An attractive pendulum consisting of a two-inch crystal ball suspended on a fused silica rod is the focus of an experiment to measure possible effects of conscious intention on an analog physical system. The pendulum is enclosed in a clear acrylic box, and provided with a computer controlled mechanical system to release it from the same starting height in repeated runs. A high speed binary counter registers interruptions of photodiode beams, to measure velocities at the nadir of the pendulum arc with microsecond accuracy. In runs of 100 swings, taking about three minutes, operators attempt to keep swings high, i. e., to decrease the damping rate (HI); to reduce swing amplitude, i. e., to increase the damping rate (LO); or to take an undisturbed baseline (BL).

Over a total of 1545 sets, generated by 42 operators, the HI \- LO difference is significant in the direction of intention for five individuals, and the difference between intention and baseline runs is significant and positive for five other operators. The overall HI \- LO difference is reduced to non-significance by strong negative performances from several operators, four of whom have comparably large scores in the direction opposite to intention. Analysis of variance reveals significant internal structure in the database (main effects $F sub {4, 189}~=~ 2.845,~~p~=~.025$). Subset comparisons indicate that male operators tend to score higher than females, and that randomly instructed trials tend toward higher scores than volitional trials, especially for male operators. Trials generated with the operator in a remote location have a larger effect size than the local trials.