The Audubon Societies 247 



On the evening of Seplcniber 11, between 6. ,30 and 7. .50, hundreds of 

 Martins passed overhead, northward bound. They came in very irregular 

 numbers, sometimes only a few appearing, and again great numbers rushed 

 by on all sides. Some flew very low and others high, but they were all going 

 in the same direction. They were probal)ly headed for some roost, where 



A MARTIN COLONY 



they congregated before their departure. I made no count as to their numbers, 

 but feel safe to say that there were at least several thousand of them. 



The last Martins of the year were seen on September 21. This is an unu- 

 sually late date for their departure from this locality. In the fall of 1910 

 they were last seen on September 12. 



Question. — Can you tell me why the Martins remain with us so much 

 later in the fall than they do at points farther south? — Harry B. Logan, Jr. 

 (age, 16), Royalton, Minn. 



[It is encouraging to learn that Martins are so numerous in certain localities, when 

 the}' are so rare or entirely absent in many places. Undoubtedly, the erection of more 

 Martin-houses would aid in preventing the decrease of this friendh' and valuable species. 



By April 29, 191 2, the house represented in the accompanying illustration contained 

 eight pairs of Martins, a good beginning for the present season. 



Just why this colony of Martins, with those of the adjoining neighborhood and 

 from points presumably farther north, should remain later in the fall than colonies to 

 the south, it is impossible to say without a closer studj' of local conditions, and a list 

 of dates of the first and last arrivals of this species along the Mississippi Valley route 

 of migration during both spring and fall. It is probable that, as in the case of some other 

 species, individuals or colonies overtake or outspeed others in making the annual jour- 



