122 GEO. H. HORN, M. D. 



It is curious in this tribe that Ardistomis should "have the elytral 

 margin interrupted with an internal plica. It thus shows considerable 

 more affinity with the HarpalinEe than do the other genera and seems 

 to be the nearest Carabine relation of the Panagaeini, in place of the 

 Cychrini as suggested by most authors. 



Sub-Family UARPALINjE. 



Middle coxal cavities entirely enclosed by the central pieces of the 

 meso- and metasternum, the epimera not attaining the coxae. Head 

 with setigerous punctures over the eyes. Thorax with setigerous punc- 

 tures at the side and posterior angle very rarely without the latter 

 and still more rarely without either. Anterior tibiae always either 

 obliquely sinuate or deeply emarginate within, the inner spur remote 

 from the apex. 



These characters seem to be the only ones in which all the tribes 

 agree. As there are many points in which wide differences occur these 

 will be left for discussion in their proper places. 



For convenience of study the sub-family may be divided in two 

 grand sections. 

 Head with two supra-orbital setigerous punctures. 



Harpalinje bisetos^. 

 Head with one supra-orbital setigerous puncture. 



Harpalin^ unisetos^. 



Small as this character may seem it is probably one of the most 

 invariable of any that have been suggested for the division of any large 

 series of genera or tribes. I have never observed an exception, although 

 Bedel * says that in two European Amara one has but one supra-orbital 

 seta (^spectabilis) and the second (^pyrenxa) none whatever. 



When two setae occur the anterior is close to the margin of the 

 eye in front, the posterior is a little remote from the eye opposite the 

 posterior margin. When there is one seta it is almost always a little 

 removed from the margin of the eye and is situated opposite the middle 

 of the eye or a little posterior to that point. 



The Harpalinae as here constituted seem to be the true development 

 of what might be called the Carabide idea of the present geological 

 period. There is evidently a close relationship in the entire series with 

 fewer breaks in the line of affinity and with very few genera that are 

 abnormal or* specially differentiated in the sense in which we observe 



Ann. Fr. 1870, suppl. p. 52, note. 



