This paper outlines some of the outstanding problems associated with t
he experimental determination of the growth modes of thin metal films
on dissimilar metal substrates, and suggests approaches for overcoming
them. Specifically, it discusses ways to differentiate between Strans
ki-Krastanov and alloy growth, the quantitative fitting of ''Auger sig
nal-time'' plots, and the determination of the stoichiometry of surfac
e alloys. A comparison of the observed growth modes with those predict
ed by the Macroscopic Atom Method shows that the results predicted by
this method must be treated with extreme caution.