2l6 WING SCALES OF LEPIDOPTERA. 



everyway. These butterflies are illustrated in Plate III., 

 figs. I and 6. 



In fig. 14 are shewn three butterflies from India, 

 Paralitica agica, Oriiioma daman's and Hcstiua uaiiia, 

 belonging to the sub-families Danaiiuc, Satyriiue and 

 Nymplialifue respectively. These insects are in no way 

 related, although found together or in the same localities, 

 and of very, much the same appearance in every way. (See 

 also illustrations in Plate V.) But the wing scales are 

 quite different in all three cases, as are also those of two 

 other butterflies, a /V/vvV/ and a J\ipilio, which simulate 

 them in the same region. 



Fig. 15 illustrates two butterflies of remarkably similar 

 appearance, a Danaid, Cadiiga tytia, and one of the 

 Papilionincc, Cadugoidcs ao^cstor. They are both light grey 

 and pearly, with black lines on the fore wings, and deep 

 chocolate markings on the hind wings. But the resem- 

 blance is produced in quite a different manner ; in one the 

 scales are thinly dispersed, light chocolate, and shaped 

 like long thin skittles ; in the other they are thickly placed, 

 even overlapping, rounded and pure white, with a sprinkling 

 of dead black scales amongst them. The effect is, however, 

 an exact similar appearance in every detail, of colour, 

 and markings. 



The conclusion to be drawn from these differences in 

 detail producing such striking resemblances is, that the 

 attribution of simulation to natural selection is the correct 

 one. Variations are constantly occurring in all directions, 

 and the scale-varieties which result in the remarkable 

 outward likenesses with [irotectivc results to the insects 

 which possess them, can be the result of no other process. 

 " For it seems manifest that these striking resemblances 

 " which are arrived at by such widely divergent means, can 

 " be reasonably explained on no other hypothesis than 

 " natural selection. . . . Under the weaning influence 



