ON HYBRIDIZATION AMONGST VEGETABLES. 



99 



roarers, and very little, if any, evil occur from breeding by a 

 roaring- horse. Many years ago Mr. Milne, of the Fulham nur- 

 sery, obtained three mules, which are well known, from Passiflora 

 racemosa set with the pollen of coerulea. The produce did not 

 require stove cultivation, like the female parent, but they have 

 been generally sterile. From their first distribution to the pre- 

 sent day they have grown in my cool conservatory, flowering 

 abundantly, but sterile, with this exception, that many years 

 ago one shrivelled, pulpless fruit was formed and ripened, con- 

 taining twelve good seeds, which vegetated. The flowers of all 

 were nearly similar to those of P. coerulea, the male. One of 

 them is growing in the same conservatory, with the flower, I 

 think, rather finer than that of the common coerulea, and it has 

 never borne fruit. The rest having been planted or left out of 

 doors at different times, have been killed by frost or neglect. 

 At the time when this dry fruit was ripened, there was a plant of 

 P. coerulea in another house in the garden, though disconnected, 

 and eighty or ninety feet distant. Therefore, either the flower 

 was fertilized by pollen of coerulea brought by an insect or by 

 accident, and two crosses by the pollen coerulea. made coerulea 

 itself from the ovary of the tender scarlet racemosa, or the mule 

 was fertilized by its own pollen, and the offspring diverged to 

 the male type, throwing off" the similitude of the female alto- 

 gether. The natural coerulea did not fruit here while I pos- 

 sessed it, and it dies when planted out at Spofforth. 



I have already stated the possibility of raising the poetic nar- 

 cissus by two or three crosses from a daffodil, and I have also 

 shown that the Gladiolus crossed from G. oppositiflorus (which 

 breeds freely with the Cape species) by pollen of G. Natalensis 

 (which will not), produces seedlings, one of which having 

 flowered, reverted a little towards the male type. Here then we 

 have the like fact. The Passifloras were produced by the pollen 

 of the mule, or by accidental access of the pollen of the male 

 parent. If by the former, these widely-separated species are 

 convertible, and a new form originates from their union ; if by 

 the latter, the male type may be obtained by repeated crosses 

 from the ovary of a very different plant; and, whichever be the 

 case, the origin from one created kind is proved in that instance, 

 and, by implication, in all cases of similar difference. I believe 

 all the Cape species of Gladiolus to be convertible ; I have found 

 no positive impediment. The rare G. abbreviatus might be 

 almost made between G. cunonius and tristis ; it has the curious 

 leaf of the latter, and a flower approaching to that of G. cunonius. 

 Mr. Plant's mule consists of the two, with the addition of G. 

 cardinalis. Perhaps I have forgotten to mention that Corbularia 

 will not breed with the other Narcissi. I have crossed the com- 



h 2 



