PRINCIPLES OF NATURAL ARRANGEMENT. 341 



in Exogens, and that they are confined to a few orders among 

 Endogens. 



495. The Leaves are subject to considerable modifications, 

 both in position, form, and structure, which are very useful in 

 classification. The general differences among the leaves of Exo- 

 gens, Endogens, and Acrogens, have already been several times 

 adverted to. The relative position of the leaves, as whether 

 alternate, opposite, or verticillate, is often a very important cha- 

 racter ; but in regard to this, as well as to other characters, it 

 often happens that it is of much greater value in some orders 

 than it is in others. Thus in Lamiacea (Dead- Nettles) they 

 are uniformly opposite : so that no plant can belong to the order, 

 in which they are alternate or verticillate. In Urticacece (the 

 Nettle tribe), on the other hand, they are constantly alternate ; 

 so that no opposite-leaved plant can belong to the order. In this 

 manner the common Dead-Netties and Stinging-Nettles may be 

 at once known from each other. But in many others, one 

 arrangement is prevalent, and yet the other sometimes occurs. 

 The degree of division of the leaves, again, is subject to consi- 

 derable uncertainty in many orders, from causes formerly adverted 

 to (. 234 9) ; yet in others a constant form is maintained ; 

 thus, leaves with teeth or jagged edges are never found in the 

 order Cinchonaceee (from which the Peruvian Bark is supplied), 

 and they are very rare in Endogens. The particular characters 

 afforded by the veining of leaves are much more constant, as 

 formerly shown (. 235), than those derived from their form ; 

 and it is probable that, as they have only been recently attended 

 to, much assistance will be obtained in classification from an 

 increased knowledge of them. A character which would not at 

 first sight appear of much importance, is afforded by the presence 

 or absence of those little dots in the leaves, which are reservoirs 

 of oily secretions ; yet these, being connected as it would seem 

 with some important differences in the general economy, are 

 extremely characteristic of certain natural orders, such as Myrta- 

 cecK (the Myrtle tribe), and Aurantiacece (the Orange tribe), 

 serving to distinguish all their members from those of other orders 

 nearly allied to them. In other orders, however, there are some 



