ON HYBRIDIZATION AMONGST VEGETABLES. 



in 



character of Echinocactus to separate it from Cereus, and that 

 I expected them to be found able to intermix. I observe that it 

 was once proposed to take the stem as the distinctive generic 

 feature, unless it should be thought better to combine them in 

 one genus. To that I say, that, while species of Cereus disagree, 

 in that some of them have fleshy, angular, thorny stems, and 

 others flat, leaf-like, smooth stems, but are proved to be of one 

 genus, we have not sound grounds for separating Echinocactus 

 on account of such a feature ; and I contend that the separation 

 is a question of fact, not of option and preference, as if the 

 botanist were a commissioner to set out the fences for an enclo- 

 sure of waste lands. I have very few Echinocacti, and have 

 had no opportunities of testing the point ; but I requested Mr. 

 Beaton, a most intelligent gardener, when he had the care of 

 Mr. Harris's rich collection, to do so, and he very soon sent me 

 seed from the mule Cereus called Jenkinsoni by Echinocactus 

 Eyresii, but they were only outwardly perfect, and did not 

 germinate. If he had remained longer there, I think he would 

 have effected the union. Mr. Beaton produced the cross between 

 Ribes sanguineum and one of the yellow-flowering species, which 

 brings that genus to my recollection. The gooseberry and red 

 or white currant are held to be of one genus, but it has not 

 been found practicable to blend them, and probably it is not ; 

 but it does not follow, because they will not breed together 

 directly, that they might not through some other species, as in 

 the genus Gladiolus G. cardinalis and Natalensis refuse to mix 

 with each other, but do freely with G. oppositiflorus. Since both 

 these crosses with Oppositiflorus make seed, I consider that, by 

 perseverance, the impracticable cross can be nearly forced, by 

 reducing both crosses one step further from the dissentient 

 parents, till the two crosses are found able to intermix, and then 

 crossing the produce upwards, step by step, with a greater ap- 

 proximation to them. I am not aware that any attempt has been 

 made to cross the acid with the sweet currants, which might so 

 perhaps yield a new and pleasant fruit, and the attempt should 

 be made. It is very difficult to foresee which combinations, 

 when made, will produce a fruitful offspring. The genus Nerine 

 strongly exemplifies that difficulty. It consists of two portions, 

 those with regular corolla and straight filaments and style, and 

 those which have them bent and the corolla distorted and in- 

 clined. There is a cross section of the genus not tallying with 

 the above-mentioned distinctions, viz. those with centripetal 

 inflorescence, i.e., beginning to flower on the outside of the 

 umbel and flowering last in the centre, and those with centrifugal 

 inflorescence. The cross between N. curvifolia of the first 

 section and pulchella of the second is very fertile, both being 



VOL. II. H 



