( xxii ) 



bird-theory by speakers whose competence and opportunities 

 for observation were quite beyond dispute. Further impor- 

 tant evidence on this point had been lately given by M. Piepers 

 (Congr. Internat. Zool., III., p. 460) who had studied the 

 question in Sumatra and Java for twenty-eight years. In 

 that time M. Piepers had seen four cases only of butterflies, 

 two belonging to the " protected " genus Euphea, being 

 attacked by birds ; and his paper referred to the fact that 

 neither Pryer, after twenty years' observation in Borneo, 

 Skertchley nor Scudder had seen or accepted such a 

 phenomenon. M. Piepers had arrived at the sufficiently 

 striking result that mimicry had nothing to do with Natural 

 selection. 



The premisses necessary to support either theory of 

 Mimicry had been unduly neglected. For example, though 

 evidence existed to show that the models were protected and 

 inedible, the proof of the edibility of mimetic butterflies had 

 not received enoiTgh attention ; it was necessary to Bates's 

 theory, and all the more so since Dr. Dixey's work on the 

 *' Miillerian " character of certain Pierinje. These theories, 

 indeed, were really working hypotheses, the object of which 

 was to suggest experimental work tending to prove or 

 disprove them ; they were not yet to be put forward dogmati- 

 cally as true and as convincing proofs of Natural selection. 

 To insist that these homoeochromatic groups owed their 

 origin merely to the educational requirements of birds had 

 led, in his view, not so much to a development as to a 

 stifling of broad speculation on and inquiry into the problem. 



The facts presented by the IthomicB (s. lat.) had scarcely 

 been touched upon in the discussion. Almost every colour- 

 type among these insects, however insignificant in appearance, 

 was represented by species of two or more of the genera into 

 which the old genus Itlwmia had been divided ; so that the 

 Ithomia. might be said habitually to exist as homoeochromatic 

 pairs. The coloration of many of these pairs, consisting of 

 nothing but a few black patches and a white or yellow patch on 

 a transparent ground was far from exhibiting the striking 

 features which one was led to believe were characteristic of 

 warning colours. 



