From: Piotr Gasiorowski
Message: 29176
Date: 2004-01-06
>> >While *h1 might explain Latin -a-, what explains the mysteryCf. also lu:men, lu:na < *leuks-mn., *louks-(m)n-ah2 (quite well
>> >disappearance of this unattested "s"? What exactly is the
>> >process? What am I missing?
>>
>> What you didn't miss earlier in nisdos > ni:dus.
>
> And is seen in the variant di- of dis-, e.g. English _digress_,
> _direct_, _divide_, _Nunc Dimittis_ (_dismiss_ comes from Mediaeval
> Latin). Its a very effective way of hiding roots, e.g. Latin
> _a:la_ 'wing'(> English 'aisle'), cognate with German
> _achsel_ 'shoulder' (and presumably with Latin a:xilla 'armpit').