Re: Why does magnesium have such a low melting point?

This is a good question and one that the answer is not easily available.
However, I did eventually find one text that offered an explanation. To 
set this up, I should point out that metals, such as magnesium, are held
together in a lattice by metallic bonds. One way of modelling this is to
assume that the nucleus of the element is a "cation" and that the valence
electron is an "anion". The result is that metals are then subjected to
the sort of analysis that you would do for any crystalline salt - that is,
the electrostatic interaction between the metal and the electron is the
binding energy.

A Group II element, being divalent, would then have two electrons as 
counter ions. A mole of cations would be expected to have two moles of
anions associated with it. As James Bowser explains in "Inorganic 
Chemistry" (Brooks/Cole Publishing Company, 1993):

"... Lithium has a comparatively low conductivity because of its high 
ionization energy relative to the others; that is, it "cheats" (on average, 
delocalizes less than 1 mole of electrons per mole of atoms), so its 
conductivity is lower than expected.
     The same rationale can be applied to the Group 2 elements, where the 
conductivities of the first two elements, beryllium and magnesium, are low 
in comparison to the others. This suggests that their higher ionization 
energies limit delocalization to fewer than 2 moles of electrons per mole 
of atoms. This also explains the comparatively low melting temperatures of 
Be and Mg. If fewer than two electrons are delocalized, then the resulting 
lattice energy must be lowered accordingly."

This is the only explanation that I found and it does make sense. Most 
texts seem to note the anomoly but do not explain it. I hope this answers
your question - although it does beg a follow-up: Why are the ionization
energies anomolously high?