FEATURES

nuclei (with new quantum numbers), in direction of other exotic nuclear systems (charmed nuclei and so on). Thephysics of A hypernucleus is generally indicated with the symbol of the parent nucleus with the suffix A, indicating that a A particle has replaced a . ~C means a nuclear system composed of 6 hypernuclei , 5 and one A particle. Three events of double hypernucleiwere observed, inwhich 2 nucleons are substitutedby Aldo Zenoni, Universitii di Bresciaand INFN Sezione di Pavia, 2 A particles. One example is ,J. He, a system composed of2 pro­ Brescia,Italy tons, 2 neutrons and 2 A particles. PaolaGianotti, INFN LaboratoriNazionali di Frascati,Frascati, The concept of hypernuclei can be extended to nuclei where a Italy nucleon is replaced by other than the A one. Experi­ mental evidence was claimed of theexistence ofE-hypernuclei,in which a nucleon is replaced by a L . This is rather sur­ ypernuclear physics was born 50 years ago, in 1953, when prising since, in nuclear matter, the L can decay through a strong H the Polish physicists M.Danysz and J.Pniewski [1] observed, interaction process (L + N ~ A + N) givingverylargewidths for in a stack of photographic emulsions exposed to cosmic rays at the hypernuclear states, unless a strongsuppression mechanism is around 26 km above the ground, the event shown in Figure 1. A at work. However, the experimental evidence of the existence of high energy , colliding with a nucleus of the emulsion, narrow L-hypernuclei is rather controversial and only the case of breaks it in several fragments forming a star. Allthe nuclear frag­ tHe is considered unambiguous. ments stop in the emulsion after a short path, but one After the very first evidence of hyperfragments in cosmic ray disintegrates, revealing the presence inside the fragment, stuck interactions with nuclei, starting from the end of the Sixties, the among the nucleons, of an unstable particle decaying weakly, the installation ofbeams ofK- mesons atparticle accelerators made it A hyperon. For this reason, this particOular nuclear fragment, and possible to study the formation of hypernuclei in the laboratory. the others obtained afterwards in similar conditions, were called Hypernuclei were produced copiouslywith little background, the hyperfragments or hypernuclei. required negative for A hyperon production being The A hyperon is a baryon, like the nucleons (proton and neu­ present in the beam. The typical reaction utilized was the 2 tron), with mass 1115.684 ± 0.006 MeV/c , 20% greater than the "strangeness exchange reaction", where a neutron hit by a K- is mass ofthe nucleon, zero charge and 1=0. It carries a new changed into a A hyperon emitting a 1t- (K- + n ~ A + rr-). The quantum number, not contained normally inside the nuclei, the experimental techniques used in these experiments were photo­ strangeness S =-1. The A hyperon is unstable and decays with graphic emulsions and bubble chambers (filled with He or heavy lifetime 263 ± 2 ps , typical of the that doesn't liquids) exposed to K- beams. These experiments mainly mea­ conserve strangeness and makes a free A mainly disintegrate in a sured the hyperon binding energies, by means of thekinematical nucleon~pionsystem. analysis of the disintegration star, and observed the principal However, since strangeness is conserved in the strong interac­ decaymodes of the hypernuclei. tion and the A particle is the lighter particle in the family of Starting from the Seventies, the study ofhypernuclei was con­ ,. hyperons (baryons with tinued with K- beams by means of counter techniques with strangeness), it can stay magnetic spectrometers and the introduction of new particle in contact with nucleons detectors (multiwire proportional chambers and drift chambers). inside nuclei and form The identification ofwell defined excited hypernuclear levels, by hypernuclei. The mere means of the kinematical analysis of the production reaction, existence of hypernuclei was one of the most important results of a first series of experi­ is ofgreat scientific inter­ ments started at CERN and continued at Brookhaven (USA). est; it gives indeed a new The physical interpretation ofthese spectra was possible in terms dimension to the tradi­ of microscopic descriptions of the A-nucleus and A-nucleon tional world of nuclei by potentials. '. revealing the existence of In the Eighties, a new technique for hypernucleus production a new type of nuclear was introduced atthe Brookhaven laboratory,bymeans ofintense matter and generating beams of high energy1t+.With this technique the A hyperon is new symmetries, new produced inside the nucleus by an "associated production reac­ selection rules, etc. tion" (1t++ n ~ A + K+). This reaction has a reduced cross - Hypernuclei represent section, compared to the"strangeness exchange" reaction, howev­ ~"...... ; . : .- the first kind offlavoured er this drawback is over compensated bythe greater intensities of j • .. the 1t+ beams. The technique was fully exploited at the KEKlabo­ ". .EJg,j,;Thefirst ratory in Japan where, for more than ten years, a great wealth of • ; ,;~ hypernuclearevent excellent hypernuclear data were produced, concerning both to ••• .r •I ,... observed ina nuclear spectroscopyand decay ofhypernuclei. .,,,( a ... ----._o£-~ ._----- .::;••O'<:..'__ __1 emulsion[1]. (A) Theinterestin these new data has triggeredthe attention ofthe I indicatesthe star ofthe primaryinteraction ofthe high energy community and, in the last few years, different I' cosmicray(p)coHidingwith a nucleus ofthe emulsion;(f) isthe laboratories aroundthe worldhave started a hypernuclear physics trackofthe produced hyperfragment;(B)is the vertex ofthe program: COSY at Jiilich in Germany, TJNAF at NewportNews in I, I decayofthe hyperfragment.Theother non-strange nuclear I USA, Nuclotron in Dubna (Russia). In particular,itisworthmen­ I fragments stop in the matter, tioningthe Italian hypernudearproject,the FINUDA experiment

~ ~ _.~ L_...~_.. .•_._"'"....._.._..__..•.._._...... _... -1 (acronim of Flsica NUcleare a DAne)at the Laboratori Nazion-

europhysics news SEPTEMBER/OCTOBER 2002 157 Article available at http://www.europhysicsnews.org or http://dx.doi.org/10.1051/epn:2002501 FEATURES ali di Frascati of INFN, which is going to start collecting data next year. The experiment, which has an ambitious physics pro­ gram, has some special features compared to traditional CHARTOFA·HYPERNUQIDES .' ~ffiJ hypernuclearphysics experiments. In fact, interestingly enough,it ,~/~~Y will operate at a e+e-collider, rather than on an extracted beam. / ~. What follows will describe it in some detail. z

lI Structure of A-hypernuclei r // ~AI AA-hypernucleus t Z is a bound state of Z protons, (A-Z-l) neu­ trons and a A hyperon. The ground state of such a system is r.:-~r:~~ '~N1~. made bythe (A-l) nucleons accommodated in the ground state of l~Cl~C '~CI

l~B the nucleus (A-l)Z and the A hyperon in its lower energy state. ~B'~BI~B The A hyperon, carrying the strangeness quantum number, is a • ~8e~Be~r.s. 3 distinguishable baryon and is not subject to the limitations ~ll~ll~ll:ll 2 imposed by the Pauli principle, therefore it can occupy all quan­ ;,H.;'H.~H.~ ~. 1 tum states alreadyfilled up with nucleons. This feature makes the ~H~" A hyperon, embedded in a hypernucleus, a unique means to explore nuclear structure. ... Fig. 2: "5egretab/e"ofthe 35 known A-hypernuclei.The : The binding energyBAof a A particle in the hypernucleus t Z 1 hypernucleusisformedby Zprotons, Nneutronsanda A hyperon. in its ground state is defined as: ! quality, due to the limited beam intensities available for hypernu­ clear studies from the existing facilities and the modest energy 2 where M core is the mass (in MeV/c ) of the nucleus (A-l)Z ,MA is resolution of the presentexperiments.Nevertheless, the gross fea­ the mass ofthe A particle and M Irypthe mass of the hypernucleus tures of the data are reasonably well reproduced by effective t Z ,experimentally measured. BAvaries linearly with A with a one-body hyperon-nucleus interactions constructed on poten­ slope of about 1 MeV/(unit of A) and saturates at about 23 MeV tial models of the hyperon-nucleon YNinteraction. for the heavy hypernuclei. This behavior suggests a simple model The starting point of these models is a good NN interaction in whichthe A particleis confinedin a potentialwell with a radius generated by the exchange of nonets of mesons, withSU(3)f con­ equalto the nuclear radius and a depth of 28 MeV;to be compared strainst for the coupling constants. Their predictions are fitted to the 55 MeV typical value of the nucleon potential well. simultaneouslyto the abundant NN data and the veryscarce YN This is consistentwith a A-nucleon interaction weaker than the free scattering data. However, until now, ithas notbeenpossible to nucleon-nucleon one. Indeed, in a meson exchange model of the obtain a reliable and unambiguous YN interaction model. interaction, the zero isospin of the A prevents the exchange of The improvement of the YNinteraction models would need isovector mesons like the 1t or the p with a nucleon and deter­ precise data on the free YN interaction,which are verydifficult to mines the lack of strong tensor components in the interaction. Therelative weakness ofthe A-nucleon interaction entails thatthe shell structure is not disrupted by the insertion of the A in the

nucleus and the lack of Pauli effects allows all the nuclear single 1.6 H'" .... \ ••.• ..1 particle states to be populated by the A. In Figure 2, the so called 1.4 "Segre table" of the hypernuclei shows the 35 hypernuclei known ·······1·· at present. ~ .. i Experiments of hypernucleus production by "strangeness "'" 1.2 ·········l..······ exchange" and "associated production" processes can produce ~

ci ··········!·············:·..··········1···········4~········1 hypernuclei in which the A populates different single particle ~

states. Tht; latter technique is particularly suitable for populating ~ 0.8 ..·······"i··.,..•....: ~ . low lying A states, thanks to the high recoil momentum trans­ ...... L _.. L t _.. ferred to the A particle in the reaction. ~0.6 A beautiful representation of this process is given in Figure 3, ,tf 0.4 where the excitation spectrum of ~Y,obtained by the"associated ~Yat production" reaction 89Y(1t+,K+) the KEK laboratory in 0.2 Japan, is shown. The spectrum demonstrates how, starting from a neutron the g912state, it is possible to accommodate a A particle ~~10 in o_3~0~~I..LlII~LI.l....::-:'-U.~.:....:...... ~s..:...... c~...... in the hypernuclear statesf, d ,p and even in the ground state s . These measurements constitute the spectacular confirmation, -BA (MeV) at a textbook level, of the validity of the independent particle model or shell model of the nucleus. In non-strange nuclei, the ... Fig. 3: Hypernuclearmassspectrum of~Yobtainedat KEKby observation of single particle states is only possible for the states the B69 experiment,with an energy resolutionof 1.65MeV.The of the most external nucleon orbits. In fact, due to the Pauli prin­ bump structures correspond to the A majorshellsorbits (s,p,d,f) ciple and pairing interactions, deeply bound nucleon single inthe bound region.Thewidths ofthe bumpsforthe po,d- and f­ particle states are so fragmented as to be essentially unobservable. orbits are significantlywider than the experimentalresolution The present experimental data on hypernuclear binding energies and the peakfor the fA-orbitissplit into two peaks. and detailed spectroscopic features are limited in quantity and

158 europhysics news SEPTEMBER/OCToBER 2002 FEATURES

obtain, due to lowhyperon beamintensities andshortlifetimes of At KEK the orbit splitting of the 3/2--1/2- doublet in ~C the hyperons. In particular,production andscatteringin the same was measured, though with a coarse resolution on yrays [5], and target are almost automatically required. At present, the experi­ a doublet spacing of 152±S4(stat)±36(syst)keV was determined. mental data on AN and rn scattering consist of not more than The splitting predicted for this transition by meson exchange 850 scattering events, in the momentum region from 200 to 1500 models is 390-780 keY, depending on the interaction used, and MeVIc. Thelowenergydata, inparticular,failto adequately define 150-200 keV by quark based models [3]. even the relative sizes of the dominant s-wave spin-triplet and Nevertheless, other recent spectroscopic dataseem to suggest a spin-singlet scattering lengths and effective ranges. larger spin-orbit level splitting on heavy hypernuclei, especially This is the reason why a systematic and detailed spectroscopic at large values of the A angular momentum. The spectrum study ofhypernuclei with high resolution experiments, with sin­ reported in Figure 3 for ~yis an example; it shows a splitting of gle and even multiple strangeness contents, would offer the thefAlevel and the width of the other levels broader than the possibility of increasing the experimental information on YN instrumental energy resolution. interaction and,for the first time, allow boththe studythe dynam­ The meson exchange models constitute the best description ics of systems carrying SU(3)rflavour symmetry, in the available, at present, for the at low energy. The non-perturbative QCD sector, andthe generalization ofthebary­ question whether they should be modified with the inclusion of on-baryon interaction to the full SU(3h family, including explicit quark effects remains open. hyperons [2]. The spin dependent parts of the YN interaction are of major Decayof A-hypernuclei interest, since theyare intimately related to the modelling ofthe In addition to information on nuclear structure and the YNinter­ short-range part ofthe interaction. Inparticular, for theANinter­ action, hypernuclei may give access to experimental information action, the spin-orbit interaction was found to be smaller than not otherwise accessible by their decays, in particular the non­ that for the nucleon, which amounts to 3-5 MeY. In fact, experi­ mesonic decay. Let us recall that a free A hyperon decays almost ments based on the hypernucleus formation kinematics, with the totallyinto a and a nucleon (A ~ p7t"(""64%),A ~ 111&(36%», best energy resolution on the nuclear levels achieved until now of with a release of kinetic energy of about 5 MeV to the nucleon, 1.5 MeY,were not able to measure any spin-orbit splitting. corresponding to a final momentum of about 100 MeVIc. The Calculations of the spin-orbit component of the AN interac­ decay may occur, in principle, with isospin change I1I= 112or tion using meson exchange theories, lead to predictions of the ill = 3/2.However, the experimental decaybranchingratios ofthe spin orbit splitting, for the 5/2+-3/2+ doublet in lBe, of 80­ A and other hyperons imply a dominance by a factor 20 ofthe 160 keY, depending on the interaction used [3]. On the other /:,.1=1/2component over the I1I = 3/ 2 one. The origin of this hand, when a quark model based spin-orbit force is used, which empiricalrule is essentiallynot understoodin a fundamental way. naturally accounts for the short range part of the interaction, The situation described above changes dramaticallywhen the much smaller values of30-40 keV are obtained. The two types of A is imbedded in the nuclear medium, since the nucleon ofthe models, in fact, predict different features for the antisymmetric mesonic decay is emitted with a momentum much less than the part ofthe spin-orbitforce. However, it needs to be said that quark nucleon Fermi momentum kf"" 280 MeV/c. Thus, the pionic models have yet to provide an extensive and satisfactory descrip­ decay modes are severely inhibited by Pauli blocking ofthe final tion ofthe YN interaction. state nucleon in all but the lightest hypernuclei and new, non­ Figure 4 reports a recent measurement of the splitting of the mesonic decay modes of the hypernucleus are introduced, 5/2+-3/2+ doublet in lBe bythe BNL-AGS E930 experiment [4], through the weak interaction of the A with the nucleons measuring y rays emitted in the nuclear transitions with the new (A + n ~ n + n + 176 MeV ,A + P~ n + p + 176 MeV). This germanium detector arrayHyperball. This newtechnique allowed process is possible onlyin hypernudei, thanks to the unique beam the energy resolution on low lying hypernuclear levels to be . of A's, stable against the mesonic decay and strong interaction, improved from a few MeV to a few keY, even if the count rate which is available inside a hypernudeus. resulting is still quite low, -200 is per month of data taking. The The study of the non-mesonic weak decay is of fundamental spacing of the two levels was measured to be 31 ± 2 keY, incom­ importance, since it provides primary means of exploring the patible with the prediction ofthe meson exchange models. four fermion, strangeness changing, baryon-baryon weak inter-

~ .:L If) (a) bound region '-.. 30 UJ C :l 25 , .--/ XBeE2 transitions 0 ~ ... Fig. 4: y-ray energyspectrum from the u k::: (5/2+,3/2+ 1/2+) 20 deexcitation of the lBe hypernudeus obtained at BNL[4].Theenergydifference 15 between the two El (5/2+·1/2+and 3/2"­ 1/2+) transitions gives information on the 10 strength of the AN spin-orbit interaction. Thesmall measuredlevel splitting of 5 31 ±2 keYis incompatible with predictions 0 obtained from mesonexchangemodels. 2200 2400 2600 2800 3000 3200 3400 3600

europhysics news SEPTEMBER/OCTOBER 2002 159 FEATURES action AN ~ NN. The non-mesonic process resembles the weak f.S = 0 nucleon-nucleon interaction, which has been studied experimentally in parity-violating NN scattering measurements. However, the AN ~ NN two body interaction mode offers more information, since it can explore both the parity-conserving and the parity-violating sectors of the f.S =1 weak baryon-baryon interaction. In the weakNNsystem the strong force masks the sig­ nal ofthe weak parity-conserving part ofthe interaction. In addition, the large momentum transfer in non-mesonic decay processes implies that they probe short distances and might, therefore, expose the role of explicit quark/gluon sub­ structure ofthe baryons. Furthermore, the fundamental question as towhether the M = 112rule,which governs pionic decay, applies to the non-mesonic weak decays may also be addressed. One ofthe most important experimental observables ofthe non­ mesonic decay ofhypernuclei is the ratio ofthe neutron-induced (An ~ nn) over the proton-induced (Ap ~ np) decay rate ... Fig. S: Afish eyeview ofthe DA«I>NEhall at lNF (Frascati NationalLaboratory),TheDA«I>NEacceleratorhasa r nntr np , which is sensitive to the isospin structure ofthe interac­ tion. In particular, it should be sensitive to the question of the circumferenceof about 190 m and two interaction regions:[Pl, on the right sideofthe picture, is presentlyoccupiedby the KLOE validity ofthe M = 112rule. experiment devotedto epviolation studies;IP2,on the left, isthe Notwithstanding the considerable physical interest, experi­ FINUDAinteraction region.The FINUDAdetector,the assembly mental data are scarce and affected by very large errors. This is with dark pink octagonal end-caps,will be movedonto the beam . mainly due to the tremendous difficulty in producing abundant­ line next autumn. 1yA hypernuclei in their ground state and detecting the products oftheir decay; in particular neutrons. However, the comparison between the predictions ofthe theory system,K- mesons can be used, as ithas been seen before, to insert and the scarce data available is particularly puzzling. In fact, "strangeness" inside the nuclei and produce hypernuclei. FINU­ whereas experimental data favour values of the ratio DA [8] is the experimentthat, nextyear,wiUmake themOst ofthis r nntrnp = 1 - 2 (although with errors of the order of 50%), the· facility andwill try to shed newand brighter light on theworld of meson exchange models, involving one strong interaction vertex hypernuclear physics. and a weak one, systematically underestimate the value of the Presently, DA<1lNEis delivering something like 150 $/s, halfof experimental ratio [6]. In this type of models the M = 112rule is which decay into a pair of opposite charged with very low generally enforced in the weakvertex. momentum, only 127 MeV/c. This low and almost monochro­ Among the mechanisms explored to remedy the puzzle, direct matic momentum value is a great advantage for hypernuclear quark model approaches have been adopted, in which the short studies, when comparing the main features of the different pro­ range region is modelled by effective four quark vertices plus duction techniques. stronginteraction corrections. These models are motivated bythe When K- or 1t+ beams are used for"strangeness exchange"or large momentum transfer of the AN ~ NN reaction. The results «associateproduction" reactions, thicktargets ofthe order ofsome 2 ofthese calculations, which yield a large violation ofthe M = 1/2 g/cm must be employed to obtain sufficient event rates. Hence, rule,provide, on the contrary; significantlarger values for the ratio the uncertaintyregarding the point ofinteraction and the energy r nn/rnp • Could this be interpreted as a sign of explicit quark straggling ofthe emitted particles limit the resolution achievable effects in nuclei or a consequence of a strong violation of the on the hypernuclear levels. The same problem occurs in experi­ M = 112rule? To answer these fundamental questions, a drastic ments using K- at rest at proton machines. Here, intense fluxes of improvement ofthe quality of the experimental data is certainly kaons are emittedfrom an externalproduction target,but the dis­ needed. tance betweenthe source and the experimental area mustbe of some tens of meters, due to radiation safety requirement. FINUDA:the Italian hypernuclear factory Therefore, kaons with momentum lower than 400-450 MeV/c The Italian National Institute for Nuclear Physics (INFN) has cannot survive such a path in a number suitable to give a beam. built, in its Frascati National Laboratory, a special accelerator, Consequently, experiments with stopped K- must be per­ DA<1lNE,dedicated to the copious production of $(1020) parti­ formed using beams of500-600 MeVIc,degraded in momentum cles, for the pu~poseof producing beams of extremely high with a moderator placed justbefore the production target. This intensityandprecise energyfor the study ofthe most rare subnu­ introduces a great uncertainty on the interaction point that can clear phenomena. negate any excellent capability ofmeasuring the momenta of the DA<1lNE(DoubleAnnularringForNice Experiments) [7] con­ emitted particles. On the contrary, the low momentum of sists of two almost circular pipes, one for the electrons and the DA<1lNEkaons, allowing them to be stopped simply in 0.1 g/cm2 otherfor the positrons,thatoverlap in two straightsections where of carbon, permits the accuracy achievable on the energy levels the beams collide head-on. The energyofeach beam is set to 510 of the produced hypernuclei to be improved up to 750 keV.In MeV in order to produce the $(1020) particle in the collisions. addition, the large acceptance ofthe apparatus,typical ofa collid­ This particle is unstable and decays, in a very short time, mostly er experiment, allows for high acquisition rates ofthe order of 80 into neutral and charged K mesons. A view of the DA<1lNEhall event/hour, for a typical hypernuclear level, at the design lumi­ can be seen in Figure 5. nosityofthe machine. Even though the main aim ofthe DA<1lNEmachine is the pre­ The FINUDA apparatus (Figure 6) is not onlyoptirnised to per­ cision study of CP and CPT symmetries in the neutral kaon form high resolution hypernuclear spectroscopy, but also to

160 europhysics news SEPTEMBER/OCTOBER 2002 FEATURES

The magnetic field of 1.10 Tesla, produced by a superconduct­ ing solenoid, is essential to bend charged particles allowing for a precise evaluation of their momentum. Each element of the FIN­ UDA detector is a small masterpiece of mechanics and electronics. Since particles involved are oflow energy, in order to perturb their properties as little as possible, onlylight materials have been employed in the construction. For the same reason the whole detector atmosphere is filled with Helium gas. In fact, if air is left in the detector, the momentum resolution /).p/Pwould be worsened from 0.3% to 1.5%. For more then 5 years, the design, assembling and installation of the FINUDA detector has involved some 50 physicists, mainly Italians,togetherwith dozens ofhighlyprofessional engineers and technicians. After such a long training period, all is nowready for the power-up.

... fig&; Agroup ofFlNUDAcollaboratorsis proudlydepicted in References front ofthe experimentalapparatus.Theleftmostladysquatting [1] M. Danysz and J.Pniewski,]. Phil. Mag. 44 (1953),348 downisone ofthe authors. [2] B.F.Gibson and E.V. Hungerford III, Phys.Rep. 257 (1995) 349 ------_._-_ ...... [3] E. Hiyamaet aL,Phys. Rev.Lelt. 85 (2000) 270 study the hypernuclear decay processes. Charged particles, emit­ ted following hypernucleus production or decay, are tracked into [4] H.Akikawa et al., Nud. Phys. A691 (2001) 134 the FINUDA cylindrical magnetic volume (1 m radius, 2 m [5] S. Ajimura et al., Phys.Rev.Lelt. 86 (2001) 4255 length) by means of four different sub-detectors, each one opti­ [6] W.M. Alberico and G. Gabarino, Lanl nud-thlOl12036, 2001 mized for a different task: silicon microstrips detect with high (To be published on Phys. Rep. 2002) accuracy the hypernucleus formation point; drift chambers and straw tubes reconstruct charged particle trajectories; plastic scin­ [7] M. Preger for the DAFNE Team: "Status of DAFNE", presented at tillators, detectors with nanosecond time response, are used to HEACC2001 - 18th International Conference on High Energy selecthypernuclear events among otherkaon-nucleus interactions Accelerators, 26-30 March 2001, Tsukuba, Japan. and to detect neutrons produced by the hypernuclear decay. [8] P.Gianotti, Nud. Phys. A691 (2001) 483c

Do.youneedaccessto world-classresearchfacilities? This EU Fifth Framework Programme offers European researchers fully-funded access to NMRC's state-of-the-art research infrastructure The EU-funded 'Enhancing Access to Research such as photonics, nanotechnology, microtechnologies Infrastructure' programme provides researchers and ICT/Bio. The programme pays researchers for the throughout the European Union member and full cost of access and laboratory use (including associated states access to NMRC's state-of-the-art training and technical support) and the full cost of facilities, equipment and expertise in areas travel to NMRC including subsistence. About NMRC NMRC was established in 1981 and is today one of Europe's premier information and communication technologies (ICT) research institutes. NMRC has a critical mass of over 200 researchers and a research infrastructure that is valued in excess of € I00 million. How to Apply For more information about the programme and facilities available, contact Paul Roseingrave at [email protected] or +353 21,490,4268.

www.nmrc.ie/access/ europhysics news SEPTEMBER/OCTOBER 2002 161