intervals, they suspend themselves, with closely folded wings. The under- 

 sides of these wings, now alone exposed, resemble the upper sides in pat- 

 tern, but are very much dimmer, the black being replaced by ochrous dusky 

 brown and the red by soft, hoary pink. In addition, the lower wings are 

 narrowly bordered on the anterior edge with pale yellow, and the slender 

 body bears spots of the same color, while the bases of the fore- wings and the 

 immediately adjacent portions of the body are marked with small red spots. 

 Half of the broad pink ( = red) wing-band shows beyond the tips of the folded 

 lower wings. The whole form of the insect thus folded and placed, though 

 without any very peculiar modifications of contour, is almost exactly like that 

 of many slender, entire leaves common in the forests it inhabits. But ob- 

 serve, it is not like a living leaf in color, nor does it (in our experience) im- 

 merse itself in a sea of foliage, there to be the single counterfeit among many 

 genuine originals. Instead, as I have said, it selects, with several compan- 

 ions, a leafless twig, in a spot where leaves are few, and together they suspend 

 themselves beneath this twig in the semblance of a row of drooping dead 

 leaves which still show traces of live color (the fine yellow lines and spots), 

 and are each marked with a partly faded pink disease-spot (the pink band 

 crossing the fore- wings) and some brighter red disease-flecks near the leaf's 

 (= the butterfly's) base. That this resemblance is not fancied, but real and 

 very potent, will, I think, be attested by anyone who studies the roosting- 

 habits of Heliconius melpomene. There are doubtless many fine cases of this 

 sort of mimicry still to be discovered. But, as we have repeatedly affirmed, 

 mimicry is not our theme in the present book, wherein we must confine our- 

 selves almost wholly to the far larger and more intricate problem oi.obliter- 

 ative coloration. Nor is it to be supposed that mimicry rather than obliteration 

 is the rule in any large group of butterflies. On the contrary, the cases of 

 out-and-out butterfly mimicry are relatively very few indeed, and scattered, 

 while 1 obliteration ' is universally and most variously achieved among them. 



Many of the butterflies which rest amid live foliage and flowers have 

 daintily detailed and at the same time £ generalized ' pictures of their varie- 



2I 5 



