A comparison was made of the effects of long-term exposure to fixed-in
terval reinforcement with unsignaled and with signaled reinforcement o
missions. When successive fixed-interval cycles ended in reinforcement
, subjects showed a clear ''scalloped'' response pattern. When reinfor
cement was omitted, but a brief signal was given in lieu of reinforcem
ent, responding in the next cycle started earlier than it did after re
inforcement. When reinforcement was omitted without exteroceptive cues
, response rates peaked near the time of reinforcement and then declin
ed to a flat but substantial level. The classic reinforcement omission
effect was observed, in that total responding was greatest after uncu
ed omission, somewhat less after cued omission, and least after reinfo
rcement. However, response rate plotted as a function of time showed t
hat the uncued-omission condition had a very different function from t
hat of the cued-omission or reinforcement conditions. A failed-discrim
ination account of the omission effect might accommodate the three fun
ctions if the discrimination is considered to be a temporal one. The t
emporal-discrimination account argues that high-rate responding reflec
ts the accumulation of subjective time proximal to the memory of the t
ime of reinforcement. The accumulation resets completely with food rei
nforcement, incompletely with cues in lieu of reinforcement, and not a
t all in the absence of cues.