CARABIDAE. 55 



Apical angles of joints 1 — 3 of anterior tarsi prolong(>d in S2)inos. 

 Eyes large. Hind angles of thorax obtuse or rounded. 



Polpochila. 

 Apical angles of joints of anterior tarsi not prolonged. Ivves small. 

 Hind angles of thorax sharply rectangular. Cratacanthus. 



Mentum not toothed. 



Posterior tarsi with the first joint a little longer than the second, 

 outer edge of middle tibic-B rather Hat, and with a double row 

 of spinules closely placed. 

 Eyes relatively small, distant beneath from the mouth ; elytra 

 with numerous dorsal punctures. Piosoma. 



Eyes relatively large, very narrowly separated from the moutli ; 

 elytra with one dorsal puncture. Agonoderus. 



Posterior tarsi with the first joint nearly as long as tlie next 

 tliree. Middle tibi?e with the spinules sparsely placed, in 

 the male arcuate and serrate on the inner side. 

 Eyes rather small ; three series of elytral punctui-es. 



Discoderus. 



The sexual cliai-acter.s are not very well marked. The males 

 have four joints of the anterior tarsi feebl}' dilated (two in Pol- 

 pochila) and rarely (Discoderus) with a few squamules beneath. 

 The latter genus has the middle tibijB distinetly arcuate and 

 serrate within. In Cratacauthus tlie right mandible of the male 

 has the basal portion which borders the clypeus more elevated, 

 while the upper edge in front of this is much depressed ; a similar 

 structure is observed in Acinopu^. 



Afjonoderus and Pogonodaptus are the only genera observed 

 in which the penultimate joint of the labial palpi is bisetose. 

 Nolhopuii and Piosuina have the ligula qiiadrisetose, and the 

 paraglosste ciliate externally at tip, the upper surface is also 

 sparsely setose in these genera. In Gralacantlius the paraglossa; 

 are very broad, and lie behind the ligula, so that when viewed 

 from the front the entire ligula has very much the ai)pearance of 

 that of a Lebiide. 



Pogonodaptus has been established on a .«niall species from 

 Texas, resembling Daptus and somewhat also Pogonus (Pogo- 

 nistes). 



(Jroup II. — Harpali. 



It is extremely difficult to draw the line with accuracy between 

 this group and the Dapti, and it is probable that other characters 

 will be found which will separate the genera, but which will not 

 allow the groups to remain as at present constituted. 



