the colours of certain Jjcpidoptera. 475 



When we also remember that, wherever they pupate 

 in a natural environment, it is almost certainly within 

 reach of small insectivorous or omnivorous mammals, 

 which can hardly have the chance of reaching the larvae 

 of hetularia, we are led to connect their more complete 

 concealment with their greater dangers. In speaking of 

 "more complete concealment," I refer to the result, 

 however brought about, — to the instinct which leads 

 them to scatter and hide we hardly know where, as well 

 as to the colour and shape. 



Mr. Bateson states that he would have been less sur- 

 prised if the golden pupae of V. urticce had been brought 

 forward as examples of warning than of protective 

 coloration. But the most prominent feature of the latter 

 is the habit of adopting a conspicuous position or 

 attitude ; for this, even more than the colour, displays 

 the organism to its enemies. In the pupse of Euploea 

 core we probably have an example in which the me- 

 tallic appearance has this significance, but it is always 

 freely exposed, and, as Mr. Minchin tells me, most 

 conspicuous, and can be seen from a great distance. 

 It is im]30ssible to say this of V. urticce as it occurs in 

 nature. Again, I have experimented with V. urt'icce, 

 and find that the most fastidious of all insect-eating 

 animals I have come across, a marmoset, devoured the 

 golden varieties, one after the other, with the greatest 

 relish. 



Mr. Bateson argues that the golden varieties cannot 

 be protective because they are conspicuous against 

 certain artificial backgrounds, which nevertheless stimu- 

 late their production. It is strange that he should have 

 employed such an argument, considering that I showed, 

 in 1887, strong reason for believing that only some of the 

 constituents of the reflected light are efiective in the 

 production of the far more perfectly concealed green 

 pupse of Pieridce. If the yellow constituent of the light 

 reflected from leaves is proved to be efficient rather 

 than the green, it by no means follows that the power 

 is not directed towards concealment, because yellow 

 backgrounds are efiective in producing green pupae. 



The same argument would deny any " ' attempt ' on the 

 part of the animal to approximate to the colour of its 

 surroundings" to the larvae of A. betularia, and the pupse 

 of V. io and the Pieridce, because all these become bright 



2 m2 



