466 
PACIFIC SCIENCE, VoL XX, October 1966 
in two training sessions. Training was continued 
for three more days (24 sessions, 438 min) 
during which time Bl consistently demonstrated 
that it had learned the association. In addition, 
it also learned to avoid the negative area with 
almost perfect performance. 
Orientation and Form Discrimination 
WHITE HORIZONTAL VS. VERTICAL RECTAN- 
GLE: Discrimination of orientation of rectangles 
was demonstrated by all sharks, four blacktips 
and one grey, presented with this problem. In 
training sessions extending over 1-3 days, clear 
signs of discrimination were shown after the 
following number of reinforcements: B3-127; 
B4-365; B5-609; B 6-1 38; Gl-120. In all 
cases, continued training and tests provided 
consistent evidence that the sharks could dis- 
criminate. 
Among the blacktips, B3 and B6 with the 
faster rates of learning were fresh sharks which 
had not been trained previously, whereas B4 
and B5 with the slower rates had been trained 
unsuccessfully to the apparently difficult circle- 
triangle problem (see below). The grey shark, 
with the fastest rate of learning, had previously 
been trained successfully to the circle-triangle 
problem. 
Only B3 learned to avoid the negative end. 
It retained the discrimination for at least 18 
days without reinforcement. A gradual extinc- 
tion of the association of the negative target 
with shock was apparent in tests conducted 
after 7, 12, and 18 days, without reinforcement. 
WHITE SQUARE VS. TRIANGLE: Two blacktips 
were exposed to this problem with uncertain 
success in one (Bl) and certain success in the 
other (B2) . Neither shark had prior training 
experience. 
Bl showed occasional signs of discrimination 
after 362 reinforcements (8 sessions in 3 days), 
but its behavior was too erratic to afford a firm 
conclusion, even after an additional 11 training 
sessions and a total of 629 reinforcements. 
An initial attempt at training B2, involving 
375 reinforcements (14 sessions in 3 days), 
was unsuccessful. When training was resumed 
after 4 days of rest, a sudden and obvious 
development of the associatoin was apparent 
after 68 reinforcements (5 sessions). Continued 
training confirmed the positive conclusion. 
Tests of retention after 5 days without rein- 
forcement were inconclusive. 
WHITE CIRCLE vs. triangle: Of three sharks 
presented with this problem, one (B4) failed 
to discriminate, another (B5) showed incon- 
sistent signs of discrimination, and the third 
(Gl) made the discrimination. 
After 103 reinforcements in 5 sessions, B4 
started to swim continuously in the center zone. 
Training was terminated because of the per- 
sistence of this behavior. 
After 727 reinforcements (17 sessions in 3 
days), B5 showed some signs of discrimination. 
However, an additional 305 reinforcements 
(10 sessions in 2 days) failed to provide further 
evidence. 
The grey shark (Gl) was trained success- 
fully to discriminate between the circle and 
triangle after 476 reinforcements (10 sessions 
in 2 days), but it did not learn the avoidance 
response, even after an additional 12 sessions 
(216 min, 380 reinforcements in 2 days). 
Color Discrimination 
The training of blacktips to discriminate be- 
tween grey and colored squares of the same 
subjective brightness (to the human eye) 
seemed to produce more hypersensitive and 
erratic behavior than was displayed in other 
discrimination problems. It was clear that some 
of the subjects discriminated between the tar- 
gets, but no adequate attempt was made (by 
substituting different shades of grey) to de- 
termine if the discrimination was based on dif- 
ferential brightness or hue per se. 
grey vs. purple: One shark (B5) presented 
with this problem failed to show any sign of 
discrimination after 611 reinforcements during 
10 sessions in 2 days. 
grey vs. blue: Both of two blacktips pre- 
sented with this problem showed infrequent 
signs of discrimination, but in neither case was 
it possible to reach a firm conclusion. 
B5 had apparently not made the discrimina- 
tion after 358 reinforcements (6 sessions in 1 
day). However, two tests conducted after a 
period of rest provided some evidence that it 
could distinguish between the targets. 
B8 was subjected to 2 days of training dur- 
