Temporal patterns of sexual behavior in rabbits as determined by an automatic recording technique.

Scientists wired male rabbits with tiny electrodes. Every thrust during mating pressed a switch. A computer logged each press, second by second, for weeks.

No one had to watch or count. The machine ran day and night and stored the data.

The rabbits thrust about 13 times every second. Each mating lasted only 1.5 seconds.

Most action happened at night. Activity came in tight clusters, then quiet spells. A clear 16- to 20-day cycle matched pseudopregnancy.

Gilchrist et al. (2018) and Lotfizadeh et al. (2020) later used accelerometers, not wires, to catch repetitive body motions in people. All three studies show machines can count fast, brief movements that human eyes miss.

Bigby et al. (2009) tested how well people agree when they watch and score the same stream of behavior. Their work warns that high-rate acts, like the 13-per-second thrusts, can make human observers disagree. The 1967 electrode fix removes that error source.

Hulse (1960) built an earlier lick-counter for rats. Both gadgets share the same idea: let the animal close the circuit and let the computer do the math.

If you study any rapid, repetitive behavior—hand flapping, stereotypy, or SIB—this paper is a blueprint. It tells you to trade stopwatches for sensors, log continuously, and look for cycles you might miss in spot checks. Modern wearables now do the same job without wires, but the logic is unchanged: let the data stream tell the real story.