NESTING HABITS OF OROPENDOLA — CHAPMAN 371 



The response of the oroj)endohis to this note is variable. At times 

 they act as one bird and phinge from their nests or perches into the 

 forest. I have even seen a bird when flyin<r toward her nest dod<T^e 

 abruptly downward in the air when hearing the alariii. On the 

 other hand, while it evidently puts the birds on tiioir «;uard, they 

 may not move when hearinp; it. 



The si<j:nifit'ance of this call is undeistood not only by oropendolas 

 but obviously by other species. On February 8, 1926, five toucans 

 (three Pteroglossus^ two Rhampha^os piscivonis) and two caciques 

 {Caciciis vlteUlnw<) that were in the nest tree dived with the oropen- 

 dolas when the alarm was «]^ven. On tho followinf^ day it induced 

 even a trespassing^ Cass/'d/'x to seek safety with her intended victims 

 fiom a suj^posed common enemy. 



The need for a lookout and for prompt obedience to this danger 

 .«:ignal was tragically illustrated on February 12, 1927. At noon on 

 that day Mr. Maunsell S. Crosbv, who was standing near the nest 

 tree watching the oropendolas at Avork, saw an eagle hawk {Splza-s- 

 tur mclanoJeucus) drop from the sky, strike an oropendola that was 

 working on her nest and bear her to a neighboring branch. The 

 hawk, which was identified by Mr. Ludlow Griscom, was still stand- 

 ing on its victim when a few moments later, attracted by the unusual 

 outcry, I reached the tree. It soon flew off with its prey to the forest. 

 This event caused tremendous excitement among the oropendolas, 

 their united cries of alarm producing the effect of a loud chorus. 

 They all left the tree and for the remainder of the day the colony 

 was completely disorganized. 



The following day the effects of this catastrophe were still evident 

 in the nervousness of the birds and the frequency with which the 

 alarm call was uttered. Normally this call may be heard three or 

 four times during a morning, but during two hours on the morning 

 of February 13 it was given at 8.50, 8.51, 8.55, 9.04, 9.07, 9.10, 9.12, 

 9.22, 9.40, 10.12, 10.15, 10.26, 10.44, and 10.50, a total of 14 times in 

 two hours. Beyond two buzzards that flew over at 9.10 and 10.44, 

 respectively, no cause for alarm was seen by me during this period. 

 The first seven times the alarm was sounded all the birds responded 

 promptly, diving to the protection of the lower growth. Later their 

 reaction was not so keen and on three out of seven signals they did 

 not respond. 



The incident illustrates the exposure to attack by a predatory bird 

 of an oropendola working outside her nest, the need for a guard, the 

 inqxtrtance of pronq>t obedience to his warning, of the quickening 

 of reacti(jns through experience, and of their decline after frequ(Mit 

 call has been made upf)n them. 



It might be imagined that the birds in their long nest bags swing- 

 ing from the tips of slender branches were immune from attack. 



