AND THEIR CULTURE. 63 



colour of many of the former, the white petals had completely resisted 

 the foreign influence, but the yellow anthers had undergone a marked 

 change, about half of them were turned to a chocolate colour, approaching 

 that of the anthers of the male parent, though not so deep. I determined 

 to try the effect of fertilisation in the third generation, and applied the 

 pollen of both Auratum and Speciosum to ten or twelve of the young 

 hybrids whose organs of reproduction appeared to remain perfect ; not 

 one of them would bear seed. In the present summer (1877) I repeated 

 the experiment on a larger scale, and fertilised about fifty flowers, after 

 removing the anthers before they ripened. Nine of these produced 

 seed pods, all of which were small and deformed except two, these two 

 contained, along with the chaff, a few seeds of promising appearance, the 

 remaining seven were full of chaff alone. The reproductive power had 

 been nearly destroyed by hybridisation repeated through three genera- 

 tions. What will result from the few seeds obtained remains to be 

 seen. Some* Lilies refuse to be fertilised by the pollen of certain other 

 Lilies. Thus, I have found that Speciosum, so readily fertilised by 

 Auratum, will give no seed to the pollen of either Brownii, Longiflorum, 

 Canadense, Tenuifolium, or Umbellatum, yet the converse does not always 

 hold true for several of these last named species will bear seed when 

 fertilised with the pollen of Speciosum. 



" A great number of subtle influences may modify the results of 

 experimental hybridisation, yet those described above were so various, 

 and extended over so many years, that the general facts to which they 

 point may, I think, be regarded as assured. An eminent botanist has 

 suggested to me that the tenacity with which Lilies fertilised by Lilies of 

 other species retain their characteristics unchanged may be explained by 

 supposing that the offspring are really no hybrids at all but results of 

 parthenogenesis, that curious phenomenon which sometimes occurs in the 

 lower order of animals, and by which a single impregnation continues to 

 take effect in several successive generations ; in other words, that a Lily 

 of which the flower was fertilised in any one year by its own pollen, may 

 bear seed in the next year, without being fertilised again. There are two 

 good reasons for believing that parthenogenesis had nothing to do with 

 the examples in question. In the first place, the Lilies subjected to experi- 

 ment were young plants that had never bloomed before ; in the next 

 place, every species fertilised by me with the pollen of another species 

 showed, with the single exception of Superbum, evidence of hybridity 

 which, though slight, was convincing. This evidence consisted in 

 markings of the stem resembling those of the male parent, in a changed 

 colour of anthers, also resembling that of the male parent, and in the 

 frequent occurrence of abortion in both anthers and pistils, with con- 

 sequent sterility. That the seedlings were really hybrids there can be no 



* I should not expect to obtain a favourable result by intermarriage between such 

 dissimilar forms as Auratum (Archelirion) and Longiflorum (Euliriori). Success, I 

 think, will more probably be obtained by hybridisation between members of the same 

 group, as, for instance, Thunbcrgianum and Oroceum, Cwididum and Washingtonianum, 

 Longiflorum and Brownii. Among the Martagon family between Chalcedonicum and 

 Martagon Album or Szovitzianum, between Excelsum and Maximoiviczii, and between 

 Kramevi and Speciosum in the Arcliclirion group. 



