localities — Oxford, Switzerland, and the Isle of Wight — with very divergent 

 results, as will be seen from the following table : — 



It should be noted that the Swiss pupse which were not taken are marked 

 ' left ' because many of them did not emerge, but were removed at the close of the 

 visit and used over again. The results thus tabulated leave no doubt about the 

 existence of an immense amount of extermination at Oxford, and, although much 

 less, a large amount in the Isle of Wight, while there was comparative immunity 

 in the two Swiss localities. This strong contrast is probably to be explained by 

 the scarcity of small birds in the latter, and their abundance in England, and 

 especially in Oxford, together with the fact that the Oxford experiments were con- 

 ducted earlier in the year Other considerations point to the same conclusion 

 viz. that birds are the enemies in question; — the inaccessible position of the pupae, 

 which were nearly always fixed at a height of about four feet from the ground ; 

 the fact that the hard caudal extremity was frequently left attached after the pupa 

 itself had been taken ; the excessive variation in mortality in neighbouring locali- 

 ties, corresponding to well-known facts in the distribution of birds. A watch was 

 kept in localities where the pupse were known to disappear rapidly, and, in a single 

 instance, a great tit was seen to creep over the bark of a tree, from which the 

 pupa was then found to have gone. 



So far as the observations extended, we inferred that the comparative freedom 

 of the Swiss pupse from the attacks of Vertebrate enemies is compensated by the 

 far greater destruction of the larvae by insect parasites. It is probable that the 

 birds which attack the English pupie benefit the larvfe by keeping down the 

 number of parasites. 



In the course of the enquiry the possibility was suggested that perhaps the 

 birds actually see the pupse being suspended, and afterwards search the spot. 

 A large number of pupse were, therefore, fixed at night by the light of a lantern, 

 but, so far as we can judge from general impressions (the analysis of the results 

 being unfinished), no difference was caused. It was also thought that, when 

 several pupte were suspended in close proximity, the birds, after finding one or 

 more, might search the neighbourhood with especial keenness, so that the chances 

 of successful concealment would be smaller than those of isolated pupae. In order 

 to test this suggestion, a number of pupae were scattered over a large area, succes- 

 sive individuals being separated by a distance of about 100 yards or more. Here, 

 too, our present impression is that little, if any, difference was produced. Both 

 these modifications of the usual conditions of experiment were made in the Isle of 

 Wight. 



The question whether there is a struggle for existence during the pupal period 

 of Vanessa urticce is answered with certainty in the affirmative as regards those 

 localities where small birds are abundant. In such places it is now proved that 

 there is a tremendous struggle with an immense mortality, in spite of the brevity of 

 the pupal stage (from ten days to three weeks in length). 



