BLASTOGENIC VARIATIONS. 163 



tinct importance as a source of variations. Thus Ben- 

 nett,* in a paper on Hybridity in Plants, makes the fol- 

 lowing remark: " There seems, however, scarcely to be 

 room for doubt that in some of our abundant wild 

 genera, such as Rubus, Salix, and Hieracium, hybridity 

 is not uncommon in nature. It has long been known 

 that in some genera, such as Passiflora, and in some 

 Orchidese, the ovules appear to be even more readily 

 fertilised by pollen of a different species. W. Focke 

 now states that this is also the case with the species of 

 Lilium belonging to the group bulbiferum, and with 

 some species of Hemerocallis; and J. H. Wilson affirms 

 the same respecting the Cape genus Albuca, also belong- 

 ing to the Liliacese." Again, Rolfe f states that Nar- 

 cissus incomparabilis is known to be wild in France, and 

 that Herbert found that on crossing a daffodil with pol- 

 len of N. poeticus, the seedlings yielded flowers identi- 

 cal with those of N. incomparabilis. Similarly, by 

 crossing the Daffodil with the Jonquil, Herbert suc- 

 ceeded in producing N. odorus. Again, Engleheart 

 has proved the hybrid origin of N. Uflorus by crossing 

 N. poeticus with the pollen of N. tazetta, he obtain- 

 ing seedlings identical with wild forms. Also he 

 reconstructed N. pulclidlus Salisb., by crossing 

 N. triandrus with the Jonquil, the seedlings proving 

 absolutely identical with the wild plant. Further 

 " Digitalis supplies some wild hybrids whose origin has 

 been artificially demonstrated. For example, D. pur- 

 purascens, Roth, has been reconstructed by crossing 

 and recrossing D. lutea and D. purpurea; and D. media, 



*Nat. Sci., vol. ii. p. 208. 



\ J. Roy. Horticult. Soc., xxiv. p. 181, 1900. 



