I cut some magnets into buckeyball-faced polyhedra and bond the pieces together so that all joints are completely flush. Assume the joints are perfectly planar-flush, and bonded with a monolayer of something with a small molecular size. Or better yet, the thing is assembled in a vacuum and the unoxidized faces bond together directly The pieces all have with the same polarity outward. What do I have? Edit:The material is samarium-cobalt with very high coercivity. Just to make isolated local domains a wee bit harder to posit as the outcome.