COLOURS OF FLOWERS. 175 



Veitcli's fine Californian introductions. In this flower there 

 is no sign of blue. Yet, if there is a genus more pre- 

 eminently blue tlian any other cj'anic race, it is surely 

 Delpltinium.^^ 



It is true that some species have never yet transgressed 

 their bounds, so that Dahlias still refuse to be bkie now as in 

 1845 ; and we are still ignorant of the reason. 



The effect of nutrition upon the colours of plants is well 

 known, in that they vary much more in a garden soil than 

 in the wild state ; and differ in colouring according to the 

 character and ingredients of the soil. Thus, as described by 

 a writer in Hovey's Magazine of Horticulture* striped Dahlias 

 will be best kept clean by planting them in a poor soil, while 

 rich soil invariably runs them. E.g. JD. var. striata formosis- 

 sima : No. 1 was planted in a poor gravelly soil, in an open 

 situation ; all the flowers but two were beautifully mottled. 

 No. 2 was planted upon a rich, cool, sandy loam ; not one- 

 half of the flowers w^ere mottled. No. 3 consisted of three 

 plants, very highly enriched ; every bloom but one was self- 

 coloured. Similar effects follow on the variegated foliage of 

 Pelargoniums, according as they are grown in a too rich soil 

 or light one.f 



" Alum is said to render the Hydrangea blue ; and some 

 saline substances, such as phosphate of iron and muriate of 

 ammonia, appear to brighten the tint of red." J It often 

 happens, however, that blue and pink corymbs occur on the 

 same plant of Hydrangea. A cutting taken from a blue 

 Hydrangea growing at Southampton, and transferred to 

 Bedfont, changed to the usual colour on blooming there. § 



Chloride of lime has been known to make a whole-coloured 



* Quoted in the Gard. Chron., 1842, p. 8. 



t Gard. Chron., 1876, p. 567. % Ihid., 1843, p. 577. 



§ Ihid., 1886, vol. xxvi., p. 118. 



