
quite unimportant. The Christmas Count areas were selected with- 
out regard to snine occurrence, they have a maximum radius of 
7-1/2 miles, and they include in their coverage all states except 
Nevada. The Winter Count areas, being many fewer in number, are 
less well scattered, but since they were selected for the specific 
purpose of counting snipe, the numbers seen per area average ten 
times as high. They were selected larzely at the convenience of 
the participants (who themselves were distributed at random with 
respect to snipe distribution}. <A series of random counts has been. 
run for the past two years throughout southern Louisiana. Since 
the results of these are comparable with those obtained through 
the regular mid-winter counts, and since Louisiana has the great- 
est concentrations of any state, there is no reason to believe 
that failure to provide a comoletely random sample has invalidated 
the results of the study. 
Table li. Sanple Size Required to Netect Snecified 
Changes in Snipe Population by itid-winter 
founts and by Christmas Season Counts 

-tid=winter Counts 

Percent Pacific Central itiss. Atlantic TOTAL 
Change 7 ae, a 
20% 28 119 92 82 ou. 
25 8: 76 59 52 60 
30 13 53 hi 37 2 
0 7 30 23 21. 2h 
Sample size (1955-56) 9 12 SO 21 92 
Christmas Season Gounts 
Percent Pacific Central iiss. Atlantic TOTAL 
Change 
15% 100 69 126 158 151 
20 56 39 71 89 86 
a 20, fe 2D h6 7 55 
30 28 18 32 Re 38 
Sample size (1955-56) 22 31 9 98 200 

