LARVAL FORMS OF COLEOPTERA 



Eac'li Avitli tliree hooks 4 



2. Exterior liook falcate and outAvardly concave; basal joint of 

 labial i)ali)ns Avitli two or three spines. 



Cichidelin i ( C in c in dela ) 

 Exterior hook straight or slightly concave toward the middle 



line; basal joint of labial palpns without spines 3 



8. Dorsal pair of ocelli snbeqnal in size 



Tetrach in i ( Teivacha ) 

 Posterior one of dorsal ocelli decidedly larger than anterior. 



.1 nt bl iicheilin i ( Am hlijch e ila ) 

 (pi. 4 B. D) 

 4. Exterioi- hook much smaller than the other two. (Tibia and 



tarsus small but separate) Omini (Onuis) 



All three hooks of about the same size. (Tibia and tarsus 

 separate {Theratesf), or fused) 



CoUyrini (CoUyris, Ctenos- 

 foma, and T her cites (?)) 

 (pl. 4 A) 



B.II. CARABIDAE 



Out of the nineteen subfamilies into which the Carabidae have 

 been divided here according to the characters of the larvae, a single 

 one, the Lebiinae, may not be natural. The evidence of a close rela- 

 tionship between the genera which have been included in it is not 

 strong and, considered as a group, its affinities to other subfamilies, 

 particularly to the Dromiinae, can hardly be traced. Furthermore, 

 because of extreme adaptation to an ectoparasitie life in all or 

 some of the larval instars, it is not always possible even to recognize 

 the larvae of some of its species as carabid larvae. In the following- 

 key the main character common to the genera of the Lebiinae ap- 

 pears rather insignificant but it sets the subfamily oif from all 

 other carabid larvae. Two of the remaining subfamilies, namely, 

 the Dromiinae and the Loricerinae, occupy an isolated position, but 

 the rest intergrades either with one or with several of the other sub- 

 families. The Dromiinae is in itself a homogeneous and natural 

 groui), and the larval form of the subfamily Loricerinae, repre- 

 sented by a singU^ genus only, is very characteristic and strikingly 

 different from other carabid larvae. The larvae of the Odacan- 

 thinae show no close relationship to the larvae of the two other 

 "Truncatij)ennes", the Lebiinae and the Dromiinae, but they 

 approach the Nebriinae. The Driptinae are closely related to 

 the Nebriinae which are rather distant from the C^arabinae, 

 according to their larvae. The Cj'chrinae are closely related both 



18 



