And if iron(III) oxide is near the melting point, it's unlikely to be magnetic and may not even be paramagnetic. This would be an interesting lab experiment: heat a sample to 1500 C exactly and test for paramagnetic properties.