Make a copy of ``poissonGe.m'' called ``poissonGeSolid.m'' in which
the expansion coefficients of the Gaussian (obtained by applying cJ()
to the sample values) are, value by value, multiplied by Sf before
being used in any future expressions. This makes copies of the
Gaussian representing the nuclear charge at all of the atomic
locations listed in X. Keeping the initialization S=[48; 48; 48], run
``poissonGeSolid.m''. This, then, gives the nuclear-nuclear
interaction energy for the germanium crystal. I find a value of about
hartree. Finally, make note of the value you
get for S=[64; 64; 64], which should be accurate to about
hartree.