LEAVES. 



68 



If wc take the 



leaves having 



the 



tinuc downward 



ceeding case, but with the paren- 

 chyma of the secondary ^eins con- 

 nected with that of the primary from 

 which they originate, we shall have 

 the parted leaf of the Geranium Ma- 

 culatum, (fig. 46.) By a still greater 

 development of the parenchyma we 

 shall have the lobcd leaf, as in the 

 Liquidamber, Acer, &;c. (fig. 47.) — 

 type represented by the Nasturtion and Hy- 

 drocotyle and a development of 

 the parenchyma only sufficient to 

 cover the veins attached to each 

 rib and we have the palmate leaf of 

 the Podophyllum peltatum, fig. 

 48.) We have a striking confir- 

 mation of the above theory of the 

 origin of lobed leaves, in the Hy- 

 drogeton fenestralis, a plant pecu- 

 liar to the island of Madagascar. 

 It is an aquatic plant bearing 

 appearance of latticework, It is a 

 skeleton leaf, the parenchyma be- 

 ing only sufficient to cover the 

 veins. The veins are fully developed 

 and regularly arranged like other 

 leaves of the same variety, but no pa- 

 renchyma is generated to fill up the 

 spaces between the veinlets. 

 ^^^ 69. It not unfrequently happens, 

 both in reticulated and parallel vein- 

 ed leaves, that the first veins after 

 entering the lamina curve downwards, 

 and if they assume 

 afterwards the upward 

 direction they form by 

 these flexures the cor- 

 date variety of leaves, 

 as in the Smilax (fig. 

 49) among the paral- 

 lel veined leaves, and in 

 some species of Poplar 

 (fig. 50) in the reticu- 

 lated. If the veins con- 

 they form the Sagittate leaves as in the 



