How did the timing between a metronome and food affect learning in Ivan Pavlov’s experiments?

In Ivan Pavlov’s experiments, the temporal relationship between the conditional stimulus (the ticking metronome) and the unconditional stimulus (food) was critical for associative learning.1

Conditioning was established most effectively when the metronome regularly preceded the delivery of food. Pavlov observed that the speed and strength of this learning were highly dependent on the interval between these two events. While conditioning could occur despite relatively long gaps, increasing this interval beyond a certain point invariably caused a decline in the animal's ability to form the association.12

Pavlov estimated that the optimum interval for this sequence was approximately 0.5 seconds. Furthermore, animals could adapt to the specific duration between the signal and the food; through "delayed conditioning," they could be trained to time their conditional response to occur after the same interval that separated the stimuli during the training process.2