( 385 ) 



The sense-bristles of tlie ventral surface have an irregular and variable position : 

 on the joints of the clnb we find here and there an irregular transverse series of 

 apical bristles, while mostly there are only some lateral bristles developed (tigs. 43, 

 44, lateral views). Dorsally the bristles of Luchdorfa are long and .strong (f 46). 



The antennae are always withcmt scaling, except on the most proximal joints ; 

 but in Sericinus we find some very long hair-like scales scattered over the dorsal 

 surface. 



Setiferons jinnctnres are niimerons, especially at tlie sides of the joints. 



c. PARNASSIINAE. 



The joints of the antennae, wliich are more or less compressed, are of two 

 different t^-pes connected by intergradations. (1) The ventral surface is without 

 groove, and the fine sense-hairs are distributed over the basal area of the joint. 

 (2) The joint has an impression of irregular shape (figs. 33, 34) ; the groove is some- 

 times nearly as regular as in Pieridae, but it is a rule that the grooves on the joints 

 of the same antenna are different in shape and size (f 34) ; on the club the groove 

 is always very irregular, often divided into several branches (f 3ii) ; the fine sense- 

 hairs are restricted to the grooves, if these are large enough. The most important 

 point is, that antennae with and without grooves occur in the same species, as an 

 examination of a series of specimens of P. njiollo from one locality proves. The 

 groove on the joints of the stalk is not mesial, but has a more exterior lateral 

 position than in Pieridae. 



The sense-bristles are ventrally not well develojied and of no regular dis- 

 tribution ; but on the dorsal side of the joints they are long and strong, on the scaled 

 as well as on the not-scaled joints (f 35). 



Setiferons punctures are very numerous, and there is scarcely a line of dis- 

 tinction between the fine sense-hairs and the hairs situated in these punctures. 



The scaling is mostly dense and rather much e.\tended. It occupies the dorsal 

 surface from the base to the apex (P. smintheMs, i-hodius) ; in some sj)ecies the stalk 

 is also fully scaled ventrally, except some distal joints ; but very often the whole 

 clnb is not scaled dorsally, and the ventral surface of the clnb and of most joints of 

 the stalk is also bare of scales. In a few forms the scaling is very scanty, the 

 antenna appearing almost not-scaled (stubbendorji). 



6. yy.^rriiALWAE. 



Under this family term I unite, in accordance with the views of many Lepi- 

 dopterists, all the Butterflies which do not belong to any of the preceding families : 

 Ilesperiiddi', Li/caeindae, Enjcinidae, Pieridae, and Papilionidae. But, by thus 

 treating Lihj/theinae, Satyrinae, etc., as subfamilies of y>/n//datlidat', 1 do not wish 

 to state as my opinion that these groups of genera will have subfamily, not family, 

 rank in my final classification of Butterflies. ISo much, however, is certain, that all 

 these Nymphalid Butterflies are, in respect to the antennae, much closer allied 

 between themselves than with any other Butterfly, and that 1 am therefore ]>erfectly 

 justified in treating here the now following groups of Butterflies as one great section 

 remarkably different from all the ^(receding families. 



The Nymj)halid Butterflies, without exception, have two ventral grooves to 

 each not-scaled joint, one at each side of the median line. The fine sense- 



