 
Summary: J. Am. Chem. SOC.1986, 108, 32413247 3241
on the accuracy of the available experimental parameters.
Some comments on the y x*f3Pzf3relationship may be
of value. For liquid metals, the surface tension was correlated"
with the plasma frequency, wp, and thereby can be correlated with
surface plasmons, so that
(6)
where Z is an effective valence, m is the electronic mass, and r
is the size of the atom. (Equation 6 may be used also to calculate
the surface tension of binary alloys.12) It was suggested earlier4
that the surface tension was proportional to a surface elasticity,
Le., with P213dependence, the twodimensional analogue to the
threedimensional bulk modulus. Consider the frequency of the
optical branch of a onedimensional lattice with two kindsof atoms
given byI3
wz = 2p(I/m + I/M)
where p is an elastic stiffness coefficient and m and M are the
masses of the atoms. When one of these is taken as an electron
this becomes
2w,2 = wpz = 3Zez/mr3
