252 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
smell, as the maxillary barblets begin to wave about more actively and finally the fish 
becomes restless. He does not find the food, however, unless in the course of his 
movements it actually touches some part of the body. 
During May and June, 1902, more systematic experiments were undertaken with 
these fish, and since these experiments are typical of those subsequently performed 
on other species of fishes 1 shall recount them in some detail. At first a few speci- 
mens were taken out in a shallow tray and the attempt made to feed them in various 
ways under close observation. They were, however, so much frightened by the 
exposure to bright daylight and by the proximity of the observer, in spite of all pre- 
cautions, that no reactions could be obtained which were at all satisfactory. A bit of 
fresh meat on a long-handled needle could be thrust slowly toward the fish as he lay 
quietly on the bottom, rubbed over his body or on the barblets, and even over the 
lips, without evoking a movement of any kind in response. The same observation 
was made with the spotted suckers. The fishes in both cases had been without food 
for several days and were very hungry, but were obviously too much frightened to 
respond to the food stimulus. 
On another occasion the same conditions were prepared, except that a few dead 
leaves were littered over the bottom of the tray. The fish when placed in the tray 
immediately sought the shelter of the leaves, and, after a suitable interval to enable 
them to become accustomed to the place, the feeding experiments were repeated. 
Selecting a fish which was entirely concealed under a large leaf, save for a projecting 
barblet, a Hit of meat on a slender wire was gently passed down into the water in 
such a way as to touch the projecting barblet. it was instantly seized and swallowed. 
This was repeated many times with several of the fishes. 
In subsequent experiments the fish were not removed from their own tank, but 
the water was drawn off so that it was only about six inches deep. Here they would 
lie under the leaves and the experiment could be continued with a minimum of 
disturbance to the fishes. The experiment of touching the barblet with meat was 
repeated hundreds of times with an almost invariable result that the fish instantly 
turned and snapped up the morsel. If the meat was merely held very close to the 
barblet it usually produced no response. The reaction was obtained equally well, no 
matter which barblet was touched. 
In a later series of experiments I found that the fish would almost always turn 
and seize the meat if he were touched at any point on the head or body. If the tail 
of the fish projected out from under a leaf and the skin near the root of a tail fin 
were touched with meat the fish would turn and seize the meat. This reaction was 
not so uniformly made at first as that from the barblets, but after a dozen or so of 
trials it followed with equal promptness and uniformity, the fish apparently requir- 
ing a little practice to learn the movement perfectly. 
The experiments last described were repeated the next day and by this time it 
was found that the fishes had become so tame that they would take the meat if 
offered to them in the open, without the shelter of the dead leaves, though not so 
certainly as when under the cover of the leaves, often taking fright from the shadow 
of the observer’s hand or from some other cause. 
In none of these cases did the fishes appear to see the bait or to perceive it in any 
way other than by actual contact with the skin at some point. If the bait were held 
