98 PB ACTIO AL GUIDE TO GABDEN PLANTS 



danger of displacement, it is better not to have to tread upon the 

 plants. 



' The " Sunk Rockery." — This is perhaps the best of all, but entails 

 rather more labour in construction. Where subsoil drainage is perfect, 

 a sunk walk may be made, not less than 10 to 12 feet wide, with sloping 

 sides. The sides may be faced with stones, as described in the " faced 

 rockery," and all or part of the excavated soil may be made into a raised 

 mound, continuing the slopes of the excavated banks, above the ground 

 level, and thus combining the facing rockery and the barrow rockery. 



' If the outer line of this portion above the ground be varied by small 

 bays, every possible aspect and slope may be provided to suit the taste 

 of every plant. However, unless drainage is perfect, a sunk walk, 

 rising to the ground-level at each end, would not be feasible. But a 

 broad walk, excavated into the side of a hill and sloping all one way, 

 could be adapted to a structure nearly similar to that described ; or the 

 ground may be dug out in the form of an amphitheatre to suit the taste 

 or circumstances. 



' But whatever the form of the rockery adopted, let the situation be 

 away from the influence of trees, beyond suspicion of the reach of their 

 roots below, or their drip, or even their shade, above. Trees which 

 shelter from only high winds are so far serviceable, and so are walls 

 and high banks. There are few alpine plants for which a storm-swept 

 surface is good, but trees are objectionable where they lessen the light, 

 which is an important element in the welfare of most mountain plants. 

 The shade and shelter afforded by the stones and form of the structure 

 itself is the best kind of shade and shelter. 



' Soil for Alpine and Rock Plants. — We now come to the subject of 

 soil, which is very important, though I attach less importance to it than 

 others do who have written on the subject. I hold that where atmo- 

 spheric and mechanical conditions are favourable, the chemical combina- 

 tion of the soil is of secondary consideration. 



' It is true that in nature we find that the flora of a limestone moun- 

 tain differs in many particulars from that of a granite mountain, and 

 on the same mountain some plants will thrive in heavy retentive soil 

 while others will be found exclusively in peat or sand. But for one 

 who is beginning to cultivate alpine plants to have to divide them into 

 lime-lovers and lime-haters, lovers of sand and lovers of stiff soil, is an 

 unnecessary aggravation of difficulties. 



' So large a proportion of ornamental plants are contented with the 

 soil which most cultivators provide for all alike — even though in nature 

 they seem to have predilections — that where an amateur has only one 

 rockery it would be too perplexing to study the partiality of every plant, 



