PURPLE ROCK CRESSES, 



i 8 I 



rience is the true guide, for there are dry- 

 spots in which the long stems of old 

 plants lose freshness, which is best re- 

 newed by topping them; in others with 

 a moist atmosphere, the plants never suf- 

 fer in this way, and grow too dense if 

 cut, with bad results in a wet season. To 

 spare the plants, it is well where seed is not 

 wanted to clip off the withered flowers, 

 and this often results in a crop of young 

 shoots suitablefor cuttings. In heavy soil 

 it is a help to raise the shoots from the 

 damp ground upon a bed of light stone 

 or broken clinkers. 



There is now a good range 

 Thei LF s f for ofcolourintheAubretias, 



Effect. . > 



from pure, rich purple to 

 mauve, rose, and pale flowers approach- 

 ing white. According to 

 their shade of colour they 

 may be grouped with 

 charming eflect among 

 Wall-flowers of rich red 

 and yellow, clumps of 

 Daffodils, Narcissus, and 

 others of the legion of 

 spring bulbs, or they may 

 be mingled with tufts of 

 Rockfoils, Sandworts, and 

 other Alpine plants in the 

 rock garden. A deep co- 

 loured kind varied with 

 little colonies of Hut- 

 chinsia is very pretty. For edgings to 

 plant borders the more compact and 

 upright kinds should be used, such as A. 

 Campbelli and purpurea, in mauve and 

 purple, and A. Leichtli?ii in rose. An 

 attractive spot to clothe with these pur- 

 ple hangings is the face of a sunk fence, 

 with a backing of good soil behind . I call 



to mind such an one, bordered on either 

 hand with a solid bank of purple flowers, 

 stretching away in a long winding vista, 

 spanned here and there by an arching 

 pergola, with just enough of variation 

 to prevent monotony, and wondrously 

 beautiful in the mingled play of light and 

 shade ; yes, and a second, marking the 

 entrance to a wayside house-of-call, to 

 which the passers throng — who knows 

 how largely for the joy of passing in re- 

 view those sheets of colour which other- 

 wise were banks of naked soil. 



There has been some ad- 



As to Kinds. • i i i 



vance with the Aubretia 

 of late years, and many named kinds, 

 mostly seedling forms of A. dehoidea, 

 are now offered ; in fact, there are more 



AUBRETI AS IN THE ROCK GARDEN. 



names than distinct kinds, though some 

 fine seedlings have been raised. These 

 garden varieties are of greater value than 

 the seven or eight wild kinds of doubt- 

 ful distinctness, and probably only forms 

 of one species. In our brief descriptions 

 we therefore neglect artificial distinc- 

 tions between kind and kind. Since all 



m 3 



