Subject: What is the nature of an atomic collision?

Imagine.
There are two playground-sized elastic balls.  The two identical 
elastic balls approach one another at the equal speed.  Upon impact, 
kinetic energy becomes potential energy until kinetic energy equals zero.  
As the elastic balls recoil, stored potential energy becomes kinetic 
energy and the balls move away from each other at their original speed of 
convergence.  Assume no energy is lost.
I have chosen this elastic quality because it is what I observe in 
those desktop thingamajigs with five suspended steel balls.  When 
molecules hit each other, do they abide by this same kinetic quality of 
elasticity (ie: are such descriptions correct for collisions involving 
atoms)

What about collisions involving electrons or quarks.

~~~~~
 
I looked at a previous answer,  http://www.madsci.org/posts/archives/
942939482.Ph.r.html,  unfortunately, it only left me more confused on my 
original
question.  There is talk of both elastic and inelastic collisions.  I
understand the theoretical part, but I am wondering which one actually
occurs.  If both, then why sometimes one and sometimes another.  And perhaps
most interestingly, what fundamental properties lead to elasticity or
inelasticity.  I am interested in only a single collision and exactly what
happens.  Here is a snippit from the article cited to me:

        In a perfectly elastic collision kinetic energy is conserved, so the
change in total kinetic energy is zero in any reference frame. This is not,
however, a very satisfying example, so for fun, let's go to the other
extreme, and assume a perfectly inelastic collision

But which one is it?  Elastic or inelastic? or both, and why!  If that still
sounds the same, then I will gladly withdraw the question.

Re: What is the nature of an atomic collision?