56 



FLORA AND SYLVA, 



on an edge rockery, and very difficult to eradicate ; but it has a near ally named 

 G. lancastriense, said to be a variety of it. It seems to be found wild nowhere 

 but in Walney Island off Furness in Lancashire, and is entirely different from 

 the type, both in colour and habit. It does not run underground or root deeply, 

 but spreads moderately, and is easily restrained, growing close on the surface. 

 The pale pink flowers are quite as large as those of the type, and I consider 

 it an excellent rockery plant. It seems to come quite true from seed, and, from 

 a gardener's point of view, it is a separate species. I have never seen a white 

 variety of it or any intermediate form. 



G. silvaticum^ a rare native with pale purple or bluish flowers — rarely white 

 — and an erect bushy habit, is a fairly good shrubbery plant. There is said to 

 be a double variety, but what I have seen for this seems identical with the 

 double G. pratense. For dry shady corners where not much else will grow, the 

 dusky G. phceum may find a place, the flowers vary from nearly black to dark 

 brown. The seedlings must be weeded round it, or it soon forms a little forest 

 too dense to flower. G. E?idressi has flowers of a gay pink, which it produces 

 for at least three months. It is not out of place on a rockery, creeping close 

 amongst the stones almost like ivy ; but in such a position it requires to be con- 

 fined within limits. It is a useful plant, not becoming shabby at any time or 

 place, and, as it roots as it runs, it is very easily managed. It does well amongst 

 stones in a semi-wild place. Similar in leaf but having a dark spot in the middle, 

 and not running, though multiplying freely from self-sown seed, G. striatum 

 may be mentioned, suitable for any dry corner. The flowers, small for the size 

 of the plant, are nearly white, beautifully veined with dark purple. This and the 

 last mentioned are the only two which I think cross in gardens, for, much as 

 their types differ, I have found plants amongst them which belonged to neither, 

 but seemed intermediate in character. 



G. macrorhizo?i, sometimes offered as G. balkamim, has large radical tufts of 

 deeply lobed leaves with a rather disagreeable odour. It spreads rapidly by means 

 of close lateral tufts. The flowers grow on twin stems, each branch having a 

 terminal bunch of four or five flowers of red purple ; the petals are veined with 

 lighter stripes. The cultivation requires no instructions. I omit to mention seve- 

 ral other kinds of very moderate merit which I grow or have grown in the mixed 

 borders, but three rockery species must be mentioned, G. argenteumfi. cinereum, 

 and G. subcaulescens. The first is one of the choicest of alpines and deserves 

 the best place that can be given to it, amongst the upper ledges of the stone 

 heap, where there is plenty of rock and little soil. It grows only a few inches 

 high with shining silvery leaves, and forms a compact stemless tuft covered with 

 flowers, large for its size, with veined petals varying from pink to white. In a 

 suitable place a plant will live for several years, but can only be renewed by seed, 

 for which the birds are eager competitors. Often a seed flies into a little crevice 



