ground that if a part of the leaves of each whorl in Galiace'se 

 were leaf-like stipules, they must bear some fixed proportion 

 to the true leaves : for instance, suppose the whorl to consist 

 of two leaves, each having two stipules, the whole number 

 of parts in that whorl should be six, and in all whorls the 

 number of parts should be some power of three ; while in 

 fact no such regularity in proportional numbers can be found 

 in a considerable part of the order. To this it might have 

 been added, that the admitted leaves are so entirely the same 

 as what M. DeCandolle conceives to be stipules, that no 

 difference whatever can in general be found in their form, 

 colour, anatomy, or degree of developement. 



Such reasons have however not proved satisfactory to 

 Botanists, who with one accord appear to range themselves 

 upon the side of M. DeCandolle, and recently the question 

 has been more particularly agitated by one of the most dis- 

 tinguished writers of this country. 



Mr. Bentham, in an article on Crusea rubra, published 

 in the Botanwt, page 82, after entering at some length and 

 with great skill into a discussion of the arguments employed 

 on^ both sides the question, has decided in favour of the 

 opinion of DeCandolle, that a part of the apparent leaves of 

 Galiaceous plants are stipules. The grounds upon which he 

 has arrived at this conclusion are essentially the following ; 



1. That the foliaceous organs in Galiacese, if viewed as 

 consisting entirely of leaves, do riot bear that relation to the 

 angles of the stem which is usual in Dicotyledons; but that 

 the relation becomes apparent if only two of them are taken 



and the rest as stipules. (DeCandoll 



fluenced by the consideration that it is onlv two of 



J 



apparent leaves which have buds in their axils ; but Mr 

 Bentham does not advert to this.) 



2. That in a number of cases, especially in Asperula 



pposite leaves are much 



3. That m Spermacoceae and other tribes, the stipules 

 connected with the petiole of the leaf into a sheath, and 



t this sheath exists in Galiaceae. 



4. That the number of parts in each whorl is not neces- 



