( 413 ) 



4. Sense-bristles snbaiicestral, or specialised as in Pierinae or as in certain 

 Lycaenidae. Connection with Pieriduc and Li/cacnidae ; no connection 

 with Papilionidae and Ni/mpludidae. 



IV. Pieridae. 



1. Scaling snbaucestral, or specialised as in other families. Dorsal side never 



eutirelj- withont scaling. 



2. Fine-sense hairs specialised as in Enjchiidae, Init more liiglily so. Pieridae 



cannot be the stock from which any other family branched off. 



3. Configuration of ventral snrface specialised, of the Erycinid type (Pierinae), 



or of a type foi;nd nowhere else {Dismorphiinne^ : one-grooved and 

 three-grooved antennae. No other family can be derived from the 

 Pieridae. 



4. Sense-bristles specialised as in certain Erycinids, sometimes obliterated. No 



connection with Xi/mphaUdac and Papilioiddac. 



V. Papilionidae. 



1. Scaling subancestral, or specialised as in other families, or reduced as 



only among Nymphulidae. Connection between I'n iiiUonidae and 

 Nymphalidae. 



2. Fine-sense hairs ventrally generalised, or specialised of a type peculiar to 



tlie family, or specialised as in Symphalidae. Development of setiferous 

 pnnctnres on dorsal and ventral snrface as in Nymphulidae. Pa iiilionidae 

 cannot be a derivation from Nymphalidae, Pieridae, or Erycinidae. 

 Close connection between Papilionidae and Xytnphalidac. 



3. Configuration of the ventral surface ancestral, or sjiecialised, of a type 



peculiar to the family or similar to that of Nymj>halidae. Conclusions 

 as before. 



4. Sense-bristles snbancestral, or specialised ; specialisation sometimes in tlie 



same direction as in Nymphalidae. Conclusions as before. 



V I. Xymplui I id a e. 



1. Scaling subancestral, or specialised as in other families ; loss of scales as 



only in Papilionidae. Clonnection between Nymphalidae and Papilio- 

 nidae. 



2. Fine sense-hairs specialised as in certain Papilionidae, sometimes nearly 



all obliterated ; on dorsal and ventral surface appearance of setiferous 

 ]iunctures as among Papilionidae. Close connection between Nympha- 

 lidae and Papilionidae ; Symphalidac cannot have given origin to any 

 other family. 



3. Configuration of ventral suri;u;o of peculiar type, but resembling that of 



certain Papilionidae : two-grooved antennae ; carinae found nowhere 

 else in Butterflies. Conclusions as before. 



4. Sense-bristles specialised, of a peculiar constant type, recalling that of 



certain Papilionidae. Conclusions as before. 



It will lie seen from this summary that the Lycaenid antenna is the most 

 generalised and the Nymphalid antenna the most specialised of all Butterfly 

 antennae, and it will further be noticed that certain specialisations are common to 



