248 THE VARIATION OF ANIMALS IN NATURE 



acceptability was assigned to each species in the following way : 

 All species removed during the first interval were given a 

 rating of ioo ; any left at the end of the experiment (i.e. when 

 birds ceased to visit the table for food) were given a rating of o. 

 By a simple arithmetical calculation (the methods will be 

 found in Appendix B, p. 380, of the original paper) species 

 eaten during the intermediate periods were given appropriate 

 ratings between o and 100. Eventually the ratings obtained for 

 any one species in different experiments could be averaged to 

 obtain a mean value for the species. The experiments were 

 conducted over two seasons, and involved more than 5,000 

 insects of 200 species, and over 2,000 bird visits. 



Some of the more important conclusions are the following : 



(1) The majority of insects are more or less palatable, or 



are at least occasionally eaten. 



(2) That, ceteris paribus, large insects are more favoured 



than similar forms of smaller size. 



(3) That a number of species with conspicuous black and 



yellow markings or brilliant metallic colours x are 

 very unacceptable. Of the species having a rating 

 of 25 or less twenty-four have this type of coloration, 

 while seven are of other types. 



(4) That none of the insects with a rating of 60-100 have 



these conspicuous patterns ; or, at least, when the 

 pattern is present, it is hidden in the resting position. 



(5) That a number of other types of colour-pattern, con- 



spicuous to human eyes, do not appear to be asso- 

 ciated with a lower (or much lower) than average 

 acceptability. This is of some importance, because 

 some of the types are the same as or similar to 

 species which have been hitherto regarded as specially 

 protected. 



(6) That species which have a procryptic pattern are 



usually very acceptable. Again, there are a few excep- 

 tions (e.g. moth with ' dead grass ' pattern, p. 354). 



(7) That some of the most strikingly marked and un- 



acceptable species are those which feed (usually as a 

 larva) on plants of the families Asclepiadaceae and 

 Apocynaceae, which have acrid or poisonous juices. 



1 Only one species involved (22 specimens). 



