366 



ECOLOGY. 



staminodium is so wrapped around the flattened style that the anther lies 



closely pressed against the face of the style, near the margin opposite that 



on which the stigma lies. 



667- The walls of the anther locules which lie against the style become 



changed to a sticky substance for their entire" length, so that they cling 



firmly to the surface of the style 

 and also to the mass of pollen 

 within the locules. The result is 

 that when the flower opens, and 

 this staminodium unwraps itself 

 from the embrace of the style, the 

 mass of pollen is left there de- 

 posited, while the empty anther is 

 turned around to one side. 



668. Why does the flower de- 

 posit its own pollen on the style ? 

 Some have regarded this as the act 

 of pollination, and have concluded, 

 therefore, that cannas are neces- 

 sarily self pollinated, and that 

 cross pollination does not take 

 place. But why is there such evi- 

 dent care to deposit the pollen on 

 the side of the style away from the 



Fig. 467. 



Canna flowers with the perianth removed to .... ■ -, tt • ■, .t 



show the depositing of the po.len on the style by stlgmatic margm ? If we visit the 

 the stamen. cannas some morning, when a 



number of the flowers have just opened, and the bumblebees are humming 

 around seeking for nectar, we may be able to unlock the secret. 



669. We see that in a, recently opened canna flower, the petal which 

 directly faces the style in front stands upward quite close to it, so that the 

 flower now is somewhat funnelshaped. This front petal is the labelbwt, and 

 is the landing place for the bumblebee as he alights on the flower. Here 

 he comes humming along and alights on the labellum with his head so close 

 to the style that it touches it. But just the instant that the bee attempts to 

 crowd down in the flower the labellum suddenly bends downward, as shown 

 in fig. 468. In so doing the head of the bumblebee scrapes against the 

 pollen, bearing some of it off. Now while the bee is sipping the nectar it is 

 too far below the stigma to deposit any pollen on the latter. When the bum- 

 blebee flies to another newly opened flower, as it alights, some of the pollen 

 of the former flower is brushed on the stigma. , 



670. One can easily demonstrate the sensitiveness of the labellum of 

 recently opened canna flowers, if the labellum has not already moved down 

 in response to some stimulus. Take a lead-pencil, or a knife blade, or even 



