THE SENSE OF SMELL IN BIRDS 647 
notably large for all four birds. This is especially true of No. 24, 
a bird which gave other suggestions of discrimination in its behay- 
ior than are indicated in the tables. The percentages are also 
large in the violet sachet powder series, but not much significance 
can be attached to this fact because of the small number of exper- 
iments in this series. 
It is of course conceivable that the food itself might have had 
an odor for the doves, even though none was apparent to the 
writer. However, if there had been any olfactory stimulation 
by the food, all four of the birds probably would have shown sim- 
ilar responses to the stimulus. That failure to find the food 
through a hypothetical odor from the food material was not due 
to disturbances produced by the odors which were employed in 
the experiments, was indicated in a short series of control exper- 
iments when no odors were used. In this series, the food was 
found entirely by chance, apparently. 
There was an unfortunate tendency for the doves to enter 
chamber A first by habit. After this compartment had been 
entered, if it was empty, they would go through the usual hunt for 
the food and would ordinarily show no further tendencies to enter 
a second chamber by habit. The extent to which the birds made 
their first entrance by habit is indicated in the following tables. 
Tables showing the number of times each chamber was entered 
first : 
TABLE 6 
Cologne series 
Nos62) “A; 8:8, 14 C76De5 
NOMS O) AL 35 Belt @a 2m) 
IMIG), GAP ANS MSIE OE (Ce aie ID) 3 
NWa.2t A, 12: 'B, 4) ses 
TABLE 7 
Musk series 
INOS 62, A, Tk Bela G14 Dre 
Non 306 Ay 323) BY 4 Cree? 
No: 927 A, 383. By 43/@F2a5 D5 
Now24> Ass 23 1 BiiS Crd sD eS 
