
According to Rohrman (1939), the resistance of aluminum to the 
atmosphere is “excellent,” but halides and chlorides exert a dissolving 
action on the oxide film that ordinarily protects the surface of this 
metal; the rate of solution in distilled water is negligible. As far 
as T can determine we can expect salt-water birds to lose their bands 
at a faster rate than fresh-water birds, but the degree of difference 
between the two remains to be determined. 
Most writers on avian longevity have not had to face the 
question of band durability. In the survival studies of small song-~ 
birds (Nice 1937; Lack 19l3a, 1943b; Farner 1945), it seems to have 
no importance. Kortlandt (19:2) concluded that it importantly dis- 
torted the mortality curve derived from the recovery of banded 
European cormorants. Marshall (197) and Paynter (1947) seem to have 
overlooked this possibility in their analyses of longevity in the 
American herring gull. 
In the United States, some veteran banders, who retrap large 
numbers of water birds over a period of years, frequently add a second 
band to the leg of an old bird wearing a fragile band put on some years 
before. O. L. Austin, BE. W. Ehmann, and M. D. Pirnie might be mentioned 
in this connection. Dr. Austin systematically rebands 9-year-old common 
terns regardless of the condition of the original ring at that age 
(Austin 1942). When TI began my study, A. S. Hawkins (in litt.) warned 
me to be cautious in interpreting waterfowl survival past the fifth 
year. 
Loss of legibility.--While the loss of a bird band’ is a 
serious weakness in some retrapping programs, illegibility is as poten- 
tially serious in others that depend on the general public to report 
recoveries to a central office. A band with the name and address of 
the Fish and Wildlife Service missing obviously has very little chance 
of ever being reported. As to which occurs first--illegibility or 
band loss--I do not know; it may vary interspecifically. 
In table 5 I have summarized the years of use on 17 illegible 
bands that the Fish and Wildlife Service had to etch with acid in 196 
and 1947. ‘The results may approximate what takes place in nature, but 
the table is potentially biased by variations in the number of birds 
banded in previous years, and by the inclusion of bands mutilated by 
gunfire. I conclude that band loss or its equivalent is potentially 
a serious factor that mst be constantly watched for in survival 
studies of adult birds. Its occurrence on a significant scale can 
often be detected by calculations of turnover and productivity. Its 
occurrence on a small scale is prcebably less easily perceived. 
2h 
