to territory unknown to them, 
and then set free. Many spe- 
‘cies have returned hundreds 
of miles in a few days after 
such sorties. A study of this 
type of navigation might help 
explain natural migrations. 
The gannet, a large 
white sea bird, which nests 
onan island in the Gulf of St. 
Lawrence was chosen for our 
firstexperiments. Seventeen 
of these birds were carried 
inland to northern Maine, a 
locale 215 miles from their 
nests, 100 miles from the 
nearest salt water and def- 
initely new to them. After 
releasing the birds, we fol- 
lowed their flight in a small 
airplane in such a way that 
we could watch a particular 
gannet for hours at a time. 
We kept the plane 1500-2000 
feet above the birds, which 
did not seem to be at all * 
disturbed. We found that Figure 2. The author holds a long-billed 
62.5 percent of the birds re- dowitcher after attaching a radioactive tag 
turned home in from 1 to 4 to its left leg. The tag consists of an alu- 
8 minum capsule less than 1/4-inch in di- 
pe aa aa erent ameter, with the radioactive material 
sealed into its center. 
of the several other wild spe- 
cies studied in the same way. Nine of the gannets were followed from 
the airplane for from 25 to 230 miles, while the remaining 8 were used 
as controls. The performance of the two groups was nct appréciably 
different. 
Figure 1, showing the paths of these birds, demonstrates graph- 
ically that the homing flight was not necessarily direct. Some birds 
which showed the most marked deviations nevertheless reached their 
nests within 2 to 3 days. Furthermore, mathematical analyses of more 
extensive homing experiments reveal that it is quite possible that other 
species which have returned from several hundred miles may also have 
found their way by exploration. Yet before airplane observations had 
been made it was almost universally believed that birds flew roughly 
straight home in such experiments--deviating no more than a few de- 
grees from the direct path. s 
Gannets, however, are not among the birds which naturally carry 
out the longest and most impressive migrations. For this reason it might 
be argued that they simply did not display the highly developed powers 
of navigation characteristic of plovers and sandpipers, for example. To 
obtain evidence on this comparative aspect of bird navigation a group 
from Cornell University, based at the Arctic Research Laboratory at 
Point Barrow, spent the summer of 1948 studying birds in arctic Alaska. 
Bg i ie 
