Return from the South. 



The weather affects their return very much, but generally by 

 April 5th. several individuals can be seen. The first arrivals 

 usually come singly, are males in adult plumage, and stay only 

 a few hours, then disappear as quickly as they came. This is 

 nearly always the case when they put in an appearance before 

 March 28th., but when March weather continues blustering and 

 keeps the birds away until April, they come in small bands, and 

 later in large numbers. Sometimes we have disagreeable wea- 

 iher in April after the birds become comparatively common, 

 but unless freezing temperature is experienced, they do not ap- 

 pear to be affected, and not a few times have I seen an old male 

 bird thrust his head out of his door, surrounded by little heaps 

 of snow on the porch, and twitter in seeming contentment. 



In the spring of 1900 we had a period of frosty nights late in 

 April after a number of brds had arrived, and one morning 

 while the air was yet keen, I noticed the absence of my birds 

 and investigated the cause which I found to be a "pack" of the 

 birds into a central room of house No. 2. One of the birds had 

 become wedged crosswise the door and was dead. After re- 

 moving this one I liberated sixteen live birds. They were pack- 

 ed in a room six by six by seven inches. At other times I have 

 observed numbers of the birds entering a single room on clear, 

 cold evenings. 



Following are my observations on the arrival of the Martins 

 at Waynesbnrg for twelve years : 



1891 — First seen April 8, four birds; common April 17. 



1892 — First seen March 31, one bird. 



1893 — No record. 



1894 — First seen April 2, two birds. 



1895 — First seen March 30, four birds. 



1896 — First seen March 28. 



1897 — First seen March 21, six birds at the various bird 

 houses in and near Waynesbnrg. One adult male remained at 

 mv box most of forenoon. For a week afterward the weather 



