A LIST OF BUTTERFLIES OF THE PALNI HILLS. 391 



D. plexippus, and D. septenfrionis, Junonia hierta, J. orithya, J. lemo- 

 nias, Atella phalantha, and a small skipper. Sometimes large 

 numbers of dragonflies accompany the butterflies in the southward 

 flight. 



It has been noted that on reaching a bank of fog near the 

 precipices on the south side of the Palnis a specimen of P. hector 

 was turned back by it. The instinct impelling a southerly 

 direction of flight was evidently operating even with a contrary 

 wind and clouded sky, but the fog was baffling, and the butterfly 

 had to retreat before it, returning however again and again right 

 up to the edge of the wall of mist. 



Towards the end of November 1.908 two opposing streams of 

 Gatopsilias, with a few Danaids and Papilios were observed flying 

 in almost opposite directions, one stream moving south and the 

 other passing through it approximately in the direction north by 

 east. It was noticeable that the butterflies showed not the slightest 

 indecision or confusion at the crossing point, each individual 

 maintaining its original direction of flight quite independently of 

 the others. It is evident, therefore, that the instinct is not in any 

 way dependent on numbers, but operates separately in each in- 

 dividual. 



A feature of the autumn migration of 1909 was the very large 

 number of the skipper Pamara mathias which accompanied the 

 other species. 



These migrations appear to occur over the whole of the grassy 

 upland region of the Palnis, where very few of the species breed. 

 Apparently the butterflies ascend from the plains on one side of the 

 range, cross the hills, and descend again on the opposite side ; but 

 I have been unable to discover how far the flights are continued in 

 the plains. P. d&moleus has been observed migrating in large 

 numbers a dozen miles or so to the south-east of the Palnis at the 

 same time that the migration was in progress on the hills. 



Co-operation by observers in other stations is much to be desired 

 in order to trace out the whole course of a migration, and it is 

 possible that systematic observation over extended areas might 

 throw much new light on the origin of this mysterious instinct. 



