males were caught on other singing grounds, most of which had been used 
off and on throughout the spring. When Prescott peninsula was retrapped 
in late May two birds were caught on new singing grounds closely adjacent 
to those where birds had been caught in April. Both of these birds were 
unbanded individuals suggesting they had moved in subsequent to the 
April trapping. 
Between March 29 and April 14, 19 birds were trapped on 18 singing 
grounds in the 3.8 miles of road regularly censused by a special student 
(Figure 1). Attempts to trap two additional birds failed. One would 
assume that many of these trapped individuals were migrants since the 
census suggests April 8-10 as the height of migration. Six of these 19 
were recaptures from the spring of 1950 so were presumed to be residents. 
Two additional birds were taken as repeats in late May trapping leaving 
the possibility that 11 were migrants. Seven birds were captured be- 
tween May 16 and May 22. Three were repeats of 1950 birds caught in 
the same sites. Two were repeats of birds banded in April, both of 
which were first caught over a mile away, while two were new birds. 
During the spring 23 individual birds were caught on singing grounds 
in this census route. Due to the fact trapping was carried on over a 
long interval of time the trapping results in this instance help little 
in interpretation of the census results. During three checks between 
April 17 and April 26 only 10 singing males were counted (Table 3). 
The count built up towards the end of April, reaching a second peak 
May 8. What evidence was gathered in late trapping suggested some new 
arrivals, and a good deal of shifting by local males. In most instances, 
however, throughout the peninsula new birds, as well as repeats, were 
taken on old singing sites and census runs revealed practically no new 
singing grounds occupied. 
As the writer has interpreted these data the count of May 8 or 10 
would seem the best indicators of the breeding population. Likewise in 
the Newburyport area (Table ) the count of May 6 or May 12 should be 
used. These results suggest the importance of running census counts 
late in the breeding season. In this area in 1951 the first week of 
May appeared to be the best time for the annual inventory in spite of 
the fact it coincides with what is probably the height of the hatch. 
Censuses wil] be taken regularly throughout the spring, however, for 
several years to discover whether the pattern evident the past two 
seasons is consistent. 
Another extremely variable factor in censusing by counting singing 
males is the perceptiveness of different personnel in hearing either the 
"peent" note or the flight song of males. A few tests were run here 
demonstrating that in a group of observers of otherwise normal hearing 
there were great differences in their ability to hear the frequency of 
the male woodcock's notes. Some could hear "peents" clearly but could 
scarcely pick up the flight notes. There were cases of the reverse. 
Counts varied as much eas 20 percent depending on the observer. There- 
fore, it is of great importance that the same observers census the same 
routes every year insofar as possible. 
ho 
