Substantial progress on some of the most fundamental issues about
 activity in galaxies requires the analysis of complete and unbiased
 samples, revealing the physical quantities underlying the observed
 emission. We propose the first detailed 2.5-200um spectrophotometry
 of the active galactic nuclei (AGNs) in the Local Universe.
 This will be done by observing the 12um Active Galaxy Sample, which
 has been selected to be bolometric flux limited. In fact the 12um
 flux carries a constant fraction of the bolometric flux for various
 classes of active galaxies, from blue quasars to dusty Seyfert 2s.
 The analysis of a representative, unbiased sample to a well-defined
 bolometric-flux limit, combined with complete multi-frequency data
 (from X-rays to radio) are both required to address the fundamental
 issues of galactic and nuclear activity and of the underlying physical
 connections between the various classes of AGNs towards a unified
 scheme. For instance, we will be able to address the question if
 Seyfert 2s are really dusty Seyfert 1s whose broad line regions are
 completely obscured. Principal component analysis and other statistical
 methods will be used to find the "dominant correlations" underlying the
 observed physical properties of AGNs and form the basis of a new
 classification system - more physically based than the traditional
 emission-line scheme.  Specifically, we propose to observe the 12um
 AGN Sample with 2.5-12um PHT-S spectrophotometry,60-200um PHT-C
 photometry and CAM mid-IR imaging. PHT-S will measure very accurately
 the emission-line spectrum, including high-ionization forbidden lines,
 H and He recombination lines, molecular transitions, dust features, and
 the continuum energy distribution. The far-IR photometry will uniquely
 detect and measure the occurrence and shape of the far-IR turnover
 providing the best available test of nonthermal models. The broad-band
 mid-IR imaging will separate nuclear and galaxy luminosities and
 measure the distribution of hot dust.