sarily some power of three, but that, taking two of the parts 

 for leaves, it is immaterial by what number of similar parts 

 those two are separated, because the intermediate processes 

 are analogous to the setae of Spermacocese, the number of 

 which is variable. 



Perhaps this question is more important in appearance 

 than in reality, for in some respects it is a mere difference 

 about words ; stipules being rudimentary leaves, and leaves 

 developed stipules. It is however connected with some 

 points of speculative interest, especially as regards syste- 

 matic Botany, and therefore I avail myself of the present 

 opportunity of stating what I conceive to be the objections 

 to Mr. Bentham's line of argument, and why I still retain 

 my original opinion upon the subject. 



1. With regard to the relation borne by the leaves to the 

 angles of the stem, it is to be observed that if those foliaceous 

 organs only which are opposite the angles are said to be 

 leaves in Galiaceae, and the rest stipules, then we must sup- 

 pose that Labiate plants have no leaves, but stipules only, 

 for in that and similar orders the apparent leaves are never 

 opposite the angles of the stem, but are always placed be- 

 tween them. Nor do I find that the number of angles in 

 the stem of verticillate plants necessarily corresponds with 

 the number of their leaves ; for example in Dysophylla 

 stellata, where the Whorls often consist of ten parts, the stem 

 has still but four angles. Neither can it be admitted that 

 bodies which do not form branches in their axils are there- 

 fore not leaves. All foliaceous organs of whatever kind, and 

 especially stipules, possess that power or not, according to 

 circumstances, as is too well known to require particular 

 proof. Besides DeCandolle's statement is not sustained by 

 fact ; for in Asperula the uppermost branches, bearing- 

 flowers, are frequently produced alternately with the leaves 

 that form the node from which they spring, and consequently 

 must, in such cases, arise from the seat of one of the sup- 

 posed stipules. It is more probable that the developement 

 of branches from a portion only of the leaves, is connected 

 with the form of the stem, and the relation which the leaves 

 bear to each other. If the form of the stem requires an 

 alternate developement of a pair or a triplet of opposite 

 October, 1838. x 



