PROTECTIVE COLORATION OF LEPIDOPTEROOS PUPiE. 285 



brassicae showing similar results, but in a less marked degree, and 

 about -10 pupiu of /'. rapae, showing marked results. Those of P. 

 brassicae were in most cases samples only, but the rest of the pupae 

 of this species obtained, 50 to 60 in number, were in glass-bottomed 

 boxes. An interesting result was that pupje from larvae of P. brassicae 

 and P. rapae, placed so that clear glass surrounded them on all sides, 

 top and bottom, considerably resembled those placed in absolute dark- 

 ness, being mostly slightly darker, but the colouring in either case 

 being not very pronounced. It appeared as if light rellectod from an 

 object of decided colour (including black and white) seemed necessary 

 to produce a very marked result. In conclusion, he said that when 

 he compared the pupiP of P. ntacliaon and /'. napi on black paper or 

 dark sticks with those on natural green surroundings, such as carrot- 

 tops and green cabbage leaves, he saw no room to doubt that the 

 adaptation in colour is in many instances protective. Thepup;o exhibited 

 were in most cases attached to the objects on which they had pupated, 

 but where this was impossible they were fastened with loops of 

 thread stretched from tiny pins ; and made a very attractive exhibition. 



Professor Poulton congratulated Mr. Merrifield on his success, and 

 pointed out that in the experiments he had conducted on these lines 

 he had dealt with Vancssid pupro, the existence of which in this stage 

 was of short duration, and that, although they had given excellent 

 results in a way, they were not so marked as in the pupiis exhibited, 

 and the observations and notes had had to be obtained in the course of 

 the few days that the pupal stage lasted. With Papilionid and Pierid 

 pupie, that went through the winter, the observations could be made at 

 leisure. He had no doubt that the response in colour to the environ- 

 ment was of protective value to the species. 



Mr. Bateson said that he had lately bred Pieris napi on a large 

 scale, and, as regards that species, he could quite confirm Mr. Merrifield's 

 observations. His own experiments had been undertaken with a 

 different object, but he had frequently noticed that there was a fairly 

 close correspondence between the colour of the pupne and that of the 

 substances to which they were attached, though exceptions were not 

 uncommon. He had observed that in this respect there was con- 

 siderable difierence between dillerent families of larvjc. In his 

 experiments the offspring of each pair were kept separate, and in some 

 families many larvie pupated on the food-plant, while in others 

 scarcely any did so. In some families the pupal colour conformed in 

 nearly every case, while in other families the unconforming pupje 

 were numerous. The pupitj are often clustered together in groups, 

 and the individuals composing such clusters sometimes agreed with 

 each other in conforming, or in not conforming. He did not purpose, 

 then, to discuss the general proposition that the colour-resemblance 

 was in these cases a protection ; but he pointed out that if a larva had 

 the power of forming a green pupa when attached to a green plant, 

 the conformity of colour would only make the pupa less conspicuous 

 so long as the plant remained green. For the summer brood of /'. 

 naiii, for instance, this resemblance might perhaps be etYective, though 

 in this case the pupal period was very short, only two or three weeks. 

 But in order that the resemblance may beneiit pupie which over- winter, 

 it is necessary that the green plant to which green pup;e are attached 

 should remain green in the winter. His own pupiu had turned greeu 



