Dear Oliver;
        A perennial discussion on the list.  The way i understand it, to be
brief, is that a subspecies is a geographically separated population
assembleage (ie, collection of populations which interbreed at least
occasionally) which shows a good deal of consistant, systematic
difference from the nominate form of the species - for example, compare
Neptunea decemcostata decemcostata with N. dec. lyrata - the former
Arctic boreal, the latter Pacific.  They are the same species, but you
can clearly see they are systematically and constantly different in
form. (interestingly enough, one sometimes sees a decemcostata s.s. that
resembles the lyrata subspecies, but so far as i know, never the other
way around - this may mean the species originated in the Pacific.)
"varieties" and "forms" are sometimes considered the same, but the way i
understand it, a variety is more different from the nominative form, and
is often semi-isolated geographically - halfway to being a subspecies
sometimes!   A "form" is usually found with the nominative form, and can
be named for something as trivial as the number of stripes on the
specimen, or the pattern or even the color! (forma "alba" has been
applied to albino specimens of many species, for example).  In almost
any given population, there will be many "forms", some named, others not.
        For the human species, there are no "subspecies" - even the natives of
Australia, who have been separated from the rest of humanity for a LONG
time, are so similar in genetics and abilites to everyone else that they
could, if one was to use the same terminology, only be assigned the
status of a "variation".  In some cases, as with the differences between
the Slavic peoples and the Germanic ones, the designation "form" would
be more appropriate.  That said, the designation of "Race" is
scientifically-speaking not very useful at all: the amount of variation
within any given ethnic or "racial" group so completely overwhelms and
overshadows the variation between such groups, and the overlap in all
cases is so great, that the small genetic clusters responsible for
trivial things such as skin color don't mean a lot when it comes to
differentiating groups in our species. If only we could get that solid
scientific truth to penetrate the thick skulls of racists and xenophobes
all over the planet - all the peoples of the earth are so very similar
in everything that counts, that any form of discrimination on the basis
of a few minor differences is the height of foolishness and only breeds
hatred and suffering. (for example, an Arabic American is virtually
always **far** more like any other American than he or she is like an
ultra-radical "Islsmic" terrorist!!!)

From the surprisingly dry North,
Ross.