Partial Variability and Vegetative Propagatiun. 147 



by the same causes. This alone would suffice to show 

 how intimately the two phenomena are connected. 



A further fact, to which much too little attention has 

 been paid, is the gradual accumulation of some peculiarity 

 by means of selection of a character in a plant propagated 



Figs. 30 and 31. 



Othonna carnosa, la grown in moist earth and 16 in dry; 

 the former with long leaves forming no rosettes and very 

 strong; the latter covered with rosettes of short stout thick- 

 leaves.^ 



Othonna crassifolia. Ila grown in damp soil and 116 in dry, 

 lie a flower, lb and 116 were branches hanging over the 

 edge of the pot. 



by vegetative methods. An example of this has been 

 given by Darwin :^ 



The well-known English plant-breeder Salter ef- 

 fected a considerable improvement in certain variegated 



' Othonna carnosa is very like O. crassifolia, but has consider- 

 ably larger leaves and somewhat larger flowers. But its relation fo 

 damp air and so forth is the same as that of O. crassifolia. 



^ J Variations of Ajiimals and Plants, I, p. 443, 444. 



