First results

During its first four years of operation, the HERMES experiment has lived up to the swiftness of its classical namesake. Primarily a rerun of previous experiments, the first HERMES investigations measured the scattered electron (positron) using targets of polarized helium-3 (1995) and polarized hydrogen (1996­97).

The resulting "overall" spin structure of the neutron and the proton reconfirmed the "spin puzzle". Good agreement with previous data from experiments at CERN and SLAC provided an essential and unique cross-check with a completely novel technique.

For a few weeks every year, HERMES has been collecting data from unpolarized gases to study further properties such as the quark flavour content of the nucleon sea, an important domain for testing models of nucleon structure. The latest HERMES data give the flavour asymmetry of the light quark sea, clearly showing an excess of down antiquarks over up antiquarks in the proton sea, in good agreement with data from experiments using a completely different approach at much higher energies.

First insights into quark spin

HERMES is now hard on the track of the "missing" spin. Preliminary results present the polarized parton distributions from data taken in 1995 and 1996, with statistics equalling that of the Spin Muon Collaboration (SMC) experiment at CERN, stopped two years ago.

During 1995 and 1996 running, HERMES has measured asymmetries where the struck quark gives a positively or negatively charged hadron recorded by the detector. Three quark polarizations were extracted: the up and down valence quark polarizations, and the average polarization of the quark sea, dominated by up antiquarks.

These preliminary results reveal that the spins of the up valence quarks point in the same direction as the overall proton spin, whereas the down valence quark seems to carry a spin pointing in the opposite direction, the spin contribution increasing with the amount of the proton's momentum carried by the quark. The sea quarks appear to slightly favour alignment with the proton, though the data are still consistent with no dependence on the direction of the proton spin.

The HERMES extraction can be compared with the SMC results. Even though the HERMES definition of flavour independence of the sea polarization varies slightly from that of SMC, the difference was verified to be less than 10%. Both experiments are comparable in their extraction of the up and down valence quark values, with the HERMES data possibly indicating a more positive contribution from the sea quarks than does SMC.

Analyses now focus on including the 1997 proton data, approximately 2.5 times larger than the sample used for the preliminary results. This will give the polarized quark distributions of the proton with unprecedented precision. HERMES is now turning to the study of the neutron with a polarized deuterium target operation for 1998 and 1999, expected to be comparable in size to the hydrogen data set.

For the detector, fresh possibilities will be opened up by the new dual radiator Ring-Imaging CHerenkov (RICH) detector (November, page 10) installed during the winter 1997/98 shutdown to replace the original threshold Cherenkov detector. The new RICH identifies protons, pions and kaons over nearly its entire kinematic range, obviating model dependence.

Flavour tagging of kaons will give HERMES direct access to the strange quarks inside the nucleon sea ­ where no experiment has been before. Since no heavier quarks (charm) are found in the nucleon sea at such low energies, these quarks must somehow involve the gluons inside the nucleon, for instance through photon­gluon fusion.

The new RICH and a recently implemented muon wall will allow investigation of such events and thus open the way to another missing part of the spin puzzle, the contribution of the gluons.

But HERMES physics will not end there. Other goals include study of further spin structure, quark­gluon correlations inside the nucleon, nuclear effects and investigating the transverse spin of the partons instead of the longitudinal part. But that will be a different story...