8 
of snow fell. Later, Mr. Watson examined the bird houses, 
and found three of the birds dead within. 
Twenty-six observers, from the different counties of Mas- 
sachusetts, report martins as having disappeared; three re- 
port them as nearly extinct ; five, as rare ; eight, as rare and 
decreasing; one, “ as usual.” In response to letters of in- 
quiry sent out later to these and others, it was learned that 
nearly all the reports referred to migrating birds. Further 
extensive correspondence leads to the belief that we have 
no fully authenticated record of the breeding of the purple 
martin in Massachusetts in 1904, except in five localities. 
Mr. Robert O. Morris speaks of four small colonies in or 
near Springfield, which are still in existence, but one of 
these has decreased one-half in numbers. Miss Emily B. 
Adams, also of Springfield, speaks of two colonies, prob- 
ably some of the same, but says the birds are being gradu- 
ally driven from their boxes by the English sparrows. Mr. 
E. H. Mosher writes from Shawmut post-office in Hew Bed- 
ford that a single pair of birds reared their young there. 
Mrs. Mary R. Stanley of Horth Attleborough, in the same 
county, and near the Rhode Island line, says the martins 
are nearly extinct there, but are still breeding at West Attle- 
borough, where her brother saw two birds feeding their' 
young. Col. John E. Thayer says martins are still breed- 
ing at Lancaster; and Mr. William Holden states that a few 
pairs of birds occupied, and probably bred in, one bird house 
in Leominster. Capt. A. B. C. Dakin of Concord states that 
a single pair of martins were resident at a neighbor’s bird 
house, but failed to raise any young. This may be accounted 
for by the fact that the English sparrows, which are notori- 
ous for killing young martins, occupied the same domicile. 
Mr. Fred. C. Dodge says that martins, which arrived ten 
days later than usual, occupied a small nesting box near his 
house in Beverly, and thinks that they probably bred. 
We have records, therefore, of martins breeding in but 
three counties in the State, — Hampden, Bristol and Worces- 
ter, — with the probability that they bred in Essex County. 
What prospect is there that the species will eventually 
increase in numbers and reoccupy its old breeding places ? 
