102 Mr. C. F. M. Swynnerton on 



witlionf. discomfort when these are in great strength rela- 

 tively to the amount of food contained in the stomach — 

 when even the Danainre may stimulate them. It is impossible 

 to go into the sul)ject more fully here. Suffice it to add that 

 just as the taste, smell, or even sight of certain foods that 

 we can eat with impunity may in themselves, by imconscions 

 association, be sufficient to promote a flow of the gastric 

 secretions and give us thereby a feeling of appetite and 

 eagerness, while the mere sight or smell of things that 

 disagree with ns produce a feeling of disinclination for them, 

 so, too, with birds, choice probably becomes, with experience 

 and practice, instinctive. A bird, seeing a well-known and 

 lately experienced Amauris when not in a position to eat it 

 with impunity, probably does not have to cast back in his 

 memory for his previous experiences of it, or to calculate 

 how full his stomach is. He merely feels disinclined for it. 



One point, discussed at the end of experiment 541 and in 

 528, and illustrated abundantly throughout the experiments 

 on L. lei{comeIas (and in other experiments not described 

 here), is worth mentioning again. It is that a rapidly- 

 digesting bird is able to go on eating the most nauseous 

 insects indefinitely, with frequent short intervals for subsi- 

 dence, provided that no higher-grade insects are available to 

 carry the filling of its stomach well beyond the point at 

 which those nauseous insects are usually refused, and to keep 

 it there. 



The actual preferences shown in these experiments may 

 be summarized as follows: — Irrisor preferred the Noctuid 

 moth Spiling omor pi la to the pleasantest Acridians and Locus- 

 tids, and these to larval migratory locusts. The locusts were, 

 in the first experiment, preferred to dead and drying butter- 

 flies, but the latter were far better relished in subsequent 

 expeiiments in which they were offered alive. In fact, in 

 experiment 502, there was an indication that a larval migra- 

 tory locust was, at any rate, not letter liked than Byhlia— vl 

 butterfly as low-graded as Keptis. Amongst the butterflies 

 themselves, Ilypolimnas, Hamanumida, Catopsilia, the three 

 Papilios — dardanus, demodocvs, and It/ceus — the larva of 



