When, say $\text{N}_2$, is in the wild, it regularly collides with other $\text{N}_2$.
I assume these collisions happen as, having become sufficiently close, the atomic electrons are repelled from the other molecule's electrons, setting up two dipoles (whose dipole moments are in opposite directions, so repel). I conjecture the $|$dipole moment$|$ $\propto r^{-3}$, the situation being somewhat analogous to the tidal effect.
TL;DR:
Regardless, there is some effective volume that the molecules take up such that, if one molecule wanders into another's effective volume, its path is considerably perturbed*. How does one calculate it, especially for diatomic molecules? Is $\text{N}_2$'s effective volume in collisions with $\text{N}_2$ noticeably different from its collisions with $\text{O}_2$?
*I feel I should clarify 'considerably'. I suppose a reasonable definition would be that if two $\text{N}_2$ molecules are travelling in opposite directions along parallel lines $d$ apart, they are considerably perturbed if the angle $\theta$ that they are both thrown off their original line by satisfies $\theta<\frac{\pi}{100}$. Note that if $d\gg$ the bond lengths, then the effective volume is essentially spherical and so $\theta$ will not be a function of the molecules' orientations, only of $d$.
If anyone's interested, I'm interested in this to work out particles' mean free paths in air.
