From: Jens Elmegaard Rasmussen
>While both Eichner and Melchert say this, I do not find it compelling as
> Some more details: The numeral is frequently written as 7-an which could
> point to a form PAnatolian *siptan- which could however not to back
> to PIE *septm. (because PIE *-m. securely gives Hitt. -un so /s^ipta-/
> could not be direct descendant either). For the same phonological
> reason s^iptamiya- cannot be an intra-Hittite derivation from a
> Hitt. cardinal *s^iptam but must go back to PIE if it is *septm. + *iyo-.