THE THEORY OF GARDEN DESIGN 149 



problem is simplified when he has to decorate a given 

 space instead of painting pictures at large. 



^Esthetic problems are always most successfully 

 solved when they are not purely aesthetic; and it is be- 

 cause the problem of garden designs has become purely 

 aesthetic that it now seems so difficult. If the designer, 

 instead of asking himself where he should place his 

 herbaceous border and where his rock-garden or his 

 rosary or his plantations of flowering shrubs, were to 

 consider how best he could contrive places of coolness 

 and shade for the summer and sheltered sunny walks 

 for the winter, he would find that his aesthetic and 

 horticultural problems were beginning to solve them- 

 selves. Flowers he would use as decoration and, 

 using them so, he would soon discover a principle 

 for their arrangement. Trees and shrubs he would 

 employ mainly for use, to give shelter and shade; 

 and therefore he would avoid the random planting 

 of them now so common. He would also avoid ex- 

 cesses of formalism, since he would not clip those 

 trees or shrubs that were planted for shade, but only 

 those which needed clipping that they might grow 

 close for shelter. He would be very sparing in his 

 use of what are called " ornamental conifers," now 

 so often misused by formalists and naturalists alike. 

 He would not plant a row of Thujas in front of a yew 

 hedge because he wished to advertise the fact that he 

 was a formalist; nor would he dot them anyhow about 

 a lawn for no reason whatever. Monkey Puzzles he 

 would leave for gardens where there are monkeys to 



