352 THE PLACE OF MIMICRY 



Heliconines, &c. Again, the fact that every form in the 

 genus is mimetic leads us back to a mimetic common 

 ancestor. Now there is no difficulty in understanding 

 such an evolutionary history in a group of closely related 

 Miillerian mimics. The common ancestor possessed 

 some special protection in the way of unpalatability, &c, 

 and adopted the Warning Colours of other more power- 

 ful combinations, also specially defended. The species or 

 forms to which the common ancestor gave rise inherited 

 the defensive quality, spread into various districts, and 

 joined the local Synaposematic combinations. It is very 

 different when we attempt to reconstruct this history on 

 the hypothesis of H. W. Bates, and try to imagine qualities 

 in an ancestor which would determine universal Pseud- 

 aposematic Resemblance among the descendant species. 

 Upon the whole, therefore, it is probable that Protogonius 

 is specially protected like so many other Nymphaline 

 genera, and that the defence holds against the majority 

 of enemies that attack butterflies upon the wing. Against 

 these all the advantages of Miillerian Resemblance would 

 be gained. Such enemies, put on their guard by the 

 well-known Warning Colours, would learn to avoid this 

 member of the group with a lessened waste of life from 

 experimental tasting. Against special enemies that 

 would devour them in spite of their special defence, the 

 cryptic colouring would be a valuable safeguard, making 

 the insects extremely difficult to mark down when they 

 came to rest. The concealed under surface is also of 

 importance — probably of supreme importance — as a pro- 

 tection against the class of enemies that hunt for insects 

 in a state of complete repose. 



Before passing to the consideration of the next group, 

 I desire to state that the interpretation of these two 

 difficult cases is put forward in no confident or dogmatic 

 spirit. The balance of evidence appears to me to 

 suggest the conclusion here stated. It is impossible 

 to predict whether the further knowledge, so greatly 

 needed, will bring with it confirmation or revision. 



The chief Old World examples which manifest the 

 same interesting combination are to be found in the little 



