86 . TJniversity of California Puhlications. [Entomology 



One of the many fundamental points upon which there is a 

 conspicuous absence of agreement is the question of the 

 division line between groups. In the taxonomy of existing 

 organisms the question is of less practical significance than in 

 paleontology, or in the study of phylogeny. The view that in 

 nature there are really no division lines is a statement of a 

 fact that, though true, is an evasion of, rather than an answer 

 to, the question, for as long as there are real differences in 

 nature, there are real divisions, though neither of these be 

 absolute. 



These divisions are by no means to be interpreted as indi- 

 cating absolutely isolated, unapproachable structural types, but 

 do signify diverse lines of evolution. All organisms would 

 probably be proven to have a single common ancestor, if the 

 course of the phylogeny of each group could be traced back to 

 the beginning, and probably one would not have to go back to 

 the very beginning of organic life to reach the individual from 

 which all existing organic groups have originated. Likewise, in 

 any group, as insects, the immediate descendants of the first 

 representative of that group probably do not represent the sep- 

 aration of the ancestors of the two oldest orders. When this 

 first division did occur, it is conceivable, indeed probable, that 

 these two ancestral individuals did not differ as much from 

 each other as they did from other forms then existing. A 

 study of insects of that time would hardly have suggested the 

 grouping that the descendants of those insects have revealed 

 in their subsequent development. 



For this very reason a study of Silurian or Devonian in- 

 sects without a knowledge of the insects of later times, if such 

 a thing were possible, would quite likely result in an entirely 

 different system of arrangement from that which we use at the 

 present day for existing insects. 



Scudder's classification of the Paleozoic insects is the nearest 

 approach to a system of this sort. His Palseodictyoptera brings 

 together in one order a number of forms which had already 

 differentiated to an extent sufticient to make them recogniz- 

 able as the probable ancestors of different orders. Because of 

 their evident similarity to each other, and difference from the 

 existing members of the orders toward which they point, they 

 are considered as not yet belonging to those orders. 



Brauer ('86), on the other hand, would place the ancient 



