Re: why is there so many differences in the appearance of the modern human race

Israel,

The exact causality of these differences is not known. Science proceeds by 
the formation of alternate, testable hypotheses, and then the elimination 
of some hypotheses by experiment. Unfortunately, experiments on human 
evolution 1) would take a long time, a lot longer than a 5 year Ph.D. 
thesis, and 2) might not pass ethical review. So I can't give you an 
answer backed by empiricism-- and beware of anyone who claims to be able 
to give you a definitive answer!

What I can do is lay out some general possible explanations. Which is the 
correct explanation for any particular trait, I don't know.

In general, variations between populations can occur for a number of 
reasons:
1) natural selection-- some environmental / ecological factor is present 
in one region but not another, and so selects for different traits in the 
two regions. This is the force that you identify in your question.
2) drift-- don't forget selection only operates because random mutations 
occur. These random mutations will occur even for traits that are not 
selected for / strongly selected for. Thus, 2 separate populations might 
diverge simply by random chance, a phenomenon called "drift".
The converse can also happen. A trait that confers an advantage in natural 
selection does not always arise in a population. Remember, it first has to 
appear by chance.
3) sexual selection-- a slight preference, arising by chance, for a 
particular trait can become a virtuous cycle as those with the trait are 
able to attract more/better mates, and those with a preference for the 
trait are selecting better mates. Any one of the traits you identified 
could simply be selected for by preferences that first arose by chance.

If you're deeply interested in the mathematics of this stuff, look for a 
basic textbook in the field called quantitative genetics. These guys look 
at quantifying just how strong drift can be, how much gene flow there has 
to be between populations to eliminate these kinds of differences, and so 
on.

PS- Evolution is still a testable hypothesis-- experiments are routinely 
run-- but it is best to use fruit flies with multiple generations a year, 
rather than humans!