262 ESSAY ON CLASSIFICATION. 



these are not to be traced so uniformly in all classes as 

 the former, they are in reality only limitations of the 

 other kinds of divisions. 



A class, in which one system of organs presents a 

 peculiar development while all the other systems coincide, 

 may be subdivided into sub-classes ; for instance, the 

 Marsupialia when contrasted with the Placental Mam- 

 malia. The characters upon which such a subdivision is 

 founded are of the kind upon which the class itself is 

 based, but do not extend to the whole class. An order 

 may embrace natural groups of a higher value than 

 families, founded upon ordinal characters, which may yet 

 not determine absolute superiority or inferiority, and 

 therefore not constitute for themselves distinct orders; as 

 the characters upon which they are founded, though of 

 the kind which determines orders, may be so blended as 

 to determine superiority in one respect, while with refer- 

 ence to some other features they may indicate inferiority. 

 Such groups are called sub-orders. The order of Testu- 

 dinata illustrates tins point best, as it contains two natural 

 sub-orders. 1 A natural family may exhibit such modifi- 

 cations of its characteristic form that upon these modifi- 

 cations subdivisions may be distinguished, which have 

 been called sub-families by some authors, and tribes or 

 legions by others. In a natural genus, a number of 

 species may agree more closely than others in the par- 

 ticulars which constitute the genus and lead to the dis- 

 tinction of sub-genera, The individuals of a species, 

 occupying distinct fields of its natural geographical area, 

 may differ somewhat from one another, and constitute 

 varieties, etc. 



1 See my Contributions to the Natural History of the United States, vol. i, 

 p. 308. 



