2 Introduction We wish to extend the utilization of the two arm spectrometer of experiment 302, which will be set up in the downstream end of Proton West, by performing a high sensitivity charm search. This will require the addition to the original design of a third spectrometer arm close to the target region and downstream ~ filters in each of the two arms. The basic limitation in searching for charmed objects produced in hadronic channels has been the high background level from conventional

more »

... rong interaction sources. We propose to enhance the signal to back ground ratio by capitalizing on the existence of charm levels (D* ,D; ~ ,A ) with mass separation only slightly larger than a pion mass. c c As an example we *± discuss in detail the search for D • The charmed vector mesons D*± have been detected at SPEAR l ) and are seen to decay o ± predominantly to D TT final states with a measured Q value of 5.7 ± 0.5 MeV. In the limit of zero Q value the pion has the same velocity as the DO in the laboratory. A symmetric double arm spectrometer, as that of E-302, selects, through their two body final states, DO,s within a restricted momentum interval. Thus the pion accompanying the DO,s * ~ from D decay are well collimated and have a central momentum of ---x PD' ~ In the proposed experiment we will require a pion in the appropriate kinematic region as a constraint in the selection of the events, both at the trigger and analysis level. We have performed a similar experiment at BNL using 10.5 GeV/c pions. We measured crB = 7 + 17 nb for the reaction *-+ 11 +p-D +A (1) In the BNL experiment the requirement of the extra pion in the trigger reduced the trigger rate by a factor of 30. Reconstruction reduced the background another factor of 6 for a total reduction in background of 180. Superior momentum analysis of the soft pion, proposed in this experiment, should permit additional off-line discrimination against background of another factor of 5, leading to an overall background suppression of 1000. To make the extrapolation from BNL to Fermilab energies we have assumed that the production of background follows the usual scaling laws. Countering the substantial decrease in background is a suppression in signal which arises from losses in the competing D* decay channels of y + Dand TI 0 + D • With the Q-value at 5.7 MeV, this loss should not exceed a factor of 2.2) With the background reduced by almost -1000 and the signal reduced by ~ 2, the net gain in the ratio of signal to background is expected to be about 500. Of course, here we measure cr