ES;PLANAT]ON OF FORM AND COLOURING. 307 



unusually tough and though she gave it a sufficient squeeze to have crushed the 

 average butterfly there was no evidence afterwards on the thorax itself of its 

 having done so. Not five minutes later she again crushed and rejected Euxanthe 

 wahefieldi and refused Papilio lyceus without tasting. 



Five minutes later she readily ate Papilio lyceus, crushed very thoroughly and 

 appeared greatly inclined to swallow Euxanthe toakefieldi but in the end rejected 

 it, refused without tasting Papilio demodocus, tasted and rejected Papilio lyceus, 

 readily accepted and after thorough crushing ate a Papilio demodocus, tasted and 

 rejected N^eptis agatha and refused without tasting Euxanthe wakefieldi and Precis 

 archesia (wingless). Three minutes later she subjected to very thorough and very 

 prolonged crushing Euxanthe toakefieldi, during which she twice nearly swallowed 

 it, but each time brought it forward into the bill again and subjected it to renewed 

 crushing. Finally she rejected it with a certain amount of -hesitation. 



I now offered a particularly large Neptis agatha which was accepted, but, on 

 being tasted, at once and emphatically flung it away. The bird then readily ate a 

 Papilio lyceus and refused without tasting the Neptis agatha and a Papilio 

 angolanus. Two minutes later she barely tasted, then dropped Eicxanthe 

 wakefieldi; five minutes later she tasted and Te]ected Papilio a^igolanus, jnst took 

 in its bill and dropped Euxanthe ivakefieldi, readily enough ate a Papilio lya;iis, 

 crushed very thoroughly and swallowed with no sign of dislike Euxanthe wakefieldi, 

 tasted and at once rejected Papilio angolanus, readily ate a Papilio lyceus, after it 

 the angolanus reoffered ; crushed and rejected the next Papilio angolconus, readily 

 ate a Papilio lyceus, and refused persistently without tasting a Papilio hippocoon, 

 a Precis cebrene, a Papilio lyceus, and a Neptis agatha. 



[SUMMAEY : — 



1. Salamis nehidosa (and, if not the result of stimulation, 



Salcdmis anarcardii). 



2. Precis natalensis f., P. dardanus 5 f> hippocoon (and, if 

 not the result of special stimulation, P. antheus, P. jjolicenes, 

 and Precis archesia). 



P. cebrene. 4 3. Papilio lyceus and P. dernodocus. 



4. Euxanthe ivakefieldi and probably P. angolanus. 



5. Neptis agccthcc. 

 1^6. Terias and M. rueppelli. 



H. loahlbergi. 



Note cct the time : "I had unfortunately no Eurytela, dry ope to offer with the 

 Euxanthe loakefieldi, but ain certain that the latter was not more distasteful to the 

 roller than is Pc^jyilio cingolanus, which nowadays roughly equals Eu7'ytela dryojye ; 

 possibly less, for there could be no doubt at all that the bird suspected both its 

 appearance* and the toughness of its thorax and so probably tended to place it 

 lower than its actual taste warranted. It appeared to be disliked (or suspected ?) 

 distinctly more than Papilio lyceus and veiy distinctly less than Neptis cigatha."] 



* White-spotted thorax and bodj^, which both in its bulging shape and in the yellow 

 colour of its fluff is highly leininisceut of Amaurls dominicanus. The latter often has 

 yellowish areas due to bareness. 



