72 Journal New Yokk Entomological Society. [Vol. vii. 



to Scymnus. Again, our representatives of Cryptognatha are likewise 

 widely separated from the Cryptognatha of Mulsant, and form in reality 

 one of the most isolated types of the family, the special character re- 

 lating to the presternum, which caused LeConte to associate them, 

 being of subordinate value and liable to appear in any tribe ; it exists, 

 for instance, in Stethorus of the Scymnini, and in Nipus of the Crano- 

 phorini, though not the distinguishing feature of that remarkable type. 

 In Zagloba of the Scymnillini it also tends to reappear. Again the 

 genus Rhyzobius is tribally distinct from Scymnus in the structure of 

 the eyes, antennae and epipleurce. 



The character relating to the anterior coxal cavities, announced by 

 LeConte, is apparently of no significance even if wholly true, as it 

 would bring together genera with no special affiliation otherwise, and 

 the character made use of by Mulsant to separate Coccinellini from 

 Cariini is of no value, there being no tribal difference between Cocci- 

 nella and Synonycha, in spite of their general dissimilarity of habitus. 



The abdomen is composed throughout of five segments, but the 

 genital armature sometimes becomes distinct and assumes the form of 

 a sixth segment. This character is very useful in the classification of 

 the tribes related to Chilocorini, and of the compact Coccinellidas hav- 

 ing narrow epipleura;, as will appear; it generally affects both sexes 

 and is particularly developed in the Hyperaspini. The Hyperaspini 

 of Chapuis include several distinct tribes, and those with but five 

 ventral segments should be removed, the retractility of the legs and 

 epipleural depressions not being tribal characters necessarily, but ap- 

 pearing in several tribes with the legs generally free. 



The tarsi in this family are in reality 4-jointed, the third small and 

 generally forming a rigidly anchylosed basal lobe of the last, but it is 

 sometimes free or partially so. The second is lobed beneath, the 

 lobe truncate at tip and hollowed on its upper surface, not bi lobed as 

 stated by Crotch (Rev. Cocc, p. 53). 



In the following pages I have made use of all generic types, for- 

 eign and native, which have been accessible to me, and regret that 

 my exotic material might not have been more extensive. Where 

 names not belonging to the fauna of the United States are introduced 

 they are preceded by an asterisk. 



Crotch employs the name affinis Rand., for the species venusta 

 and notulata, but in error, as affinis, of Randall, is simply a synonym 

 of Hyperaspis binotata Say. 



