/ 



A, 



into 



of 



-er per. 

 • power 



f 



getables 

 rpofe of 

 lygamia 



organs. 

 ,ve both 



only fe- 



ret. In 

 efs both 

 ave nei- 

 remain. 

 3rets are 

 female 



:he male 



; of time 

 like the 



f 



,rto 



this 



thec 



m 



fern 



ales 

 ales 



) 



? 



5 or P^^' 



r^rf' 



, by ^: 



Sect. XX. 3. 



NATURAL CLASSES. 



569 



foning as well as by obfer 



be lefs mutable than oth 



as- 



the fliortnefs of the filaments of dodecatheon meadia, cyclam 



fo 



m, bora 



& 



fufchia. and oth 



As mentioned in Botanic Gar 



den, Vol. II. note on 

 filaments are e? 



iviead 



Th 



n\ 



th 



flower of meadia th 



ceedingly fhort compared to the ftyl 



d feem 



have been in that circumflance immutabl 



Whence it became 



ceflary, firfl to furnifh them with long anthers, which ftand pointed 

 towards the diftant fligma apparently endeavouring to reach i 

 condly, it was necefTary to bend the flower-flalks, when th 



Se 



\ 



open into thofe graceful curves 



hich 



ft 



the 



ncommon 



beauty both of th 



flo 



d of the fufch 



that the ftigma by 



handng down immediately beneath the anthers might th 



falls, the prolifi 



farina. 



And that this was the evident defig 



of the curvature of the flower-ftalk appears from its rifing again, and 

 becoming quite ere£t, as foon as the impregnation of the pericarp is 

 accompliftied. Thirdly, as the flower thus becomes perpendicularly 

 pendent, it was necefTary to refle£l: the petals for the purpofe of ad- 

 mitting light and air to the fexual organs. 



We may reafon from this ftrufture of the meadia, that all this ap- 

 paratus of long ere£t anthers to approach the ftigma ; of bending the 

 flower-ftalk, that the fexual organs might become pendulous j and 



d 



air 



might have all 



uld have erown as long as th 



G 



then of reflefling the petals to give light 



been fpared, if the filaments alone cc 



ftyle ; as occurs in moft other flower 



flowers the filaments are the moft unchangeable parts of then 



that hence the comparative length of the filaments in refpefl 



ftyle would afford the moft immutable mark of their effential cha 

 radter, or for the purpofe of claffification. 



Another apparent inftance of the great unchangeablenefs of th 



And that therefore in thefe 



I ; and 

 to the 



length of the filaments exifts in the hemerocallis ful 



^a, tawny day- 

 h I obferve the ftyle is crooked, or bent into a zigzag, 



about the middle of it, evidently for the purpofe of fhortening it, 



4 D that 



\ 



