326 ANGIOSPERMAEDICOTYLEDONES 



middle. The alar laminae possess longitudinal folds, and arch inwards so as to 

 touch each other at the middle of the curvature, in front of the tip of the carina. 

 Near the posterior end of the upper margin of each ala there is a long conical 

 process, which fits into a corresponding depression in the upper margin of the 

 carina. Still further back there are two deep alar folds, and also a shallower one, 

 the inner surfaces of all which are so intimately united with the carina that the 

 alae cannot be separated without tearing them. The rounded posterior lobes of 

 the alar and carinal laminae are directed backwards, and lie above the staminal 

 tube. Their elasticity causes this tube to return into the carina. The upper filament, 

 as in V. sativa, is fused with the others, permitting access to the nectar only by 

 means of passages at its base. There is abundant secretion at the usual place. 

 The claws of the petals are in contact with one another, and 13 mm. long. They 

 are enclosed by the calyx-tube for a distance of 6-7 mm. The stipules serve as 

 extra-floral nectaries, but only a few of them are functional. 



Visitors. Kirchner, in Wurtemberg, saw a humble-bee (Bombus lapidarius 

 Z. 5) on cultivated plants. 



769. V. Faba L. (Sprengel, 'Entd. Geh.,' pp. 357-60; Herm. Muller, 

 'Fertilisation,' pp. 206-7; Darwin, Ann. Mag. Nat. Hist., London, Ser. 3, ii, 1858, 

 p. 460; Mattei, Justs bot. Jahresber., xvii, (1889) 1891, p. 480; Knuth, 'Bliitenbiol. 

 Beob. a. d. Ins. Riigen.') The fragrant flowers of this species are white with a black 

 spot on each ala. Their mechanism agrees, according to Hermann Muller, with that 

 of V. sepium, but they are much larger. The nectar, however, is more easily 

 accessible, for the vexillum and alae are less firmly held together, and the carina 

 is more readily depressed. The swollen processes on the under-side of the vexillum 

 are also wanting, so that the claw of this petal (13-16 mm. long) is but loosely 

 held by the calyx-tube. On the other hand, the two involutions connecting the 

 alae and carina are present, but the union is less firm than in V. sepium, and 

 the backwardly directed alar processes are much more feebly developed. The alae 

 and carina are but slightly elastic, so that if forcibly depressed they do not return 

 to their original position. 



Darwin found that when insects were excluded the plant was only one-third 

 as fertile as when they were allowed to visit the flowers. If, however, the flowers 

 protected from such visits were shaken, healthy seeds were abundantly produced. 

 Mattei describes an Indian variety as adynamandrous. 



Only long-tongued bees are able to suck the nectar legitimately, and in 

 doing this to effect cross-pollination. Short-tongued bees either collect pollen 

 on visited flowers with exposed anthers (simultaneously bringing about crossing) 

 or steal nectar by perforation. Bombus terrester Z. 5 (proboscis 7-9 mm. long) 

 is notorious in this respect, and very seldom tries to suck legitimately. The 

 honey-bee either steals nectar through the holes made by B. terrester, or collects 

 pollen. 



Visitors. I observed the following bees in the Island of Riigen. 



1. Apis mellifica Z. 5, stealing nectar as above mentioned; 2. Bombus terrestei 

 Z., perforating the flowers ; 3. B. hortorum Z. 5, very freq., skg. ; 4. B. rajelhu 

 K. j, skg. 



