222 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY, 



garden for at least half a century, flowering not only in neglected 

 corners, but in the richest flower borders. Yet cultivation 

 never called forth in it its power of producing orange-coloured 

 flowers, until I introduced, five years ago, the orange-coloured 

 variety from the Pyrenees, where it is common. Both yellow 

 and orange seedlings are now produced by it. 



I may state generally that I have never observed in my 

 garden any distinct change in colour in plants raised from seed 

 of the normal type which could not be accounted for by some 

 cross with another plant in the garden. 



(8) The third class of changes we observe in cultivation is 

 the most frequent and the most remarkable. Without any 

 intentional crossing having been practised, we find that seed- 

 lings show characters differing from those of the seed-parent. 

 This may arise sometimes from the development due to cultivation, 

 but it is caused in most cases by a spontaneous cross with the 

 pollen of some different variety of the same species, or by spon- 

 taneous hybridising with a different species. The results of 

 these crosses are so various, and the mixed characters so 

 puzzling, that botanists generally dislike committing themselves 

 to name garden flowers, but if pressed to give a plant a specific 

 name they naturally choose that to which the plant shows the 

 nearest affinity, and I find that this is as likely to be the pre- 

 sumed pollen-parent as the seed-parent. It often happens also- 

 that in gardens two closely allied species seem to combine into 

 one, their offspring losing the distinctions which the plants 

 maintain in their respective homes. I shall mention several 

 cases where this seems to be the case. I find that no rule can be 

 laid down concerning the fertility of these hybrids. Some seem 

 entirely barren ; others produce fertile seeds sparingly ; whilst 

 in some genera they seem quite fertile, though seldom constant- 

 There is often a tendency to approach nearer to one parent in 

 each successive generation, affording a presumption that the 

 hybrid is fertilised not by its own pollen, but by that of one of 

 its parents. I have rarely seen an instance in which two 

 distinct species combine into an intermediate garden form which 

 is constant from seed, and sufficiently different from either 

 parent to deserve to be retained as a garden species. Lychnis 

 Haageana, if really a hybrid, is a noteworthy example of such a 

 plant. Under this head I speak of plants which have sprung up 



