CLASSIFICATION OF THE RHYNCHOPHOROUS COLEOPTERA. 389 



"which no entomologist ever refers without finding original ma- 

 terial by which he can profit ; a remarkable instance of the good 

 results to be obtained by a careful and intelligent study of a very 

 limited fauna. The RhjMichophorous series is divided as follows : 



Segments of the abdomeu immovable, 2d and 3d nearly equal 



ISOTOMA. 



Brucliidee, AntliribidiB (including Urodou), Eliiuomacerida;, Atte- 



labidfe. 

 Abdomen with the 1st and 2d segments connate, the I'emaiuiug three 



movable, the 2d usually much longer than the 3d . Axisotoma. 

 ApionidEe, Curculionidaj, Cossouidaj (including Calaudra), Tomicidaj. 



From a survey of the difterent schemes of arrangement which 

 have been thus briefly reviewed, it is evident that while the prin- 

 cipal types of the Rhjnichophorous series, and the main divisions 

 of the great family Curculionidiie have been clearly perceived, the 

 attempts to define tliese important forms have failed in a greater 

 or less degree, on account of the want of proper subordination in 

 the characters made use of: all of them natural, all of them im- 

 portant, though in a less degree than supposed by the expounder 

 of each particular system. 



To supplement the memoirs above referred to, there came in 

 more recent times the beginning of a systematic studj^ of our 

 species of Curculionid;i3 by Dr. George H. Horn, a careful and con- 

 scientious study of the Calandridae and Cossonidae and of some 

 Mecorhynch genera of the United States.* In the introductoiy 

 remarks he observes : — 



" One character is mentioned in the following pages that ap- 

 pears to have escaped notice. In most if not all of the genera of 

 Mecorhynques, the males have eight and the females seven dorsal 

 abdominal segments. The Calandrides and Cossonkles appear not 

 to possess this character, as also all the Brachyrhynques which I 

 have had time to examine." 



The value of this original observation of Dr. Horn is very great, 

 but the limitation which he has placed upon it, though correct as 

 regards the Calandride and Cossonide types, is erroneous as regards 

 the Brachyrh^'ucs, which have the abdominal sexual characters 

 precisely as in the genera in which he first observed them. So too 

 have the Brenthidffi, and all the anomalous sub-families of Curculi- 



* Contributions to a Knowledge of the Curculionidce of the United States. Proc. 

 Am. Philosophical See. 1873, 407. 



