288 RALLID^ : RAILS AND THEIR ALLIES. 



not ascertained with precision, but probably do not 

 differ materially from those of our other Rails. The 

 bird has been found in October, and even so late as 

 November. Many specimens have been secured of late 

 years during the migrations, in Southern New England. 

 It is also known to breed with us. Nidification is 

 substantially the same as that of other Rails. A set 

 of eggs is described by Coues as consisting of six, differ- 

 ing from those of P. Carolina in the ground color, which 

 is a rich, warm buffy-brown, marked at the greater end 

 with a cluster of reddish-chocolate dots and spots ; the 

 shape as in those of P Carolina, but the size smaller, 

 being from 1.15 X 0.85 to 1.05 X 0.80. 



A letter from Mr. Geo. Bird Grinnell to Dr. Merriam, 

 gives a glimpse of the interesting habits of this diminu- 

 tive Rail : " I was working a young setter on Snipe, on 

 a piece of wet meadow near Milford, Conn., and several 

 times during the early part of the day was annoyed 

 by the pertinacious way in which the dog would trail up 

 some bird which neither he nor I could start. At 

 length, during one of these performances, I saw the 

 puppy grasp at something in the bog before him, and 

 immediately a small Rail rose and fluttered a few yards. 

 Noticing its small size, and the fact that it had some 

 white on its wings, and seeing from its flight that it was 

 a Rail, I shot the bird before it had gone far, and when 

 it was brought by the dog I was delighted to see that it 

 was P. noveboracensis, a species I had never before seen 

 alive. During the day several more individuals were 

 secured. The next opportunity I had of looking for 

 these birds was, I think, Oct. 14. That day my brother 

 and I secured eight in an hour or two. They were 

 ridiculously tame, and would run along before the 



