4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 5 1 



stirpes, mainly founded, however, on certain of Robineau-Desvoidy's. 

 Schiner thus largely adopted Robineau-Desvoidy's stirpes in those 

 divisions which he did recognize, but bunched his genera along with 

 those of Rondani, Robineau-Desvoidy's reviser. The eight taxo- 

 nomic divisions adopted by Schiner generally obtained throughout 

 the epoch. 



Rondani's system, unlike Robineau-Desvoidy's, took little note of 

 habits, and, while less detailed, was more secure from being founded 

 primarily on external anatomical characters. But these characters 

 were liable to misinterpretation in certain cases. 



Brauer and von Bergertetarnm inaugurated the present and fifth 

 epoch in [889, which is destined to hold out for a greater degree of 

 radicalism than its predecessors. They approached the subject 

 largely in a new way, greatl) lessening the difficulties of classifica- 

 tion in the superfamily by recognizing a large number of sections 

 which correspond to the subfamilies and tribes of the present paper. 

 At the same time, they greatly multiplied the number of genera, 

 whereby they were able to present comparatively concise diagnoses 

 of these, as well as of their sections. 



They adopted Robineau-Desvoidy's plan of grouping the forms 

 into many small divisions, hut they did not feel hound, as did he, to 

 adhere to any definite scheme of life habits for indicating taxonomic 

 limitations. In the main their divisions were made on quite original 

 lines. However, many of Robineau-Desvoidy's old stirpes are still 

 recognizable, now more or less revised, restricted or enlarged, and 

 the\ must be considered as the original foundation of our present 

 subfamilies and tribes. Brauer and von Bergenstamm's characters 

 were better chosen and represent a more exhaustive study of the 

 subject, as would naturally follow from their having enjoyed the 

 greatly superior advantages derived from marked increase in biologic 

 progress since the time of Robineau-Desvoidy and Rondani, and 

 access to the greatly enriched collections of material drawn from all 

 parts of the globe. 



Until quite recently Brauer and von Bergenstamm's system has 

 been followed rather indifferently — in some cases enlarged upon, in 

 some revised — by students of the group contemporaneous with them 

 and continuing in the work since their time. The general trend of 

 sentiment now, however, is strongly in their favor, recognizing, as 

 it does, the necessity of a subdivision of the superfamily into many 

 subfamilies, tribes, and genera, so as to allow of more careful and 

 concise diagnoses. While it is true that a middle course between 

 the two extremes of conservatism and radicalism is usually the best 

 one to follow, the present superfamily furnishes a notable exception 



