Smith on Purple Martin Migration. 41 



A MIGRATION FLIGHT OF PURPLE MARTINS IN 

 MICHIGAN IN THE SUMMER OF 1905. 



FRANK SMITH. 



In the Bulletin of the Michigan Ornithological Club, Vol. 

 V. (pp. 77-78), the writer described an extensive flight of 

 Sparrow Hawks which took place August 30th, 190-4, at 

 Macatawa, Mich., a summer resort located on the east shore 

 of Lake Michigan, southwest from Grand Rapids. More than 

 a thousand hawks passed the point of observation within a 

 period of about seven hours. Thev were flying southward 

 along the lake shore, while a moderately strong wind was 

 blowing from the northeast, and had been blowing frcin a 

 similar direction during the preceding day and night. Un- 

 fortunately no effort was made to determine whether similar 

 numbers of hawks were passing over the territory farther 

 inland, and so there was merely an assumption that a con- 

 centrated stream of these birds was passing along the shore, 

 and that this concentration was due to the flight of the hawks 

 with the wind until they reached the shore, which they then 

 followed in preference to continuing their original direction, 

 which would have taken them over the lake. 



On August 15, 1905, at the same locality, the writer had an 

 opportunity to watch an extensive flight of Purple Martins. 

 Again -a moderately strong wind was blowing from the north- 

 east, which direction it had held during the previous night. 

 Actual counts of the numbers of individuals passing south, 

 at various intervals between 9 a. m. and 12 :30 p. m., gave an 

 average of thirty-two per minute. As the whole width of 

 the bird stream was too great to be under observation at one 

 time, there must have been more than ten thousand indi- 

 viduals which passed the point of observation on that day. 



On this occasion it was determined tO' ascertain whether or 

 not the great numbers were limited to a narrow area along the 

 lake shore. Advantage was taken of an interurban car going 

 to Holland, which is about six miles inland, and there obser- 

 vations were made for comparison. Between 11 :03 and 11 :30 



