THE STUDY OF LIVING NATURE 203 



strongest evidence in favour of his theory. The illustra- 

 tions will be chiefly selected from old subjects in which 

 something new may still be found. The examples must 

 be limited, and therefore I propose to select them from 

 a single Order of insects — the Lepidoptera, or butterflies 

 and moths. /^* il,,^J 



Some points in the Resemblance of Butterflies to Dead 



Leaves. u*XUl £&* 



One of the finest and best-known examples of conceal- 

 ment for the purposes of defence is found in the under- 

 side coloration of butterflies of the genus Kallima, so 

 graphically described by Wallace. 1 Among the most 

 interesting details of the resemblance is an oval transparent 

 patch on each fore-wing, which allows the light to pass 

 through and produces the effect of a hole in the apparent 

 dead leaf. In another part of the fore-wing of Kallima 

 the appearance of a hole is produced as an artist would 

 paint it, by the use of white body-colour. Of these two 

 methods found on the same wing, the former is undoubtedly 

 the more recent and more highly-specialized method ; for 

 when the transparent ' window ' is examined under the 

 microscope, scattered opaque white scales can still be 

 seen in abundance over its surface, not thickly placed so 

 as to prevent the passage of light, but witnesses to an 

 earlier and less perfect representation of light shining 

 through a hole. 



In many species a hole is suggested by means of the 

 more primitive method alone. One of the most remark- 

 able markings possessed by any British insect is the white 

 4 C ' or ' comma ' on the under surface of the hind-wing 

 of the ' Comma ' butterfly, Polygonia (Grapta) C-album. 

 Many years ago I came to the conclusion that it repre- 

 sents, in bright, strongly-reflecting ' body-colour ', the 

 light shining through a semi-circular rent in a fragment 

 of dead leaf — the rent produced when a little segment of 

 leaf has broken away along a curved line, but still remains 

 connected with the rest across the chord of the arc. 

 Unless such a segment remains precisely in the plane 

 1 Essays on Natural Selection, London, 1875, pp. 59-62. 



