ANGIOSPERMS. 



487- 



botanists, usually consists of a dense coarse-grained protoplasm in which o-rains of 

 starch and drops of oil may be recognized. When the grain bursts in water, the 

 fovilla escapes in masses connected by mucilage and often in long vermiform 

 threads. The surface of the extine is commonly found coated with a yellow oil, or 

 of some other colour, often in evident drops, which renders the pollen viscid and 

 adapted to be carried by insects from flower to flower; in only a comparatively 

 few cases is it quite dry and powdery, as in Urticaceae and many Grasses, where it is 

 projected with violence from the anthers or simply falls out. 



At the time when the pollen-grains are nearly mature, and the flower-bud 

 is preparing to open, the wall of the pollen-sacs undergoes a further development \ 

 The outer layer of cells or epidermis always remains smooth-walled (see Fig. 351, 

 p. 489) ; the inner layers or endothecium are also smooth if the anther does not 

 dehisce. If on the other hand it opens by recurved valves (Fig. 331 k, p. 475); 

 the cells of the innermost layers only of these valves are provided with thicken- 

 ing-bands (or are fibrous) ; while, when the pollen-sacs dehisce longitudinally, the 

 whole of their endothecium contains fibrous cells. There is usually only one such 

 layer, sometimes several ; in Agave ame7-icana as many as from eight to twelve; 

 The thickening-bands of the fibrous cells which project inwards are usually wanting 

 on their outer wall ; on the side-walls they are generally vertical to the surface of the 

 pollen-sac ; on the inner wall they run transversely and are united in a reticulate 

 or stellate manner. Since the epidermal cells contract more strongly when the 

 ripe anther-walls dry up than those of the endothecium which are provided with 

 thickening-bands, they exert a force which has a tendency to make the anther-wall 

 bulge outwards and give way at its weakest point. The modes in which the 

 anthers open are very various, and are always intimately connected with the other 

 contrivances which are met with in the flower for the purpose of pollination with or 

 without the agency of insects. Sometimes only a short fissure (pore) is formed at 

 the apex of each anther-lobe, as in Solanum, Ericaceae (Fig. 33-2, p. 4'75), &c., 

 through which the pollen of both the contiguous pollen-sacs escapes ; but more com- 

 monly the wall gives way in the furrow between the two sacs (the suture) along its 

 whole length, the tissue which separates them becoming at the same time more or 

 less destroyed, and thus both pollen-sacs dehiscing at the same time by the longitu- 

 dinal fissure (Fig. 351). It is this phenomenon that has given rise to the erroneous 

 description of these anthers as being bilocular; but if nomenclature is to have a 

 scientific basis, they must be termed quadrilocular, in contrast to the really bilo-' 

 cular anthers of Asclepiadeae and the octilocular ones of many Mimosese. Some- 

 times again the anther-lobes open at the apex by a pore which results simply from 

 the destruction of a smafl portion of tissue at this spot (Hofmeister). In other 

 respects \ve still want a detailed and comparative investigation of these processes, 

 which are very various and of great physiological importance; only the addi- 

 tional remark need be made here, that it is very important from a systematic point 

 of view whether the anthers open inwards towards the gyn^^ceum (introrse), or 

 outwards (extrorse), the diff'erence depending on the position of the suture and 

 hence on that of the pollen-sacs on the inner or outer side of the filament. , , 



» Compare li. v. Mohl, Vermischte Schiiften, p. 6:. 



