150 LEFIBOPTERA INDICA. 



between the two broods of wet-season butterflies (May — June and August — Septem- 

 ber), I have never observed any difference. But between sjDecimens of tlie wet and 

 dry-season broods there are in many genera very perceptible differences. There is a 

 difference in size, the wet-season specimens being generally smaller, and there are 

 minor differences in the angulation of the wings and in the tone and purity of the 

 colouring below. But the most remarkable difference is in the presence of large eye- 

 like spots on the underside of the wet-season forms, either absent or greatly reduced 

 in those of the dry season. To give an example, the Sarju Valley was on September 

 20th full of ragged specimens of a strong-ocellated butterfly {Mycalesis mineiis proper), 

 which a week later was succeeded by swarms of a larger, more angulate, non-ocellate 

 insect {Myc. visala), precisely similar in its structure, and especially in those very 

 complex organs, the prehensores, which in general mark by strong differences the 

 slightest specific variations. In the same way Melanitis Leda was succeeded by the 

 non-ocellate Mel. Ismene, and Junonia Asterie by the non-ocellate J. Almana. Such 

 facts are best studied in countries where the wet and dry seasons are well marked, 

 and there is none better than Travancore, where in 1882-3 I first observed these 

 curious metamorphoses. Hitherto I have only found them to exist in Junonia, 

 TplitJiima, Melanitis, and in the following groups of ilf?/caZests — Galyslsme, Orsotricena, 

 Telinga, Gareris, Virapa, and Samanfa. Many genera of the Satyridse seem to pro- 

 duce only the two wet-season broods, and so fail to show what variations the weather 

 would produce in their markings. Again, some dry countries produce only the dry 

 forms, and some wet countries only the wet forms. For instance, I believe that 

 Sindh produces only the dry-weather form Junonia Almana, while the wet climate of 

 Ceylon and Singapore produces only /. Asterie. In dry Orissa, the wet and dry 

 forms of Orsotricena Medus and 0. Buneka are tolerably distinct, and the latter is 

 the commoner of the two. Bat on the damp Malabar Coast, of the representative 

 forms Orsotricena mandata and 0. mandosa, the latter, the dry-weather one, is scarce 

 and imperfectly differentiated. The cause of this seasonal dimorphism would be a 

 curious subject of study, but I have no suggestions to offer myself. Mr. deMcdville 

 has recently observed to me that perhaps, owing to the rank vegetation of the rainy 

 season, the butterflies are better concealed, and have therefore been free to assume 

 these handsome spots, under the influence, perhaps, of sexual selection. This view 

 is, I think, somewhat strengthened by certain facts. The dry season forms are all 

 more or less leaf-like in shape and colouring, while in the wet-season ones there is no 

 such mimicry. This argues greater exposure to danger in the dry season. Eemem- 

 bering, however, the remarks of Darwin and Wallace on the protective nature of the 

 ocelli of birds, borne out by the testimony of all sportsmen, and observed by me in 

 the case of the common wild Peacock, I am inclined to think that the ocelli are a 

 direct protection to the insect during tlie rains. Certainly that is likely to be the 



