>Why is Elmet out of the picture? Because it's Brittonic (or a substrate thereof)? If so, it's only out of the picture if the ancestry of himilizzi is straightforwardly Germanic.
> --- In firstname.lastname@example.org, "Torsten" <tgpedersen@> wrote:
> > --- In email@example.com, "dgkilday57" <dgkilday57@> wrote:
> > > --- In firstname.lastname@example.org, "Torsten" <tgpedersen@> wrote:
> > > >
> > > > [...]
> > > >
> > > > Unless of course you accept that *himil- etc is not Germanic.
> > >
> > > But if not Germanic, then what? It would have had to be
> > > borrowed before Grimm's Law as *kemil- or whatever, and that
> > > gives no advantage that I can see over assuming a derivative of
> > > inherited *k^em-, regardless of the difficulty in finding exact
> > > morphological parallels for the whole set of words. I am not
> > > ruling out the possibility of a loan, but mere l/n-suffix-
> > > alternation does not raise a red flag. In fact, by chance I
> > > ran across two Gmc. examples of -n > -l in loanwords: OHG
> > > <kumil> beside <kumin> from Lat. <cuminum> 'cumin', and Westf.
> > > <laemmel> from Lat. <lamina> 'layer'. But all these illustrate
> > > is that the suffix-alternation in inherited words was
> > > sufficiently common to be generalized. They do not make such
> > > an alternation diagnostic of loanwords.
> > We can leave -l/-n alternation out of consideration as a
> > diagnostic, but we still have the association with the -et-
> > collective suffix, otherwise most connected with tree and plant
> > names of dubious ancestry, cf Udolph
> > http://tech.groups.yahoo.com/group/cybalist/message/57163
> > and with settlements which would have been caught in the middle
> > between a northern and a southern branch of Germanic expansion
> > from Przeworsk (Suevi)
> > http://tinyurl.com/ydcsm68
> > (thus Venetic?)
> Regarding the map, I will need to look in detail at the English
> place-names cited by Udolph (unless Brian has already done so).
> Anyhow, this suffix also forms denominative abstracts like OE
> <the:owet> 'servitude', so I would not bet too heavily on
> <himilizzi> as a collective, now that <Elmet> is out of the picture.