21 



sidered to be a family markfortheCicindelidsin distinction from 

 the Carabs but even if the number of ocelli in this way looses 

 its great importance as a family character, it surely cannot 

 loose all its value, and it will certainly, not only within the 

 Lebiini group, but also wherever it may be found, prove to 

 be of considerable value as a character within the groups. In 

 order to accentuate the fact I refer to my work about 

 the ocelli and their position, see illustration I, fig. 10, which 

 shows the position of the ocelli of Lebia chlor ocephala, while 

 fig. 7 shows that of Odacantha melanura and which at the 

 same time represents the position of the ocelli of the group 

 Dromiini. 



Our knowledge of all the species has hitherto been 

 too confined to give a perfect system, even within the Ca- 

 rabs belonging to the Tnincalipennes section, and any system 

 would be of no use as soon as new species were found and 

 the following classification must therefore be considered as 

 quite provisional. (I have included 2 exotic genera which 

 I do not doubt belong to the group Odacanthini). 



The Carabs belonging to the Trimcatipenues section are 

 divided as is known, in the following way: 



a. Brachytinidae : J* S and $ with 7 ventral segments. 



b. Lebiinidae: ^ and $ with 6 ventral segments. 



Group : Lebiini. 



The larvae with i ocelli on either side, clypeus without teeth. 

 r Genus: Lebia. 



Group : Odacanthini. Group : Dromiini. 



The larvæ with 6 ocelli on either side, clypeus with teeth. 

 Genus: Odacantha. Genus: Dromius. 



— Ophionea. — JJemetrias. 



■ — Casnonin. 



