SUBFAMILY IV. CONOCEPHALINJE. 537 



(1'JOOa) and myself (1903, 382) in our former treatment of the 

 genus, viz., the presence or absence of spines on the under side of 

 the hind femora, is unreliable as it is variable in the same species. 

 They have used instead the form of the male cerci as their prin- 

 cipal differential character and have supplemented this with other 

 characters pertaining to the form of lateral lobes of pronotnm, 

 form and size of ovipositor of female, etc. Their keys are open 

 to the objection, that they contain a number of merely compara- 

 tive terms such as "form more robust" and "ovipositor with gen- 

 eral form less arcuate," as against "form more slender" and "ovi- 

 positor with general form more arcuate." Such terms, as I long 

 ago discovered by personal experience, are confusing and mis- 

 leading to a student who has only a few specimens at hand. Aside 

 from this, the keys in their synopsis are so much better than any 

 previously known that I have used them as the basis for the one 

 given below, modifying them here and there in sequence and by 

 the addition of any fixed character which, in my opinion, would 

 make them more simple and more easily understood by the be- 

 ginner. On account of the great confusion in the synonymy of a 

 number of the species, it is impossible in giving the distribution 

 and notes on habits to always refer each record made by past 

 collectors to the species to which it rightfully belongs. I have, 

 however, where following the synonymy as given by R. & H., as- 

 signed the notes and records of synonyms to the species under 

 which they are placed by those authors. 



Fifteen species and two varieties of Orclicliniitni are herein 

 recognized as occurring in the territory covered by this work. 

 For convenience of treatment these are first separated into tw r o 

 groups as follows: 



KEY TO CROUPS OF EASTERN SPECIES OF ORCIIELIMUM. 



a. Cerci of male of moderate length, the apical portion behind the me- 

 dian tooth not much if any longer than the basal portion in front 

 of it (Fig. 179); ovipositor in the majority of the species not dis- 

 distinctly more than half as long as hind femora and (except in 

 gladiator) with upper margin always evidently and regularly 

 curved. GROUP I. p. 537. 



aa. Cerci more elongate, the apical portion behind the tooth distinctly 

 longer than the portion in front of it (Fig. 183); ovipositor 1 (except 

 rarely in continuum) distinctly more than half as long as hind fe- 

 mora, its upper margin either straight or curved. GROUP II. p. 551. 



GROUP I. THE AGILE GROUP OF ORCHELIMUM. 



This group as defined above contains eight species. In general 

 its members are larger and more robust than those of Group II. 



