34 BUTTERFLIES OF MOUNT PENRISSEN. 



(32) Cycmiria strqpkis (Druce). 



(33) Cyaniris planta (Druce). 



This species leads up to C. haraldus (Fab.) placed by some 

 authors in a separate genus — Lyccenopsis. 



(34) Kacaduba sp ? 



A couple of females only were taken, and though I am 

 strongly of opinion that they will constitute a new species. 

 I defer a description until I can procure a male. The 

 colouration and markings are much the same as in N. aluta 

 (Druce.) but there is no basalband on the underside of 

 the forewing as in that species. 



(35) Arhopala similis (Druce). 



This is the variety a of A . agesiase (Hew. ) 



(36) Tajuria isaeus (Hew.) 



(37) Biduanda sp ? (Nov.) 



I am pretty confident that this species of which one female 

 example was taken, will prove to be new, but I must 

 defer a description of it until I can obtain access to larger 

 collections and more recent literature. The species ap- 

 pears to be related on the one hand to Biduanda thesmia 

 (Hew.), on the other to Biduanda Jaritsoni (Druce.), judg- 

 ing at least by the pattern of the wings on the under- 

 side. If ever the phylogeny of the Lycanida? comes to 

 be traced, the wing patterns on the underside must care- 

 fully be taken into consideration ; from a study of these, 

 extending over three years, I have come to the conclusion 

 that the more highly organised the butterfly, the more 

 the underside wing-pattern tends to disappear from the 

 discs of the wings and concentrate at the margins and at 

 the anal angle of the hind-wing, in some cases disappear- 

 ing even from these areas, as for example in Bornean ex- 

 amples of Loxura atymmvs. A reason for this concentra- 

 tion of wing-pattern may perhaps be sought in the follow- 

 ing considerations : — the typical Lyecrnid wing-pattern on 



