90 



ON HYBRIDIZATION AMONGST VEGETABLES. 



scarlet ■£ cardinalis mule. Few of their ovules were at all fer- 

 tilized, and the greater part of those were chaff, but a few appa- 

 rently good seeds were amongst them, which will probably 

 vegetate. G. Gandavi itself has ripened its seed three successive 

 years, and one from the first batch of its descendants is now in 

 flower. It preserves the cross-bred type, and might be accounted 

 a distinct species, if I did not know that it was raised from 

 oppositiflorus by pollen of Natalensis. It reverts, however, a 

 little towards the male parent, the purple stripes of the female 

 parent being less strongly marked than the original mules, and 

 the flower is scarcely as large or bright coloured, following the 

 course I have observed in other cases, that seedlings from a 

 cross-bred plant by its own pollen sometimes degenerate in the 

 size or brilliancy of their flowers. 



We must next turn to the Gladioli of the northern hemi- 

 sphere, there being, however, a plant of intermediate position, 

 G. eequinoctialis, on the heights of Sierra Leone. The northern 

 Gladioli are all purplish, with a tendency to rose-colour, and in 

 a few cases to blue, excepting the whitish and the white vari- 

 eties. They peremptorily refuse to breed with the Cape species ; 

 and, although I will not say that the cross is impossible, I have 

 failed in so many attempts that I have abandoned them. But 

 although the northern species differ somewhat from those of the 

 Cape territory, and agree with G. Natalensis in a more direct 

 presentation of the flowers to the front from an erect stalk, there 

 are a great many different local forms of them, with a great 

 general similarity of aspect and intermediate forms, which almost 

 defeat the attempts to distinguish them specifically, but furnish, 

 with a similarity of flower, a strange diversity of seed — the 

 winged or foliaceous margin of the African species being con- 

 spicuous in Byzantinus, communis, Boucheanus, and some 

 others ; totally disappearing in G. segetalis, Fischerianus, and 

 some others ; curtailed in some varieties, and almost obsolete in 

 others, of G. communis. The gradual curtailment of that margin 

 in varieties of communis, as well as the close resemblance of 

 G. segetalis to them, shows that the separation of the latter as a 

 genus is preposterous. But there is another strange circum- 

 stance connected with them, which tends to show how species 

 originate. G. Byzantinus will grow and increase greatly in 

 almost any soil or situation ; G. segetalis is very apt to die at 

 Spofforth — I supposed it tender and fearful of wet. The white 

 Gladiolus commutatus of Bouche, communis albus of the Dutch, 

 but in truth a white G. segetum, always dies at Spofforth — I 

 believe they perish because the soil, however good for barley, is 

 too light for them. Near Trieste and in Santa Maura G. sege- 

 talis engrossed strong yellow ploughed land that lay flat and 



