646 R. M. STRONG 
5. RESULTS OF EXPERIMENTAL STUDIES OF THE SENSE OF 
SMELL IN RJNG DOVES 
The methods which were employed in this work have been 
described on pp. 632-9, where it will be noted that a labyrinth 
was used, and the birds were compelled to find their food in one 
of four chambers which were selected by the experimenter at 
random. 
If the doves had entered the chambers entirely in hit or miss 
fashion with no clues or habit preferences which would influence 
their choice of a chamber, they would be expected, according to 
the law of error, to distribute their first choices equally among 
the four chambers, and the compartment which contained the 
food would be entered first 25 per cent of the time. What ac- 
tually happened is shown in the following tables: 
TABLE 2 
Cologne series 
No. 62,19 per cent; No. 30, 28 percent; No. 92, 333 per cent; No. 24, 24 + per cent 
TABLE 3 
Musk series 
No. 62, 20+ per cent; No. 30, 28.9 + per cent; No. 92, 31 per cent; No. 24, 25 
per cent 
TABLE 4 
Violet sachet powder series 
No. 62, 50 per cent; No. 30, 40 per cent; No. 92, 28 per cent; No. 24, 75 per cent 
TABLE 5 
Oil of bergamot series 
No. 62, 37+ per cent; No. 30, 31 + per cent; No. 92, 36+ per cent; No. 24, 47 
per cent 
It will be seen in the table that when cologne and musk were 
used, the percentages of correct first entrances were not signifi- 
cantly different from what might be expected according to the 
law of error. With oil of bergamot, however, the percentage is 
