•\2() MR. C. F. M. SWYNNEHTON ON THE 



high degree of unpleasantness, for she was unmistakably hungry and would I feel 

 sure have eaten a Terias, judging from the avidity with which she now accepted 

 and swallowed down a Mycalesis campina. It is interesting in this connection to 

 note that on my commencing the above experiment tlie roller was hopping about 

 on the ground searching for food and picking up an occasional working termite. 

 Nevertheless, she failed to see a solitary larval grasshopper which was 

 lying motionless on the floor. She usually eats it right up to the end when 

 offered by the forceps, yet now she passed over it two or three times without 

 seeing it. I have noticed this difficulty in distinguishing motionless objects 

 frequently before on the part, not only of the roller, but of my various other 

 insectivorous birds, and one sees exactly the sa.me thing in buck and other wild 

 mammals. Had the grasshopper moved it would probably have at once been 

 snapped xip. I now fed the bird by hand and somewhat later in the day offered 

 her a Myc. campina which she readily ate. She then accepted readilv a Pachnoda 

 impressa, but absolutely failed to eat it, as the beetle slipped with a snap from her 

 bill each time she applied any pressure. She then ate readily three smallish 

 grasshoppers, including the one that had been lying in the cage all day ; 

 refused, then tasted slightly and i-ejected a Myc. camjyina. I then reoffered the 

 Otitoniid, but she again failed to crush it owing to its extreme slipperiness and 

 toughness. She commenced by refusing it without tasting, but on my continuing 

 to offer it, she made quite a number of attempts to crush it, sometimes succeeding 

 in holding it in her bill while she struck it several times against the perch, but it 

 always slipped out in the end and the bird finally refused to have anything more 

 to do with it, edging away whenever it was offered. She then ate readily a 

 E'lirytela Marbas, but refused and on persuasion tasted slightly and rejected 

 a Myc. campina. 



IM. campina preferred to Z. elegans and A. esebria (or whatever this was taken 

 for) ; gi'asshoppers and B. hiarbas eaten after the bird had failed to eat Pachnoda. 

 The defence afforded the latter by gloss and toughness was well illustrated ; also 

 the roller's memory and capacity for recognizing even a wingless Acrseine body by 

 the incident of the A. esebria.'] 



Exp. 110. — November 28. Hungry ; she at first refused without tasting a 

 Pachnoda impressa, evidently remembering her experience of yesterday, but on 

 my persisting in offering it, accepted it several times in succession. It each time, 

 however, slipped out of its bill and the bird only twice or thrice ari-ived at the 

 stage of banging it against the perch — for she seemed to realize that this was 

 the correct procedure here, crushing with the bill being out of the question. 

 She finally refused to have anything more to do with it. 



After an interval I offered her another Cetoniid, Rhabdotis aulica, a somewhat 

 laro-e metallic-green species. She at once accepted it, but only to go through once 

 more tlie tantalizing performance which she had experienced in connection with 

 the Pachnoda, with the result that the Rhabdotis too was soon given up as a 

 bad job, the bird showing irritation whenever I brought it near to her. She then 



