THE DESCENT OF THE RIO GY-PARANA 259 



of the tropics; whether in the fever-stricken Choco on the 

 west coast of Colombia, at the base of Duida on the 

 Orinoco, or in the wilds of Matto Grosso, the ravaging 

 hordes seemed always the same. One moment they hurry 

 along in solid formations, the next side-lines have been 

 thrown out in all directions, covering many square yards 

 of ground. Not one leaf or crevice escapes the alert scout- 

 ing-parties, which ascend even to the top of the tallest 

 tree. When a victim is discovered the news in some mys- 

 terious way is flashed to the main column, and battalions 

 of reinforcements immediately rush to the encounter, charg- 

 ing the prey and clinging with vise-like mandibles to any 

 part of its body that offers a hold. Usually the approach 

 of the devastating host is preceded by a swarm of panic- 

 stricken insects, crawling, hopping, and flying in their en- 

 deavors to escape destruction; large, hairy tarantulas crawl 

 to the tops of bushes and leap from leaf to leaf, only to be 

 discovered and routed, until in despair they spring to the 

 ground, which by this time is one surging mass of ants, 

 where they are despatched in short order. I have seen 

 scorpions and centipedes eight inches long suffer a similar 

 fate; no living thing seems to escape the avalanche of de- 

 struction. Flocks of ant-birds usually follow in the wake 

 of the army, feeding upon the ants and upon the insects 

 that have been driven from their hiding-places. One of the 

 questions that naturally arises in this connection is how 

 the callow young of birds escape from the ants, as caged 

 birds are not immune from their attacks, and dead or 

 wounded birds placed near the army's line of march are 

 quickly discovered, torn to shreds, and carried away. 

 While in British Guiana I had been watching the nest of 

 an ant-wren containing two helpless young, placed in the 

 crotch of a tree a few feet above the ground, for several 

 days; one morning the whole region was swarming with 

 ants and the nest was empty; however, not long after, and 

 also on subsequent days, both parent birds were seen con- 

 tentedly carrying food into a thicket fifty yards away. A 



