Log entry time 17:30:00 on September 24,1998

Entry number 211

A few weeks ago I asked Rob to determine the D+/H+ ratios with a cut on -t.

The idea was to enhance the L- response without losing much statistics. Here

are his results for parallel kinematics at Q2=1.6:



-t cut    		      epsilon=.27    	 epsilon=.62   	  comments

---------     	  -----------     -----------     --------



none	 	     .901+-.014      .911+-.006       WRONG YTARGET CUT!!



-t < 0.25    	   .828+-.014       .873+-.006      



-t >=0.25   	    1.442+-.067     1.243+-.027



The "no cut" result is wrong since the Ztgt cut of +-1.7cm was not used. The

official value is in the August 17th Fpi analysis report. This whole table will have

 to be redone with the right Ztgt cut. 



 Due to the correlation between -t and MM, it is not surprising that the

ratio in the low -t region is lower and that in the high -t region is higher.

Potentially more interesting is the fact that the epsilon dependence in the

low -t region is much more significant than without the cut. However, the 

acceptance corrections could wipe this out. 



I don't want anyone to get side-tracked on this, but at this setting the integrated

(epsilon*L)/T strength may  only be 1-2 when you put the experimental epsilons in. 

A -t cut of this sort should yield an equivalent or better result in terms of statistical

signficance.

  <\pre>