+ 
miles 
reach southern Illinois in the fall are almost 2 weeks later, 
in early September (Fig. 24). 
Most of the Swainson’s thrush population passes 
through Illinois in September, with peak numbers in the 
north probably in early September, and incentral and 
southern Illinois in mid-September to late September. 
Small numbers are found until mid-October throughout the 
state. A few may linger into November (Ford 1956), but 
winter records (Ford 1956, Ridgway 1889, and Norton & 
Spitzer 1965) are accidental or erroneous. 
During ‘their night migrations, Swainson’s thrushes are 
among the most common casualties at television towers. 
There is good correlation of dates between our highest cen- 
sus counts and the highest kills in central Illinois (Fig? 24 
and 25). As in the spring, the lowest counts of Swainson’s 
thrushes in the fall were in the south. The tower kill data 
for Cape Girardeau, Missouri on the Mississippi River 
also indicate that fewer thrushes migrate through southern 
24 
Fig. 23.—General distribution of the S 
son’s thrush. 
Illinois. Paul Heye (mimeographed report) found the 
tember kills of migrants there to be generally lower 
those in central Illinois, and the percentage of thrust 
be much lower (1 percent, versus 14 percent in centra 
nois). 
In our censuses, the fall counts of Swainson’s and 
Catharus thrushes were consistently much below tho 
spring. This is contradictory to expectations, and a 
the findings of Annan (1962) whose eraphs for nortl 
ern Illinois just as consistently show much higher po 
tions in the fall. His data show a ratio of 1 bird 1 
spring to 11 inthe fall, versus our counts for then 
which show 4 birds in the spring to 1 in the fall. Thi 
serious discrepancy cannot be resolved without bette 
sus methods. . 
Our censuses showed the same spring to fall ratic 
1) of Swainson’s thrushes for east-central I}lino! 
northern Illinois, but very different ratios for west-¢ 
