337 



to the size and structure of the scales themselves, parti}' to 

 the mode in which the scales are inserted into the wing. The 

 scales constituting the oval patch are large, about six or seven 

 times as long as broad, and with sides roughly parallel. Those 

 of the silky area are small, oval or pear-shaped, attached by 

 the broader end. The scales of the oval patch are set up at 

 an angle and overlap ; those of the silky area are arranged like 

 a mosaic and are closely appressed to the surface of the wing. 

 The silky area and chalky patch together occupy about three- 

 quarters of the area of the forewing, and a little less than half 

 that of the hindwing. The object of the silky patches, with 

 their pavement-like arrangement of scales, may be to ensure 

 the smooth sliding over one another and the close fitting to- 

 gether of the surfaces which are apposed during rest. When 

 the silky patches coincide, the chalky patches coincide also, 

 and the insect is in the normal position of repose. It is 

 perhaps conceivable that the sliding together of the two 

 smooth surfaces may give rise to a tactile impression which 

 assists the butterfly in assuming the appropriate attitude. An 

 interesting point is that the ordinary pattern of the wing is 

 not continued on the silky areas, being confined to those 

 regions which are visible while the insect is at rest. This is a 

 strong argument in favour of the \-iew that the wing-pattern 

 has been evolved under the influence of natural selection. 



When the wing has been denuded of scales, the position of 

 the scent-patch is still clearly indicated by the more cloudy 

 appearance of the wing-membrane in that situation. On 

 examination this is seen to be due to the presence of the special 

 sockets which serve for the attachment of the scent-scales. 

 These sockets, which are considerably larger in every dimension 

 than those belonging to the ordinary scales of the wing, are 

 confined, as has been seen, to one surface — the lower surface of 

 the forewing, tlic upper surface of the hindwing. They are 

 very easily distinguisliable from the ordinary sockets on micro- 

 scopic obserx'ation, not only b\- their superior size, but also 

 from the fact that they focus at a difi'erent ]vvv\. 



In Aoncptcron vir'^o Bates, the male has similar ()\al patches 

 in the corresponding situation on the fore- and hindwing. The 

 silky area surrounding the oval patches extends e\'en larthcr 

 43 



