336 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



lessly in their rock gardens merely because tliey are found in the Alps. ,i 

 This is another instance of the unintelligent imitation of nature. The j 

 white Veratriim or the Globe flower often look well enough on the slopes i 

 of a great mountain, but near the smaller alpine plants in a rock garden i 

 they look as incongruous as Madonna lilies or German irises. Docks 

 and nettles grow astonishingly high among the Alps, but no one, I t 

 suppose, would therefore put them in a rock garden. It is useless ; 

 to waste bold rock-work upon large lowland plants. The plants will 

 dwarf any rocks possible in our gardens. If they are grown in the 

 rock garden at all they should be upon its outskirts ; and so should 

 large shrubs like Cistus cyprius or Cistus laurifolius, or the larger ! 

 brooms or Barberries. In my opinion the smaller plants should be 

 grown among the larger and more massed rocks, as we find them in i 

 the Alps. Here, too, should be the smaller alpine shrubs. Where 

 rocks are massed boldly little plants do not look meagre, and the rocks ! 

 are a foil to their delicate beauty. If some easy-growing plants are to ^ 

 be grown among the smaller alpines, so that vegetation may not seem 

 too sparse, the best and safest are the smallest houseleeks, such as | 

 Sempervivum arachnoid eum and S. Laggeri. These, though they in- 

 crease fairly fast, will not smother plants near them; and they do not 

 seem out of scale with the smallest alpine. The alpine toadflax is a 

 dangerous plant for the purpose, as it grows at a great pace in wet 

 summers and will cover a small neighbour very quickly. 



If once we make up our minds to observe scale in our planting we 

 shall find the problems much simplified. Scale does not depend so 

 much upon the size of a rock plant as upon its whole character and the 

 size of its leaves. Thus Lithospermum prostratum and Hypericum 

 reptans are not out of scale with small alpines, however large they may 

 grow. But Campanula carpatica is out of scale because it has large 

 leaves and the habit of a herbaceous plant. It again is a plant for the 

 outskirts of a rock garden, and so are the larger pinks, because their : 

 flowers are borne on long stalks and their whole scale is large. 



As in building rock-work the main lines should be laid dowm first 

 with large rocks, so in planting the main effect should be secured first 

 with drifts and masses of the more easily grown plants, combined with 

 appropriate shrubs. Such drifts should follow the lines of the rock- 

 work, and where one drift passes into another the plants may be mixed 

 so that the change may not look too abrupt. It is worth noting that 

 prostrate and low-growing plants look much better in long drifts than 

 plants which have upright stalks. A line of these always looks artificial 

 on rock-work. Like Lombardy poplars in a landsca}^, they should be 

 planted singly so as to make a sharp contrast with the masses of pros- 

 trate plants. Nothing, for instance, could look more ineffective than a 

 long unbroken line of Saxifraga pyramidalis in flower; nothing better 

 than occasional heads of this varying a drift of Lithospermum pro- 

 stratum. If it is massed at all it should be massed not in a drift but in a 

 compact patch behind some drifts of a prostrate plant, so that its stalks 

 will bend ovfer and make a cloud of blossom. And here I may remark 



