First email from Jim:

Date: Wed, 01 Sep 2004 13:31:24 -0500
From: Jim Birchall 
To: Juliette Mammei , Roger Carlini ,
     Klaus Grimm , Mark Pitt ,
     Neven Simicevic , Greg Smith ,
     Dave Mack , Norman Morgan ,
     Mike Finn , David Armstrong ,
     Yongguang Liang ,
     Allena Opper ,
     Shelley Page , Tony Forest 
Subject: Variations on the primary "Boston" collimator

Here are some results with changes made to the primary collimator and 
the cleanup collimator blanked out.

Jim

Jim's attachment:
CollimatorStudy.pdf

Second email from Jim:

Date: Wed, 01 Sep 2004 18:55:08 -0500
From: Jim Birchall 
To: Dave Mack 
Cc: Juliette Mammei , Roger Carlini ,
     Klaus Grimm , Mark Pitt ,
     Neven Simicevic , Greg Smith ,
     Norman Morgan , Mike Finn ,
     David Armstrong , Yongguang Liang ,
     Allena Opper ,
     Shelley Page , Tony Forest 
Subject: Re: Variations on the primary "Boston" collimator

Dave,

I was already working on an extra column (new table attached) that 
includes an uncertainty due to B(Q^2). The last column gives the 
running time for a combined uncertainty of 3.4% on Qw due to counting 
statistics and an uncertainty in B(Q^2) that contributes a 2% 
uncertainty to Qw at Q^2 = 0.03. The 3.4% and 2% numbers are from the 
TDR. Running time is decreased a little at low Q^2, unchanged at Q^2 = 
0.03, and increased a little at higher Q^2, but there's no big changes.

The numbers in the table are for a beam polarization of 85%. 80% was 
used in the TDR.

Jim

    [ Part 2, Application/PDF  45KB. ]
    [ Unable to print this part. ]


    [ Part 3: "Attached Text" ]



On 1 Sep 2004, at 4:01 pm, Dave Mack wrote:

>
> 	Dear Jim et al,
>
> 	Jim's results remind me of a very old issue: we can't actually
> optimize the experiment unless we have some way of incorporating the 
> Q^2
> dependence of the uncertainty due to the hadronic contribution, 
> B(Q^2). We
> know we have an experiment near Q^2 = 0.03, but without incorporating 
> the
> presumably increasing hadronic uncertainty, I don't know how to choose 
> (on
> Jim's "Summary of Results" page) between a 2060 hour run at Q^2 = 
> .0293,
> and a 1690 hour run at Q^2 = .0317. The latter is of course very 
> tempting.
>
> 	Is anyone close to having the required parameterization?  Given
> the uncertainty in the strange quark contributions, I'm sure that 
> can't be
> rigorously done. But I suspect one could derive a useful approximate
> expression. I'd be happy to tackle it, but I'm reluctant since Mark or
> Mike could probably do something "good enough" before I had time to 
> make
> tea.
>
> 	regards,
>
> 	Dave
>
>


Jim's attachment:
Untitled.pdf