96 



ON HYB1UDIZATION AMONGST VEGETABLES. 



the various species of Fuchsia; but the natural forms of F. 

 virgata, gracilis, globosa, and discolor, appear to me so superior 

 to the mixed shapes produced by gardeners, that I have been 

 unwilling to deteriorate them by intermixture. In one Fuchsia 

 of a mixed race I was struck with the appearance of the green 

 colour of the fruit of F. fulgens, but more vivid, though in all 

 other respects it seemed to conform with those that bear red 

 berries. As the natural green fruit of Fuchsia is agreeable to 

 the taste, it is possible that a hardier green-fruited race may be 

 obtained with the same quality. The supposed sterility of mules 

 has very much deterred cultivators from trying how far plants 

 that bear palatable and wholesome fruit may be intermixed, and, 

 excepting Mr. Knight's experiments, very little has been done 

 in that branch. Our climate does not suit experiments, to ascer- 

 tain whether the lemon, orange, shaddock, citron, and lime are 

 indefinitely convertible, and, if not, exactly what insurmountable 

 impediments occur ; but I believe no one has even tried to 

 blend the very highly-flavoured Fragaria viridis with the larger 

 and more fruitful kinds, and gardeners adhere to the chance of 

 improved seedlings from the most approved individuals, though 

 they are aware that size, without flavour, is unsatisfactory. It 

 is remarkable that in some genera bearing eatable fruit the 

 crosses are usually fruitful, and in others not — a circumstance 

 which requires deep investigation. The few mule Passifloras 

 raised seem indisposed to make seed, and still more to fill the 

 fruit, if formed, with succulent pulp ; but it does not follow 

 that the case will be such with all. P. quadrangularis bears a 

 large, rich-flavoured fruit in a stove under peculiar treatment ; 

 edulis, a better-flavoured fruit in a greenhouse; laurifolia and 

 maliformis very different fruit in a stove ; a small species sent 

 to me by Mr. Maclean, from Chorillos above Lima, under the 

 name of suffruticosa, bears a delicious fruit of the size of a 

 gooseberry in a stove ; P. alata, nearly akin to quadrangularis, 

 has a vile fruit ; casrulea, no better ; but it remains to be tried 

 whether no hardier species of passiflora will give a fertile, suc- 

 culent, and well-tasted fruit by intermixture with the best, 

 tenderer kinds. In the genus Cactus it is so. Well-flavoured 

 mules are obtained at once from C. Ackermannianus, of which the 

 fruit is very bad, by crossing it with C. speciosissimus or Phyl- 

 lanthoides. It is remarkable, that in the section Cereus the 

 mules are as fruitful, and have the fruit as juicy, as the natural 

 forms, however dissimilar. Nothing can be more unlike each 

 other in the same genus than the two species I have last named, 

 yet they breed willingly together, and the fruit of the mule 

 differs from that of either parent as much as the flower. I stated 

 (Amar. p. 345) that I could not see a single point in the generic 



