94 NATURAL SCIENCE [February 



speaking, inaugurated with the hindwings, and that these usually 

 exhibit the effects more plainly than the forewings in one and the 

 same specimen. We should logically expect this to be the case 

 from the entire direction of specialisation in insects, and we may 

 regard it as arising ultimately from mechanical causes. 



5. The general parallelism of the evolutionary movement of the 

 neuration in the two main directions throughout the lepidoptera, as 

 explained by me, exposes its secondary nature. 



We may now apply these observations to the classification of the 

 Palaearctic and Nearctic butterflies especially, premising that the 

 facts recorded here do not furnish everywhere the basis for a new 

 system, but are mainly of use to throw light on the phylogeny and 

 correct the sequence of all the groups and forms. We endeavour to 

 exhaust the capabilities of the neuration as a guide to the under- 

 standing of the butterflies. We use the characters of the neuration 

 especially as the readiest index to decide upon the position of genera, 

 in order that we may avoid the common mistake of deriving the 

 generalised from the specialised genus, a mistake we believe to find 

 in the recent ' Handbook ' of Mr Meyrick, and elsewhere. Com- 

 mencement is made with the more specialised groups in each family, 

 and the sequence is the one recommended to be followed in books 

 and collections. 



A. Papilionides. Vein IX of forewings present, a last more or less curved and 



shortened, free and longitudinal vein running from base of wing to internal margin. 

 Fam. I. Parnassiidae. Vein IV2 of forewings inclines towards tlie cubitus. 

 No cross-vein between cubitus and vein VII. 



Sub-fani. 1. Parnassiinae. Vein IVi springs from the radius ; the radius 



is specialised and four-branched : IIIi -fo, III3, III4, Ills. 

 Sub-fam. 2. Zcrynthianac (Thaidinae). Vein IVi springs from cross- 

 vein ; the radius is five-branched. Vein III3 shows a diverging 

 movement, and has slipped forward to beyond the cell. 

 Fam. II. Papilionidac. Vein IV2 maintains a central position on forewings ; 

 on hindwings it inclines towards the cubitus. A cross-vein between cubitus 

 and VII. Radius five-branched, generalised. 



B. Hespeeiades. Vein IX of forewings absent ; vein VIII forms a loop at base of VII, 



sometimes reduced to a scar, again vanished. This character is not exclusive of the 

 moths; Sphingides, Saturniades, etc. 



[Group Pieri-Nymphalidae, characterised by the more unequal S2iacing between 

 the veins and the special line of progress in the disintegration of the 

 media and its system, which runs parallel to the course pursued in the 

 Papilionides, as well as in most other lepidoptera, by the branches not 

 remaining stationary, but moving to radius or cubitus, and ultimately fusing 

 with the systems of the latter primary veins. ] 

 Fam. III. Picridac. Radius three- to five-branched. This family diff"ers from 

 the succeeding four by its specialisation in two directions. It differs from 

 the brush-footed butterflies by the movement of IVi, which leaves the 

 cross-vein and emerges from radius in the Pierinae ; the Nemeobiidae repeat 

 this movement on hindwings. 



Sub-fam. 1. Pierinae. The first branch of radius springs from upper 

 margin of cell. The absorption of veins gives no tribal characters. 

 The mass of genera belong here. 



