364 MANTON COPELAND 
By testing a resting fish with carmine suspended in water, I 
was unable to discover any evidence of a current passing through 
the olfactory chambers, neither of the intermittent type dependent 
on respiratory movements common in many fishes, nor of the 
continuous kind produced by cilia, as described by Parker (’10) 
in the catfish. If, however, a colored solution is gently forced into 
one of the olfactory apertures by means of a pipette, it readily 
passes through the chamber and out the other aperture. I am 
led to conclude, therefore, that the forward locomotion of the 
puffer forces water through the anterior openings of the nasal 
chambers and out the lateral ones, and at that time conditions are 
most favorable to the stimulation of the olfactory cells by odorous 
substances. That, in truth, the puffer is seldom at rest, when in 
captivity at least, is readily apparent after a few hours observa- 
tion of its habits in a large aquarium. The elevated position of 
the nasal chambers is well adapted to the formation of water 
currents through them by forward locomotion. When a fish is 
swimming rapidly back and forth in an aquarium the olfactory 
organs become directed backward slightly, as it progresses through 
the water, but become erect as it turns in its course. At first, I 
believed that this inclination of the nasal organs was indicative 
of the force of impact of the water against them. I subsequently 
discovered, however, that this is not necessarily so, as the same 
result could be obtained by making a threatening gesture in front 
of the fish. 
Preliminary tests to determine whether the puffer would react 
to concealed food were begun upon eight to twelve fishes, which 
occupied one of the large observation aquaria of the Station. The 
method of experimentation was essentially like that of Parker 
(710, 711), and Sheldon (’11). Two cheese cloth packets of similar 
appearance, one containing meat of the smooth dogfish, Mustelus 
canis, and the other filled with cheese cloth, were suspended some 
distance apart in the aquarium. ‘The presence of food in one of 
the packets could be detected by the fish only through the stimu- 
lation of its chemical sense organs by material emanating from 
the meat. In several tests the packet containing meat was 
quickly seized and bitten open, whereas the other, although some- 
