398 VnivcrsHy of California Publications in Zoology [Vol. ii 



numbers are always counted than truly exist, owing to the com- 

 minuted condition of some, makes it evident that no exaggeration 

 is possible here. Consequently it affords dependable evidence as 

 to the numbers of weed seeds, insects, etc., taken by a bird. On 

 the other hand, the percentage-of-volume method can be depended 

 upon only to furnish an idea of the comparative quantities of 

 the different kinds of food. Personal error in estimating has 

 to be allowed for in this method, for whereas a certain insect 

 might be to the eye of one person ten per cent of the volume, it 

 might represent fifteen per cent to the eye of another. 



The furnishing of complete data as to the bird whose stomach 

 is examined (date, locality, kind of field, collector, etc.) should 

 afford information, first, as to the variation in the amount of 

 food taken by birds during the day, month, and year, and second, 

 the food preference of birds in a given locality and in dift'eretit 

 localities. The record of the exact time of day, the month, and 

 year w-hen the bird was collected furnishes the basis for the first, 

 the record of the habitat, as, for example, the kind of field, 

 orchard, or vineyard, the ba.sis for the second. 



An attempt has also been made to improve the method used 

 in determining the economic status of a bird. As has already 

 been pointed out, the economic status of a bird was originally 

 determined by inference. A bird in the grain field must be 

 eating grain and therefore is injurious. Experience has taught 

 that such reasoning is fraught with error. And further exper- 

 ience has taught us that even though a bird may cause consid- 

 erable damage, j'et because of its usefulness as a weed-seed 

 destroyer, as an insect destroyer, or as a bird important in keep- 

 ing the balance in nature most suited to man, it may be more 

 beneficial than harmful. At one time the total good accom- 

 plished by a bird was held to inhere in the number of injuri(His 

 insects it destroyed. Today, although we still retain this idea, 

 we see a little further and conclude that a bird may be beneficial 

 because it destroys insects (almost all insects being potentially 

 destructive), and not because it chooses a particular class of 

 insects arbitrarily classified as harmful by man. 



An attempt has been made to arrive at the average volume 

 of food taken bv the meadowlark, bv determining the volume 



