162 BULLETIN 17 9, UNITED STATES NATIONAL MUSEUM 



This apparently makes for ease of catching. On bright days I found it diffi- 

 cult to see insects when they were flying high in the glaring light, yet when 

 they chanced to fly between me and a dark background, sighting became easier. 

 Hence, it seems probable that the increased visibility of insects sighted below 

 the phoebe's perch might be another factor influencing the number of flights 

 directed downward. 



A complete reversal of this flight direction came with the approach of dark- 

 ness. * * * In the dim light the insects could not be sighted near the 

 ground among the long shadows, yet hundreds were visible in the air above. Of 

 117 flights noted during the approach of darkness, 79 per cent were directed 

 upward. The number of upward flights increased with increasing darlmess 

 as shown in a recorded sample of flights. 



Usually no hesitation is apparent before an insect is pursued. Yet the 

 directness of flight as well as the usual capture of but a single insect on each 

 trip indicates that the insect is sighted from the perch rather than while flying. 

 Where the bird feeds over a lawn, insects are sighted again and again from 

 ten to twenty feet from the perch; the bird flies directly to the exact position 

 on the lawn where the insect is located, plucks it up without the slightest 

 indication of uncertainty and returns to the perch. 



Where slow flying insects are numerous in the air, two insects are occa- 

 sionally taken in one flight, the second presumably sighted in the air 

 but probably after the first, since a series of insects, regardless of their 

 abundance, never was taken without the bird returning to its perch. 



When insects were scarce, on rainy days and early on cold mornings, hover- 

 ing was noted over grassy areas or alongside walls. Evidently these hoverings 

 were a means of sighting insects hidden in the tall grass. These hovering 

 flights are not the regular method of feeding, nor does the bird regularly 

 range away from the perch in this manner ; instead, they seem to be an adapta- 

 tion to unfavorable conditions where it is necessary to search out insects 

 that can not be sighted from the perch, and less frequently for detecting in- 

 sects already sighted from the perch but lost from the bird's view after 

 flight began. 



The method of searching from a perch not only enables the phoebe to locate 

 insects with a minimum of energy, but the rejection of unsuitable species 

 apparently occurs before leaving the perch. Common yellow cabbage butterflies 

 (Eurymus) frequently were observed to pass slowly in jerky flight, unmolested, 

 within a few feet of a phoebe. * * * 



In late spring and early summer the adult phoebes were observed to pass 

 up dozens of checkered fritillary butterflies (Argynnis) . * * * The young 

 phoebes fresh from the nest were not so selective as their parents in foraging. 

 On three occasions the young were observed to snatch a fritillary and swallow 

 it, gulping down the large wings as well as the body. These incidents in- 

 dicate that the older birds become conditioned against this insect. * * * 



The majority of insects caught were small enough to be swallowed in flight. 

 But when forms larger than the house fly were taken, the catch was carried 

 back and tapped against the perch several times as if to smash and kill it 

 before swallowing. * * * 



After catching large insects the birds exhibit noticeable preference for 

 substantial flat perches against which they can break up their prey. 



The two western species of Sayornis are sufficiently distinct in their 

 manner of flight to enable them to be distinguished when seen only 



