The E2 cross section calculated by Langanke and Shoppa for the RIKEN experiment on the Coulomb dissociation of B uses F2 nuclear matrix element from one specific model. Other nuclear models predict a considerably smaller E2 cross section (by approximately a factor of 3 to 4), and Langanke and Shoppa appear to assume the most optimistic scenario, predicting a large E2 cross section. We also note that Barker has already criticized the nuclear model used by Langanke and Shoppa. A model independent chi-square analysis of the RIKEN data suggest the best fit for the current RIKEN data is obtained with F1 amplitudes only. The upper limit (90% confidence) on the E2 component derived from our chi-square analysis is considerably smaller than that used by Langanke and Shoppa. The model dependent analysis of Langanke and Shoppa should not be considered as a correction to the RIKEN result, as claimed, and their quoted Si7 (0) is not substantiated. PACS number(s): 25.20.x, 25.70.z, 27.20. +n In a recent publication Langanke and Shoppa (LS) [1,2] calculated the E2 cross section for the RIKEN experiment on the Coulomb dissociation of B [3].The measured cross section of the Be(p, 7) B reaction includes contributions from s and d waves (Ssi), p waves (SM, and Ss2), and f waves (Ssz), where the p-wave cross section is dominated by a resonance at E, =632 keV. All these amplitudes contribute to the measured Coulomb dissociation of B, with the El component being dominant, and the (small) E2 component being enhanced due to the large virtual photon fiux (especially at large angles, 8~4'). In this Comment on the work of LS we concentrate on the E2 cross section of the Coulomb dissociation of B and ignore the M1 cross section, even though the M1 appears to contribute to the 600 keV angular distribution data of Motobayashi et al. [3] at a level comparable to that of the E2 ((10%), which is however smaller than the quoted accuracy of the RIKEN experiment (15 -20 %). Recently one of us (M.CJ.) notified LS of a number of mistakes in their original paper [1] which led to the publication of an Erratum [2] with a correction of Fig. 2 of LS [1]. We first note that LS state [1]: "We find that the ratio [of the E2/E1 Coulomb dissociation cross section] is robust against this [energy and angular] averaging, " but in their Erratum [2]they correct for the fact that they neglected to average over the energy resolution of the RIKEN experiment. A more severe problem in the LS paper is the fact that they ignored the angular resolution of the RIKEN experiment. Since the angular distribution of the E2 Coulomb dissociation cross section is different than that of the E1 cross section, the predicted E2/E1, after averaging over the angular resolution of the RIKEN experiment, is different than that predicted by LS. In fact, the acceptance of the RIKEN detector is such that the so-called efficiency (i.e., relative number of particle detected convoluted with angular resolution) is not the same for E1 and E2 cross sections. This invalidates the basic assumption of ...