LILIACEAE 



443 



Herm. Miiller (Vosges and Alps), the humming-bird hawk-moth (Macioglossa 

 stellatarum Z.); also lo other sp. of Lepidoptera in the Alps. Knuth (gardens at 

 Kiel) saw the same hawk-moth flying continuously from flower to flower, and 

 pollinating a large number. Delpino observed a hawk moth, probably Deilephila 

 euphorbiae L. 



2723. L. bulbiferum L. (Sprengel, op. cit., p. 189; Herm. Miiller, ' Alpen- 

 blumen,' pp. 45-7 ; Focke, ' Beob. an Feuerlilien ' ; Neubert, Tagebl. 52. Vers. D. 

 Natf. in Baden-Baden, 1879; Kerner, op. cit., II, p. 461; Knuth, op. cit.) — The 

 flowers of this species are adapted for pollination by butterflies. In spite of the 

 flame-coloured perianth, which gleams very brightly in sunshine, the odourless 

 blossoms very rarely attract butterflies to the nectar abundantly secreted in the nectar- 

 grooves of the perianth leaves. Anthers and stigma mature simultaneously and are 

 at the same level ; the latter is slightly bent down below the former. A lepidopterid 



Fig. 395. Liliiini bulbi/erufn, L. (after Herm. Miiller). A. Partly dissected flower (x 5). 

 B. Basal part of a perianth leaf (x ij). C Transverse section through the base of a perianth leaf 



( X 54). h^ nectar-groove ; hi}^ hairs closing do. ; /, ridges bearing do. ; w, nectary ; j/, stigma. 



alighting on the lower perianth-leaf and probing thence for nectar will first brush 

 against the stigma and then touch the anthers, thus regularly effecting cross- 

 pollination. 



Should insect-visits fail, automatic self-pollination is sometimes possible by 

 contact of anthers and stigma ; this, however, is rare or ineffective. Neubert says 

 that the species is always self-sterile. Focke (Ost. Bot. Zs., VVien, xxvii, 1878) 

 describes the variety croceuvi Chx. as self-sterile, and infertile even when dusted with 

 pollen from plants of the same origin, being only fertile when the pollen comes from 

 those of different origin. Crossing is effective in the variety Buchenavii Focke, which 

 as regards fruit is a sub-species between L. bulbiferum and L. croceum. In explana- 

 tion of this behaviour on the part of croceum, Focke supposes that the self-sterile 

 plants have all arisen from a single specimen by vegetative means. Kerner states 

 that the variety sets fruits and gerniinable seeds regularly, but produces no bulbils in 

 ihe axils of the foliage leaves ; this occurs almost regularly in the type species, which, 



