SPECIAL METHODS OF PLANTING 77 



diameter and these mostly carry too large an amount 

 of root to be " poked " through an existing hole or 

 crevice satisfactorily. 



The opportunity of obtaining small seedling plants 

 is rare, and when one is so able unless the ancestors 

 of them are known, they are very liable to grow into 

 anything but the true longi folia, since these plants 

 hybridize so readily, that impure blood is frequently 

 met with, and such is very apparent in this case, since 

 longifolia when really true, unlike almost all other 

 members of the encrusted saxifrages, never produces 

 offsets, and reproduces itself solely from seed. It will 

 easily be understood that under these circumstances 

 unless the inflorescence is carefully gauzed, very 

 unreliable progeny will usually result. 



But to return to the formation of rock for the 

 reception of these plants. It follows that if plants with 

 a fair quantity of root are to be inserted in such a way 

 as to protrude through a crevice, the easiest way to 

 achieve that end is to place them in a position as the 

 face of rock is being built. To do this, I laid the first 

 course, of irregular stone pieces, on a well-rammed 

 foundation, and filled in behind them with very stony 

 soil, largely mixed with old mortar rubbish. 



The saxifrages were then taken from their pots, 

 the crocks removed, the lower roots carefully loosened, 

 and then the plants laid each into a chink or crevice 



