158 ROBIN ROOSTS. 



larger than it would have been had the sky 

 been clear. 



On the afternoon of the 29th I again 

 counted the arrivals at the eastern end; but 

 though I set out, as I thought, in good sea- 

 son, I found myself once more behind time. 

 At 6.30 robins were already dropping in, 

 notwithstanding the sky was cloudless. In 

 the first five minutes eighteen birds ap- 

 peared; at sunset 818 had been counted; 

 and at 7.30, when I came away, the figures 

 stood at 1267. "The robins came more 

 rapidly than last night," I wrote in my note- 

 book, "and for much of the time I could 

 keep watch of the southeastern corner only. 

 My vision then covered much less than a 

 quarter of the circuit; so that if the birds 

 came as freely from other directions, at least 

 five thousand must have entered the wood 

 between 6.30 and 7.30. As long as it was 

 light they avoided passing directly by me, 

 going generally to the left, and slipping into 

 the roost behind some low outlying trees; 

 though, fortunately, in doing this they were 

 compelled to cross a narrow patch of the 

 illuminated western sky. I suspect that the 

 number increases from night to night. Be- 



