SHORT-BILLED MARSH WREN 273 



likely to be all on the same pitch, and in even time, like the beginning 

 of a song sparrow's song. Then the song is a simple 'tip-tip-tip- 

 trrrrrr'. But when there are four or five notes there is likely to be a 

 change in pitch, or a pause after the first note, giving a result like 



'tip tip - tap - trrrrrr'. The trill usually ends the song, but 



there is sometimes a still lower terminal note, making it end 'trrrrr- 

 tup'. The pitch of my records varies just an octave, from C " ' to 

 C " ", but no one song I have recorded covers a whole octave. The 

 lengths of songs in my records vary from 1% to 2% seconds." 



Francis H. Allen watched one at close range, as he sang, and says 

 (MS.) : "When he uttered the first notes of his song he raised his tail, 

 sometimes perpendicular to his back or even pointing forward, some- 

 times not so far, and sometimes hardly at all. With the last notes 

 of the song, the tail would go back to a position about horizontal with 

 the body. Often, though not always, it was jerked in time to the 

 notes, that is a couple of emphatic jerks at its highest point, simul- 

 taneous with the two emphatic opening notes of the song, and then a 

 quavering fall with the closing trill." 



FaU. — Mr. Saunders's observations in Connecticut indicate that 

 the fall migration begins early. He found birds singing in August 

 in a place where he "was very sure no such bird had been the previous 

 May and June, a tall grass area back of the salt marshes near Fairfield 

 Beach. On July 26, 1941, 1 found a bird singing in a similar locality 

 back of the beach, a place I passed or visited frequently throughout 

 the year. In the next few days I found several birds in this general 

 vicinity, and by August 9 the birds were abundant all through the grass 

 areas back of the beach, and I heard the song in many different places. 

 The birds continued abundant all through August but began to de- 

 crease early in September, and the last one was found September 20. 

 Evidently fall migration can begin in July, at least in some years." 



Field marks. — As the short-billed marsh wren is oftener heard than 

 seen, its peculiar and quite characteristic song is the best means of 

 identification. Its haunts are different from those of the long-billed 

 marsh wren, and it is almost never seen in the cattail swamps. If one 

 can get a good look at it, which is not easy, it can be distinguished 

 from the long-billed species by the streaked crown and by the absence 

 of the white line over the eye and the absence of the black back patch. 

 The shorter bill is not very conspicuous in life. 



'Winter. — The 1931 Check-list does not extend the winter range of 

 this wren beyond the southern border of the United States, but Dr. 

 Sutton (1940) took one and heard others on April 18, 1939, in south- 

 ern San Luis Potosi, Mexico; he also saw one in the American Mu- 

 seum of Natural History that was taken in Tamaulipas on March 22, 

 lf^88. Probably the species winters regularly in at least noithern 

 Mexico. 



