Someone wrote:-
> But early Germanic *xarvað- would have become Common Slavic
> **xorvod- > **xravod- in dialects related to Croatian.
> Besides, where is the "-ian" part? One would definitely expect an
> ethnonymic suffix like *-ên(-in)- or *-ak- in a name derived from a
> mountain range.
At the time of the big migrations, all sorts of peoples were moving
into and across the area, and likely local place names were passed
about repeatedly with more or less distortion among and between
native and immigrant speakers of many sorts of dialacts and
languages, and exact etymological routes might not be reliable. As
well as regular phonetic rules, there would be "Ypres-to-Wipers"-
style familiarizing distortions, and plain simple mishearings. For
example, in Old Norse the Biblical name [Elizabeth] ended up as
[Ellisíf] = "old-age - wire", and [Jerusalem] as [Jórsala] (compare
Uppsala in Sweden). I can still imagine some migrating Slavs picking
up Germanic [harvað] (a form attested in a Viking-age Old Norse poem)
directly or indirectly and after a few passages between speakers of
different dialects it might end up as [hrvat] used for itself by a
group of Slavs who lived for a while in the Carpathians.