2 authors, including:

incorrect. Because of an error in our database, the mb(Pn) values due to the formula from Evernden (1967) were plotted instead of the intended values of mb(Pn) due to the formula of Denny et. al. (1987). This error only affected the earthquakes in Figure l a and the corrected Figure l a is given below. This error does not change any of the conclusions in the original paper. The corrected Figure la still shows good separation between earthquakes and explosions, independent of magnitude. Best fit regression lines were calculated separately for the earthquake and explosion populations and are shown in Fig. l a. The two populations now have the same slope and show a separation of 1.13 log moment units (lmu). One event treated as an earthquake falls quite close to the explosion population in Figure la. This event is the May 14, 1981 event, which Taylor (1994) determined to be a large mine collapse at Gentry Mountain, Utah. It was not treated separately from the earthquakes in the original paper but here it was excluded from the earthquake regression line fit This event has some similarities to the explosions since it occurred at a very shallow depth of approximately 200 m, and is expected to radiate relatively more of its seismic energy as P waves than an earthquake