208 MxVRSH PLANTS 



modification of some of the outer stamens. The 

 real stamens, which are numerous, lie more internally 

 still, and then, in the centre of the flower, we find 

 several carpels quite free from one another. Honey- 

 glands, even more highly modified, occur in the flowers 

 of several other members of the same family, Ranun- 

 culacese, as, for instance, in the Hellebores (Christmas 

 Eose, etc.), where they are fairly large and horn-Hke 

 in shape, and in Eranthis, the Winter Aconite, where 

 they are tubular. 



The presence of honey-glands and a copious 

 supply of honey impHes that this flower is fertilised 

 by insects ; yet the dome-shaped roof of the converging 

 sepals never opens. How, then, do the insects manage 

 to get inside? The mystery can be explained by 

 anyone w^ho will take the trouble to watch these 

 flowers for a few minutes on a bright sunny day. 

 Before very long, one or more small flies will be seen 

 to alight on the sepals, push their w^ay between them, 

 and disappear bodily into the interior. Later on 

 they will be seen to creep out again. It is these flies, 

 attracted by the honey, which serve as pollen carriers 

 from flower to flower. No bee can get at the honey 

 because of the roof of sepals. Thus we have here an 

 interesting contrivance for cross-pollination. 



Although the nectar, copiously secreted in the 

 honey-glands, is strictly reserved for certain small flies, 

 other insects, which are unbidden guests, make 

 determined attempts to reach the honey and to rob 

 the flower. Their chief difificulty is that their bodies 



