40 The Rev. P. Keith on the Classification of Vegetables, 



verely censured, seeing that the sexual system is actually and 

 professedly artificial. 



But in the system of Jussieu several important traits of affi- 

 nity are already determined before the class is fixed. All 

 plants composing the classes of the first grand group are al- 

 ready connected by the link of their being cotyledonous, that 

 is, by a character founded on the structure of the embryo, and 

 its mode of growth. All plants composing the classes of the 

 Dicotyledons and Monocotyledons are further connected by 

 their being respectively exogenous or endogenous, accordingly 

 as they belong to the former or to the latter division, that is, 

 by a character founded on the structure of the stem and its 

 mode of growth. These characters are evidently and essen- 

 tially natural. The Dicotyledons are subdivided into minor 

 groups, upon the ground pf their being furnished with a calyx 

 and corolla, or with a calyx only, a character found to be of 

 the greatest importance in bringing together natural orders, 

 though not infallible ; and the subdivisions are distributed into 

 sections, upon the ground of their being polypetalous, mono- 

 petalous, apetalous, or anomalous. If these last characters 

 are not absolutely natural, they are at least absolutely neces- 

 sary to give facility to the investigations of the student, and 

 are to be admitted till better characters are discovered ; and 

 if you say that they are founded on number merely, it is not 

 exactly so. It is upon structure rather than upon number, 

 a character of more value. For the distinction lies between a 

 corolla, the petals of which are free, and a corolla, the petals 

 of which are united, or it depends upon the absence of a co- 

 rolla altogether. 



Lastly, from the several sections, the classes themselves, 

 which are fifteen in number, derive their immediate origin, upon 

 the principle of the insertion of the stamens, as being hypogy- 

 nous, perigynous, or epigynous, characters evidently affinial*, 

 and very available to Linnaeus in the circumscription of hisl2th, 

 13th, and 20th classes; for how. without their aid, could he 

 have brought together so many plants connected by natural affi- 

 nities where his styles and stamens might not have been easily 

 counted? Number is no doubt the work of nature, as well as 

 other characters, but it is found to be liable to great mutabi- 

 lity, by an abortion, or by an undue multiplication of parts, 

 and is consequently not to be depended upon in the circum- 

 scribing of orders or of genera. It is but a fallacious mark 

 at the best, if taken by itself; for although the genera belong- 

 ing to a natural order may have all the same number of styles 

 and stamens, yet all plants having the same number of styles 

 and stamens do not belong to the same natural order. Hence 



• [See Lond. and Edinb. Phil. Mag., vol. v. p. 206, note.— Edit.] 



