MAKSH-WRENS. 99 



essentially from those of the long-bills. Those I have 

 found were all in similar localities, and if the birds were 

 not seen, could only be recognized by the fact that the 

 eggs are white. 



What I take to be a curious error in Barton's " Frag- 

 ments of Natural History" reads as follows: "Mota- 

 cilla Troglodytes? (marsh - wren) commonly continues 

 with us the whole year; in the winder -time taking 

 shelter in our houses, stables, etc." Certainly both spe- 

 cies of marsh-wrens are strictly aquatic, and we have 

 seen that they are migratory. Here we have no evi- 

 dence of a change of habits, but a confounding of the 

 marsh-wren with the Carolina wren. This is the more 

 interesting, inasmuch as, in Wilson's time, and later, 

 when Audubon wrote, the latter was held to be a rare 

 bird in this vicinity, and my own inference that a re- 

 cent colony liad become established as far north as 

 Philadelphia, falls. Certainly, Carolina wrens are resi- 

 dent species in the strictest sense ; and how they could 

 have been confounded with the aquatic marsh-wrens is 

 not readily comprehended. 



A word with reference to a feature common to both 

 marsh-wrens. Given clear skies, as the moon fulls in 

 May, June, July, and August, these birds forego sleep, 

 unless they take short diurnal naps, and with that same 

 restlessness that marks their movements "from the 

 rising of the sun even unto the going down of the 

 same," these wrens are ready to greet the rising moon 

 and sing to her a sweet sibilant serenade until the day 

 dawns. With singing -birds generally, such perform- 

 ances at night are occasional or uttered while dreaming, 



