Email from Roger:

From carlini@jlab.org Thu Sep 23 11:28:30 2004
Date: Mon, 20 Sep 2004 00:46:17 -0400
From: Roger Carlini 
To: finn@physics.wm.edu
Cc: 'Jim Birchall' , 'Juliette Mammei' ,
     'Klaus Grimm' , 'Mark Pitt' ,
     'Neven Simicevic' , 'Greg Smith' ,
     'Allena Opper' , 'Tony Forest' ,
     'Norman Morgan' , 'Mike Finn' ,
     'David Armstrong' ,
     'Yongguang Liang' ,
     'Shelley Page' , 'Dave Mack' ,
     'Richard Jones' 
Subject: Re: Inelastic rates on the Cerenkov bar

Hi Mike:

I think your not understanding the point I was trying to make. I am not talking about anything related to overlapping 
the sector images. I am talking about plugging the upper corners (large theta and phi) of say the 9-8-22 collimator 
concept to possibly cut down on events accepted by the reference "bar" detector which have both large theta's and phi's. 
The collimator does not have to have a constant theta acceptance as a function of phi. It's just a thick plate of metal 
into which we can cut any shape hole desired. Doing this acceptance "tweaking" at large +-phi's might allow for a bit 
narrower (not shorter) bar being used with the basic 9-8-22 design and thus reduce the inelastics a bit - as its the 
fringe fields that create the moustache. It might be worth a try in the simulation.

As for overlapping of the images. This was dropped because it would require highly sculptured detectors to suppress 
inelastics - and I was not suggesting this concept.

Best Regards

Roger