Possible tests for the presence of an isotensor current

Abstract

The possibility that the electromagnetic current has an isotensor component has never been excluded. To be sure, if an I ~ 2 electromagnetic current existed, it would have rather peculiar properties provided one wants to maintain the usual structure ot the charge operator. (In particular, the elastic electric isotensor form factors must vanish at zero momentum transfer.) However unattract ive the existence of isotensor currents may be from the theoretical point of view, the question ultimatel~ has to be answered experimentally. In recent works, BLIN-STOYLE analysed possible isotensor electromagnetic effects in nuclear physics (1) with the conclusion that an isotensor admixture-i f present-is at most of the order of ~ 20~o; while SANDA and SHAW claim from their analysis of pion photoproduetion off nucleons (2) that there is a substantial isotensor admixture present (of the order of 20 ~o). While the results of ref. (1.2) are not inconsistent with each other, in our opinion, they are not quite conclusive. In nuclear physics on the one hand, various effects tend to suppress an isotensor contribution, mainly due to the fact that nucleons in a nucleus are almost on their mass shells. On the other hand, the quanti tat ive analysis of Sands and Shaw is based on some dynamical calculations which may not be entirely reliable. In view of the importance of the problem, in this note we want to suggest some further--essential ly quali tat ive--tests to decide whether the electromagnetic current has an isotensor component. Specifically, we propose that double pion photoproduction and electron-positron annihilation into 7:+re-T: ° can provide adequate tests. Assume that an isotensor current exists. Our tests depend crucially on its charge parity. If the isotensor current had C = ~1 (abnormal charge parity), its presence could be perhaps best tested in the processes ~o__> ~+7:-~0 or ~o~ r:+7:-y. Such tests have been performed by a number of groups and the (( Rosenfeld Tables ~> (3) give a small world average ((1--2)% with a large error) for the asymmetry parameter. (The a n a -