Feb. lo, 1876] 



NATURE 



291 



lisation would be prevented if the sepals and petals were 

 as much reflexed as they are in L. Martas;on ; for butter- 

 flies would sit down on them and suck the honey out of 

 the channels at their base, without touching the stigma or 

 anthers. But in this species only the ends of the sepals 



and petals are spread apart, whilst, as far as the sexual 

 organs extend, the leaves of the perianth diverge bat so 

 slightly that a butterfly, when inserting its proboscis into 

 the nectary, can scarcely avoid touching the stigma and 

 anthers ; and, the pistil being situated nearest to the 



Fig 85. 



Fig. 85.— Lateral view of the flower after the halt of its sepals and petals and one of the two lateral anthers have been removed and the undcrlip 

 somewhat depressed : seven times natural size. Fig. 85*. — Transverse section of the spur, behind the line m n. Fig. 85. 



lowermost petals and sepals on which the butterflies alight 

 and suck, the stigma here also will be commonly first 

 touched and thus fertilised by pollen-grains of flowers 

 previously visited. 



Although, by the contrivances now described Lilium 

 bulbiferum, from its very conspicuous flowers, is very likely 

 to be cross-fertilised by butterflies in sunny weather, still 

 in rainy periods many flowers may wither without having 

 received any visit from a butterfly. Hence the possibility 

 of self-fertilisation appears to be indispensable both to L. 

 bulbiferuin and to L. Mariagon. In both the anthers 



and stigma are simultaneously developed to maturity, and 

 are often found in contact with each other ; and self- 

 fertilisation may thus be effected in case cross-fertilisation 

 by visiting Lepidoptera is wanting. 



4. Direct obser\'ation of the visitors proves that Z. 

 Martagoti is really fertilised by Sphingidce, for instance 

 by Macroglossa stellaiarivn, as observed by myself (see 

 Nature, vol. xii. p. 50), and by Sphinx eiiphorbice, as 

 observed by Federico Delpino ;' and that L. bulbtfenun 

 is really fertilised by butterflies, for instance by Polyom- 

 viatiis virgaurecF, L., P. hippothoe, L., var. eurybta, Ochs., 



//.f 



Fig. 86-88 — Lilium bulbiferum. Fig. 86. — Lateral view of the newer after 

 the half of the perianth has been removed ; natural size, a, anthers ; 

 St, stigma; n, nectary. 



and Ar^yn7iis aglaia, L,, all of which (July 20, 1875) I 

 found repeatedly in the flowers, not only sucking the 

 honey, but also resting, displaying their wings in the sim- 

 shine, and sometimes a male of Palyornmatus virgaures, 

 L., sitting by the side of a female of the same species. 



Fig. 87. — Basal portion of one of the leaves of the perianth, f, entrance into 

 the nectary ; n, nectary : magnified twice. 



It is a Striking fact that all these species of butterflies 

 are of nearly the same splendidly red colour as the flowers 

 they visit. I do not know whether this fondness has 

 been effected by natural selection, agreement of colour 



I Written to me in a letter of May 1873. 



