Douglas (1992) has presented evidence based on deconvolution analyses
of British array data recorded from explosions at the Shagan River Tes
t Site (STS), USSR and the Nevada Test Site (NTS), USA from which he h
as concluded that the source time dependence characteristics of explos
ions at these test sites is significantly different from that associat
ed with the Mueller-Murphy (M-M) seismic source model. The objective o
f this note is to demonstrate that the M-M model is in fact consistent
with a very large variety of short-period P-wave observations from ex
plosions and to argue that the alternate source time function proposed
by Douglas (1992) is inconsistent with both theoretical and observati
onal constraints on the explosion source process. More specifically, n
etwork-averaged teleseismic P-wave spectra corresponding to NTS, Frenc
h Sahara and STS explosions in granite are used to illustrate the fact
that the M-M source model can provide remarkably good descriptions of
the observed amplitude levels and spectral shapes over the short-peri
od band extending from at least 0.5 to 2.5 Hz, as well as seismic esti
mates of explosion yield which agree very closely with corresponding y
ield estimates obtained from other, independent sources.