TRANSPARENCY IN WINGS OF LEPIDOPTERA. 117 



that it was always so, and, although the Af/i/rta is now about five times 

 as abundant as the Euafira, the latter must have had a start of the 

 former in developing its transparent patches. 



EsTHEMopsis SEKiciNA. — Unfortunately the only specimen available 

 for observation has a considerable amount of fungus growth on it. 

 The same general statement can, however, be made as with the other 

 species, viz., that the transparency apparently arose earliest in the 

 hindwing, then in the central area of the forewing, and last of all in the 

 apical patch. The normal shape of the scales is more or less rectangu- 

 lar, with irregular teeth at the upper end. In the transparent portions 

 the scales have first lost the teeth, and have then become rounded at 

 the upper end, and become shorter, and finally almost circular. 

 But, as before, the normal scales evidently first became less closely 

 packed together. Having become rounded the scales are seen to be 

 transparent. The remarkable thing in connection with this insect's 

 transparency is that there are two layers of scales, the one overlaying 

 the other, and both have become transparent, the normal black 

 scales being completely transparent and colourless, the overlapping 

 blue scales having become bluish transparent. Some scales are to be 

 found that are only partially transparent, the transparent portion 

 being of the bluish tinge. It is likely, therefore, that these scales 

 were black and blue, the latter portion having become changed only. 

 Some of the scales of the cilia are two-coloured without any change to 

 transparency. 



losTOLA DmsA. — 111 the cell of the hindwing the scales are trans- 

 parent, and smaller than in the transparent portions of the forewing. 

 They are long, and without any tooth-like projections, and are 

 thinly distributed. In the transparent portion of the cell of the fore- 

 wing the scales are still a considerable distance apart, and without 

 tooth projections are quite transparent, and are also reduced in size, 

 but scarcely so much so as in the hindwing. The dark, almost black, 

 scales are usually tridentate, the plain, non-toothed, scale is, therefore, 

 in all probability an advancement. Near the inner margin is a long 

 streak of blue scales ; these are semitransparent, or partly transparent, 

 and generally have indications of being bidentate, though blunted con- 

 siderably. The broad subapical transparent patch of the forewing has 

 the transparent scales much less numerous than in the cell of the 

 forewing, while the small apical patch shows the scales to be only 

 semitransparent, and to be closely packed together, directly pointing 

 to the latest and least development towards transparency. A <? and $ 

 examined. The $ has the apical patch less developed than the $ . 



PsEUDARBESSA DEcoRATA. — Although not nearly so transparent as 

 the other species, the interest attaching to the present instance is 

 every bit as great, for here it is possible to guess pretty accurately the 

 order in which the methods of obtaining transparency arose. To the 

 unaided eye it is noticeable that the central portion of the hindwing, 

 or that lying within the discoidal cell, is the most transparent portion. 

 Under a :^-lens the scales, which are all of the same shape, are more 

 or less themselves transparent, and are spaced further apart than in 

 the contiguous portions of the wing. In the semitransparent patches 

 of the forewing the scales have arrived at the same degree of trans- 

 parency, but are closely packed together, and this alone accounts for 

 the very much less transparent look they have to the naked eye. It 



