[en] The direct photon measurement can provide information about the thermal evolution of the hot dense nuclear matter produced in ultra- relativistic heavy-ion collisions. However, the measurement of the transverse momentum distribution of the thermal photons is essential to obtain detailed information, in particular, about the short-lived initial hot phase. To date there has been no confirmed observation of direct thermal photon radiation in the transverse momentum region below about 3 GeV/c. On the other hand, the upper limit on the direct photon yield in S + Au reactions at large transverse momenta determined by the WA80 experiment has been shown to imply that the initial temperature of the system must be relatively low (< 250 MeV). Such low temperatures imply that the system has access to a large number of degrees of freedom, consistent with QGP formation or with a hadronic system consisting of the full spectrum of known resonances up to large mass, but inconsistent with a hadronic gas of only the lowest lying resonances. It is an interesting question how a hadronic gas consisting of the full spectrum of resonances could come into chemical equilibrium already during the initial phase of the interaction. Preliminary results from WA98 for Pb + Pb reactions indicate that there is no significant integrated direct photon excess in central collisions compared to peripheral collisions. This conclusion is based on the variation of the ratio of the electromagnetic to total transverse energy with centrality, which is consistent with no increase. The analysis of the photon data obtained with the WA98 leadglass detector is underway. Comparison of the preliminary inclusive photon spectrum to recent theoretical predictions for the thermal direct photon signal indicate that it may be observable for the case of a hadron gas, but is likely below the expected level of sensitivity for the case of QGP formation