Type-III echoes observed at auroral latitudes by coherent radars have
two distinguishing spectral characteristics: they are relatively narro
w and have small Doppler shifts. It has never been entirely clear whet
her this is characteristic of a new mechanism of plasma instability or
whether these echoes are just a subset of normal Type-I (Farley-Bunem
an) spectra with unusually small Doppler-shifts. In this paper, a larg
e number of spectra gathered during two 50-MHz radar observing campaig
ns in 1981 and 1983 have been examined in order to see whether there i
s any statistical basis for postulating a separate class of echo and,
if so, to determine which properties these echoes have. From the data
presented, it appears that Type-III echoes are, indeed, a separate cla
ss. In agreement with previous observations, the data show that these
echoes do not appear to be quantified in Doppler velocity but are spre
ad over a range-of velocities. There is a strong dependence on aspect
angle, with Type-III echoes appearing preferentially on radar paths wh
ere the aspect angles are significantly different from 'perfect' aspec
t. For these high aspect angle paths, there appears to be no marked de
pendence of the probability of occurrence of Type-III echoes on the or
ientation of the radar beam with respect to the ambient electric-held
direction. Finally, there is some suggestion that Type-III peaks are o
ften accompanied by harmonically related spectral peaks.