THE VARIABILITY IN CULTIVATION OF HARDY FLOWERING PLANTS. 227 



A. apennina, though their flowering time overlaps. A. alpina 

 and A. sulfur ea keep quite true to colour from seed. In the 

 same way Ranunculus is a constant genus, though botanists 

 have given the ambiguous name "hybridus" to a good and 

 constant alpine species. Gentiana is constant ; G. acaulis and 

 G. verna, G. septemfida and G. asclepiadea, which flower 

 together side by side, continue true. I should a year or two ago 

 have said the same of Poppies ; but I have a puzzling set of 

 Poppies, in which P. orientate seems to have crossed with 

 P. rupifragum, which is the seed-parent. These are absolutely 

 barren, and are still under investigation. 



(4) The fourth of the questions I raised, whether cultivation 

 increases or decreases the seed-bearing power of flowers, admits- 

 of no general answer. Seed-bearing often depends upon a 

 combination of accidents independent of cultivation. An un- 

 usually hot and early season, a change of soil or position, may 

 often alter our conclusions. Luxuriance of growth is not in 

 itself, as some suppose, an obstacle to fertility, unless it causes 

 the flowers to be double ; and it is well known that even in some 

 forms of doubling the seed-bearing organs retain their power. 



Looking to my own experience of plants remarkable in Edge 

 garden for luxuriance of growth : Lilium Martagon, often 6 feet 

 high, with fifty flowers ; florists' Delphiniums, 10 feet high ; 

 Veronica longifolia, 7 feet high — all these continue fully fertile. 

 For obvious reasons, early-flowering varieties are often fertile, 

 and late-flowering forms barren in the same species ; though, 

 on the other hand, some plants flower continuously through 

 summer, and only the latest flowers produce any seed. Of 

 varieties of the same species, one may produce plenty of seed 

 every summer, w T hilst another is persistently seedless. 



Two well-known garden plants, supposed to be luxuriant 

 developments produced by cultivation, but both of them of 

 obscure history, may seem to favour the belief that luxuriance 

 tends to sterility ; for I have tried in vain for many years, both 

 in my own garden and in others, to find a seed on either of 

 them. One is called Helianthus multiflorus ; it exists in several 

 forms, both single and double, and is said to belong to the species 

 H. decapetalus. The other is a beautiful and very robust herba- 

 ceous Veronica, the best of its class in cultivation. It is re- 

 ported to have come from Japanese gardens, and is referred by 



