ATTACKS BY BIRDS 



121 



collectors that they have never observed the destruction of butterflies by birds, it is refreshing 

 to find a frank expression of opinion to the contrary, and here we find Dr. Franz Doflein, 

 who has a large experience, expressing his surprise that naturalists who have had collecting 

 opportunities in the tropics should have failed to observe what he himself evidently regards 

 as a common occurrence. Instancing the facility with which such facts may be overlooked, 

 Marshall points out how rarely ornithological works mention facts of this kind, and yet 

 Parkinson Curtis, as recorded in * The Entomologist' of 1903, records that a kestrel observed 

 on Ballard Down, near Swanage, was continually swooping down and catching and eating 

 butterflies on the ground. In the course of an hour some thirty-six specimens were captured 

 in this manner, whilst the bird was observed to continue the pursuit for at least five hours. 

 Here a single bird in one day may reasonably be assumed to have destroyed 180 butterflies, 

 and as the observation was confirmed in succeeding seasons by Mr. Curtis, and independently 

 by Mr. Colthrup, it is evident, as the author indicates, that butterflies are a common food 

 of the kestrel, and yet its habit in this respect has been quite overlooked. The garden warbler 

 is further instanced by Professor Kennel as having been observed by him to feed its young 

 ' almost exclusively on butterflies all day long '. Further instances of a similar kind are 

 adduced, showing that absence of direct evidence for the fact is by no means necessarily 

 evidence against it. A long list of instances is then given of attacks on butterflies in the 

 Palearctic region. No less than thirty-four species of birds are recorded on seventy-eight 

 occasions as having been seen to feed on butterflies, and thirty-eight species of the latter were 

 observed to have been attacked. As an instance of the kind of statement made by the 

 opponents of mimetic theories, Marshall quotes the remark of H. P. R. in ' Country Life 

 March 14, 1908, p. 384, which reads as follows : ' No one has been able to adduce any examples 

 of a bird eating a butterfly, beyond a few cases where the food was either the Meadow Brown 

 or the Large Heath [E. janira or E. tithonus), or the Green Hairstreak — out of thousands 

 of observers, entomologists and ornithologists.' The mass of evidence in the Palearctic 

 region alone is a sufficient reply to this surprisingly confident and inaccurate statement. It 

 may be added that out of the seventy-eight occasions above referred to, these three butterflies 

 were observed to constitute the prey in only six instances. In tropical regions the available 

 records are naturally fewer, not because such attacks are of less frequent occurrence — they 

 are undoubtedly much more common, but because the number of naturalists who have had 

 an opportunity of recording such facts is comparative^ small. Nevertheless, in Africa, 

 thirty-one species of birds, together with certain cases where the bird was not identified, 

 are recorded as having attacked thirty-one species of butterflies on forty-five occasions. 

 It is worthy of note that, even in this comparatively small number of observations, the 

 protected nature of certain dominant forms is evident. There are only four attacks on 

 species of the genus Acraea and none at all on Danaine butterflies, further proof, if any were 

 needed, of the distastefulness of these groups. 



Out of nine recorded attacks on Nymphalinae only two are in connexion with a species 

 having a mimetic association, and these are both Atella phalantha, which is synaposematically 

 associated with Pseudargynnis hegemone. Out of the five recorded attacks on species of the 

 genus Papilio none of these were mimetic, though one was the male of dardanus cenea. 

 These facts should be borne in mind, since it has been stated by those who deny the value 

 of warning colours, that when birds have been seen to attack butterflies they attack the 

 so-called protected species as frequently as any others. The records of attack in the Indo- 

 Malayan region are equally instructive. Thirty-one identified and eight unidentified species 

 of birds have been seen to attack fifty-one species of butterflies on seventy occasions^ Of 

 1200 o 



