220 Mr. H. W. Bates on the Papilionid© 



vules. Pupa secured by the tail and a girdle across 

 the middle in an upright position, 



Subfam. 1. Papilionince. 



Subfam. 2. Pierince. 

 Family 3. Lyccenidce. Six perfect legs in 5 ; four in S , the fore 

 tarsi wanting the tarsal claws, but densely spined 

 beneath. "Wing-ceUs (except in Eumoius*) not closed 

 by perfect nervules. Pupa seciu^ed by the taD. and a 

 girdle across the middle. 



Family 4, Erycinklce. Six perfect legs in § ; four in S > the fore 



tarsi consisting only of one or two joints, and spineless. 

 Subfam. 1. Erycinince, Pupa recumbent, flattened beneath, 



secured by the tail and a girdle across the middle. 

 Subfam. 2. Stalachtince. Pupa not flattened beneath, secured 



rigidly by the tail in. an inclined position, without 



girdle f. 

 Subfam. 3. Lihytheinm. Pupa suspended freely by the tail. 



Family 5. Nymphalklce. Fore legs imperfect in both sexes : in the $ 

 wanting the tarsal claws ; in the S the fore tarsi aborted, 

 consisting only of one or two joints. Pupa suspended 

 freely by the taU. 

 a. Lower disco-cellular nervule, especially of the liiud wing, more 

 or less atrophied. 

 Subfam. 1. NympJialhue (^Nymplialklce, Ageronidce, Eurytelidce, 

 and MorphkJce (part.) of authors). 

 h. Lower disco-cellular nervule perfect. 

 Subfam. 2. Brassol'moi. 

 Subfam. 3. Satyr ince. 

 Subfam. 4. Danaince. 

 Subfam. 5. HeUconinceX- 

 Subfam. 6. Acrceincn. 



* This genus constituted a distinct family in the system of Boisdural (Spec. 

 Gen.) ; it was incorporated with the Lycaniidaj by Westwood in Doubled. & Hew. 

 Gen. Diurn. Lep. If the lower disco-cellular nervule prove to be aborted con- 

 stantly in the numerous aberrant genera of Lycii^nidae of Eastern Asia, the genus 

 Eumceus might form a subfamily of Lycaenidaj founded on the closure of the 

 wing-cells. 



t The4;ransformations of Sfalachtis have not yet been recorded. The statement 

 here made is founded on unpiiblished observations of my own on St. Calliope. 



I The constitution and affinities of the subfamilies Danain(B, HcliconincB, and 

 AcraincB will be discussed in a paper on those groups which I shall publish 

 shortly. 



