Haukur Thorgeirsson wrote:

> It seems to have some
> significance that in the North 'sunna' and 'máni' do not have the same
> genders as 'sol' and 'luna'. In the South the sun is a harsh (masculine)
> master but in the North it's a kind mother.

Heil Haukur,

this is exactly the explanation I used to come up with for a long time.
But if you take a close look at the other Indo-European languages the
case is very confusing and doesn't seem to show any systematic pattern
which makes the above explanation somewhat dubious, at least:

Lithuanian:
Sun: fem. - Moon: masc. (originally neut., but Lithuanian lost neut.
completely).

Slavic:
Sun: neut. - Moon: fem. and masc. (They have two words for the moon)

Indic:
Sun: neut. (since oldest times). Later a fem. word is derived from that.
Moon: masc.

Proto-Indo-European:
Sun: neut. There seem to have been several words for the moon, the most
important one masc., at least one more which is fem.

Old-Irish:
Sun: fem. Two words for the moon. One originally neut., after the
decline of Old-Irish neut. it became masc. The other one masc. In
contemporary Irish both words for the moon are fem.

Cymric (Wales):
Sun: originally fem., today masc., two words for moon, both of them fem.

Old Continental Celtic:
Sun: gender unknown, probably fem., Moon: masc.

All Germanic languages:
Sun: fem., Moon: masc. But in English the genders changed under Romance
influence in the late Middle Age, so the sun is referred to as "he" and
the moon as "she".

Greek and Latin:
Sun: masc., Moon: fem.

Hope this helps,

Greetings and best wishes,

Kurt Oertel