PROPAGATION 145 



the case of S. Cotyledon and its varieties, 

 hybrids of *S. longifolia, S. Griesbachii^ and 

 others of a Hke habit, the rosette from which 

 the flower-spike springs dies after flowering, 

 but round it will appear a number of offshoots, 

 each of which will have developed roots by 

 the autumn. They can easily be separated 

 from the main stem, and will have formed 

 sturdy plants by the following year. 



The Androsaces also are easily increased 

 by division. Take a plant of Androsace 

 Chumbyi: after flowering, a number of small 

 woolly rosettes will appear attached to the 

 parent plant by thin wire-like stems. Each 

 of these in a month or more will develop 

 roots and form a complete plant which can 

 easily be separated. Androsace lanuginosa^ and 

 its variety Leichtlini^ will, on careful examina- 

 tion, be found to have made roots on some 

 of their trailing stems, and these will readily 

 grow when separated from the parent plant 

 and potted. 



Few families lend themselves better to 



division than do the Primulas ; in fact, it is 



essential to their successful cultivation that 



10 



