We propose to use SWS on board the ISO satellite to search for hidden
broad line regions (BLR) in a sample of Seyfert 2 galaxies from
intermediate resolution observations of the emission lines Brackett
alpha (4.05 micron) and Pfund alpha (7.46 micron). The observations
will allow us the detection of any broad component obscured up to 180
magnitudes in the visible (Goodrich et al. 1994). This represents a
significant improvement, w.r.t. previous works, in the penetration
of the molecular/dust torus around the BLR and continuum source
predicted by the unification models and therefore the proposed
observations will constitute one of the most powerful tests of such
models. It is expected to extend considerably the sample of
"hidden" Seyfert 1 known to date. Additionally, we will determine the
reddening of the NLR and of the BLR (if detected) and, from the
comparison of the reddening of the detected BLR with that inferred
from the column densities of intervening material obtained from
soft X-ray observations, we will be able to deduce several important
properties of the gas such as the mean dust-to-gas ratio along the
line of sight.  Also very important information will be extracted
from the analysis of the widths of the profiles. We will search for
the presence of systematic increases in the width at longer
wavelengths in both the NLR and the BLR by internal comparison of
our data and with the optical lines. This will provide direct
evidence of velocity stratification in both line-emitting regions
as a function of optical depth, which can be compared with current
models of cloud velocity distribution present in the literature.