As an aid in the study of the time lag of spark breakdown, devices have been constructed to apply a rectangular voltage wave and to measure its length which is from one to 5000 microseconds in duration. Oscillograms of switching transients and test‐gap voltages in this investigation indicate that many previous investigators have been deceived in assuming a rectangular wave of applied voltage, and that others may have introduced errors by assuming the gap voltage to be influenced by reflections. It is possible to apply an essentially rectangular voltage by (1) maintaining the spark which closes the circuit through the applying switch before the contacts meet mechanically, and (2) by making the test‐gap circuit aperiodic. The length of the applied voltage wave is determined by a vacuum tube circuit designed to pass a constant current output as long as the bias to the first tube exceeds a predetermined value. The output current is passed through a ballistic galvanometer whose reading is a measure of the time during which the gap voltage remains above the datum level. A thyratron circuit has been arranged to supply the timer with rectangular waves for the purpose of calibration. Cathode‐ray oscillograms showing errors of 5 percent to 10 percent indicate that the accuracy of the apparatus is sufficient for most breakdown studies.