LILIACEAE 441 



elongates, so that its anther is at the same level as the still receptive stigma, and then 

 dehisces, while the other five remain shorter and have already scattered their pollen. 

 Automatic self-pollination by fall of pollen from the latter is excluded, because the 

 papillose stigmatic surfaces are situated on the inner side of the stylar lobes. 



Besides these normal flowers occasional blossoms with a gamophyllous perianth 

 occur at Wulfshagen. The circumference of these is the same from stalk to tip, so 

 that they are cylindrical ; they are easily distinguishable from normal flower-buds, as 

 the latter are conical in form. Humble-bees cannot pollinate these abnormal flowers, 

 as the entrance is too small to admit them. It is therefore possible that this is 

 a cleistogamous accessory type of the normal flower ; I have, however, been unable 

 to determine whether such flowers are fertile. 



Visitors. — Knuth makes the following remarks on these. — 



During one hour on the rsth of May, 1887, I observed in more than 20 

 cases Bombus terrester L. j and 5. This humble-bee alights on the outside of a 

 perianth-leaf, then creeps round on its lower margin into the inside of the flower 

 and climbs up on the inner surface of the leaf until it can lick nectar conveniently. 

 In doing so it brushes against the already receptive stigma in younger flowers with 

 its back and dusts it with pollen brought from older flowers, thus crossing separate 

 stocks. In older flowers also cross-pollination is ensured by insect-visits, because the 

 stigma projects a little beyond the anthers, and is therefore brushed against first 

 by a humble-bee visitor, which in climbing higher dusts its back with fresh pollen. 



This was the most usual way in which the humble-bee behaved in visiting 

 a flower, but I also observed another method ; the insect did not then creep up 

 on the inner surface of the perianth leaf, but chmbed up the style and the stamens 

 and sought in vain at the base of the latter for nectar. It thus brushed against the 

 stigma and anthers with the lower-side of its body and effected crossing. 



2718. F. imperialis L. (Sprengel, 'Entd. Geh.,' pp. 189-91 ; Herm. Miiller, 

 ' Weit. Beob.,' I, p. 275; Knuth, ' Bloemenbiol. Bijdragen'; Borbas, Ost. Bot. Zs., 

 Wien, XXXV, 1885.) — Borbas observed heterostyly in this species. 



Visitors. — The following were recorded by the observers, and for the localities 

 stated. — 



Borgstette (Nassau), the honey-bee. It uses the stigma as an alighting-platform, 

 creeps over the anthers into the nectariferous base of the flower, and then flies 

 straight away to another flower, upon the stigma of which it deposits pollen. Knuth 

 (Kiel Botanic Garden), the honey-bee, freq. Loew (Berlin Botanic Garden), the 

 honey-bee and 3 other bees — i. Anthophora pilipes Z'. 5, skg. ; 2. Bombus hortorum, 

 L. 5, do. ; 3. Andrena fulva Schr. 5, po-cltg. (' Blutenbiol. Beitrage," II, p. 68). 



2719. F. kamtschatcensis Ker-Gawl. — 



Visitors. — Loew (Berlin Botanic Garden) observed the Muscid Calliphora 

 erythrocephala 3Ig., creeping down to the nectaries, and creeping out again with 

 its thorax yellow with pollen. 



2720. F. latifolia Willd.— 



Visitors. — Loew (Berlin Botanic Garden) observed the bee Andrena fulva 

 Sckr. 5, creeping right into the flowers, po-cltg. 



2721. F. lutea Mill. — 



Visitors. — Loew (Berlin Botanic Garden) observed the humble-bee Bombus 

 terrester Z. 5, creeping into the flowers, po-cltg. 



