The superconductor is supposed to consist of alternating layers of two
kinds: (1) layers with an attractive electron interaction and an effe
ctive mass of usual magnitude; (2) layers without interaction and with
a large effective mass. The overlap between the layers is assumed to
be small, its energy t being much less than DELTA. It is shown that su
ch a model explains the most peculiar property found in experiments on
electronic Raman light scattering: different threshold values for the
Raman satellite measured at two different polarizations of the incide
nt and scattered light. The tunnelling conductance G(V) = dJ/dV is ana
lyzed for the same model. In order to fit the qualitative features of
experimental data, it is assumed that the tunneling probability to the
normal layers is much smaller than that to the superconducting layers
. The conductance is calculated for the case t much less than DELTA. A
brief analysis is given for the case t congruent-to DELTA, which prov
es that such an assumption definitely contradicts the experimental dat
a for BSCCO. The possible nature of the electronic states in the norma
l layers is discussed.