NUTHATCHES AND TITMICE. 75 



such as seasonal changes necessitate, and as these do not largely 

 affect insects, which constitute the great bulk of the bush tit's food, it 

 follows that the variation in diet from one month to another is not 

 great. The smallest quantity of animal food was in March, when 

 it amounted to 53 percent, but the percentage was almost exactly the 

 same for November. One stomach taken in April contained nothing 

 but insects and spiders, and 11 collected in June contained no vege- 

 table food. Probably examination of a greater number collected in 

 these months would result differently. While the material available 

 for the present investigation is not so extensive as could be desired, it 

 is sufficient to indicate beyond reasonable doubt that the relative pro- 

 portions of animal and vegetable food in the diet of the bush tit vary 

 little from season to season. 



Animal food. — The largest item in the insect portion of this bird's 

 food consists of bugs (Hemiptera), which amount to over 44 percent 

 of the whole. The gnatcatchers are the only birds yet investigated 

 whose diet is made up so largely of this order of insects. Moreover, 

 the particular families of Hemiptera so extensively eaten by the bush 

 tit are the two that are most destructive to the interests of horticul- 

 ture — namely, the plant-lice (Aphididse), and bark-lice, or scales 

 (Coccidse). The last amounts to nearly 19 percent of the year's food, 

 and are eaten in every month. The greater number are consumed in 

 July, 46 percent; June follows second in rank, when they constitute 

 33 percent of the food of that month. The large black olive scale 

 (Saissetia olece) was identified in 44 stomachs, but other species also 

 were found. The question is often asked, Does any bird feed upon 

 the San Jose scale ? While the writer is not prepared to give a posi- 

 tive affirmative answer from direct evidence, there can be no reason- 

 able doubt that this insect is often eaten by birds. It must be borne 

 in mind, however, that the so-called San Jose scale is one of the 

 smaller species, and its distinctive characters are so minute that after 

 it has been taken into a bird's stomach, mixed with other food, and 

 more or less digested, it is impossible to determine its identity. It 

 is easy to ascertain that a pasty mass in a bird's stomach is composed 

 of scales partly digested, but to identify the species is quite another 

 matter. The olive scale and others of its genus, on the other hand, 

 are so large and their shells are of such structure that they can often 

 be identified, at least generically, even from fragments. 



While the San Jose scale was not positively determined, another 

 species of the same genus, the greed}" scale (Aspidiotus rapax), was 

 found in 4 stomachs, and scales not specifically identified were found 

 in 113. Of a total of 353 stomachs, 158 held scales; several were 

 entirely filled with them, and in quite a number upwards of 90 per- 

 cent of their contents consisted of these insects. No other family of 

 insects was identified in so many stomachs. As it is certain that the 



