Comments on atmospheric neutrino oscillation scenarios with large nu_mu <-> nu_e transitions

The evidence for $\nu_\mu$ disappearance in the Super-Kamiokande atmospheric neutrino experiment and the negative searches for $\nu_e$ disappearance in the CHOOZ reactor experiment can be easily reconciled by assuming oscillations with large amplitude in the $\nu_\mu\leftrightarrow\nu_\tau$ channel and small (or null) amplitude in the $\nu_\mu\leftrightarrow\nu_e$ channel. It has been claimed, however, that some oscillation scenarios with large $\nu_\mu\leftrightarrow\nu_e$ mixing can also be constructed in agreement with the present data. We investigate quantitatively two such scenarios: (a) threefold maximal mixing; and (b) attempts to fit all sources of evidence for oscillations (solar, atmospheric, and accelerator) with three neutrinos. By using mainly Super-Kamiokande data, we show that the case (a) is disfavored, and that the case (b) is definitely ruled out.