i 4 BIRDS OF THE 



May, when absence from the city for the 

 summer has terminated the records for the 

 season, but not until the spring migration 

 had almost ended and there would be little 

 more to record, yet failing to obtain the 

 June records, such as they might be. Not 

 half a dozen days in the eight seasons sub- 

 sequent to the year of beginning have the 

 visit and the record been omitted after the 

 season had opened. The observations have 

 been systematically carried on with a grow- 

 ing interest and a realization that the ma- 

 terial gathered was supplying valuable 

 data in relation to several aspects of bird- 

 migration. 



Usually two complete rounds of the Gar- 

 den have been made each morning in order 

 that the full census of the day might be 

 obtained. The second enumeration of flocks 

 and recording of individuals has commonly 

 confirmed the first, or if not, such jnodifi- 

 cations in numbers have been made as the 

 second round indicated to be a truer record. 

 Naturally facility in accomplishing this 

 with an approximation to accuracy was 



