I20 LANDSCAPE DESIGN 



stances, circles may be turned into ellipses in this way, squares made 

 oblongs, without the deviation from the simpler form being apparent. 

 Where a formal grove, a wall or gate, even a building sta^ids be- 

 tween one open area and another, the best development of the shapes 

 of the open areas may produce a distorted shape of the intervening mass. 

 In the case of the grove, this may of course readily be managed. In 

 the case, however, of so definite things as buildings and even gateposts, 

 many instances can be found, particularly in Italian work, of excellent 

 effects" produced by curiously distorted shapes, which can be known 

 to be distorted only by measurement, since from no one place can the 

 two unrelated sides be seen at the same time. It is extremely difficult 

 to judge whether two straight lines upon the ground are exactly parallel, 

 if they are nearly so and converging in perspective. It is, if anything, 

 still more difficult to judge whether one line meets another at right 

 angles on the ground, unless one line is carried by the junction point 

 so that the eye may judge the similarity of the two supplementary 

 angles. It is therefore easy with proper precautions to construct a 

 garden or other formal design which shall appear to be symmetrical 

 and rectangular while really being neither. If a comparison of the 

 measurements which might betray the deceit is skillfully made impos- 

 sible or difficult, the departure of the total shape from symmetry may 

 be very great without being noticed. In part this is due to the fact 

 that a man is very likely to take it for granted that a formal scheme 

 is symmetrical and rectangular, and only some striking discrepancy will 

 call a contrary fact to his attention. One out of many notable instances 

 of this is at Isola Bella, in the break in the direction of the axis of the 

 scheme on the stairway, which is so arranged that no comparison of 

 the direction of the two lines can be made. The change of direction 

 is about seventeen degrees, but almost every one assumes that the 

 whole scheme lies upon one axis. 

 Illusions of It is possible therefore by slightly converging the boundary lines of 



^^^ a garden, perhaps by making paths somewhat smaller at a distance 



than they are near at hand, to give an exaggerated appearance of 

 length to a garden through its apparently great diminution in perspec- 

 tive. A similar exaggeration of the effect of size may be brought about 

 in informal design by using trees in the distance which, while apparently 



