98 



ON HYBRIDIZATION AMONGST VEGETABLES. 



centripetal ; that between N. curvifolia and undulata of the 

 second is absolutely sterile, because the latter differs in being 

 centrifugal. The mule, however, follows the male in being 

 centripetal, but is quite barren. A difference, therefore, which 

 had been overlooked by botanists till I observed it, seems to be 

 the outward mark of the impediment which makes the mule 

 unfruitful. Cross-breeding, therefore, shows that feature to be 

 more important than it had been thought. 



The blending of colours amongst cross-bred plants is rather 

 capricious. The golden-flowered Swedish turnip crossed with 

 the yellow-flowering white turnip did not give an intermediate 

 shade of colour, but some plants with golden and some with 

 yellow flowers. The orange and the blue Anagallis gave a 

 pale, dull pink. The scarlet G. cardinalis with a white G. 

 blandus, and with the yellowish G. angustus and G. oppositi- 

 florus, gave a purplish rose-colour. Rhododendron ponticum 

 and Rhodora Canadensis by the yellow Azalea very much dis- 

 carded colour, instead of taking an intense mixed hue. The 

 yellow and red-flowered Ribes, however, gave a dull rust-colour. 

 It generally requires two crosses from the scarlet to fix the 

 bright colour in the mule. G. Natalensis impresses its colour 

 much more strongly on G. oppositiflorus than G. cardinalis does, 

 being more nearly akin ; whether from that cause or not I will 

 not presume to guess. 



Hybrid Alstrcemerias have been raised, but with difficulty, 

 and they seem delicate. I believe it was a mistake to suppose 

 Mr. Van Houtte's fine collection of many-coloured Alstrcemerias 

 to be cross-bred. Poeppig had long ago informed us that A. 

 haemantha sported in the Subandine meadows of Antuco with 

 every shade, from white to citron, orange, rose-colour, vermi- 

 lion, and crimson ; and I believe his plants to have been ob- 

 tained by seed from that quarter, which I had long vainly tried 

 to obtain. I have raised a Bomarea between variabilis and 

 acutifolia, which seems to be fertile. I do not believe it would 

 be possible to cross a Bomarea with Alstrcemeria. I think I 

 may venture to say that, as far as I have observed, the prevailing 

 disposition of cross-bred vegetables seems to assimilate more to 

 the male than to the female parent, though the appearance may 

 possibly be sometimes the reverse, and often strictly intermediate ; 

 but, as far as I have seen, if we obtain a cross between a hardy 

 and a tender species, the produce, where the male is hardy, will 

 be much more hardy than where the female is hardy and the 

 male tender. This is very important and very conspicuous in 

 cross-bred rhododendrons. I do not think it is so with animals, 

 where the cases seem analogous. Having bred many horses, I 

 have found mares that were roarers almost invariably produce 



