242 



ORGANOGRAPHY. 



generally tapers in an almost imperceptible manner from the 

 base to the apex, it is then described as filiform, as in the 

 Rose ; or when it is very slender, as in most Grasses and Sedges, 

 it is capillary {figs. 488 and 490). In the latter case, the filament 

 instead of supporting the anther in the erect position as it 

 usually does, becomes bent, and the anther is pendulous. At 

 other times the filament becomes enlarged, or flattened in vari- 

 ous Avays. Thus in some cases, it is dilated gradually from 

 below upwards like a club, Avhen it is clavate or club-shaped, as 

 in Thaliclrum ; or it is slightly enlarged at the base, and tapers 

 upwards to a point like an awl, as in the flowering rush 

 (JButomus umbellatus); in other cases it is flattened at the base 

 (fig. 491), but there is no marked tapering, the rest of the fila- 



Fig. 490. 



Fig. 491. 



Fig. 492. 



Fir/. 400. A locusta of Wheat (Tritiaim) consisting of several flowers, the 

 stamens of which have very long capillary filaments, and versatile pendu- 

 lous anthers. The anthers are bifurcated at each extremity, and resemble 



Bomewliat the letter x in form Fig. 4!>1. Three of the stamens of the 



Tamarix gnllicn, with their filaments flattened at the base and united 



with one another Fig. 402. Pistil of Campanula, with a solitary stamen 



arising from the summit of the ovary. The filament is flattened. 



mcnt assuming its ordinary rounded form, as in Tamarix gallicn 

 {fig 491); or the whole of the filament is flattened, and then it 

 frequently assumes the appearance of a ]ietal, when it is de- 

 scribed as petaloid, as in the Watcr-I.ily {Nt/mphaa) {figs. 438, 

 e, and .'iOO), and in the Canna and allied plants. 



Sometimes the filament is toothed, as in tlie Allium {fig. 493), 

 or for/ml, as in Cranibe {fig. 494) or furnished with various 

 aj)pendages, as in the Borage {fig. 495), and Zygo])hyIlum, in 

 which ease it is said to be appnidiculate. Tliese ajipeiiduges 

 are evidently of the same nature as the scales, &c., i)reviously 

 described as occurring on the corolla. 



Length, Colour, and Direction.— The length of the filament 

 varies niuch. TIius in tlie Borage and plants generally of the 



