Comparison of the high resolution X-ray ROSAT HRI image of A2218 with the optical HST image shows several interesting correlations. The X-ray emission within a 1 radius core is resolved into several components; the central dominant galaxy does not coincide with either of them or the emission centroid. The major X-ray peak is an elongated feature which coincides with optical arcs at r=20 from the cD. We speculate that this may be lensed X-ray emission, for example, of the same object lensed in the optical. Alternatively, this feature may be a merger shock, or a gas trail of an infalling subgroup. Two other X-ray enhancements are close to the two major mass concentrations known from the lensing analysis. Both lensing and a merger are likely. Previous X-ray derivations of the A2218 mass used the data with angular resolution that blurred the features mentioned above into a broad constant core. As the HRI data show, such a core does not exist. Because of this, the hydrostatic estimate of the projected mass within the lensing radius can in principle be increased by a factor of 1.4 (and the mass within a sphere by a factor of 2.6) compared to previous analyses. However, for a merging cluster, the hydrostatic analysis is generally inapplicable. Most other lensing clusters are more distant than A2218 and obtaining adequate X-ray data for them is even more difficult. Together with the likely overestimation of mass by the lensing analysis (as in the simulations), oversimplification of the gas model resulting from inadequate resolution may account for the lensing/X-ray mass discrepancy as suggested for A2218. (ABRIDGED)