CLASSIFICATION OF THE ANIMAL KINGDOM 2 



of various authors, not on the basis of the reputation of those authors 

 but on my judgment of the soundness of their supporting arguments or 

 on my analysis of the data they present. In none of the larger groups 

 has the work of any single author been accepted without modification. 

 Several considerations have influenced the decisions embodied in 

 this classification. 



First, a false picture is given by a simplified classification, because the 

 existing diversity is one of the principal features of the animal kingdom. 

 Therefore, no groups should be combined merely for the sake of sim- 

 plicity. 



Second, although the previous item would seem to require coverage of 

 the groupings at all possible levels, to show the extreme range of division 

 and subdivision, this is not in fact possible. Not only are there many 

 conflicting groupings at certain levels, such as of phyla or orders, but 

 there is no practical way to show these groupings in a general classifica- 

 tion. It is a compromise that is believed to be effective to subdivide the 

 phyla only into classes, subclasses, and orders. Other possible groupings, 

 such as subphyla and superorders are referred to in the notes. 



Third, two groups which are so distinct at any level that they cannot 

 be described in common terms must be separated at that level. (For 

 example, Pterobranchia and Enteropneusta; see the Notes on the Taxa.) 



Fourth, groups which cannot be distinguished at any particular level 

 by the type of characters used for their neighbors must be combined at 

 that level. (For example, the sometime classes of Nematoda.) 



Fifth, the discovery of groupings within a class, for example, does not 

 justify the creation of new classes for each of the subgroups. The proper 

 level for the new groups can only be determined by comparison with 

 neighboring parts of the classification. 



Sixth, although uniformity in the form (endings) of names at each 

 level would unquestionably be helpful, it cannot now be attained with- 

 out adding greatly to the total of name forms and synonyms. The sys- 

 tems so far proposed are so diverse as to introduce further confusion of 

 their own. None of the systems has been widely enough accepted to be 

 entitled to adoption throughout the Animal Kingdom. None has been 

 so widely accepted on a world basis, even in one group, as to indicate 



