Re: First iron swords on mass scale
Baltic and Slavic have related terms: Lithuanian gel(e)z^is,
OPrussian gelso, Slavic z^elEzo < B-Sl *gele(:)Zo-. I'm not sure how
to analyse them (*gWelh-eg-??).
C.D. Buck suggests a connection with the root *g^hel- "yellow; gold"
(Slavic *zol-to > zla:to ~ zoloto "gold"). Maybe a borrowing from a
centum word akin to Greek khalkos "copper".
In fact, the *gHel- connection is what most Slavicists assume, but
I'm sceptical. Iron is neither yellow nor gold-like.
The exact form of the B-S etymon is indeed problematic. We have
Slavic z^ele^so and Lith. gelezis, OPr. gelso (I suspect /gelzo/, in
view of the German orthography of Old Prussian). The Slavic and
Baltic terms are not quite compatible, apart from initial *g(w)el-.
Actually, the Slavic word is z^ele^zo and the Lithuanian one
gelez^is. Lithuanian has eliminated the neuter gender, so the
different stem is scarcely surprising. Slavic z^ (before a front
vowel) = Baltic g (PIE labiovelar or unsatemised velar), and Slavic z
= Lithuanian z^ = Old Prussian z, orthographic (satem treatment of *g
(H)). The only trouble is the vowel-length difference in the second
syllable. Before IE unaspirated voiced stops Balto-Slavic shows vowel
lengthening, the precise conditions for which are not yet known
("Winter's Law"). The apparent mismatch may be due to the failure of
Winter's Law to apply in this particular Baltic form for reasons we
don't fully understand at present (there are also a few Slavic words
in which it "should" but doesn't apply). With this single
reservation, the match is pretty good.
O. Trubachev suggests other etymology for Common Slavic *z^ele^zo :
from of the root *z^el- meaning 'clot' (also 'knob, gland', from
whence, eg., *z^IlvI 'tortoise') >'swamp-ore'>'iron eliquated from
such ore'. That would reflect the real technology adopted by Slavs.
To be honest, I can't recall his treatment of -e^z-o-/-e.z^-i-, but
everything seemed neat when I was reading that.
Now you can easily found cognates and develop PIE form :).