<3 



printed in the first volume of the Transactions of the 

 Horticultural Society. In the fifth volume of the 

 second edition of the Hortus Kewensis, which was 

 published in 1813, the varieties of D. mperflua, there 

 named, are Purpurea, Lilacina, and Nana ; the latter 

 being taken from a double variety, figured in An- 

 drews' " Botanical Repository/ 5 but which is cer- 

 tainly not particularly entitled to be considered as a 

 dwarf plant. No varieties of D.frustranea are given 

 in the Hortus Kewensis. 



Mons. cle Candolle, in his essay on the genus, has 

 observed, that it is not probable we shall ever see a 

 blue one, since the variation is from purple to yellow. 

 He considers blue and yellow to be the fundamental 

 types of the colours of flowers, and that they mutually 

 exclude each other : yellows pass readily into red or 

 white, but never into blue : and in like manner, blue 

 flowers are changed by cultivation into red and white, 

 but never into yellow. (Hort. Trans, i. & iii., Gard. 

 and Flor. ii. 66.) 



Until about forty years ago, no variety was known 

 that did not possess a tinge of purple in its blossoms, 

 and it was even doubted whether a blossom entirely 

 untinged with purple could be produced. {Hort. 

 $oc. Trans, vii.) 



When Mr. Sabine wrote on the dahlia in 1818, the 

 single varieties only were abundant ; the number of 

 double ones was very limited, but they rapidly in- 



