ti NATUllAL IfTSTOBY OF PLANTS. 



are liii^lier up on this basilar portion of the stem, each becoming 

 in turn a leafy aerial branch ending in an inflorescence ; later 

 on, swelling at the base, and then giving up its juices to supply 

 the evolution of the buds axillary to its first leaves. Thus the 

 rhizome of the Columbine gradually ramifies, while the primitive 

 root, after having at lirst formed a tolerably large tap-root, gradually 

 becomes hollow and withers, till it has entirely disappeared, and the 

 ])lant is wholly nourished by the adventitious roots, which appear 

 at each period of vegetation at the base of the ascending axes. 

 Many perennial Ranuncidacece resemble this, and belong to the group 

 of plants with definite axes evolved successively. 



Xanthorkiza apiifolia Lher.,' though of very different habit,and with 

 very small flowers which at first sight hardly recall those of the 



Columbine," presents nearl}^ the same floral 



organization, and may be considered a 



,1^.- "' ^ smaller type of this,^ from which the only 



'^M^'^' ^^vl-J essential point of difference is the smaller 



^\{ number of staminal whorls in the former. 



Ix^ ---^ j The perianth (fig. 13) is formed of five 



Xanthorkiza apiifolia. caducous scpals, as a rulc quincuucially 



^'^- ^^' ^^^* ■^*' imbricated in the bud, and of five smaller. 



Flower. Petal. 



fleshy, glandular petals, contracted at the 

 base into a narrow claw, and dilated above into a cordate somewhat 

 concave limb (fig. 14). 



The stamens are often ten in number, arranged in two whorls, so 

 that five are opposite the sepals and five opposite the petals ; but 

 we often find the parts of one whorl more or less abortive. Each 

 stamen consists of a hypogynous filament, and a basifixed flattened 

 bilocular anther, dehiscing by two longitudinal lateral clefts, rather 

 turned inwards than outwards. The gyna^ceum is often formed of 

 five free carpels opposite the petals, each composed of a unilocular 



' XauUiorliiza MAi<«ir, ex Schkeb., Gen., 727, ^ Payer {Organog., 217). " There are," says 



n. 1S51.— Lamk., ///., I. 854; DC, P/Wn, i. 65. he, " only very mmute tlifterences between the 



— Spach, Unit, a Bfff., vii. 407. — Endl., Oen., flowers of Aquilegia and XanthorUza, con- 



n. 4803.— A. Gray, UL, t. 17.— B. H., G'f w., 9, sistiiig, as they do, in the different nnmber of 



n. 2'.>. — H. Un. ^i^/rtn-vOHfa., iv. 4-4. — Xanthorkiza whorls in the androceum on the one hand, and 



rtpii/b/tV7,Liirii.,.S7/r/>. iVo».,79,t.38.— DPHAM., in the form of the petals on the other; and I 



Arbr., lust cd., iii. t. 37. can with difficulty understand why botanists have 



- This has been generally placed near Actma placed these two genera in different sections. 



or I'nonia. A.-L. de .IrssiKi' {Gen. 234) says 

 on tliis plant — " Ciniirifiigee qffiiiis." 



