To explain the T-T' mass splitting of the recently discovered T resonances, we interpret them as bound states of a quark and an antiquark which are not members of the fundamental (2) representation of color SU(2). A quark in the 8 (or maybe 5) representation is consistent with the data. We discuss some of its properties and the phenomenological consequences that follow from our suggestion. (Submitted to Phys. Rev. Lett.) * Work supported by the Department of Energy.