376 



REPORT OF NATIONAL MUSEUM, 1901. 



made of long- slender poles stuck into the ground and tied above to a 

 ridgepole; this was covered with horizontal and overlapping layers of 

 palm-leaf thatch, the strips being tied by vines to each rafter (rig. 2). 

 A door was left in the center of each side; these were closed with strips 

 of thatch. Low half-moon ridges of earth were heaped up inside of 

 each door to prevent the entrance of water. A little room about 6 feet 

 square was cut off at one end by walls of thatch. Five little heaps of 

 ashes and rirebrands around the walls and several posts for hanging 

 hammocks showed where the different families made their homes 

 (fig. 3). 



Their method of cultivation is like that of the Jamamadi already 

 described, but their fields are much smaller and less carefully tended. 



e> 



<a 



9 



<9 

 d 



o 



C 7/| 



O 



c 





o o 



a 



Fig. :!. — Ground plan of Hypurina house, a, ioot of rafter.s; b, mound of earth inside 

 e, fireplaces; d, posts for hammocks; e, room; /, door. 



>f door; 



They depend much on hunting, and especially on fishing, for a living. 

 The family of the Hypurina, Pedro Bom, w T ere living on fish when 

 visited. One of their methods of fishing is with a basket attached to 

 a spring pole; this is set and baited under water. The fish caught are 

 dragged out of the water head down, so that they can not escape 

 (figs. 4 and 5). 



They frequently all leave their villages on extended hunting and 

 fishing excursions, living during this time in open sheds of thatch 

 (papira), which are built wherever they stop for the night. Their 

 canoes are each made of a single piece of bark, like those of the 

 Jamamadi (fig. 12), but they are more rude and the ends are left open. 

 They are probably never used for navigating the Purus, serving only 

 for the narrow streams flowing into it. Several villages are accus- 



