52 CLASSIFICATION OF INSECTS. 



ancestral primitive form. In any natural classification, 

 therefore, these cases of extreme specialization by reduc- 

 tion, although properly regarded as degraded forms, should 

 stand at the head or be mentioned last in their respective 

 groups. 



Precisely the same statement must be made with regard 

 to the parasites. These creatures living upon or in other 

 animals have attained their fitness for such habitats by 

 gradually losing to a greater or less extent the parts and 

 organs which their ancestors, who lived exclusively in the 

 outside world, once possessed. The wingless parasitic 

 insects are in most cases rightly regarded as probably the 

 direct descendants of allied but winged forms. It is hard 

 to understand the development of the fleas, for example, 

 in which the young in their embryonic stages and trans- 

 formations resemble the similar stages of some true flies 

 (see Packard, Entomology for Beginners, note to p. 115), 

 unless these animals can be considered the modified de- 

 scendants of winged, dipterous insects, which, on account 

 of their peculiar habits, have become specialized by reduc- 

 tion and lost their wings, and in large part their dipterous 

 characteristics in later stages. The position of an animal 

 in a table of classification should be determined by its 

 place in the evolution of the group, — the most generalized 

 or primitive first, the more specialized by addition or com- 

 plication next, and the still more specialized, whether by 

 reduction or additional complication, next, and the most 

 specialized last in the series. Whether the last be a para- 

 site reduced to an extremely degraded structural condition 1 

 or not is a matter of no essential importance, its place in 

 the table or book being a matter determined, so far as 

 practicable, by its natural aflinities. 



1 Meyrick (^Trans. Ent. Soc, Lond., 1884, p. 277) maintains 

 that when an organ has wholly disappeared in a genus, other 

 genera which originate as offshoots from this genus cannot 



