An empirical examination of Bradford's law and the scattering of scientific literature