( xxxiii ) 



the purposes of this discussion, much as I should have pre- 

 ferred to keep even this until the experimental evidence had 

 seen the light. 



The experiments. — Difficulties and likely causes of error 

 (at first mainly suggested by Mr. Marshall's critical brain) 

 were taken severely in hand from the very outset, while further 

 complicating factors were noted and duly guarded against 

 for the future as they came to light. Experiments on captive 

 birds were checked by a series of observations and experi- 

 ments on wild ones. It was realised that results obtained 

 from a few offerings or even a few dozen experiments could 

 hardly be regarded as entirely trustworthy, and, in the main 

 experiments alone, probably not far short of 17,000 butterflies 

 were offered from first to last to the various animals employed, 

 as well as much prey of other orders. The refusals and re- 

 jections, duly given a value by the eating immediately after- 

 wards of other species, or by the knowledge that the animal 

 was hungry, and frequently confirmed by re-offering and re- 

 offering, must alone have run into a few thousands. 



I mention all this simply to indicate that the work was 

 attemptedly thorough and probably reliable, and that the 

 brevity of some of the statements I am about to make should 

 not be taken as a measure of the evidence on which they are 

 based. 



B. Bearing of results on certain objections to the Selectionist 

 view. 



Alleged indiscriminateness ; discrepancies between treatment 

 of same prey by wild birds and tame. — (1) Unless through 

 sheer impossible hardness, size, etc., there is practically no 

 such thing as " inedibility." In the early morning, or after 

 the ejection of a pellet, a bird may quite readily eat Acraeinae 

 or even Danainae. As it fills up somewhat it refuses such 

 very low-grade prey, but still eats other species, which in 

 turn it rejects when slightly fuller — and so to repletion- 

 point. 



In view of this it is unsafe to deduce either " palatability " 

 or " indiscriminateness " from an acceptance — including the 

 finding of an insect in a wild bird's stomach or " unpalata- 



PROC. ENT. SOC. LOND., I. 1915 C 



