J. E. 1NGPKN ON SCALE EVOLUTION. 97 



few pigment granules. A more developed hair showed traces of 

 longitudinal markings. 



4. — A sharp, slender, swallow- tailed scale, with strong longi- 

 tudinal lines, and short transverse lines between them. 



5. — A wider scale, with the sides convex and the ends blunter. 



6. — A shorter, wider scale, with two rounded ends. 



7. — A similar one, with three rounded ends. 



8. — A longer and narrower form, with two bluntly pointed 

 ends. 



9. — A leaf-shaped scale, with a sharp point. 



10. — A paddle-shaped scale, longer and narrower than the last, 

 and also sharply pointed. 



11. — A scale longer, blunt at the end, and apparently thicker. 



12. — Longer, narrower, and more solid in appearance. 



13 and 14. — Scales or hairs at the margins of the wings, long, 

 bifurcated or trifurcated. All these varieties show strong longi- 

 tudinal markings. 



The scales on the orange and yellow borders to the hind wings 

 and on the white spots are of an ordinary type, similar to 6, 7, and 

 8, but they have not the strong longitudinal markings charac- 

 teristic of the others. 



15. — In a long oval depression on the upper front margin of the 

 hind wings is a bundle of what I believe are considered to be scent 

 hairs. These differ from any of those already described. They 

 average a quarter of an inch in length, and the largest are nearly t ^qo 

 of an inch in diameter. They are brown, cylindrical, slightly tapering 

 towards the rounded ends, and are firmly attached to the wing. 

 They lie on a bed of cream-coloured scales, somewhat of the shape 

 of No. 6, but with very obscure markings, These hairs have a cell 

 structure suggestive, in miniature) of that of the deer, with a 

 distinct medulla. This structure is probably well known to those 

 who have studied the characteristics of scent hairs. On the under 

 margin of the front wings are a number of scales, which, when the 

 wings are closed, are in contact with the scent hairs, and appear 

 to have a somewhat similar structure, but less developed. The 

 striae are only on one side, on the other are traces of cell-structure 

 and pigment dots. 



The attachment of the hairs first described, 1, 2, 3, etc., to the 

 wing, is very pretty — the socket looking like the neck of a bottle 

 with a thickened rim. 



