We have endeavored to show the utility of this software for the analysis of emission line spectra in some important astrophysical contexts. We believe these tasks can greatly facilitate deriving an accurate, first-order approximation to a correct physical picture of many kinds of nebulae. Indeed, given the paucity of the extant data for most objects (see, e.g., the catalog of Kaler, Browning, and Shaw 1995) that picture is likely to be the best one can hope to do for the vast majority of nebulae. The tasks in this package can make use of whatever diagnostic information is available, and they are designed to operate on data from a large number of objects in sequence, which makes the nebular package ideal for archival research programs. Finally, this package is supported and is widely available to the astronomical community, and draws upon some of the most recently published atomic data.
Useful as this software is, however, it is not difficult to imagine circumstances in which the traditional diagnostics would be difficult to interpret. In Seyfert galaxies, where the source of ionization is extremely strong, where the range of gas temperatures and densities can span several orders of magnitude, and where multitudes of partially photo-ionized clouds contribute to the observed emission line spectrum, the inferences from the nebular tasks are not likely to be very illuminating, and the derived ionic abundances will be plain wrong. And in complex radiation fields the various, exotic radiative-transfer mechanisms and subtle optical depth effects (c.f. § 2.2 ) will not be accounted for with this simple 5-level atom approximation. For these and other applications, one must turn to appropriate photoionization models. But even in such cases, we believe that an analysis with the nebular utilities is a useful starting point. Indeed, the nebular package is most useful for bridging the gap between hand calculations for a few diagnostics, and the complex and detailed information one must provide as input to constrain a photoionization model.
Finally, the program does not treat the effects of temperature fluctuations
on the derived ionic abundances. Peimbert (1967) developed a formalism for
the determination of the RMS fluctuation in T
from various emission line
and continuum diagnostics. Ionic abundances are then calculated based on
a mean temperature T
and RMS fluctuation parameter 
. While we have
(at present) no such capability to do this in nebular, the program
can be used to derive temperatures that are useful in determining the extent
of such fluctuations in a given nebula. Moreover, since the physical cause
of such fluctuations is currently unclear, we submit that our multi-zone
approach reduces the effects of such T
fluctuations (also fluctuations in N
) by deriving mean temperatures for multiple ionization zones-when
permitted by the spectrophotometric data in hand.