Afaik, science doesn't know this yet. I've seen an interesting theory that noticed that if we take the navier-stokes equation of an ideal flow, insert the loren

Afaik, science doesn't know this yet. I've seen an interesting theory that noticed that if we take the navier-stokes equation of an ideal flow, insert the lorenz factor there, we can relatively easily derive both magnetic (maxwell) and gravity equations. If that guess is correct, the nature of magnetism is a flow of 'water' that assumes curious shapes, forms stable structures (particles) and binds them together. The nature of that 'water', however, will remain a mystery. Problem is such theories are impossible to verify today.