COLOURS OF FLOWERS. 181 



generation . . . the coloured portions begin to predominate. 

 ... I have never been able to observe a single instance of 

 variegation coming directly from the coloured original. The 

 contrary, however, takes place with regard to the dottings ; 

 these come directly from the coloured tjpe." The variegated 

 varieties the author had succeeded in fixing were Gomphrena 

 glohosa ; Antirrhinun Tnajus, Convolvulus tricolor, Neifnopliila 

 insignis, Porfulaca grandiflora and BelpMnium Ajacis. 



Other florists have found that by crossing whole-coloured 

 flowers pure white seedlings may result. 



Abutilons have an instructive and, in part, a somewhat 

 similar history. No hybrids were raised from the old bronze- 

 red and striped form, which was usually barren, until the 

 white "Boule de Neige " was introduced. Mr. George 

 crossed this with " Duke of Malakoif." The white one had 

 itself previously thrown up every shade of dingy white ; but 

 whether by being spontaneously crossed or uot, does not 

 appear to be known. Some of the colours of the seedlings 

 of this cross were pale and dark pink, pale orange, bright 

 carmine, salmon, orange-red, etc.* 



Somewhat analogous results w^ere obtained by Mr, Veitch 

 with Rhododendrons imported from Borneo. Thus a cross 

 between the larger-flowered B. Javanicum, which is orange- 

 coloured, with the smaller white narrow-lobed B. Jasmini- 

 florum, gave rise to the rose-coloured " Princess Royal." A 

 further cross of the last with the parent B. Jasminiflorum 

 eliminated the red colour ; the offspring, however, retained 

 the form and size of the corolla of the " Princess Royal." 

 It is called " Princess Alexandra." 



In the above-mentioned the effect of the white has been 

 to separate the tints; so that from the old Bronze-red 

 Ahutilon JDarwinii we get yellows and reds of different 



* Gard. Chron., 1878, p. 792. 

 18 



