^°'i9lf ^"^1 McAtee, Contents of Bird Stomachs. 455 



reckoned in the same way, amount to 23 per cent. It can hardly 

 be claimed that this is the most accurate method of calculating 

 the relative amounts of food found in a bird's stomach. Birds 

 are fond of eating a great many different things, the aggregate 

 quantity of which may be small, just as human beings eat a little 

 butter and sugar at nearly every meal, but never make a whole 

 dinner of either. To illustrate, in an examination of 2258 stomachs 

 of the Crow Blackbird, corn amounted to 35 per cent of the food 

 by bulk, but when reckoned by the number of times taken it 

 aggregated 52 per cent." 



Other illustrations from Professor Beal's work show still greater 

 diversity between percentages obtained by these two methods. 

 For instance spiders were found in 26 per cent of 389 stomachs 

 of Hylocichla ustulata, but composed only 1.82 per cent of the 

 stomach contents by bulk. The same data for Hylocichla guttata is : 

 spiders were found in 49 per cent of 514 stomachs, but made up 

 only 7.35 per cent of the total contents. Ants, much smaller 

 creatures on the average than spiders, while found in fewer stomachs 

 (249 as contrasted with 254) of the Hermit Thrush, compose a 

 considerably' larger volume of the food, namely 12.54 per cent. 

 This fact is just contrary to the normal expectation, and would 

 never be guessed from figures showing merely the frequency of 

 occurrence. The misleading nature of such figures is further 

 shown by the fact that wild fruits found in 243 stomachs somewhat 

 fewer than held ants, compose 26.86 per cent or more than twice 

 as large a proportion of the total bulk of the food. Furthermore 

 caterpillars, occurring in 268 or 52 per cent of the stomachs, and 

 hymenoptera (other than ants), found in 136 or 26 per cent, form 

 nearly equal percentages (within a fraction of one per cent) of the 

 total subsistence. 



Beyond showing the futility of Gilmour's particular variety of 

 the numerical system, these instances prove, that frequency nota- 

 tions, no matter whether they refer to systematic or to economic 

 groups, do not indicate the importance of these groups in the diet 

 of the species concerned. Hence they do not suffice for the needs 

 of economic investigations. 



Let us see what other objections can rightfully be lodged against 

 the numerical system. In the first place the adherent to this 



