40 PALM TREES 



from their favourite stations on the summits of the 

 loftiest trees of the forest. 



When he wishes to make a " gravatana n he searches 

 in the forest till he finds two straight and tall stems of 

 the " Pashiuba miri M of such proportionate thicknesses 

 that one could be contained within the other. When 

 he returns home he takes a long slender rod which he 

 has prepared on purpose, generally made of the hard 

 and elastic wood of the " Pashiuba barriguda," and 

 with it pushes out the pith from both the stems, and 

 then with a little bunch of the roots of a tree fern, 

 cleans and polishes the inside till the bore becomes as 

 hard and as smooth as polished ebony. He then care- 

 fully inserts the slenderer tube within the larger, 

 placing it so that any curve in the one may counteract 

 that in the other. Should it still be not quite correct, 

 he binds it carefully to a post in his house till it is per- 

 fectly straight and dry. He then fits a mouth-piece of 

 wood to the smaller end of the tube, so that the arrow 

 may go out freely at the other ; and when he wishes to 

 finish his work neatly, winds spirally round it from end 

 to end, the shining bark of a creeper. Near the lower 

 extremity he forms a sight with the large curved cutting 

 tooth of the Paca (Ccelogenus paca), which he fixes on 

 with pitch, and the gravatana is then fit for use. 



These tubes are never less than eight and are often 

 ten or twelve feet long, and on looking through a good 

 one, not the slightest irregularity can be detected from 

 one end to the other. The bore is generally not large 

 enough to admit the tip of the little finger, so that the 



