“…It was concluded that new experiments with radio frequency sources were needed, and several groups modified the existing break spark test equipment to reduce its capacitive loading effect on radio frequency sources: experiments were than undertaken over a range of frequencies to determine whether these had an influence on the power required for ignition. Early experiments were undertaken with 50-W source impedances, as seemed a logical choice, but the results showed somewhat surprisingly high powers to be required and an unexplained frequency dependence [3], [4].…”