CRUCI FERAE 



77 



to 4 mm. The cordate shape of the base of the calyx indicates even externally 

 the position of the nectaries. The insertion of each short stamen is surrounded 

 by a moderately large swelling that secretes nectar, of which a large drop is produced 

 on each side, so that as much as half the corolla-tube may be filled with it. The 

 four long stamens are surrounded at their bases by a much smaller swelling that 

 does not secrete nectar. The two sepals enclosing them are therefore but slightly 

 gibbous. 



The anthers of the four long stamens are placed immediately beneath the 

 opening of the flower, and dehisce by slits about 5 mm. long, which are turned 

 inwards. The anthers of the two short stamens are equally long, but their filaments 

 measure only 2-3 mm., so that they do not reach the stigma, which stands about 

 8 mm. up the corolla-tube. The four long stamens therefore serve for self-pollination, 

 which is automatically effected by the fall of pollen, or is due to insect-visits. The 

 short stamens — which also dehisce introrsely — serve for cross-pollination : the 



--«' 



Fig. 27. Matthiola incana, R. Br. (from nature). (l) Flower; natural size. A', pouch at the base ot 

 the calyr. (2) Stamens and pistil after removal of sepals and petals, the stamens spreading ; natural size : 

 a\ anther of a long stamen ; a", anther of a short stamen; j, stigma; w, the ridge-ltke nectary surrounding 

 the base of the shorter stamen. (3) Stamens and pistil ; twice natural size : references as in (2). 



proboscis of an insect endeavouring to get at one of the secreting nectaries, will 

 be dusted with some of the pollen from the adjacent short stamen, and will transfer 

 it to the stigma of another flower of the same species. 



Visitors. — In the jjarden of the Ober-Realschiile at Kiel, I observed a butterfly 

 (Vanessa urticae Z.) visiting cultivated plants and sucking nectar, and as it visited 

 several flowers in succession cross-pollination must have been effected. The 

 proboscis (14-15 mm. long) of this insect can reach right down into the nectar- 

 secreting base of the flower. In the same place I also noted Pieris sp., skg. 



169. M. annua Sweet. (Nobbe, Bot. Centralbl., xxxii, 1887, p. 253; Knuth, 

 ' Bloemenbiol. Bijdragen.') — F. Nobbe observed in cultivated plants that when the 

 seeds germinated very rapidly (3-4 days), most or somedmes all of the flowers 

 were double, but that — on the other hand — plants raised from seeds which germinated 

 slowly usually bore fertile single flowers. The same investigator also found that 

 if a variety of annual stock naturally inclined to produce double flowers was crossed 

 with one chiefly bearing single flowers, the mongrel exhibited the peculiarities of 

 the parent from which pollen was taken, not so much as regards the colour of the 



