lEJSn GAEDENING 



3^ 



spring and summer, in full aim, iie^Jt close 

 under cloches, and the soil never allowed to 

 become dry. Watering is necessary, some- 

 times almost hourly. The cuttings root 

 tremendously fast, but a short delay in water- 

 ing, and good-bye to the lot! Cuttings taken 

 with a " heel " should have the heel slightly 

 l^ared down. Cuttings without a " heel '" 

 should l)e cut oft just below a " join " or ridge 

 where two leaves were sprouting. 



Laijci'niij. — .Most woody stemmed Alpines 

 can be layered, and if they are difficult as 

 " cuttings " layering should be tried. Select 

 stems near the ground, " snick " them with a 

 knife just below a " join," cut about half 

 through and then up at right angles, and peg 

 down the cut portion in a depression of light 

 sand}- soil, and cover well over with the same. 

 Some Alpines are very difficult to propagate at 

 all, Liinuiii salsuluidca prusfrata especially. 

 Some of the rarer AccnifJiuliinons and Erinacca 

 [)ungcn>< are almost equally difficult. With 

 them 1 have succeeded by a dodge I call " pot 

 layering."' Before " snicking " them the bark 

 on the side to be snicked is scraped with 

 a i^enknife for one-eighth of an inch and left for 

 a week. The snick is then made, but the upper 

 cut omitted (I do not know why some plants 

 object to the upper cut, but my experience is 

 that the harder the wood the more likely the 

 layer \^ill wither if upper cut) ; the layer is then 

 drawn tlu-ough a small thumb pot, imtil the 

 snick " is about i inch inside the pot. Crock, 

 sufficient to prevent soil coming out at bottom, 

 is put in, and the layer tightly potted with 

 sandy soil. The pot is then sunk and covered 

 with soil, as an ordinary layer, and usually, in 

 time, roots well, and the rooted layer is then 

 freed from the parent b}^ cutting off just below 

 the ]>ot. This is the only method by w'hich I 

 have succeeded with the Linum. 



Certain hardwoods — like dwarf Co/()H('((«fc/«, 

 Ericas, Andromeda irtraguna^ &c. — seem to 

 root even more readily if in moss, and I never 

 trouble to layer these, but just sprinkle moss 

 well over the branches on the ground and leave 

 it there, and they soon throw out roots. 

 CijUhus strikes badly, and varies from seed 

 but increases readily by grafting on seedling 

 r.aburnum in spring. Tufted plants, like 

 Dimiflnis, root readily if fine soil is worked 

 into the tufts as a top-cb-essing, and Drahas 

 likewise. ]\Iost Geraniiiin.fi can be increased 

 by root cuttings. Rare Semperviiuniis not dis- 

 ])osed to make offsets can be made to do so by 

 stripping of! the heart of the rosette, when 

 youngsters will be given off around the bare 

 stem. 



MURUAY HORNIBROOK. 



}'/iutu hii] 



'II. d- 5, Thompson 

 LaVATER.4 AR.BORE.4 



Pruning Roses. 



\\'iTii the excejjtion of Kamblers, ^larch is the 

 principal month for pruning Eoses. It is cus- 

 tomary to begin with the Hybrid Perpetuals 

 and Hybrid Teas, dealing with the Tea-scented 

 varieties towards the end of the month. To 

 give complete and absolute directions for ])run- 

 ing Eoses is a difficult matter on i)aper, so 

 uuich dei)ends on the vigour and age of the 

 Eoses. One rule may safely be observed, and 

 that is — all newly-planted Eoses should be cut 

 hard back to within a couple of buds or eyes 

 of the base, and weak shoots s'hould be removed 

 entirely. Any Eoses planted late this spring 

 should be cut back hard, whether they be 

 climbers or bushes. 



The general collection of Hybrid Perpetuals 

 should be dealt with according to the vigour of 

 each plant. All varieties do not grow equally 

 well in all gardens ; some flourish in certain 

 soils and localities better than in others, and 

 they must be pruned accordingly. Weak 



