Hi Paul,

"...when at inferior conjunction, Venus too is approximately as
bright as magnitude -3."

I commented on Venus and Mercury vanishing without actually
having checked the matter (I made a rash assumption that they would
get vanishingly faint, too). For these objects, the magnitude is
based on empirical formulae from the _Explanatory Supplement to the
Astronomical Almanac_, which give a passably decent magnitude
even when the objects are near the sun.

"...The lunar magnitude ought not to be a problem here: when the
Moon is really new... its magnitude is still as bright as magnitude
-3, due to Earthshine."

Ah, a good point. My current formula for lunar magnitude doesn't
account for this.

The _Explanatory Supplement_ gives magnitude formulae for all
planets and some planetary satellites, covering the range of phase
angles seen from Earth. It doesn't give such formulae for the moon,
less "significant" satellites, outer planets as seen at small
phase angles (for example, Mars as seen from Jupiter), or Earth
as seen from any place off Earth. The _Explanatory Supplement_
didn't have to worry about such oddities.

For small, rocky satellites, I used the same formulae as for
asteroids. For the Moon, Earth, and other objects, I derived
a formula based on a diffusely-reflecting sphere. This is a lousy
assumption for the Moon, but was the best I could do at the time.
Through the wonders of on-line resources, I may now be able to do
an on-line literature search (something I couldn't do back when I
derived the diffuse-sphere formula) and find something better.
Even if I can't, adding in brightness due to earthshine shouldn't
be too difficult.

-- Bill