THE LEAVES OF VASCULAR PLANTS. 241 



b}' a simple leaflet, as is seen in several species of 

 Ononis, &c. These different examples may prove, in 

 some degree, how easy it is to confound the limb, pro- 

 perly so called, wdth the bordered petiole. But this 

 confusion is of but little importance ; for it may be said 

 that the longitudinal nerve of the limb being a prolonga- 

 tion of the petiole, all limbs are only bordered petioles. 



2d. There are some Dicotyledons, such as the greatest 

 number of UmbelHferge and Ranunculaceae, in which 

 the fibres that ought to form the petiole, instead of 

 being distributed at their origin in a compact bundle, 

 arise side by side in a transverse series, which occupies 

 either all the circumference of the branch, or a remark- 

 able portion of it. The base of the petiole is then flat, 

 and more or less sheathing ; but soon the petiolary 

 fibres come nearer together, and form bundles as usual, 

 and the upper part of the petiole does not differ from 

 those with a round base. The sheath, although flat, 

 preserves the characters of the petiole ; it has few or 

 no stomata, and does not always decompose carbonic 

 acid gas. 



This expansion of the base of the petiole takes place 

 in a very high degree in the upper leaves of Lepidium 

 perfoliatum, Bupleurum perfoliatum, &c. where it en- 

 tirely takes the appearance of a foliaceous limb. 



In the upper part of the stems of the Umbelliferge, 

 these petiolary sheaths are frequently seen to exist, 

 although they have not been able to produce either the 

 foliaceous limb, or, sometimes, the cylindrical part of 

 the petiole. If one found an umbelliferous plant, which 

 had only these sheaths, he would be inclined to give 

 them the name of leaves, although they evidently are 

 sheathing petioles ; thus it is that we call leaves, in 

 Lathyrus Nissolia, true petiolary sheaths, which, when 

 they are entirely devoid of the limb, dilate still more 



VOL. I. R 



