In a recent paper (Nature, 1996, submitted) we claimed that the 'bump'
in the extensive air shower size spectrum near 10(6) particles is due
to heavy nuclei from a comparatively local 'source'. The energy spect
rum of this single source is of the shape advocated by Berezhko et al.
(JETP 82 (1996) 1) for supernova remnants (SNR) and is characterized
by, approximately, an E-2 spectrum up to an energy E-max followed by a
rapid fail above. The SNR model makes specific predictions for E-max
as a function of nuclear charge. If, as is likely, the CR nuclei are f
ully ionized, we must identify the 'bump' in the paper submitted to Na
ture with the CNO group of nuclei. We have, accordingly, searched for
the corresponding bump due to iron at a bigger shower size. Analysis o
f the world's data so far leads us to claim its detection, although no
t, yet, at as high a level of significance as the first bump. Prospect
s for augmenting the size spectrum technique, for studying what we cal
l this new branch of spectroscopy, are examined. (C) 1997 Elsevier Sci
ence B.V.