REPRODUCTION. 223 



corollas, in which the hood of the lower lip is closely appressed to the 

 upper lip {Linaria). Anton v. Keener, who has made a special 

 study of alpine plant-life, has added to botanical literature two 

 important volumes, entitled ' ' Protective Arrangement of the Pollen 

 against Premature Liberation and Germination," and " The Pro- 

 tection of Flowers against Undesirable Guests." In reference to 

 the latter work (Innsbruck, 1879), I will state that it was above all 

 the author's desire to show clearly the suitable adaptations of the 

 various floral arrangements and to strengthen our teleological con- 

 ception of nature. We do not see in it any evidence in support of 

 the theory of selection,^ as the author seemed to indicate in the in- 

 troduction to this work. Ch. K. Sprengel (1793) made important 

 discoveries in regard to the physiological significance of individual 

 floral parts, especially the corolla and its zj'gomorphy. He is the 

 discoverer of the law of the absence of self-fertilization. We now 

 know that this ' ' law ' ' is not generally applicable, since there are 

 plants with special adaptations for self-fertilization (see below, 

 cleistogamous flowers). 



B. The Stamens and Pollen-grains. 



The stamens of gymnosperms are in general quite different 

 from those of angiosperms. In the former the part which bears 

 the pollen-sacs is sometimes flat, sometimes j^eltate or cylindrical ; 

 in the angiosperms it is in general filamentous, and is known as 

 the filament. Among gynmosi^erms the number of pollen-sacs is 

 usually much greater and more variable than among angiosperms ; 

 two is the usual number in the latter group. 



The following discussion of stamens is based upon their ap- 

 pearance among angiosperms. To explain the structure and 

 function of this organ we will give the important characteristics 

 illustrated by a few typical examples. 



The elongated portion, filament^ supports the anthers (pollen- 

 sacs). Fig. 150 shows the most frequent form of dehiscence or 

 opening of the anther, that is, it splits open in its longitudinal 

 direction ; more rarely there are pores formed at the apex, or the 

 apices may open by means of transverse valves. In the Ericacem 

 we find an interesting arrangement. The anthers open by the 

 formation of pores. The hardened appendages of the anthers 



