ItANUNGULACETE. 



25 



at the base, and taper at the tip to support a basifixed two- 

 celled introrse anther, dehiscmg by two longitudinal clefts.' The 

 gynseceum consists of from three to five- free carpels, inserted 

 on a spira? near the apex of the receptacle, each composed of an 

 ovary tapering above into an acute style, which is stigmatiferous 

 only at the top and on the margins of the vertical groove, which 

 runs the whole length of the inner angle of the carpel. The ovary 

 contains two vertical rows of anatropous ovules, inserted along the 

 inner angle. The fruit is usually formed of three follicles dehiscing 

 along the inner angle (fig. 47) to set free the seeds, which have a 

 spongy, more or less rugose surface, being covered by wrinkles and 

 membranous projecting folds. The embryo is surrounded by the 

 abundant fleshy albumen. 



A. Napellus is a herbaceous plant with alternate palmatisect 

 exstipulate leaves, and blue or white flowers in terminal racemes. 

 Each flower is axillary to a bract which becomes smaller and less 

 dissected as it is higher up on the principal axis. The top of the 

 pedicel is slightly swollen, and at this point we notice, applied to 

 the calyx itself, the two lateral sterile bracts 

 which accompany the flower and have been 

 carried up with it. 



We know about a score of other Aconites 

 properly so called. But all these species 

 have not sepals formed exactly as in A. Na- 

 pelliis.'^ Thus, A. hehegynum DC, and A. 

 variegatimi L., have the posterior sepal like 

 a conical compressed helmet. A. Anthora L.* 

 has this same sepal conical and semicircular, 

 while in A. Lycodonum L.," it assumes the 

 form of a true narrow elongated spur (fig. 

 48), obtuse only at the tip. But there 

 is every possible transition between these 



Aconitiim lA/coctonum. 

 Fig. dS. 

 Flower. 



' Each cell wlieu open forms, as in the Colum- 

 bines, a plate spread out edgewise. The cleft 

 being far more interior than exterior, this 

 plate is attached to the connective, not at the 

 middle of its breadth, but nearer the inner 

 border. 



2 The number three is by far the commonest, 

 though we see in gardens flowers with five, six, 

 eight, and even more carpels. 



3 Which we must understand does not preclude 



the existence of secondary radiating rows analo- 

 gous to those observed in NigeUa. 



•• Sect. iv. Napellus DC, Syst., i. 371, (incl. 

 Cammnriim [DC, /. cH., 37 !•, sect. iii.J, Cori/- 

 llia'ba Reicub., Enchi/lodes Reichb., ox SrACir, 

 Suit, a Buff., vii. 367). 



s Sect."i. Anthora DC, Siist., i. 3G1-, Prodr., 

 i. 56. 



" Sect, ii., Li/coctomiin DC, Sj/st., i. 367, 

 Prodr., i. 57. 



