Spectra have been extracted from either the cleaned, background
subtracted and attitude corrected data cubes ( GINGALAC
database) or the cleaned, background subtracted data cubes (
GINGABGD database), and are integrated over the entire observation
interval. The top- and mid-layers are treated separately within both
databases. Where the background subtracted LAC 2-10 keV count rate
exceeds 
counts sec-1, both layers are simultaneously fit
to simple absorbed power-law and thermal bremsstrahlung models. Plots
of the spectra and best-fit models are available within both databases
(see figure 5) as are the results of the fits with their errors. In
addition, the top-layer spectra are fitted to power-law and thermal
bremsstrahlung models with and without a narrow (intrinsic width much
less than the energy resolution of 
keV) Gaussian emission
line at 
-7 keV. As for the joint fits, plots of the best-fit
models and results of the fits are available in the databases. If the
presence of an emission line is significant at the 95% level
(determined by an F-test), then the best-fit emission line parameters
are also written to the database.
The spectral files within the GINGALAC database can be used to
make more detailed fits to the data than has been performed in the
pipeline. The spectra within the GINGABGD database can be used
in two ways. First, the spectra can be used as a measure of Galactic
emission at low latitudes (
)
where this is
not already taken into account (e.g. when background data is
taken from observations at high Galactic latitudes as is always the
case with the universal background subtraction technique).
Second, to provide spectra of serendipitous sources in the Ginga background fields of view. The format of the spectral files
follows the OGIP FITS conventions (Arnaud, George & Tennant 1995) and
these files can be used within the XSPEC spectral fitting
package, which is part of the XANADU suite of software.