50 U. S. NATIONAL MUSEUM BULLETIN 217 



common on each arrow, but sometimes three or even four were used- 

 Paired feathers were preferably matched from corresponding feathers 

 of the right and left wings. If tail feathers were used, matching parts 

 were used by preference in pairs. I did not learn the principle or prac- 

 tical reason on which this practice was based. I also asked whether left 

 wing feathers were preferred for convenience of attachment by right- 

 handed arrow makers, as Bridges (1949) relates to have been the case 

 among the Ona Indians of Tierra del Fuego. I could find no com- 

 parable discrimination among the Nunamiut. 



Along the low arctic coast hawks are rare and feathers were accord- 

 ingly one of the uniquely valuable articles of the old Nunamiut in 

 their trade with the coast. Some of the old Nunamiut people were 

 specialists in catching hawks with whalebone snares set near a perch 

 or nest, and the nesting locations were established in old Nunamiut 

 knowledge. I learned that many nests were known to have remained 

 in use during the lives of the narrators and during their father's times, 

 and so far as it appeared, indefinitely. The present generation had 

 discovered a few nests which, so far as they knew were unknown to 

 their fathers, but none were thought to be at new sites. From ac- 

 counts of several older ISTunamiut men has been compiled the follow- 

 ing list of 17 nesting sites of gyrf alcons which they know in the area 

 from the Killik on the west to the Anaktuvuk on the east, an area 

 near 100 miles on each side (no nests are known on the Colville below 

 the Anaktuvuk's mouth) : 



On Anaktuvuk River: 30 miles from mouth, near Naniksruk (specimen 



obtained); 60 miles from mouth; and 90 miles from mouth, Soakpuk 



Mountain. 

 On Colville River: 25 miles above Umiat; and 4 miles above mouth of Killik 



River. 

 On Killik River: 15 miles from mouth; 45 miles from mouth; 20 miles above 



Odrivik Lake, near Irivik Cave (two nests, both occupied in 1950, very 



close together; specimens obtained); 22}^ miles above mouth; and at 



Akmalik Creek. 

 On Okpikyuak River, east of Killik. 

 On Okomilaga River: 8 miles east of preceding location; 3 or 4 miles inside 



mountain line; and at summit. 

 Between Okomilaga and Chandler River: At mountain line; and 6 miles 



north of preceding location. 



It is thought that nests do not remain unused. If one bird of a 

 pair is killed a substitute promptly joins the bereaved mate and the 

 pair continues at the site. I could not press to learn the length of the 

 delay except that replacement is thought not to be deferred until the 

 next season but to occur right away. There seems to be a reserve, gen- 

 erally unobserved, of single birds in this population which otherwise 

 exists in firmly attached pairs. 



