The X axis of the histogram is pronunciation time, on a millisecond scale, observations start at 0 and range up to 1600 ms. The Y axis is the frequency with with each interval of pronunciation time was observed. The data are collapsed across about 40 participants, which is why there are such large frequency counts.