362 



proportion of the leaves as much to be depended upon, but doubts 

 the constancy of the form of the root, while, in fact, the reverse is the 

 case. I shall therefore touch upon the distinctions between CE. pim- 

 pinelloides and CE. peucedanifolia. Both appear so similar when 

 growing, that a practised eye only can detect the difference between 

 them without looking at the radical leaves. But these, when not 

 withered up, are very discriminative, for in Ql^. pimpinelloides they 

 are bi-pinnate, the pinnulse broadly elliptical or wedge-shaped, so 

 deeply cut as to be almost pinnatifid, the segments very acute, mucro- 

 nate, with a broadly sheathing, membranous petiole; while in GE. 

 peucedanifolia there are mostly no radical leaflets at all, or if any, 

 they are ascending (not extending Jtorizontanp,3iS in pimpinelloides), 

 and they agree entirely with the stem-leaves, viz., bi-pinnate, with 

 acute, entire lanceolate or linear pinnae. But the stem-leaves are ex- 

 ceedingly variable in aspect, both in pimpinelloides as well as peuce- 

 danifolia ; frequently they are only simply pinnate, and then the pinnaj 

 are very long, nari'ow, linear and acute. Sometimes the radical leaves 

 of pimpinelloides have their pinnae almost as narrow as those of the 

 stem-leaves, and then without examining the root it becomes very dif- 

 ficult to distinguish it at first sight from CE. peucedanifolia. The 

 upper leaves of both are simply pinnate. No dependance can be 

 placed on the general involucra, which, though stated by Smith to be 

 "numerous" in CE. pimpinelloides, are so far from being so, that scarce- 

 ly one specimen in twenty has any at all, or only a single setaceous leaf- 

 let; sometimes, however, an involucre of five or six unequal leaflets is 

 present. On the other hand, CE. peucedanifolia, said by Smith to have 

 no general involucre, has very frequently one, and sometimes several 

 leaflets at the base of its umbel. The umbels in pimpinelloides have 

 from six to fourteen I'ays, while those of peucedanifolia have never 

 more than eight ;* the unfbellules are many-flowered, very dense, when 

 in perfection forming globular heads ; the external florets, which have 

 very large petals, unequal in size, are barren, on long pedicels, the 

 internal fertile florets on short swollen callous ones. In CE. peuceda- 

 nifolia the internal pedicels are not at all swollen, nor is the fruit so 

 densely clustered. This latter is a stouter and taller plant than pim- 

 pinelloides ; often exceeding a yard in height, with a thick, striated; 

 hollow stem, which, after flowering becomes incapable of supporting 

 itself, so that it then sinks down upon the grass beneath it, and is so 



* After flowering, however, the external barren florets fall off" in both species, and 

 then this character is not so apparent. 



