The discrepancy between mass estimates from X-ray and strong gravitational lensing analyses has recently been pointed out for several clusters, in particular A1689 and A2218. We analyze the internal structures of these clusters by applying a recent development of the method of wavelet analysis, which uses the complete information obtained from optical data, i.e., galaxy positions and redshifts. We find that both clusters show the presence of structures superimposed along the line of sight with different mean redshifts and smaller velocity dispersions than the system as a whole, suggesting that the clusters could be cases of the ongoing merging of clumps. We compute the masses of the two clusters by adding the optical virial masses of the single clumps, which are supposed to be virialized. By rescaling our masses to the same radii, we compare our results with mass estimates derived from X-ray and gravitational lensing analyses. In the case of A2218 we find an acceptable agreement with X-ray and gravitational lensing masses. For A1689, we find that our mass estimates are smaller than those from X-ray and gravitational lensing at both small and large radii. But the observed X-ray temperature in this cluster, and thus the derived X-ray mass, could be enhanced by collision-heated gas during cluster merging. The high central cluster mass obtained from lensing analysis could be explained only by adding the masses of two background galaxy groups possibly aligned with the cluster. At variance with earlier claims, there is no evidence that X-ray mass estimates are underestimated.
