-5- 
Page 6, lines 4 to 9. To my way of thinking, there is noth- 
ing puzzling about the population recoveries at all; they are ex- 
actly what one might expect considering the respective breeding 
densities each year. The only recovery percentage point that is 
significantly off its theoretical place in the curve is the one 
for 1936. In appraising population recovery, we must consider 
not so much the percent of increase as the percent of increase 
for a given population level of breeding birds. No matter how 
favorable the breeding season from the standpoint of climate, etc. 
I would be surprised to find a rate Of increase exceeding 50% for 
a breeding density that was already close to the carrying capacity 
of- the Prairie du Sac area. 
You will find inclosed the Prairie du Sac recovery curve 
brought up to date by the inclusion of the 1940 fall census, as 
given in your letter of December 11 to Walter Scott. There are, 
incidentally, several errors in the recovery percentages given 
in Table 1; the 1929 recovery should be 450% with a question mark 
instead of 45%; that for 193S should be 254% instead of 228% (my 
faulty arithmetic); for 1939, 228% instead of 320% (assuming that 
the 1939-40 fall figure is correct). This curve is our one big 
contribution to knowledge of population recovery and shows as well 
as anything could be expected to that the percent of increase on 
one area for one period of years tended very much to vary in in- 
verse ratio to the breedirg density. There is nothing whatever 
unsubstantial about the recoveries for 1937 and 1939; the percen- 
tage points fall almost exactly in the part of the curve where