We show that the dinotor is sensitive to discrete energy level corrections to the continuum approximation for the density of states. These corrections do not destroy the metastability of the dinotor, but may produce a 'magic number' effect. New metastable configurations of the dinotor may also exist, possibly creating further difficulty identifying the dinotor as the cause of the anomalon effect. PACS: 12.35.Ht; 25.70.Np; 12.35.Eq Castillejo et al. [1] have proposed that a large toroidal metastable object, which they call the dinotor, composed of quark matter may be formed in heavy ion collisions. The production mechanism [1], although still far from clear, is conceived to be a more or less coherent process whereby a nuclear system is pushed over a barrier towards a new configuration where the individual nucleons have merged to form a single toroidal object composed of quark matter. This ring will then expand, emitting soft photons and pions to loose excess energy, ending up as a torus of very large circumstance with roughly the same quark density as normal nuclear matter. The torus will have a larger geometrical cross section than ordinary nuclei, and simple arguments [-1] suggest that the geometrical cross section may be twice as large as the more conventional spherical nucleus. Spontaneous decay into nuclear fragments or unbound nucleons seems unlikely; crude semielassieal arguments [1] imply that there could be an energy barrier of up to 300 MeV to be penetrated before a single nucleon can be formed. The dinotor is robust and will tend to survive subsequent collisions. The non-localised wavefunctions suggest that the energy transfered to the dinotor during the collision will be distributed evenly among the quarks making it difficult to break off large fragments. One of the more interesting features of the dinotor maybe the presence of non-trivial spin-colour correlations in the wavefunction and hence, despite being an overall colour singlet, it may provide a colour non-singlet target for a fast hadron. A number of experimental signatures have been suggested [1], and these include long lifetimes, large interaction cross sections and low charge to mass ratios for N= Z dinotors (perhaps 20% lower than conventional nuclei) due to the extra mass per baryon of the dinotor. The charge to mass ratio of the dinotor will increase with increasing Z along the valley of stability of conventional nuclei, due to the reduction of the Coulomb repulsion caused by the more spread out geometry. If the dinotor is formed in a high angular momentum state electromagnetic cascades may easily be produced (and detected) as it relaxes, since the dinotor possesses a large electric quadrupole moment.The prediction is made within the framework of the MIT bag model [2], which in the static cavity approximation [3], confines quarks to a cavity having an additional volume energy density B, where B is the 'bag constant', above the 'normal' external vacuum energy density. Minimizing the total energy (kinetic + potential + volume energies) d...