290 



NATURE 



\^Fcd. 10, 1876 



visited, will be partly rubbed off on to the lip of the stig- 

 matic cavity (/ Fig. 85). All the other contrivances of 

 the flower are nearly the same as in the large-flowered 

 form of V. tricolor, described in detail in a previous article 

 (Nature, vol. ix. p. 47). 



It may be worth mentioning that in the lower Alpine 

 localities (for instance, near Valcava, 1,500 m., and near 

 St. Gertrud, Sulden, 1,800 to 1,900 m. above the sea-level) 

 I found a variety of V. tricolor, which, as well in the con- 

 spicuousness of its flowers as with regard to its fertilisers, 

 is intermediate between the large flowered form of V. tri- 

 color (Nature, vol. ix. p. 46, Fig. 15) and V. calcarata 

 (Fig. 82). The flowers of this variety, which is called 

 alpestris, are 25-30 mm. long, and 18-22 mm. broad ; the 

 three lower petals are yellow near their base, as in 

 our tricolor and calcarata, marked with black streaks 

 converging towards the entrance of the flower ; the two 

 upper petals are very variable in colour, white, or bluish, 

 or yellow, with a large bluish margin. The spur is also 



Fic. 82. 



Fig 8j. 



Fig. 82-83.— Kw/a calcarata.^ Fig. 82.— Whole plant, showing a flower in 



front, natural size. 

 Fig. 83 — ^The same flower, laterally viewed, showing the long slender spur. 



variable in length, but on an average remarkably longer 

 than in our V. tricolor. I found the flowers of this variety 

 frequently visited by butterflies {Polyommatus virgaiirecE, 

 L., P. Jiippotho'c, L., var. eurybia, Ochs., $, Lycaena 

 semiargus, Rott., Ar^ynnis pales, Hesperia serratula, 

 Ramb.), but only once by a humble bee {Bomius ter- 

 restris, L., -^ , sucking),, whilst our V. tricolor is generally 



I a anthers ; a^, upper anther ; a^, insertion of the removed lateral stamen ; 

 ai, lower anther ; rt/i, appendage of the upper sepal ; /', beard, i.e., tuft of 

 hirs; cb, the lateral rurface of the stigmatic knob; c'.cS, orange-coloured 

 appendages of the cjunectives ; fi^,fii, filaments ; k, knob of the stigma ; /, 

 lip, labiate appendage of the stigmatic opening; w, nectary, ?.^., honey- 

 secreting appendage of the lower filaments ; ov, ovary ; p, petals ; /', lower, 

 /=, lateral, /3, upper petal ; po, pollen-collecting hairs ; fr, protective hairs 

 (Sprengel's " Saftdecke") ; ^, sepals ; s^, upper sepal (with the appendage 

 "/') ; ■r^. lateral sepal ; sp. the uppermost part of the spur, contaicing the 

 honey ; st. stigmatic cavity ; str, streaks converging towards the entrance 

 of the flower (Sprengel's " Saftmal ") ; sty, style ; y, yellow part of the 

 lower petal. (The rest of the corolla is blue.) 



visited by Apida, more rarely by butterflies, and V. cal- 

 carata exclusively by butterflies. Thus Viola tricolor, 

 var. alpestris shows us one of the steps by which the 

 common form of this species may have been gradually 

 modified into V. calcarata. 



Another Alpine flower, remarkable from its conspicuous- 

 ness and adapted to Lepidoptera, is Lilium bulln/ertun 

 (Fig. 86-88), which I found on stony slopes of the Schan- 

 fick valley, near Chur, in the Spoel Valley, near Zernetz, 

 and, somewhat more frequently, in shelving meadows of 

 the valley of Sulden, beneath the Ortler (1,700-1,800 m. 

 above the sea-level). Although in most points of its 

 structure agreeing with Lilium Martagon, def cribed in my 

 article X. (Nature, vol. xii. p. 50), this flower may be of 

 some interest, because it shows by what slight modifica- 

 tions a sphingophilous species may be adapted to diurnal 

 Lepidoptera, or vice versd. The number and arrange- 

 ment of the parts of the flower and the structure of the 

 nectary (Figs. 87, 88) are, indeed, the same in L. btilbi- 

 fet um as in L. Martagon. That, nevertheless, the latter 

 is cross-fertilised by'SphingidEC, the former by diurnal 

 Lepidoptera, is proved by the following differences : — 



I. The flowers of L. Martagoti, being dark reddish 

 brown, and in the daytime but faintly scented, are only 

 slightly attractive to day-fliers, whilst during the evening 



Fig. 84. — Entrance to the flower : seven times natural size. 



they emit a very attractive sweet odour. L. biilbiferuvt, 

 on the contrary, by the unusual size of its flowers, with a 

 diameter exceeding 100 mm., and by their colour appear- 

 ing very splendidly red in the sunshine, is most conspi- 

 cuous in the daytime, even from a great distance ; but being 

 but slightly scented, is incapable of efficaciously attracting 

 night-fliers either by its odour or its colour. 



2. The flowers of L. Martae:on are generally bent down- 

 wards, and its sepals and petals reflexed so far as to place 

 the entrances to its nectaries m a nearly horizontal direc- 

 tion, its stamens and pistils projecting downwards, with 

 only their ends slightly bent upwards (Fig. 63, Nature, 

 vol. xii. p. 50). Thus it affords no landing-place, and offers 

 its honey exclusively to such insects as are capable of 

 inserting a long slender proboscis into the flowers, while 

 they hover in the air by very rapid movements of their 

 wings. L. bulbiferum, on the other hand, having its 

 flowers obliquely upright, and offering to their visitors a 

 commodious standing-place on the lowermost petals or 

 sepals, the honey of the lowermost nectaries is accessible 

 to every insect the proboscis of which is long and slender 

 enough to be inserted into the honey-secreting channel. 



3. Cross-fertilisation by visiting Sphingidae is effected 

 in Lilium Ma>-tagon by the pistil overtopping the anthers, 

 and therefore being first touched by the legs and under- 

 side of the visitors, and thus smeared with the pollen of 

 flowers previously visited. In Z. bulbtferuvi cross-ferti- 



