THE MILKWEED FAMILY. 229 



There are five anthers placed close together, each ter- 

 minating in a membranous appendage that projects over 

 the thickened stigma disk. 



The anthers are separated from each other laterally by 

 a deep, vertical slit, bordered on either side by a thin 

 triangular process, the anther wing. At the upper ex- 

 tremity of the slit is a minute, black body, corpusculum, 

 which, when removed by a needle, is found to be con- 

 nected by means of a delicate, curved band on either side, 

 with a flattened, yellow, and waxy pollen-mass, poUinium. 

 Longitudinal swellings on the outside of each anther indi- 

 cate the position of the poUinia before their removal. 



Each of the slits already described is continuous within 

 with the stigmatic chamber, into which the pollen must be 

 introduced in order that fertilization may take place. It 

 is obvious that this cannot happen unless the poUinia are 

 removed from the anthers, and brought into the stigmatic 

 chambers by some external agency. 



This is accomplished by bees and other insects that visit 

 the flowers for honey.^ Alighting on the umbel the insect 

 easily gets its foot caught in the lower part of one of the 

 slits, and in attempting to withdraw it, one of the claws 

 is guided into the notch in the lower end of the corpus- 

 culum. With a strong pull, the latter is removed from its 

 place, and the insect carries away with it the two poUinia, 

 which by the twisting of the delicate bands, retinacula, 

 that connect them with the corpusculum, are now brought 

 into such a position as to be readily introduced into 

 the slit leading to the stigmatic chamber of some other 

 flower. If this has been done, and the insect is strong 

 enough, it frees itself by a vigorous pull, breaking the 



1 Hildebrand and Miiller have given a full account of the process, the 

 latter writer with illustrations. Fertilization of Flowers, p. 396 et seq. 



