Use R=0.7 cone jets

HT(jets) with eta(jet)<2.0, ET(jet)>15 GeV.

HT(EM) = sum(ET(EM)) eta(EM)<2.5, veto 1.1<eta(EM)<1.5, E(T)>20GeV

DM/M, DM/sqrt(M) where M=min(2*(Mej-Me'j')/(Mej+Me'j'))

ST=HT(jets) + HT(EM)

S=sum (E(jets,EM))

Centrality=ST/S

Aplanarity (jets+EM)

Mee, M18(JH), M20(JH), M22(JH)
where M18(JH)= Sqrt [(Mej-180)**2 + (Me'j'-180)**2]/180
Mej, Me'j' are the ej invariant masses corresponding to minimum	difference 
pairs.
	
SQUAW FIT variables:
Ms, Ps, Chisq, Mejs, Me'j's, DMs/Ms, DMs/sqrt(Ms)

Two signal vs. background optimizations:

1) S^1.15/[B  + (delta  B)^2]^0.35  (this is  a  modification of  the S/sqrt(B +
   (delta B)^2)  formula for a  Poisson  statistics case -  see  Mark Strovink's
   calculations). This is "discovery optimization."

2) Fix B at  0.3 events  (75%  probability of  not  observing any  events in the
   presence of  the  background only)  and optimize  the signal.  This is "limit
   setting   optimization" since if  one is  lucky indeed  (3  chances to 1) and
   observe no events the limits do not depend on the background and its error at
   all!