Re: S mobile

Moreover, Old Chinese s- dropped from clusters without affecting the
tone on the way to several dialects, including Mandarin.

============
I'm afraid you are wrong.
From Baxter :
sui2 < snjuj

The loss you supposed is not listed among the
34 regular changes listed by Baxter.

#sn- is kept as #s-
no doubt.

Arnaud
====================

Personally, however, I'm very dubious of any connection of either of
these words with the PIE word. I could be persuaded of a connection
between the Chinese word and 'Tibeto-Burman' *glang ~ *glak ~ *grang ~
*klak ~ *m-klak 'cold', with some forms also meaning 'freeze', though
it's not the most obvious of connections. (The last two allofams are
cited by Matisoff with a final superscript 'H' - I don't remember its
significance.)
Richard.
============
Can't you see this *glang is the same as PIE *gel ?
Another word is han2 "winter" < *gol-an

You are so very dubious of any connection
that you would fail to see an elephant in your kitchen.

Arnaud
=================