Group Aphidixa, Family Aphididae 67 



tempt to break up the large genus Aphis. The lateral tubercles as rudi- 

 mentary organs have very great phylogenetic value, but are of little 

 use as generic characters. In early work on the family the author 

 also attempted to use them as a basis for generic characters in the 

 old genus Aphis, but soon came to see that they led only to artificial 

 groups that would not harmonize with other characters present. Lat- 

 eral tubercles when found in full number are present on the prothorax 

 and the first seven segments of the abdomen. Koch recognized a series 

 of species within the genus Aphis as Pleurodonten, which evidently 

 were forms with full or a large number of tubercles present. The 

 number is more commonly reduced ; we have a very large series of 

 the genus Aphis with only three, the prothoracic, one anterior abdom- 

 inal, and one posterior abdominal. The abdominal will in turn dis- 

 appear leaving only the prothoracic, the last and most persistent of 

 all the tubercles. Lateral tubercles in a more or less rudimentarv con- 

 dition are now known for practically all the tribes of the family, and 

 they may be looked upon as originally characteristic of the family, but 

 appear now as rudimentary organs that are gradually becoming lost. 

 Their persistence and conspicuous presence in the tribes Pterocom- 

 mini and Aphidini are therefore significant as indicating the close rela- 

 tionship between these two tribes. The characters are too broad to be 

 applied to genera, where they at most would indicate the degree of 

 reduction that has taken place, which varies greatly for difi:'erent tribes 

 and groups. 



x\ single or even a series of arbitrarily chosen characters will lead 

 to more or less artificial results. The same may give the phylogenetic 

 evolution of the organ or organs in question but not necessarily that of 

 the organism. The value of a given character must be ascertained for 

 each group separately. While a given character may have very great 

 value for a certain group, the same character may have little or no 

 value for a closely related group. The series of characters that we 

 may make use of are therefore not the same for all, but will dififer 

 in their arrangement and values for the different groups, be this a 

 species, genus, tribe, etc. 



The primary aim of classification is not the arrangement of forms 

 in convenient groups by which the name can readily be ascertained, 

 important as this is from a practical point of view; if this were so 

 we might as well at once accept the numerical or a Dewey system and 

 have each category fixed once for all, as has more than once been 

 suggested and attempted in the history of taxonomy. The aim is to 

 express the true phylogenetic relationship of the various categories to 



