the landscape architect is obliged to 



make rather steep slopes for a short distance, not primarily for the sake 

 of the landscape appearance but to provide for a road or some such 

 structure.* If the purpose of the designer is to make the whole com- 

 position as natural as possible, he may diminish the incongruity of such 

 grading in two ways. He may so arrange it that from the important 

 viewpoints the line of sight strikes first on a naturally modeled sur- 

 face on the hither side of the depression and strikes again on a naturally 

 modeled surface beyond it, the observer being left to infer that what 

 he does not see between these two points is like the surfaces which he 

 does see. It may be, however, that the re-graded surface cannot be 

 concealed from the observer. This will plainly be the case, for in- 



* Cf. article by J. C. Olmsted, The Treatment of Slopes and BankSy in Garden and 

 Forestf Sept. 5, 1888, with cross-sections. (See References.) 



