March 29, 1877] 



NATURE 



473 



FER TILISA TION OF FL O WERS B Y INSECTS ^ 

 XVI. 

 Alpine Species of Gentiana adapted to Lepidopiera. 

 fiENTIANA BAVARICA (Fig. 106-108), G.verna 

 ^-^ (Fig- 109- 1 11), G. nivalis (Fig. 11 2- 114), of which I 

 have examined living specimens in the Alps, and G. imbri- 

 caia, cestiva, pumila, and utriculosa, of which I have 

 examined only dried specimens,* agree so completely in 

 the structure of their flowers, and in their contrivances 

 for cross-fertilisation by insects, that they all obviously 

 belong to the same section of the genus Gentiana, and are 

 adapted to the same group of visiting insects. They all 

 possess a long narrow corolla (Fig. io6),^ which contains 

 in its lowermost portion the honey, secreted, as in our 



Fig. 106. 

 Fir.s 106-108. — Gentiana bavarica. Fig. io6. — Lateral view of a flower 

 which begins twisting and closing, the calyx and the anterior portion 

 of the corolla having been removed (3J : i). Fig. 107. — The same 

 flower viewed from above (3J : i). The margin of the circular stigma 

 is already covered by the revolving corolla. FiG. 108. — The stigma seen 

 from above (7 : i).4 



two first groups, by an annular swelling at the base of the 

 pistil 5 («, Fig. 106, III, 113). They all have the entrance 

 to the corolla-tube closed by the bi-lobed stigma being 



f Continued from p. 319. 



* For specimens of G. pumila, asliva, prostrata, Froelichii, and pur- 

 purea, I am indebted to Prof. Ascherson of Berlin. 



3 Only in G. cestiva the corolla-tube seems to be considerably wider than in 

 the other species. 



* The following explanation of the lettering applies to all the figures : — 

 a = anthers, co =: corolla, Ji =: filaments, n = nectary, o = openings con- 

 ducting to the honey, ov = ovary, p = petals, s = sepals, st — stigma, y = 

 folds of the corolla by which its twisting is made possible. 



5 I found the length ot the corolla-tube, from the nectary to the stigma, in 

 the species in question as follows : G. cestiva, 26 28 mm. ; G. vema and var. 

 brackyphylla, 23 ; G. bavarica, 20-22 ; G, utriculosa, 18-22 ; G. piimila, 

 x6-i8 ; G. imbricata, 15 ; G. nivalis, 13-16 mm. 



dilated into a circular disc {st. Fig. ic6, 109, 112), and 

 bordered at its margins with hair-like papillae (Fig. 108}. 

 In all of them small openings are visible in the fully- 

 opened flower between the margin of the stigma and the 

 inclosing corolla (Fig. 109) ; but as soon as the corolla 

 begins twisting (Fig. no), these openings are con- 

 cealed. In all of them the anthers surround the 

 stigma, thus placing their pollen on the way to the honey 

 {a, Fig. 106, 1 10, 114). They can all, therefore, be ferti- 

 lised only by such insects as have a proboscis sufficiently 

 long to reach the base of the corolla, and at the same 

 time either sufficiently slender to enter through the small 

 openings (Lepidoptera), or sufficiently strong forcibly to 

 enlarge the entrance of the flower i^humble-bees). Such 

 an enlargement, indeed, would be possible by the expan- 

 sion of the same folds between the petals {/. Fig. 106, 107, 

 109, no, 112), by which the flower is enabled to twist, 

 and to close, as soon as colder weather frightens away its 

 natural fertilisers. 



Now, looking about to discover what Lepidoptera and 

 humble-bees might be the natural lertilisers of the present 



Fig. 107. 



group, I was surprised at the fact that of all the 162 

 species of Lepidoptera which I had observed visiting 

 flowers in the Alps, Macroglossa stellatarum alone has a 

 proboscis sufficiently long (25-28 mm.) to reach the honey 

 in all the species of Gentiana in question, and that of 

 the Alpine humble-bees, even those provided with the 

 longest proboscis of all, Bombus hortorum 

 (18-21 mm.) and B. opulentus, Gerst. (22 

 mm.), are incapable of reaching the honey 

 in all the above-named species, except by 

 thrusting the whole of their head into the 

 narrow corolla-tube. Moreover, direct ob- 

 servation seems to prove that humble-bees 

 insert at the most their proboscis into these 

 flowers, but never their whole head. For instance, in 

 the Albula Pass, July 28, 1876, I saw a humble-bee flying 

 about for a long time in search of flowers ; at length it 

 hastily visited Gentiana vema, but having only once 

 thrust its proboscis in^o a flower, it flew away out of my 

 sight. Likewise near Pontresina, August 4, 1876, I saw 

 Gentiana nivalis visited by Bombus meftdcar, Gerst. $ , but, 

 after having hastily tried one or two flowers in the same 



Fig. 108. 



