Models of neutrino masses are discussed capable of explaining in a nat
ural way the maximal mixing between nu(mu) and nu(tau) observed by the
Super-Kamiokande Collaboration. For three generations of leptons two
classes of such models are found implying: a) Delta m(23)(2) much less
than Delta m(12)(2) approximate to Delta m(13)(2) and a small mixing
between nu(e) and the other neutrinos, b) Delta m(12)(2) much less tha
n Delta m(13)(2) approximate to Delta m(23)(2) and a nearly maximal mi
xing for solar neutrino oscillations in vacuum. (C) 1998 Elsevier Scie
nce B.V. All rights reserved.