FAMILY CRACIDAE 301 



Museum was taken on the Rio Changena. McLeannan collected them 

 a hundred years ago along the railroad near Lion Hill, but since 

 his day the only reports for the Canal Zone come from Barro 

 Colorado Island where a few remain. While the crested guan ranges 

 in forested level lands, it is most frequent in hill country where it 

 ascends regularly to 1,000 meters, though more common lower down. 



They are more active than the bigger curassows and are less timid. 

 Their food in large part is such forest drupes as wild figs and 

 mangabe berries, which they seek in the high trees and for which 

 also they descend to the ground when there are many fallen, dropped 

 from their own feeding, and from that of parrots, toucans, and 

 other birds. Guans range regularly at times on the forest floor, 

 often making considerable scratchings to uncover food. 



They are seen in pairs, or, where common, as many as 6 or 8 may 

 be found together. When disturbed on the ground they mount im- 

 mediately into the trees, flying with noisy wings, and then run 

 actively along the larger branches, taking flight again through and 

 over the tree crown. In the air they move with neck outstretched, 

 alternately sailing with wings stiffly spread, and flapping quickly 

 after a short distance to maintain momentum. It is common in such 

 movements to see them sail off from some high point to cross a wooded 

 valley, and in regions where the trees are tall birds moving through 

 the tree crown often fly across clearings 75 meters or more above 

 the ground. When not frightened they walk gracefully along the 

 inclined branches, but at an alarm they may freeze motionless among 

 the leaves, standing erect, or crouched on a branch with neck ex- 

 tended, when it is most difficult to make them out in spite of their 

 size and their long bodies. In the more open gallery forests when 

 wind is not blowing I have found it profitable at such times to watch 

 the leaf shadows on the ground as it possible to detect slight head 

 movements of hidden birds that otherwise would remain unnoted. 

 In our field work we often use whistles that produce a variety of 

 sounds, some shrill and penetrating, others gabbling or moaning to 

 attract such species as forest hawks and high-canopy hummingbirds. 

 The pava cimba finds these sounds disturbing, and frequently 

 betrays its presence by calling in reply. Occasionally, particularly with 

 shrill eagle calls, they may become much excited, when I have had 

 them run out on open limbs, or even descend to the ground to strut 

 about with spread wings within a few meters of my feet. 



Their usual note is a yelping call that is repeated excitedly, often 

 for several minutes. Another louder sound with a strange resonant 



