312 ANNUAL EEPOET SMITHSONIAN INSTITUTION, 1915. 



was in plain sight, on the edge of a scar in the earth from a recently 

 uprooted tree. Finallj'^, though, the tension was relaxed; the song 

 ceased. Where would it be next time? In front of me? Or would 

 the singer see me and depart for good, still a mj^stery? Even as I 

 was thinking these things a ghostly-silent little shadow sped dangling 

 past me and came to a halt about 30 feet away, half lost in the dark 

 fog, on the far side of the raw little clearing. In awful anxiety lest 

 he become swallowed up in the mist and lost to me, and with a great 

 effort not to lose the dim impression of the faintly seen bird, I moved 

 slightly for a better view. My long watch was futile, for my spirit 

 bird disappeared. I sat awhile and mourned, with a great deal of 

 invective in my heart. But soon realizing that this was futile, I de- 

 cided to practice the song I had learned. Imagine my surprise, after 

 the first attempt, to hear, close by, the loud wip-wip of yesterday 

 and to se€ it followed almost immediately by another ghost bird, 

 which had the grace to alight or stop running (I couldn't be positive 

 which) within range and in sight. This proved to be C. turdhia. 

 Although we often heard the curious protracted song later, when we 

 went to the top of the range, we never again caught sight of this 

 little-known bird, and this specimen remains unique in the whole 

 South American collection. 



The several species of true ant-thrush, Fwrnicarius^ all have char- 

 acteristic notes, combined with the same skulking, rail-like habits of 

 the foregoing. The recently described Colombian form of F. rufl- 

 fectus has two sharp whistles, the last a semitone above the first. 

 This, in our experience, was never varied. F. aruilis connectens^ from 

 the lower forest zone of the eastern foot above Villavicencio, had a 

 song the exact reverse of that of Grallaria hypole'uca' a loud note on 

 G, followed, after a rest, by a close descending scale of three or four 

 semitones. Foi'micarius^ like GrallaHa^ has a sort of clucking qual- 

 ity when heard near at hand. 



Few brush birds have more distinctive notes than the ant-shrikes 

 of Thamnopkilus and their relatives. The commonest one we en- 

 countered, T, irmlthtriatus^ has the characteristic drj^, woody, de- 

 scending scale common to many species. It strongly suggests in 

 quality the spring " rucking " of a nuthatch. It might be written 

 ruk, ruk, ruk, uk, uIj, k, k, k beginning lazily and gathering speed 

 as it descends. All these birds put much etTort into their calls and 

 sing with head up and tail down. The latter moves noticeably at 

 each note and, as with the trogons, we came to look for the vibrating 

 tail when hunting them. 



The many species have different notes, but most are readily recog- 

 nizable as ThamnophUus when any one of them becomes thoroughly 

 familiar. Until one has had real experience with tropical birds, it is 



