AMERICAN HONEY PLANTS 199 



carbon, hydrogen and starch atoms, which causes them to behave differ- 

 ently when examined by polarized light, and materially affects other of 

 their physical properties. 



The flow of the water of nectar seems to be like that of water through 

 water pores, an infiltration under pressure when root-absorption is least 

 active and leaf evaporation checked; but thoroughly and repeatedly wash- 

 ing the glands sometimes puts a stop to it. Beating rain does this as 

 effectively as experimental washing. Change of position and closing in 

 dark, rainy weather characterize some flowers, and keep the rain from 

 washing away their accumulated nectar and checking its replenishment. 

 This was Sprengel's explanation of the fringe of hair on the petals of the 

 wild geranium. In proportion as such nectar guards are effective, they 

 preserve the supply and contribute to its continuance, in proportion as 

 rain has opportunity to beat upon the nectar glands it wastes and may 

 even check the production of nectar. 



This stopping of nectar flow by washing away the secretion of the 

 glands, is connected with the affinity for water of sugars. The flow of 

 water appears to 'be started by the osmotic force of the disintegrated part 

 of the walls of the secreting cells; it is stopped when the resulting sub- 

 stance has been removed from the outer surface of the secreting cells. 



Stomata on nectariferous tissue of Xanthoceras sorbifolio. Greatly magnified. Copied 

 from Bonnier's "Les Nectaires." 



If this were all, unless the degenerating cellulose were replenished in 

 sufficient quantity there would hardly be such a thing as honey production. 

 Indeed, some extra-nuptial glands secrete a nectar containing so little 

 sugar that even ants may not be attracted by it — as is said to be the case 

 with climbing smartweeds cultivated in England, though it is not usually 

 true of such plants growing wild here, where they are at home. Com- 

 monly, however, the sugar in nectar is replenished while the secretion of 

 f uid continues. 



The passage out of sugar from a living cell is very different from the 



