Production of Mimetic ResemUanccs among Butterflies. 333 



the grave difficulties in the way of accepting any ex- 

 planation of mimetic resemblances other than Natural 

 Selection have already been very clearly set forth in an 

 admirable paper by Professor Poulton (Journ. Linn. Soc. 

 Zool., xxvi, pp. 558-612), which has been recently repro- 

 duced in his book "Essays on Evolution/' and which it would 

 be well for any would-be critics of the theory of mimicry 

 to "read, mark, learn and inwardly digest" before puttintr 

 pen to paper. * 



But to revert to the question of birds, it is clear tliat 

 the above negative statements have been generally accepted 

 without proper consideration. When a naturalist who has 

 spent some time in the tropics expresses a decided opinion 

 to the effect that birds do not normally eat butterflies, 

 because he has never observed them doing so, it is in- 

 cumbent upon us, before accepting his evidence as having 

 any real scientific value, to satisfy ourselves that he hal 

 made a systematic and thorough investigation of the 

 subject, and that his views are not based merely on casual 

 and inadequate observations. For in a matter of this 

 kind there is grave danger that absence of evidence may 

 be due simply to lack of observation. If a collector main- 

 tains that birds do not eat butterflies, we are justified in 

 asking him for a full list of the other insects which he has 

 seen captured by birds. And I venture to think that a 

 closer inquiry of this kind would reveal the fact that 

 most of the negative evidence which has been brought up 

 against the Selectionist interpretation of mimicry is really 

 of little worth. 



Some Reasons for the Paucity of Evidence. 



_ There can be no question that the published records of 

 birds attacking butterflies are limited in number, though 

 perhaps more numerous than generally supposed. If then 

 we believe that such attacks are really of comparativelv 

 frequent occurrence in certain regions, how are we to 

 account for the fact that so few observations have been 

 recorded ? Certain aspects of this question were dealt 

 with by Mr. Trimen in his Presidential Address to this 

 Society in 1897 (Proc. Ent. Soc, p. Ixxxix), when he said : 

 " I am persuaded that . . . the dearth of evidence is due 

 to the neglect of well-directed and sustained observation. 

 Little can be gained by merely noting such cases as 



