168 



always the four long stamens only burst ; the other two 

 remain closed. 



It has been repeatedly observed that the sécrétion of 

 nectar begins as soon as the stamens open. 



In connection with what was stated above, one would 

 be inchned to infer from this that flow of water from the 

 anther causes the sécrétion of nectar. If, however, with 

 Stdlaria média, the anthers are removed before they hâve 

 discharged water to the nectaries, one finds ail the same 

 the nectaries amply provided with honey, when the flower 

 opens. The same may be observed in the maie flowers of 

 Aesculus Hippocastanum. In the still nearly closed flower- 

 bud the nectary is dry yet. When the flower continues to 

 open, small drops of liquid are seen to appear on the sur- 

 face of the nectary, still before the anthers extend halfway 

 from the bud. Thèse droplets increase in size as the an- 

 thers approach the moment in which they open. By weigh- 

 ing it may be proved that the anthers hâve already lost 

 part of their original weight when the flrst droplets of 

 nectar appear on the surface of the nectary. From this 

 circumstance also one would be inclined to infer that the 

 water of the anthers cômes out again as nectar. When, 

 however, from very young buds, whose nectary is not 

 raoist yet, the anthers are removed, yet at a later stage 

 of development of the bud, sécrétion of nectar is found in 

 them as in buds that hâve kept their anthers. 



With Fritillaria imperialis I found the same; but hère 

 the sécrétion of nectar was not so'abundant as in buds, 

 the anthers of which had not been removed. In my opinion 

 thèse observations indicate that the sugar, stored up in 

 the nectaries or other sugar-containing tissues of the 

 flower, at the moment when it begins to exert its osmotic 

 action, attracts water not only from the anthers but also 

 from other parts of its surroundings. 



