We present a theorical investigation of the optical transmission of a dielectric grating carved in a tungsten layer. For appropriate wavelengths tungsten shows indeed a dielectric behaviour. Our numerical simulations leads to theoretical results similar to those found with metallic systems studied in earlier works. The interpretation of our results rests on the idea that the transmission is correlated with the resonant response of eigenmodes coupled to evanescent diffraction orders.For a few years, properties and technological applications of one-or two-dimensionnal metallic gratings have received a growing interest. In 1998, Ebbesen et al [1] reported on optical transmission experiments performed on periodic arrays of subwavelength cylindrical holes drilled in a thin metallic layer deposited on glass. These experiments renewed the motivation for investigating metallic gratings. Two attractive characteristics of their results are often cited : the transmission, which is a lot higher than the addition of individual holes contributions, and the peculiar wavelength dependance of the transmission. Further work [1][2][3][4][5][6][7] has suggested that these features arise from the presence of the metallic layer, and call for the presence of surface plasmons in order to explain these transmission characteristics. In particular, they have identified the convex high transmission regions, i.e., the regions between the minima, as regions dominated by the plasmon response. However, many questions remain and need to be answered in order to clarify the mechanisms involved in these experiments.In a recent paper [8], extensive simulations have been performed in order to understand the optical properties of a chromium layer similar to those developed in experiments. Recalling of Wood's anomalies [9], it is shown that the transmission and reflection are better described as Fano's profiles correlated with resonant response of the eigenmodes coupled to nonhomogeneous diffraction orders. Indeed, as explained by V.U. Fano [10], A. Hessel and A.A. Oliner [11] for one dimensional gratings, Wood's anomalies are related to eigenmodes grating excitation. To be accurate, they have shown that Wood's anomalies [11] may arise in two ways. The first case occurs at Rayleigh's wavelengths, when a diffracted order becomes grazing to the grating plane [12]. The diffracted beam intensity then increases just before the diffracted order vanishes. The other case is related to a resonance effect. Such resonances come from coupling between the incident light and the eigenmodes of the grating. Both types of anomalies may occur separately and independently, or may almost coincide.In our previous paper [8], as we used metal in our device, it seemed natural to assume that these resonances are surface plasmons. Nevertheless, it is important to note that our analysis did not make any hypothesis on the origin of the eigenmodes. This implies that it could be possible to obtain transmission curves similar to those found for metals, by substituting guided modes...