6 NORTH AMERICAN 



dition ; among the multitude examined I have only rioted it in three 

 species. Austini, milleporus, and umbratilis, and then in but one 

 specimen of each. It is certainly a very surprising phenomenon, and 

 everything said concerning it must he more or less opinionative ; as 

 far as I have been able to perceive, it is purely accidental. 



The structure of the mentum lias been very singularly mistaken by 

 Erichson, that author, usually so accurate, stating that the [tart is 

 triangular in outline. ; this is to be accounted for, from the fact that 

 the mentum is generally deeply seated among the oral organs, only 

 the triangular carination of its exposed surface being clearly visible; 

 this was undoubtedly mistaken for the entire mentum, although it is 

 difficult to understand how its form could have been overlooked during 

 the process of dissection. 



It will be seen that I have departed from the usual mode of treat- 

 ing the genus Stenus, by dividing it into two genera, according as the 

 tarsal joints are simple or bilobed. This will probably meet with 

 opposition and perhaps be totally discarded ; I have therefore not 

 repeated specific names in the new genus, so that they can be united 

 without causing confusion. Most of the species of Areus can be 

 recognized at a glance, without examining the tarsi, indicating a 

 more or less natural division ; and for those species which may be 

 considered connecting links, the propriety can be no less than that 

 which warrants the continuance of many genera of Carabidae, as, for 

 instance, Selenophorus. There is no species in our fauna which 

 cannot be placed at once in its proper genus, and the division is 

 certainly to be recommended on the score of convenience, having in 

 mind the fact that generic limits must always be more or less arbi- 

 trary. 



In our fauna the genus Areus is much less homogeneous than Stenus, 

 having forms such as advena, megalaps, and Sayi, which may with 

 further study be considered as possessing characters of generic value. 

 In Stenus, when the third or fourth ventral segment is modified in 

 the male, the fifth is invariably affected, while in Areus, the second, 

 third, and fourth segments are often modified, when the fifth is per- 

 fectly normal. 



Explanatory Notes. 



In a group in which the species are very closely related, long and full 

 descriptions are a necessity, and I have perferred to make each description 

 complete in itself, and all of like construction. To save time in comparisons, 

 similar parts are defined at corresponding places in each. This form is the 



