Five major findings resulting from this project are listed below.

(1) Our first set of observations on Mrk 110 (1999-2000 season) has  
revealed that the H beta line emission lags the optical continuum emission
by 23 days, while the He II 4686 line lags by only 5 days. Given the lags
and widths of the line profile variations, a black hole mass of 18 million
solar masses is measured. We have also discovered the cause of a shoulder on
the red wing of the [O III] 5007 line: contamination by  He I 5015. The 
He I line shows up remarkably well in the rms spectrum, showing
conclusively that this shoulder is not [O III] emission from the BLR as had
been previously believed. We have also determined that the Fe II emission
does not vary appreciably in this AGN. 

(2) We observed Ark 564 simultaneously with the International AGN Watch
campaign, including RXTE and HST observations. From our "engineering
quality" data on Ark 564, we have learned that despite the very rapid and
very large amplitude variations in the X-ray band (factor of 16), the
optical broad emission lines vary almost imperceptibly. By painstakingly
careful modelling, we have detected H beta / [O III] 5007 variations of
only 3.2% (rms). This tiny variation of only 3% in H beta was a complete
surprise, and illustrate just how peculiar the narrow-line Seyfert 1 AGN
are. ``Normal'' Seyfert 1's typically shown variations an order of magnitude
larger. This work has led to a Master's thesis by Luigi Gallo and several
conference presentations.

(3) We have devised a new method for interpolating over gaps in AGN light
curves. CCF lag determinations are notoriously difficult when dealing with
unequally sampled data, and this software called "flicken" helps to remove
that problem. Flicken uses Monte Carlo 1/f**2 random walks to interpolate
gaps in the light curves. The development of the code is still in progress,
but it will eventually released to the public. Flicken was used extensively
in the Master's thesis of Danielle Martino, and is part of Jessica Castora's
Master's thesis research. An early description of the program was included
in a conference proceeding.

(4) A careful re-analysis of the NGC 5548 optical light curves as published
by Wanders and Peterson (1996 ApJ 466, 17) has revealed several interesting
features: the asymmetry in the line varies on a ~ 1-3 year timescale,
suggesting the variations are dynamical in nature. They are certainly not
echo effects or pressure/density pertubations in the accretion flow. We
confirmed the lags found by Wanders and Peterson, but in addition we found
that the H beta red wing versus blue wing light curves show a lag of ~2-4
days, suggestive of inflow of material in the black hole. This relative lag
between the wings of the line is robust because this is a differential  
measurement, not a measurement with respect to the continuum variations. 
Work on a paper is in progress, led by graduate student Martino who earned
her MS in Summer 2004. Her Master's thesis is based on, and contains the
results of, this investigation.

(5) Although not a planned part of our project, we have learned a tremendous
amount about how to cope with the data limitations and calibration issues
present in the new Hobby-Eberly Telescope observations. While the data do
suffer from some calibration problems, they do demonstrate that the HET and
Low Resolution Spectrograph (LRS) are quite stable and can provide the very
high quality data necessary for 2-d echo mapping: During the course of the
several month-long observations of Mrk 817, the LRS was  physically removed
from the telescope for an upgrade. Upon its return, the LRS gave identical
line profiles as had been seen weeks before (to within <3%, the precision
of our test), demonstrating the the HET can be used for 2-d echo mapping.