350 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



white, yellow, and purplish flowers ; while R. maritimus flowers are 

 yellow, and scarcely veined. 



From the foregoing abbreviated descriptions it will be seen how 

 all the characters relied upon are variable, as Bentham observes; and, 

 judging by the figures of the siliquas, they certainly are constricted, but 

 in that of Tournefort the constrictions are much reduced under culti- 

 vation, so that it does not appear surprising that they should vanish 

 altogether. R. maritimus, being a South European type, will account 

 for the tenderness sometimes shown in the radish; so that the general 

 result appears to be that radishes have been raised in many countries j 

 from the local sub-varieties of this variety of R. Raphanistrum.'^ 



Fig. 122. — Wild Eadish {Eaphanus Ea2)lianistruin). 



"With regard to the two principal forms of the root of the garden 

 radish, the long, spindle-shaped and the globular turnip-formed, 

 M. Oarriere describes his experiments with the wild species Raphanus 

 Raphanisirum, L. (fig. 122), growing them from seed in. two very 

 different soils during five consecutive years. Some were grown at Pans 

 in a light soil, others in the country in a firmer or strong soil, argilla- 

 ceous and calcareous. At Paris the long form prevailed almost entirely. 

 It was just the contrary'- elsewhere ; the tu])erous form was most abun- 



* The last seven paragraphs are quoted from the Gardeners^ CJironicle, 

 June 25, 1898, p. 389. 



