See subject.
Metaphors have them always as spherical or spherical with “[electron-statistical location]-cloud cover.” Is this mandatory?
When they’re zipping along then do they elongate in other reference frames, then, as we all do? Physicists have to factor in fantastic mass-change phenomena.
How do spheres “elongate?”
Certainly inside atoms, electrons aren’t necessarily spherical. They’d be orbital-shaped.
For a free electron, its minimum extent is related to its momentum uncertainty, by Heisenberg. In principle, that could be different in three orthogonal directions, so instead of a simple spherical distribution, it would be ellipsoidal.
If they aren’t at the minimum uncertainty, they could be anything. Shoot an electron at an ampersand-shaped slit in a screen, and it could be bell-shaped in it’s direction of motion, and fuzzy-ampersand-shaped in the transverse directions.
None of these shapes have sharp boundaries, of course.
Shape doesn’t really have quite the same meaning at this scale, but in no real sense is either the shape of the nucleus or of the electron orbitals necessarily spherical.
The Schrodinger equation isn’t Lorentz invariant, but by substituting it for a relativistic wave equation the shape of an atom (however you choose to define it) will be subject to Lorentz-Fitzgerald contraction just like any other shape. Mass-change doesn’t enter in to it as mass is always defined as the rest mass.
A spherical atom will contract in to a prolate spheroid when viewed by someone in a reference frame where it is travelling a significant fraction of c.
Given that fundamental particles (electrons and quarks) are, to all reliable indications, point particles (or at least, so close to being point particles that we can’t tell the difference), I think it’s fair to describe their “shape”, in so far as they have one, as spherical.
Wouldn’t a point particle technically be dimensionless? Or at least be so small as to be regarded as dimensionless for all practical purposes?