The last two weeks I spent developing software to get GEn for our fourth
kinematic.  I have had some good success in getting results, though it has
become obvious there is still a great deal of work to try and understand the
data that must be done before we are done.

The software I've developed calculates the asymmetries for a set of runs given
a set of quasielastic cuts.  It proved to be useful to also keep track of the
positive particles in these counts and calculate a raw asymmetry for those as
well.

The target information was pulled from various sources.  The HWP status was
taken from the EPICS database that my code directly interfaces with.  Now that
this is available, any information stored in the databases can be directly
applied.  The target polarization (which will eventually go in the SQL
database) was taken from the HALOG and put into a lookup table.

The error bars are statistical only.  Systematics were assumed to be 0 at this
point.

Page 1
I started looking at the hydrogen runs to understand what we get out of a pure
proton sample.  There appears that there may be a small asymmetry that needs to
be taken into account for charged particles.

Page 2
More interesting is that there was a non-zero number of neutral particles.  
These were tracks in the neutron arm that did not have an associated hit in 
the veto.  For the tight elastic cuts applied, there was about 5% of protons 
going into this bin.  For more loose quasielastic style cuts, this was closer 
to 10%.  This has been confirmed by Rob in his analysis.

Page 3
Invariant mass spectrum for the H2.  This appears to be normal for both charged
and uncharged particles.  BigBite appears to be seeing elastics for these
cuts as we expect.

Page 4 and 5
Missing parallel and perpendicular momentum.  These show different
distributions for the charged and uncharged.  Essentially, they are peaking at
different places suggesting that the neutron arm is seeing different things in
space and possibly time for the neutral events.

Page 6 and 7
These are the number of clusters found in the neutron arm when we have charged
and uncharged events.  There is a dramatic number of drop of single track
events for neutral particles.  What I believe is happening is that there is a
proton elastic event that interacts in the iron at some place, fragments into
multiple clusters in separate directions.  When reconstructing these clusters
they are not associated with a veto hit due to their dispersion away from the
original track direction.  This needs to be understood.  Contamination of
protons into our neutron sample is extremely detrimental to our error bars.

Page 8
Physical asymmetry for He3 data.  Vertex dependent polarization was 
parameterized from measurements by Ameya.  There was a correction made on the 
total number of neutral particles, subtracting out a number of positive 
particles such that we get a ratio of 0.08 neutrons to protons.  This is from:

xs_n = 2e-34 cm^-2, xs_p = 5e-34 cm^2

We have 2 protons to neutron, so we expect

.2N_n = N_p 

With 100% proton detection efficiency and 40% neutron detection efficiency we
expect

0.08N_nc = N_pc (N_*c, number counted)

Page 9
We see that we have 0.22N_nc to N_pc.  This corresponds to a proton leakage
rate of just over 10%, which was observed in H2.

Page 10
Same analysis applied ot the nitrogen sample.  The first two runs were low on
statistics.  There appears to be a non-zero asymmetry.  This needs to be
understood on how to apply, since it may be significant.  Nitrogen corrections
were not applied to the He3 analysis.

For those curious, a value of: 

*EXTREMELY PRELIMINARY*  GEn(Q^2 = 1.69GeV^2) = 0.037 *EXTREMELY PRELIMINARY*

was obtained from the He3 runs.  These used a cuts of:

P_miss_par < .25 GeV
W = 0.94 +/- .2 GeV
|coplane| < .05
|inplane diff| < 0.08
p_e > 1GeV

proton_leakage = 0.1

I have no error bars for that since they are a bit more difficult to calculate
and I need errors on directional measurements of our polariziation and on
BigBite's reconstruction as well as the necessity of using a numerical
method to get GEn from Aphys (see Gregg's finite acceptance report).

There is a considerable amount of work that needs to be done:

1) Need to understand the background under the quasielastics.  Gregg's Monte
Carlo will be very useful for this.

2) Need to eliminate our proton leakage into the neutral spectrum.  I believe
this should start by looking at fragmentation of protons in the neutron arm in
the H2 data.

3) Nitrogen corrections need to be applied to the physical asymmetry.  This
shouldn't be difficult to calculate once we know the parameters of nitrogen
in the target cell and the referece cell.

4) SQL database needs complete set of variables, namely from the target 
polarization, etc).  Also, we need a conversion of beam current measured to 
actual beam current from what is available from the EPICS data in the SQL
database.  Monitoring the QE rates is probably something we want to do.

5) We need to agree on a set of cuts to apply to get what we call
quasielastics.

We should discuss any other analysis ideas to get GEn out.  The way I see it is
there are these fundamental parts of the analysis that we need to get out of
the way and understand.  Once we have these in place, the rest is just
improving error bars, such as BigBite optics, neutron arm timing, etc.

Seamus