hep-ex0105048/bp.tex
1: \documentclass{elsart}
2: \journal{Nuclear Physics B}
3: 
4: \usepackage{epsfig}
5: \usepackage{rotate}
6: 
7: \newcommand{\numucc}{$\nu_{\mu}~CC~$}
8: \newcommand{\anumucc}{$\bar\nu_{\mu}~CC~$}
9: \newcommand{\numunc}{$\nu_{\mu}~NC~$}
10: 
11: \newcommand{\bp}{$Bp~$}
12: \newcommand{\bpim}{$B\pi^{-}$~}
13: 
14: \newcommand{\bmath}{\begin{displaymath}}
15: \newcommand{\emath}{\end{displaymath}}
16: \newcommand{\be}{\begin{equation}}
17: \newcommand{\ee}{\end{equation}}
18: 
19: \begin{document}
20: \begin{frontmatter}
21: 
22: \title{\boldmath {A study of backward going $p$ and $\pi^{-}$ in 
23: \numucc interactions with the NOMAD detector }}
24: \collab{NOMAD Collaboration}
25: \author[Paris]             {P.~Astier}
26: \author[CERN]              {D.~Autiero}
27: \author[Saclay]            {A.~Baldisseri}
28: \author[Padova]            {M.~Baldo-Ceolin}
29: \author[Paris]             {M.~Banner}
30: \author[LAPP]              {G.~Bassompierre}
31: \author[Saclay]            {N.~Besson}
32: \author[CERN,Lausanne]     {I.~Bird}
33: \author[Johns Hopkins]     {B.~Blumenfeld}
34: \author[Padova]            {F.~Bobisut}
35: \author[Saclay]            {J.~Bouchez}
36: \author[Sydney]            {S.~Boyd}
37: \author[Harvard,Zuerich]   {A.~Bueno}
38: \author[Dubna]             {S.~Bunyatov}
39: \author[CERN]              {L.~Camilleri}
40: \author[UCLA]              {A.~Cardini}
41: \author[Pavia]             {P.W.~Cattaneo}
42: \author[Pisa]              {V.~Cavasinni}
43: \author[CERN,IFIC]         {A.~Cervera-Villanueva}
44: \author[Dubna]             {A.~Chukanov}
45: \author[Padova]            {G.~Collazuol}
46: \author[CERN,Urbino]       {G.~Conforto}
47: \author[Pavia]             {C.~Conta}
48: \author[Padova]            {M.~Contalbrigo}
49: \author[UCLA]              {R.~Cousins}
50: \author[Harvard]           {D. Daniels}
51: \author[Lausanne]          {H.~Degaudenzi}
52: \author[Pisa]              {T.~Del~Prete}
53: \author[CERN]              {A.~De~Santo}
54: \author[Harvard]           {T.~Dignan}
55: \author[CERN]              {L.~Di~Lella}
56: \author[CERN]              {E.~do~Couto~e~Silva}
57: \author[Paris]             {J.~Dumarchez}
58: \author[Sydney]            {M.~Ellis}
59: \author[LAPP]              {T.~Fazio}
60: \author[Harvard]           {G.J.~Feldman}
61: \author[Pavia]             {R.~Ferrari}
62: \author[CERN]              {D.~Ferr\`ere}
63: \author[Pisa]              {V.~Flaminio}
64: \author[Pavia]             {M.~Fraternali}
65: \author[LAPP]              {J.-M.~Gaillard}
66: \author[CERN,Paris]        {E.~Gangler}
67: \author[Dortmund,CERN]     {A.~Geiser}
68: \author[Dortmund]          {D.~Geppert}
69: \author[Padova]            {D.~Gibin}
70: \author[CERN,INR]          {S.~Gninenko}
71: \author[Sydney]            {A.~Godley}
72: \author[CERN,IFIC]         {J.-J.~Gomez-Cadenas}
73: \author[Saclay]            {J.~Gosset}
74: \author[Dortmund]          {C.~G\"o\ss ling}
75: \author[LAPP]              {M.~Gouan\`ere}
76: \author[CERN]              {A.~Grant}
77: \author[Florence]          {G.~Graziani}
78: \author[Padova]            {A.~Guglielmi}
79: \author[Saclay]            {C.~Hagner}
80: \author[IFIC]              {J.~Hernando}
81: \author[Harvard]           {D.~Hubbard}
82: \author[Harvard]           {P.~Hurst}
83: \author[Melbourne]         {N.~Hyett}
84: \author[Florence]          {E.~Iacopini}
85: \author[Lausanne]          {C.~Joseph}
86: \author[Lausanne]          {F.~Juget}
87: \author[INR]               {M.~Kirsanov}
88: \author[Dubna]             {O.~Klimov}
89: \author[CERN]              {J.~Kokkonen}
90: \author[INR,Pavia]         {A.~Kovzelev}
91: \author[LAPP,Dubna]        {A.~Krasnoperov}
92: \author[Dubna]             {D.~Kustov}
93: \author[Dubna,CERN]        {V.~Kuznetsov}
94: \author[Padova]            {S.~Lacaprara}
95: \author[Paris]             {C.~Lachaud}
96: \author[Zagreb]            {B.~Laki\'{c}}
97: \author[Pavia]             {A.~Lanza}
98: \author[Calabria]          {L.~La Rotonda}
99: \author[Padova]            {M.~Laveder}
100: \author[Paris]             {A.~Letessier-Selvon}
101: \author[Paris]             {J.-M.~Levy}
102: \author[CERN]              {L.~Linssen}
103: \author[Zagreb]            {A.~Ljubi\v{c}i\'{c}}
104: \author[Johns Hopkins]     {J.~Long}
105: \author[Florence]          {A.~Lupi}
106: \author[Florence]          {A.~Marchionni}
107: \author[Urbino]            {F.~Martelli}
108: \author[Saclay]            {X.~M\'echain}
109: \author[LAPP]              {J.-P.~Mendiburu}
110: \author[Saclay]            {J.-P.~Meyer}
111: \author[Padova]            {M.~Mezzetto}
112: \author[Harvard,SouthC]    {S.R.~Mishra}
113: \author[Melbourne]         {G.F.~Moorhead}
114: \author[Dubna]             {D.~Naumov}
115: \author[LAPP]              {P.~N\'ed\'elec}
116: \author[Dubna]             {Yu.~Nefedov}
117: \author[Lausanne]          {C.~Nguyen-Mau}
118: \author[Rome]              {D.~Orestano}
119: \author[Rome]              {F.~Pastore}
120: \author[Sydney]            {L.S.~Peak}
121: \author[Urbino]            {E.~Pennacchio}
122: \author[LAPP]              {H.~Pessard}
123: \author[CERN,Pavia]        {R.~Petti}
124: \author[CERN]              {A.~Placci}
125: \author[Pavia]             {G.~Polesello}
126: \author[Dortmund]          {D.~Pollmann}
127: \author[INR]               {A.~Polyarush}
128: \author[Dubna,Paris]       {B.~Popov}
129: \author[Melbourne]         {C.~Poulsen}
130: \author[Zuerich]           {J.~Rico}
131: \author[Dortmund]          {P.~Riemann}
132: \author[CERN,Pisa]         {C.~Roda}
133: \author[CERN,Zuerich]      {A.~Rubbia}
134: \author[Pavia]             {F.~Salvatore}
135: \author[Paris]             {K.~Schahmaneche}
136: \author[Dortmund,CERN]     {B.~Schmidt}
137: \author[Dortmund]          {T.~Schmidt}
138: \author[Melbourne]         {M.~Sevior}
139: \author[LAPP]              {D.~Sillou}
140: \author[CERN,Sydney]       {F.J.P.~Soler}
141: \author[Lausanne]          {G.~Sozzi}
142: \author[Johns Hopkins,Lausanne]  {D.~Steele}
143: \author[CERN]              {U.~Stiegler}
144: \author[Zagreb]            {M.~Stip\v{c}evi\'{c}}
145: \author[Saclay]            {Th.~Stolarczyk}
146: \author[Lausanne]          {M.~Tareb-Reyes}
147: \author[Melbourne]         {G.N.~Taylor}
148: \author[Dubna]             {V.~Tereshchenko}
149: \author[INR]               {A.~Toropin}
150: \author[Paris]             {A.-M.~Touchard}
151: \author[CERN,Melbourne]    {S.N.~Tovey}
152: \author[Lausanne]          {M.-T.~Tran}
153: \author[CERN]              {E.~Tsesmelis}
154: \author[Sydney]            {J.~Ulrichs}
155: \author[Lausanne]          {L.~Vacavant}
156: \author[Calabria]          {M.~Valdata-Nappi\thanksref{Perugia}}
157: \thanks[Perugia]           {Now at Univ. of Perugia and INFN, Perugia, Italy}
158: \author[Dubna,UCLA]        {V.~Valuev}
159: \author[Paris]             {F.~Vannucci}
160: \author[Sydney]            {K.E.~Varvell}
161: \author[Urbino]            {M.~Veltri\thanksref{mike}}
162: \thanks[mike]              {Corresponding author. {\it Email address:} veltri@fis.uniurb.it}
163: %\author[Urbino]            {M.~Veltri\corauthref{cor}}
164: %\corauth[cor]              {Corresponding author.}
165: %\ead{veltri@fis.uniurb.it}
166: \author[Pavia]             {V.~Vercesi}
167: \author[CERN]              {G.~Vidal-Sitjes}
168: \author[Lausanne]          {J.-M.~Vieira}
169: \author[UCLA]              {T.~Vinogradova}
170: \author[Harvard,CERN]      {F.V.~Weber}
171: \author[Dortmund]          {T.~Weisse}
172: \author[CERN]              {F.F.~Wilson}
173: \author[Melbourne]         {L.J.~Winton}
174: \author[Sydney]            {B.D.~Yabsley}
175: \author[Saclay]            {H.~Zaccone}
176: \author[Dortmund]          {K.~Zuber}
177: \author[Padova]            {P.~Zuccon}
178: 
179: \address[LAPP]           {LAPP, Annecy, France}                               
180: \address[Johns Hopkins]  {Johns Hopkins Univ., Baltimore, MD, USA}            
181: \address[Harvard]        {Harvard Univ., Cambridge, MA, USA}                  
182: \address[Calabria]       {Univ. of Calabria and INFN, Cosenza, Italy}         
183: \address[Dortmund]       {Dortmund Univ., Dortmund, Germany}                  
184: \address[Dubna]          {JINR, Dubna, Russia}                               
185: \address[Florence]       {Univ. of Florence and INFN,  Florence, Italy}       
186: \address[CERN]           {CERN, Geneva, Switzerland}                          
187: \address[Lausanne]       {University of Lausanne, Lausanne, Switzerland}      
188: \address[UCLA]           {UCLA, Los Angeles, CA, USA}                         
189: \address[Melbourne]      {University of Melbourne, Melbourne, Australia}      
190: \address[INR]            {Inst. Nucl. Research, INR Moscow, Russia}           
191: \address[Padova]         {Univ. of Padova and INFN, Padova, Italy}            
192: \address[Paris]          {LPNHE, Univ. of Paris VI and VII, Paris, France}    
193: \address[Pavia]          {Univ. of Pavia and INFN, Pavia, Italy}              
194: \address[Pisa]           {Univ. of Pisa and INFN, Pisa, Italy}               
195: \address[Rome]           {Roma Tre University and INFN, Rome, Italy}      
196: \address[Saclay]         {DAPNIA, CEA Saclay, France}                         
197: \address[SouthC]         {Univ. of South Carolina, Columbia, SC, USA}
198: \address[Sydney]         {Univ. of Sydney, Sydney, Australia}                 
199: \address[Urbino]         {Univ. of Urbino, Urbino, and INFN Florence, Italy}
200: \address[IFIC]           {IFIC, Valencia, Spain}
201: \address[Zagreb]         {Rudjer Bo\v{s}kovi\'{c} Institute, Zagreb, Croatia} 
202: \address[Zuerich]        {ETH Z\"urich, Z\"urich, Switzerland}                 
203: 
204: \clearpage
205: 
206: 
207: \begin{abstract}
208: \noindent Backward proton and $\pi^-$ production has
209:  been studied in \numucc interactions with carbon nuclei. 
210:  Detailed analyses of the momentum distributions, of the production rates,
211:  and of the general features of events with a backward going particle,
212:  have been carried out in order to understand the mechanism producing 
213:  these particles. The backward proton data have been compared with the 
214:  predictions of the reinteraction and the short range correlation models.
215: \end{abstract}
216: 
217: \begin{keyword}
218: Neutrino interactions, cumulative production, intranuclear cascade,
219: short range correlations
220: \PACS 13.15.+g \sep 13.85.Ni
221: \end{keyword}
222: \end{frontmatter}
223: 
224: 
225: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
226: \section{Introduction}
227: \label{intro}
228: It is a well established experimental fact that in the high energy 
229: interactions off nuclei there are particles emitted backwards, with 
230: respect to the beam direction, which have energies not allowed by the 
231: kinematics of collisions on a free and stationary nucleon.
232: Backward going protons are commonly observed while, in absence of
233: nuclear effects, their production is forbidden. High energy mesons,
234: whose production in the backward direction is only allowed up to a given 
235: momentum, are detected also, at momenta above such a limit.
236: Since a long time \cite{baldin} this effect has been used 
237: as a powerful tool to investigate nuclear structure. The models 
238: proposed to explain the origin of these particles (also called
239: cumulative in the literature) can be divided essentially into two
240: categories: models based on the intranuclear cascade mechanism
241: and models based on the cumulative effect of groups of correlated 
242: nucleons/quarks.\\
243: \noindent In the {\em intranuclear cascade models}, the
244:  production of particles in the kinematically forbidden region
245: (KFR) can be seen as the result of
246: multiple scattering and of interactions of secondary hadrons,
247: produced in the primary $\nu$-nucleon collision, with the other nucleons 
248: while they propagate through the nucleus \cite{kopel}. 
249: The reinteraction or intranuclear cascade (INC) models usually rely on 
250: Monte Carlo methods to make their predictions \cite{ranft}, \cite{ferrari-ranft}. 
251: The importance of the intranuclear cascade mechanism is that it can provide 
252: information on the space--time evolution of the hadronization process. 
253: Experimentally one observes that the cascade is restricted to slow particles 
254: only, while the fast ones do not reinteract inside the nucleus.
255: The currently accepted explanation for this effect is the ``formation zone''
256: concept \cite{stodolsky}-\cite{eliseev}. This is the 
257: distance (or the time) from the production point which is required for 
258: the secondary hadrons to be ``formed'', i.e. to be able to interact
259: as physical hadronic states. Since the distance/time required, 
260: due to the Lorentz time--dilation factor, 
261: is proportional to the energy of the secondary, 
262: the INC process is restricted to slow hadrons which have formation 
263: lengths smaller than the nuclear radius.\\
264: An advantage of neutrino (or more generally lepton)
265: induced interactions with respect to hadronic processes is the fact that
266: the projectile interacts only once, avoiding the complications
267: related to the projectile rescattering in the target. Natural drawbacks
268: are the facts that the equivalent of the projectile energy
269: in  the hadronic processes, i.e. the hadronic jet energy  in the leptonic 
270: processes, is not fixed but varies from event to event and suffers from 
271: systematic uncertainties, especially related to the reconstruction of neutrals.\\
272: Another feature of the neutrino--nucleus interaction is the fact that, due to 
273: the extremely small neutrino--nucleon cross-section, the interaction can 
274: practically take place anywhere in the nucleus. This is not the case in
275: hadron--nucleus (and also photon--nucleus) interactions where, due to the large
276: cross--section, the interaction takes place essentially on the first 
277: nucleons along the particle path, i.e. on the surface of the nucleus. 
278: As a consequence not all the nucleons participate in the scattering and 
279: the total hadron--nucleus cross--section is smaller than $A$ (the number 
280: of nucleons in the nucleus) times the total hadron--nucleon cross--section. 
281: This effect is known as {\it shadowing} \cite{arneodo}.\\
282: In the {\em correlated nucleon/quark models}
283: the backward particles are produced in the collisions off 
284: structures with mass larger than the mass of the nucleon. 
285: These structures are formed,
286: at small interparticle distance, under the action of the short range
287: part of the nuclear force. They may either be described as fluctuations of the
288: nuclear density \cite{burov} or as clusters of a few correlated nucleons 
289: \cite{fs1}-\cite{fujita2}. In any case 
290: these structures represent the effect of gathering two or more nucleons in
291: small volumes with a radius of the order of $0.5 \div 0.7~fm$. 
292: The nucleons in these structures can acquire high momenta and 
293: the fast backward going particles can be seen as a direct manifestation 
294: of the  high momentum tail of the Fermi distribution. 
295: For instance, the classical mechanism of Fermi motion of nucleons
296: inside the nucleus can explain backward proton production only up
297: to about $300~MeV/c$, well below the observed limit.
298: Nucleons can also loose their identity in these structures and larger masses
299: can be reached if quarks in neighbouring nucleons stick
300: together, forming multiquark clusters \cite{carlson,multiq}.
301: The momentum distribution of quarks in the cluster is, 
302: in this case, responsible for the observed spectra of cumulative particles.\\
303: \noindent These two classes of models  are not mutually exclusive. 
304: Indeed some features of the data can be explained by both, and 
305: the experimentally observed production in  the backward hemisphere
306: \cite{bosveld} can have contributions from both mechanisms.\\
307: \noindent While the production of particles in the KFR has been investigated
308: over a very wide range of incident energies and on many different nuclear targets
309: using hadron beams \cite{fredriksson}, the data on backward particle
310:  production in $\nu~(\bar\nu)$ induced reactions are far less abundant.
311: Up to now only five bubble chamber experiments have studied backward protons 
312: (and in only one experiment backward pions as well) using bubble chambers filled 
313: with deuterium \cite{matsinos}, heavy liquid fillings \cite{berge}-\cite{matsinos}
314: and in one case a hybrid emulsion--bubble chamber technique \cite{dayon}.\\
315: Early hadronic experiments had found that the invariant cross-section
316: for backward particle production can be parametrized as $e^{-BP^2}$ where $P$
317: is the particle momentum. It was also found that the slope parameter $B$
318: is almost independent of the type of incident particle, its energy and target
319: type, a fact known as ``nuclear scaling''. Neutrino experiments have confirmed 
320: these properties.\\
321: Here we present a detailed study of backward going protons (\bp) and
322: backward going $\pi^{-}$ (\bpim) produced in charged current neutrino 
323: interactions in the NOMAD detector. In Sec.~\ref{nomad} 
324: and Sec.~\ref{data} we describe the experimental apparatus and the data treatment.
325: In Sec.~\ref{invslope} we report the invariant cross section
326: distributions of \bp and \bpim. In Sec.~\ref{kine} we discuss the kinematical
327: properties of events with a backward particle. In Sec.~\ref{bpy} we give the
328: production rates and study their dependence on the atomic number using the 
329: results of other neutrino and hadron experiments as well. 
330: Finally in Sec.~\ref{data&model} we compare the data with the predictions 
331: of theoretical models.
332: 
333: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
334: \section{NOMAD detector and neutrino beam}
335: \label{nomad}
336: For the study of backward particles presented here, the most important 
337: component of the NOMAD detector is the active target. 
338: Its main features are highlighted here, 
339: while a detailed description of the full detector (shown in Fig.~\ref{detector}) 
340: can be found in Ref. \cite{nom1} and \cite{nom2}. \\
341: The target was designed with two conflicting requirements. It had to be 
342: as light as possible in order to minimize photon conversions, 
343: multiple scattering and secondary hadronic interactions, 
344: and it had to be as heavy as possible in order 
345: to produce a large number of neutrino interactions.\\
346: This conflict was resolved using an active target (2.7 tons) of 44
347: drift chambers (DC), $3\times3 m^{2}$ each \cite{nomad_dc}, 
348: perpendicular to the beam axis. The target mass is provided by the chamber 
349: structure having an average density of $0.1~g/cm^{3}$.
350: The drift chamber composition is reported in Table~\ref{table:dc_comp}, 
351: where it can be seen that carbon and elements with nearby atomic numbers
352: represent over 90\% of the total weight.
353: For  this reason we consider this study as a measurement of backward production
354: in $\nu_{\mu} C$ interactions.  
355: The chambers are placed inside a $0.4~T$ magnetic field and provide a momentum 
356: resolution which can be parametrized as:
357: 
358: \begin{displaymath}
359: {\sigma_{P} \over P} = {0.05 \over \sqrt L} \oplus {0.008 P \over \sqrt{L^5} }
360: \end{displaymath}
361: 
362: \noindent where $L$ is the track length in meters and $P$ the momentum in $GeV/c$.
363: The first term is the contribution from multiple scattering and the second 
364: comes from the single hit resolution.
365: 
366: \begin{figure}[htb]
367: \begin{center}
368: \vskip -0.5cm
369:    \epsfig{file=fig/nomad3.eps,angle=-90,width=120mm}
370: \end{center}
371: \caption{{\it Side view of the NOMAD detector. }}
372: \label{detector}
373: \end{figure}
374: 
375: \begin{table}[htb]
376: \centering
377: \begin{tabular}{crrrr}
378: \hline
379: Element & Z   & Weight (\%)  & p (\%) & n (\%) \\ 
380: \hline
381: H/D & 1  & 5.14  & 5.09  & 0.05 \\
382: C   & 6  & 64.30 & 32.12 & 32.18 \\
383: N   & 7  & 5.92  & 2.96  & 2.96 \\
384: O   & 8  & 22.13 & 11.07 & 11.07 \\
385: Al  & 13 & 1.71  & 0.82  & 0.89 \\
386: Si  & 14 & 0.27  & 0.13  & 0.14 \\
387: Cl  & 17 & 0.30  & 0.14  & 0.16 \\
388: Ar  & 18 & 0.19  & 0.09  & 0.10 \\
389: Cu  & 29 & 0.03  & 0.01  & 0.02 \\ \hline
390: &   &       & 52.43 & 47.56 \\ \hline
391: \end{tabular}
392: \caption{{\it NOMAD drift chamber composition, showing the
393:  proportions by weight of atoms, and of protons and neutrons. As can be 
394: seen the target consists mainly of carbon and is practically isoscalar.}}
395: \label{table:dc_comp}
396: \end{table}
397: 
398: \noindent The target is followed by a transition radiation detector
399: (TRD) \cite{trd}, a preshower (PRS) and an electromagnetic
400: calorimeter (ECAL) \cite{ecal}. A hadron calorimeter and two muon
401: stations are located just after the magnet coil. The neutrino interaction
402: trigger \cite{trigger} consists of a coincidence between two planes of
403: counters located downstream of the active target, in the absence of a signal from
404: a large area system of veto counters upstream of the NOMAD detector.\\
405: \noindent The CERN-SPS wide band neutrino beam is produced by $450~GeV/c$ 
406: protons incident on a beryllium target. Neutrinos are produced in the 
407: decay of secondary pions and kaons in a $290~m$ long decay tunnel at 
408: an average distance of $625~m$ from the detector. The relative beam 
409: composition is predicted to be 
410: $ \nu_{\mu} : \bar\nu_{\mu} : \nu_{e}: \bar\nu_{e}= 1.00:0.0612:0.0094:0.0024$ 
411: with average energies of  23.5, 19.2, 37.1 and 31.3 $GeV$ 
412: respectively \cite{gianmaria}.
413:  The average energy of muon neutrinos interacting in the 
414: apparatus is about 41 $GeV$.\\
415: \noindent Referring to Fig.~\ref{detector} the coordinate system adopted
416: for NOMAD has the $X$-axis into the plane of the figure, the $Y$-axis 
417: upwards and the $Z$-axis horizontal, approximately
418: along the direction of the neutrino beam, which points upward, at an angle of
419: $2.4^{\circ}$ with respect to the $Z$-axis. Its origin is in the center of 
420: the front face of the first DC along the beam.
421: 
422: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
423: \section{The data sample}
424: \label{data}
425: 
426: \subsection{Event selection}
427: \label{evsel}
428: This study is based on the NOMAD full data sample collected between 
429: 1995 and 1998. 
430: In this analysis the event selection requires a primary vertex with at 
431: least 2 tracks inside the fiducial area defined by 
432: $\vert X \vert <130~cm$ and $ -125< Y <135~cm$. Along the $Z$ direction
433: a cut $Z_{VMIN} < Z < 400~cm$ is imposed. $Z_{VMIN}$ is $5~cm$ for the  1995
434: and 1996 data and is $35~cm$ for the data collected in 1997 and 1998. 
435: In the second period
436: the first DC module was removed to install NOMAD--STAR \cite{star}
437: and the fiducial volume was slightly reduced. \\
438: To select a \numucc event a negative $\mu$ attached to the primary vertex, 
439: identified by the muon stations and having a momentum of at least $3~GeV/c$,  
440: was required. Under these conditions the total \numucc sample consists of 944019 events.
441: 
442: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
443: \subsection{Track selection}
444: \label{tracksel}
445: Only tracks attached to the primary vertex 
446: are used in the search for backward particles. The track is required to 
447: have at least 8 DC hits,
448: %% for an acceptable uncertainty on the momentum reconstruction, 
449: the distance of the first hit from the primary vertex
450: to be less than 15 cm, and the relative error on the reconstructed 
451: momentum to be less than 0.3.
452: For the \bp search, since the identification method makes use of the
453: momentum--range relation, tracks are required to range out in the DC volume.
454: To avoid escaping tracks we require in addition the 
455: last hit of the track to be in a reduced fiducial volume defined by
456: $\vert X \vert <100~cm $, $\vert Y \vert < 100~cm $ and $Z_{MIN} < Z < 375~cm$, 
457: where $Z_{MIN}=30~cm$ for the 1995 and 1996 data sets and  $Z_{MIN}=60~cm$ for the rest. 
458: Finally no secondary vertex must be found close to the last hit of the track.
459: 
460: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
461: \subsection{MC sample}
462: \label{mc}
463: In order to estimate proton and pion reconstruction efficiencies and 
464: background we use a large sample of \numucc Monte Carlo (MC) events.
465: These events are generated using a modified version of LEPTO 6.1 \cite{lepto}
466: and JETSET 7.4 \cite{jetset}.
467: The target nucleon motion is simulated using the Fermi momentum distribution 
468: proposed by Bodek and Ritchie \cite{bodek}.
469: The generated events undergo a full detector simulation based on 
470: GEANT \cite{geant} and are subsequently reconstructed.
471: MC events and tracks are selected in the same way as are experimental
472: data yielding a final MC sample consisting of $\simeq 2.5 \times 10^6$ events.
473: The MC data used for this analysis do not include the simulation of 
474: ``nuclear effects'' such as the intranuclear cascade or correlations
475: \cite{nom2}.
476: When needed, specific correction procedures to the MC events 
477: are applied using the experimental data distributions (see Sec.~\ref{datacorr}).
478: 
479: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
480: \subsection{Backward proton identification}
481: \label{bp_id}
482: 
483: Fig.~\ref{lenvsp} shows the experimental distributions 
484: of length vs. momentum for positive (left) and negative (right) 
485: tracks, going backward with respect to the beam direction, and satisfying
486: our selection criteria. 
487: 
488: \begin{figure}[htb]
489: \begin{center}
490:  \vskip -0.5cm
491:  \epsfig{file=fig/lenvsp.eps,width=120mm}
492: \end{center}
493: \vskip -.2cm
494: \caption{{\it Distributions of length vs. momentum for positive 
495: (left) and negative (right) backward going tracks selected as described
496: in the text. The line indicates the position of the cut (Eq.~\ref{lcut}). }}
497: \label{lenvsp}
498: \end{figure}
499: 
500: \noindent Two distinct populations are clearly visible on the positive sample plot.
501: Protons, having a shorter range than $\pi^+$, tend to accumulate in the
502: lower right part of the plot while the $\pi^+$'s tend to populate 
503: the left--hand side. Comparing the two plots we see that the lower right
504: part of the negative sample is much less populated than the corresponding 
505: region of the positive one, these tracks being mainly $\pi^-$. 
506: The separation between protons and pions
507: is also visible in Fig.~\ref{two_peak} where the momentum distribution of
508: positive (dots) and negative (solid line) backward tracks is shown for 
509: different intervals of track length. 
510: In this figure the heights of the negative distributions have been 
511: normalized to the positive pion peaks.\\
512: \noindent We identify as a proton any positive 
513: backward going track which passes our selection cuts and has length $L$:
514: \be
515: L \le 2000~(P_{REC}+0.150)^{3.6}~~cm
516: \label{lcut}
517: \ee
518: 
519: \noindent where $P_{REC}$ is the reconstructed track momentum in $GeV/c$.
520: The cut position was optimized by using the MC results of forward tracks
521: and is shown in Fig.~\ref{lenvsp}.
522: 
523: \begin{figure}[htb]
524: \begin{center}
525:  \vskip -0.5cm
526:  \epsfig{file=fig/lenvsp_slice.eps,width=120mm}
527: \end{center}
528: \vskip -.5cm
529: \caption{{\it Reconstructed momentum distributions for positive 
530: (dots) and negative (solid line) backward going tracks for different intervals
531: of the track length. The heights of the negative distributions have been 
532: normalized to the positive pion peaks.  }}
533: \label{two_peak}
534: \end{figure}
535: 
536: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
537: \subsection{Data corrections for the \bp analysis}
538: \label{datacorr}
539: The reconstruction procedure used the energy loss of a pion for the track
540: fit. Instead of refitting proton candidate tracks with the correct energy 
541: loss, an empirical correction was applied to the reconstructed proton momentum
542: $P_{REC}$. This correction, which was obtained from MC events, amounts to
543: $+180~MeV/c$ at $P_{REC}=200~MeV/c$ and becomes negligible above $500~MeV/c$.
544: This procedure gives the true proton momentum $P$.\\
545: The raw number $N_{p}^{raw}$ of identified protons in a given momentum
546: bin must be corrected  for reconstruction efficiency $\epsilon_{REC}$,
547: stopping efficiency $\epsilon_{STOP}$, identification efficiency $\epsilon_{ID}$
548: and pion contamination in the data $\pi^{+~data}_{cont}$. All these
549: corrections will be described below. They are functions of either the
550: reconstructed or the true proton momentum.
551: Since the MC used for this analysis does not properly account for 
552: \bp production, some of these efficiencies can only be obtained 
553: by correcting the MC efficiencies with the data, or directly from data
554: themselves. The ``true'' number $N_{p}$ of protons is then:
555: \begin{equation}
556: N_{p} ~=~ N_{p}^{raw} ~~~~{1 \over \epsilon_{REC}}~~~~{1 \over \epsilon_{STOP}} 
557: ~~~~{1 \over \epsilon_{ID} }~~~(1 - \pi^{+~data}_{cont} )
558: \label{p_corr}
559: \end{equation}
560: \noindent with an average overall correction factor of 3.9~.
561: \vskip 0.1 cm
562: {\em a) Reconstruction efficiency $\epsilon_{REC}$ }\\
563: This is the ratio between the number of 
564: reconstructed and generated protons in a specific bin of the true 
565: momentum $P$ and angle $\theta$ (measured with respect 
566: to the beam direction). Since this quantity depends only on the detector
567: geometry and composition, we assume that the predictions obtained using
568: forward going protons are applicable to the backward ones with the
569: replacement $\cos\theta \rightarrow -\cos\theta$. An average value of 0.48
570: is obtained for $\epsilon_{REC}$, the distribution of which is 
571: shown on the left in Fig.~\ref{effric}.
572: \vskip 0.05 cm
573: {\em b) Stopping efficiency $\epsilon_{STOP}$}\\
574: This is the ratio between the number of backward protons fully 
575: contained in the target ($Bp_{STOP}$) and all the reconstructed ones (\bp). 
576: Its average value is 0.42 in the momentum range used in this analysis.\\
577: \begin{figure}[htb]
578: \begin{center}
579:  \epsfig{file=fig/effric_plot.eps,width=120mm}
580: \end{center}
581: \vskip -.5cm
582: \caption{{\it Reconstruction efficiency (as described in the text) as a 
583: function of momentum and angle between the 
584: track and the nominal beam direction. 
585: The plot on the left is for protons, the one on the right for $\pi^{-}$. }}
586: \label{effric}
587: \end{figure}
588: \noindent The protons are extracted from the sample of positive backward tracks by 
589: subtracting the backward $\pi^+$ content. 
590: The stopping efficiency $\epsilon_{STOP}$ is computed from the following 
591: quantities:
592: \bmath
593: \epsilon_{STOP} = {Bp_{STOP} \over Bp} = 
594: {B^+_{STOP}~-~~B\pi^{+}_{STOP} \over B^+~-~B\pi^{+}}
595: \emath
596: \noindent where $B^+$ ($B^+_{STOP}$) refer to 
597:  backward (backward contained) positive tracks. $B\pi^{+}$ and $B\pi^{+}_{STOP}$ 
598:  are the analogous quantities for backward positive pions.
599: They are evaluated from the data as follows.\\
600: In the NOMAD isoscalar target the length distributions of $\pi^+$ and $\pi^-$,
601: in each momentum bin, are the same apart from a scale factor. Therefore
602: the number of backward $\pi^{+}$ can be computed by the number of backward
603: $\pi^{-}$ once the relative population of $B\pi^{+}$ and $B\pi^{-}$
604: is known. The ratio of the two populations is measured from a clean sample
605: of pions obtained by selecting tracks with a sufficiently longer length than the
606: corresponding proton range.
607: This ratio has comparable values for contained and not contained
608: pions. It is almost constant in the momentum range of interest with
609: a weighted average of $R~=~1.82\pm0.05$.\\
610: The pion content in the backward sample is therefore:\\
611: \bmath
612: B\pi^+ ~=~ R \times B\pi^{-}~~~;~~~B\pi^{+}_{STOP} ~=~ R \times B\pi^{-}_{STOP}
613: \emath
614: \noindent where the negative pion sample ($B\pi^{-}$, $B\pi^{-}_{STOP}$) is
615: taken to be the sample of negative (negative contained) backward tracks.
616: This relies on the fact that the MC simulation shows that the $e^{\pm}$ 
617: contamination in the backward region is at the level of a few percent above 
618: $\approx 200~MeV/c$.
619: \vskip 0.05 cm
620: {\em c) Identification efficiency $\epsilon_{ID}$ }\\
621: This is the fraction of \bp with track length smaller than the length 
622: cut. Above $P_{REC} \simeq 300~MeV/c$ this quantity is $1$.
623: The results of forward going protons are applicable also in this case.
624: \vskip 0.05 cm
625: {\em d)  Pion contamination $\pi^{+~data}_{cont}$}\\ 
626: This is the fraction of  $\pi^+$'s in the sample of identified protons,
627: which amounts to  $\approx 8\% $ above $P_{REC} \simeq 250~MeV/c$.
628: To estimate this contamination in the data, the MC prediction ($\pi^{+~MC}_{cont}$) 
629: for forward particles has been corrected using the backward data 
630: distributions as follows. 
631: For each momentum bin the length distributions of protons
632: in the data and MC can differ in normalization but not in shape 
633: since the track length depends on the mechanism of energy loss which is well 
634: reproduced by the MC. The same applies to $ \pi^{+}$.
635: The length cut applied on the momentum--length plane (Eq.~\ref{lcut})
636: defines in the data and MC two populations: the $\pi$--like tracks, (the tracks
637: above the cut) and the $p$--like tracks (those below the cut). 
638: To account for the difference in the populations 
639: of protons and  pions in the data and MC the quantity $ \pi^{+~MC}_{cont}$
640: has been weighted by the double ratio of the $\pi$--like tracks (data/MC)
641: and $p$--like tracks (data/MC):
642: \bmath
643:  \pi^{+~data}_{cont}~=~{ N_{\pi^{+}}^{data} \over N_{p}^{data} }~=~
644:  { N_{\pi-like}^{data} / N_{\pi-like}^{MC} \over 
645:    N_{p-like}^{data} / N_{p-like}^{MC} }~~ \pi^{+~MC}_{cont} 
646: \emath
647: \noindent The above correction is evaluated as a function of the
648: reconstructed momentum.\\
649: 
650: 
651: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
652: \subsection{Backward $\pi^-$ identification}
653: \label{bim_id}
654: 
655: \noindent As already pointed out the $e^-$ component is negligible in 
656: the negative backward track sample. 
657: We therefore assume that any negatively, backward going, 
658: charged track with $P > 0.2~GeV/c$ is a $\pi^-$. 
659: The smallness of the contamination and the fact that it is not necessary 
660: to look for stopping tracks to identify the \bpim reduce the number of 
661: corrections to be applied to the data  to only one, the reconstruction 
662: efficiency $\epsilon_{REC}^{~\pi^{-}}$.For this quantity we use the MC 
663: predictions obtained for forward going $\pi^-$.\\
664: The true number of negative pions is then obtained from the raw number of 
665: identified negative pions $N_{\pi^-}^{raw}$ in a given momentum bin 
666: according to:
667: 
668: \begin{equation}
669: N_{\pi^{-}} ~=~ N_{\pi^-}^{raw} ~~~~{1 \over \epsilon_{REC}^{~\pi^{-}} }
670: \label{pim_corr}
671: \end{equation}
672: 
673: \noindent In this case the average value of the correction is 2.
674: 
675: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
676: \section{Backward $p$ and $\pi^-$ invariant momentum distributions}
677: \label{invslope}
678: 
679: \noindent The inclusive spectrum of backward particles is typically represented 
680: using the normalized invariant cross section $ (1/\sigma_{TOT})~E~d^3\sigma/dP^3$,
681: where $E$ is the energy of the backward going particle. The invariant
682: cross section is usually \cite{matsinos} parametrized by an exponential form as:\\
683: 
684: \be
685: {1 \over N_{ev} }~{E \over P}~{dN \over dP^{2}}~=~C~e^{-BP^{2}}
686: \label{bparam}
687: \ee
688: 
689: \noindent where $N_{ev}$ is the total number of events, 
690: $C$ is a constant and $B$ is the slope parameter. In previous experiments 
691: $B$ has been found to be almost independent of the projectile type and momentum, 
692: and of the atomic number of the target.
693: This behaviour has been termed ``{\em nuclear scaling}'' \cite{fredriksson}.\\
694: The inclusive spectrum of \bp and \bpim is shown in Fig.~\ref{slopes}a and 
695: \ref{slopes}b respectively, together with the exponential fit. 
696: The values of the measured slope parameter $B$ are reported in 
697: Table~\ref{table:slope_table}. The first and the second errors are statistical 
698: and systematical, respectively.
699: \begin{figure}[htb]
700: \begin{center}
701:    \epsfig{file=fig/slopes.eps,width=160mm}
702: \end{center}
703: \vskip -.5cm
704: \caption{{\it Invariant momentum distributions for backward going protons 
705: (a), $\pi^-$ (b) in $\nu_\mu CC$ events and protons in $\bar\nu_\mu CC$ events (c)}}
706: \label{slopes}
707: \end{figure}
708: 
709: \begin{table}[ht]
710: \centering
711: \begin{tabular}{llcll}
712: \hline
713:                  &          & $\Delta P~(GeV/c)$ & $~~~B (c^{2}/GeV^{2})$ & $~~~C (c^3/GeV^2)$  \\ \hline
714: Backward $p$     & \numucc  & $0.37 \div 0.70$   & $10.54\pm0.20\pm0.5$ & $4.08\pm0.19\pm0.5$ \\ 
715:                  & \anumucc & $0.37 \div 0.70$   & $10.79\pm0.78$       & $2.71\pm0.54$       \\ 
716: Backward $\pi^-$ & \numucc  & $0.32 \div 0.85$   & $10.03\pm0.28\pm0.3$ & $0.17\pm0.01\pm0.02$\\ \hline \\
717: \end{tabular}
718: \caption{{\it Fit ranges and values of the slope parameter $B$, for backward protons and
719: $\pi^{-}$, as obtained from the exponential fit to the invariant cross section.
720: The first and the second errors are statistical and systematical, respectively.}}
721: \label{table:slope_table}
722: \end{table}
723: \noindent The systematic uncertainty was estimated by slightly changing
724: the values of all the cuts used, by varying by a small amount the correction functions
725: and also by changing the size of the fiducial volume. The values of the slope parameters
726:  measured in this experiment are found to be compatible with the
727: results obtained in other neutrino (see Fig.~\ref{bnu} for the \bp case) and
728: hadronic experiments.
729: \begin{figure}[htb]
730: \begin{center}
731:    \epsfig{file=fig/bnu.eps,width=120mm}
732: \end{center}
733: \vskip -0.5cm
734: \caption{{\it 
735: The value of the slope parameter $B$ obtained in neutrino experiments. 
736: For NOMAD the statistical and systematical error have been added in quadrature. 
737: The full circles are the results from $\nu_{\mu}~CC$, 
738: the open circles from \anumucc and  the cross is for \numunc interactions. 
739: For BEBC $\nu$ and $\bar\nu$ results have been combined.}} 
740: \label{bnu}
741: \end{figure}
742: \noindent The invariant cross section for \bp is larger than the one for
743: \bpim by about one order of magnitude but the values of the slopes are
744: similar. The kinematic ranges of the two invariant distributions are also different.
745: To be identified \bp have to stop inside the target volume; the rather low
746: density of the NOMAD target (see Sec.~\ref{nomad}) restricts to $\approx 0.7~GeV/c~$
747: the maximum useful momentum value.\\
748: \noindent The invariant cross section and slope parameter for \bp in \anumucc events 
749: are given in Fig.~\ref{slopes}c and Table~\ref{table:slope_table} respectively.
750: During normal operations \anumucc events were also collected due to the small
751: $\bar\nu_{\mu}$ component of the dominantly $\nu_{\mu}$ beam \cite{nom1}.
752: A dedicated $\bar\nu$ run yielded an additional sample of \anumucc events
753: included for this analysis.\\
754: Antineutrino events are selected under the assumptions that the efficiencies and 
755: pion contamination are the same as those used for the neutrino events,
756: but requiring a positively charged muon instead of a negative one.
757: The final sample consists of $61134$ events containing 1764~\bp.
758: 
759: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
760: \subsection{Energy dependence of the slope parameter}
761: 
762: \noindent
763: In Fig.~\ref{behad} we show the slope parameter $B$ for protons as a function
764:  of the visible hadronic energy $E_{HAD}$ and the $Q^2$ of the event.
765: The visible hadronic energy is defined as: 
766: \bmath
767: E_{HAD}= \sum E_{c} + \sum E_{n} 
768: \emath
769: \noindent where $\sum E_{c}$ is the sum of the energies of reconstructed charged tracks
770: (assuming the mass of the pion if the particle type is not explicitly identified).
771: The sum includes all the charged tracks attached to the
772: primary vertex by the reconstruction program, those belonging to a 
773: pointing $V^{0}$, the ``close hangers'' i.e. 
774: isolated tracks not associated to the primary nor to a secondary vertex
775: and having their first hit in a box around the primary vertex of size
776: $\vert X \vert$, $\vert Y \vert <10~$ cm and $-5<Z<20~$ cm, and the ``pointing
777: hangers'' i.e. the hangers which have an impact parameter of less than 10 cm
778: when linearly extrapolated back to the z-plane of the primary vertex.
779: $\sum E_{n}$ includes tracks associated to secondary vertices corresponding 
780: to interactions of neutral particles and the energy of neutral particles 
781: reconstructed in the ECAL. The reconstructed neutrino energy $E_{vis}$
782: is taken to be the sum of the muon energy $E_{\mu}$ and of $E_{HAD}$.
783: The square of the four--momentum transfer $Q^2$ is 
784: $Q^2 = 4~E_{vis}E_{\mu}sin^2\theta/2$, where $\theta$ is the muon angle 
785: with respect to the neutrino direction.
786: \begin{figure}[htb]
787: \begin{center}
788:    \epsfig{file=fig/behad.eps,width=120mm}
789: \end{center}
790: \vskip -.5cm
791: \caption{{\it 
792: The value of the slope parameter $B$ shown as a function  
793: of the hadronic energy $E_{HAD}$ (top) and of $Q^2$ (bottom). }}
794: \label{behad}
795: \end{figure}
796: \noindent No significant dependence of the slope $B$ on either quantity is observed, 
797: in agreement with the expectations of ``nuclear scaling'' as observed in hadronic 
798: beam experiments. The range covered by NOMAD is similar to the one covered by different
799: experiments with hadronic beams.
800: 
801: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
802: \subsection{Angular dependence of the slope parameter}
803: The angular dependence of the slope parameter $B$ is usually given with respect
804: to the direction of the exchanged boson, approximated by the hadronic jet direction
805: \cite{matsinos}. This allows a comparison with the results obtained in 
806: hadron beam experiments. The angular dependence of $B$ 
807: for \bp and \bpim is shown in Fig.~\ref{bcosthe} as a function of the angles with 
808: respect to  the beam and to the hadronic jet direction. 
809: Although these two angles are highly correlated, there are small differences in 
810: the slope values. If the beam direction is used $B$ is systematically smaller by 
811: $\approx 0.5\div1 (GeV/c)^{-2}$, in the \bp case, while for \bpim the difference is less 
812: pronounced. As can be seen from Fig.~\ref{bcosthe}, 
813: larger values of $B$ are preferred at increasingly backward directions.
814: This behaviour has already been observed in neutrino \cite{matsinos}, 
815: photon \cite{alanakyan} and hadron \cite{bayukov,komarov,yulda}
816: nucleus experiments.
817: \begin{figure}[htb]
818: \begin{center}
819:    \epsfig{file=fig/bcosthe.eps,width=120mm}
820: \end{center}
821: \vskip -.5cm
822: \caption{{\it 
823: Angular dependence of the slope parameter $B$ for $p$ (left) and $\pi^{-}$ (right).
824: The value of $B$ is shown as a function of the particle angle with respect to the 
825: hadronic jet direction (full circles) and with respect to the beam 
826: direction (open circles). }}
827: \label{bcosthe}
828: \end{figure}
829: 
830: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
831: \section{Kinematical properties of events with a backward particle}
832: \label{kine}
833: 
834: Table~\ref{table:kine_table} shows a comparison of the average values of some
835: kinematical quantities for \numucc events with and without \bp or $B\pi^{-}$.
836: $<C>$ and $<N^{\pm}>$ are the average event charge and charged multiplicity 
837: ($\mu^{-}$ included), $\nu = E_{vis}-E_{\mu}$, $x = Q^{2}/2M\nu$ is the 
838: Bjorken scaling variable and $W^{2} = M^{2}-Q^{2}+2M\nu$ is the square of the 
839: invariant mass of the hadronic system, $M$ being the nucleon mass.\\
840: \begin{table}[htbp]
841: \centering
842: \begin{tabular}{crrr}
843: \hline
844:              & CC~(no~$Bp$, no~$B\pi^-$) & CC~($Bp$) & CC~($B\pi^-$) \\ \hline
845: $<C> $       &  0.46              &  2.11$\pm$0.02    &   0.08$\pm$0.02 \\ 
846: $<N^{\pm}>$  &  5.13              &  6.63$\pm$0.03    &   7.28$\pm$0.03 \\ 
847: $<E_{vis}>$  & 40.75              & 38.9$\pm$0.4      &  39.5$\pm$0.5   \\ 
848: $<E_{HAD}>$  & 12.32              & 11.4$\pm$0.2      &  11.7$\pm$0.2   \\ 
849: $<Q^{2}>  $  &  7.07              &  5.5$\pm$0.1      &   5.9$\pm$0.1   \\ 
850: $<x>      $  &  0.32              &  0.26$\pm$0.002   &   0.28$\pm$0.003\\ 
851: $<W^{2}>  $  & 17.19              & 16.4$\pm$0.2      &  17.5$\pm$0.3   \\ \hline
852: \end{tabular}
853: \caption{\it {Comparison of average values of some kinematical variables in events
854: with and without an identified \bp or \bpim. The values of $<C>$ and of 
855: $<N^{\pm}>$ are corrected for the track reconstruction efficiency.
856: Only statistical errors are shown. They are negligible for the first data column.}}
857: \label{table:kine_table}
858: \end{table}
859: \noindent For events with a backward particle the average values of the
860: charged multiplicity are consistent with the presence of 
861: extra tracks as expected in both models. 
862: The average value of the event charge in the \bp sample ($<C>=2.11$) is 
863: larger than in the no~$Bp$, no~$B\pi^-$
864: samples as measured in the data ($<C>=0.46$). It is also larger than in the total 
865: MC sample where no nuclear effects are present ($<C>=0.31$). 
866: This observed increase is not entirely due to the bias introduced by demanding 
867: a positive particle, since an average value of $<C>=1.21$ is obtained when 
868: requiring a backward positive track in the MC events.
869: This can be expected from both mechanisms outlined in the Introduction.
870: In the collisions of secondary hadrons with nucleons inside the
871: nucleus the total event charge will increase by one unit in interactions 
872: off protons while no extra charge will be produced when scattering on neutrons.
873: Also in the framework of the short range correlation model (in its simplest form of
874: two--nucleon correlations) the mechanism of pair breaking predicts extra
875: tracks in the final state. The mechanism is of course symmetric for $p$ and $n$
876: but backward going neutrons are not detected.\\
877: \noindent The events with a \bpim show an average charge smaller by $\simeq 0.4$
878: units with respect to the value where no \bpim is present.
879: Here again this effect cannot be entirely interpreted as the result of the
880: bias associated with the requirement of at least one negative backward track, 
881: since the MC in this case predicts $<C>=-0.41$.
882: Therefore also in the sample of events with a \bpim  an increase in the 
883: average value of the event charge is observed.\\
884: From Table~\ref{table:kine_table} it appears that the average values of the 
885: kinematical variables, other than $<C>$ and $<N^{\pm}>$, for events with either 
886: a \bp or a \bpim are systematically lower than for events without identified 
887: backward particles.
888: 
889: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
890: \section{Backward particle rates}
891: \label{bpy}
892: 
893: The \bp rate has been compared with the results
894: of the other $\nu$--nucleus experiments in order to study a
895: possible atomic number dependence. In NOMAD, where \bp are identified in the momentum
896: interval from $370 MeV/c$ to $700 MeV/c$ we obtain, after correction,  the 
897: number of events reported in Table~\ref{table:nombpy_table}.
898: 
899: \begin{table}[ht]
900: \centering
901: \begin{tabular}{crr}
902: \hline
903:  N. of \bp or \bpim & N. of \bp ev. & N. of \bpim ev.\\ \hline
904:   0              & 904212        & 939617\\ 
905:   1              &  37634        &   4238\\ 
906:   2              &   2168        &    164\\ 
907:   3              &      5        &      0\\ \hline
908: \end{tabular}
909: \caption{\it {The number of events, after correction for efficiencies,
910:  as a function of the multiplicity of backward protons 
911:  in the momentum interval $370\div700 MeV/c$, and of backward negative pions
912:  in the momentum interval $350\div800 MeV/c$.}}
913: \label{table:nombpy_table}
914: \end{table}
915: \noindent These figures correspond to an average number of 
916: \bp per event, $<N_{Bp}> = (44.5 \pm 0.5)\times 10^{-3}$,
917: where the error is statistical only.
918: \begin{table}[htbp]
919: \centering
920: \begin{tabular}{lrrrrr}
921: \hline
922:                      &       &                &             &                &                  \\
923: Experiment           & $<A>$ & $\Delta~P$~~~  & $<E_{\nu}>$ & $ <N_{Bp}> $   & $ <N_{B\pi^-}> $ \\ 
924:                      &       & (GeV/c)        & (GeV)       & ($\times 10^{-3}$)~~ & ($\times 10^{-3}$)~~\\ \hline
925: BEBC $D_2$ \cite{d2} & 2     & 0.35$\div$0.80 & 50          & 6.3   $\pm$ 0.7 & 1.0 $\pm$ 0.3 \\ 
926: NOMAD                & 12    & 0.37$\div$0.70 & 41          & 44.5  $\pm$ 0.5 & 4.8 $\pm$ 0.1 \\ 
927: BEBC $Ne$ \cite{matsinos} & 20 & 0.35$\div$0.80 & 50        & 97.7  $\pm$ 3.7 & 6.8 $\pm$ 0.9 \\ 
928: 15foot \cite{berge}  & 20    & 0.20$\div$0.80 & 50          & 137.5 $\pm$ 15  & \\ 
929: SKAT \cite{skat}     & 52    & 0.32$\div$0.70 & 10          & 185.8 $\pm$ 10  & \\ 
930: E--564 \cite{dayon}  & 80    & 0.30$\div$0.80 & 60          & 331.1 $\pm$ 47  & \\ \hline \\
931: \end{tabular}
932: \caption{\it {Comparison of \bp and \bpim rates for neutrino-nucleus 
933: experiments. The average mass number $<A>$, the momentum interval 
934: $\Delta P$ used in the analysis, the average event energy $<E_{\nu}>$ and the  
935: \bp and \bpim rates are listed for each experiment.
936: The errors are purely statistical. }}
937: \label{table:bpy_other}
938: \end{table}
939: \noindent The backward proton yield obtained in other $\nu$--nucleus experiments is
940: shown in Table~\ref{table:bpy_other}.
941: The values, which were extracted from the 
942: original references, are not directly comparable because
943: of the different momentum intervals used in the different experiments.
944: \noindent In order to study the $A$ dependence the \bp yields measured in 
945: NOMAD and E--564 \cite{dayon} were extrapolated to the interval from 350 to 800 MeV/c  
946: assuming the dependence given in Eq.~\ref{bparam} with the measured slopes.\\
947: The extrapolation for NOMAD gives:
948: 
949: \begin{displaymath}
950: <~N_{Bp}>_{350\div800 MeV/c}~ = 
951:  \left[ \rule{0mm}{5mm} ~52.8 \pm 0.6 (stat.) \pm 7 (syst.)~ \right] \times 10^{-3} 
952: \end{displaymath}
953: 
954: For E--564 we obtain $<~N_{Bp}>=(234\pm33) \times 10^{-3}$.\\
955: The $A$ dependence for experiments most directly comparable to NOMAD
956: is shown in Fig.~\ref{bpyield}a. In the range $A=20\div80$
957: it has been parametrized as $ <N_{BP}>\propto A^{\alpha}$, 
958: where $\alpha=0.68\pm0.12$ \cite{dayon}.
959: The same parametrization with a similar power law was found to describe
960: \bp production in $\pi^{+}$ and $K^{+}$ collisions with $Al$ and $Au$ 
961: nuclei at 250 GeV/c \cite{na22}.
962: It is evident from Fig.~\ref{bpyield}a that this simple power law does not
963: describe the NOMAD data taken at a lower $<A>$.
964: \begin{figure}[htb]
965: \begin{center}
966:    \epsfig{file=fig/bpyield.eps,width=130mm}
967: \end{center}
968: \vskip -.5cm
969: \caption{{\it The average number of \bp per event (a) and of \bpim per event (b)
970: in the momentum range from  $350\div800 MeV/c$ 
971: in neutrino experiments as a  function of the mass number $A$.
972: The line shown in (b) is the result of the fit described in the text. }}
973: \label{bpyield}
974: \end{figure}
975: \noindent The \bpim rate was directly measured in the same momentum range
976: used for the \bp analysis. Its average value is found to be: 
977: \begin{displaymath}
978: <~N_{B\pi^-}>_{350\div800 MeV/c}~ = 
979:  \left[ \rule{0mm}{4mm}  ~4.8 \pm 0.1 (stat.) \pm 0.3 (syst.) \right] \times 10^{-3} 
980: \end{displaymath}
981: \noindent In this case a good fit to the two BEBC points and to the NOMAD value 
982: can be obtained using the form $A^{\alpha}$ giving $\alpha=0.83\pm0.25$, 
983: as shown in Fig.~\ref{bpyield}b.
984: 
985: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
986: \section{Comparison of data with models}
987: \label{data&model}
988: \subsection{$E_{HAD}$ and $Q^2$ dependence of \bp and \bpim rates}
989: 
990: The \bp and the \bpim rates have been studied as a function of the hadronic energy
991: $E_{HAD}$ and of $Q^2$. In both cases, shown in Fig.~\ref{bpy_ehad}  
992: and \ref{bpimy_ehad} a decrease of the yield with increasing 
993: $E_{HAD}$ and $Q^2$ is observed.
994: As pointed out in the Introduction, this can be interpreted in terms of the 
995: ``formation zone'' concept. 
996: The larger $E_{HAD}$ and $Q^2$, the larger is the average energy of the outgoing 
997: partons therefore resulting in hadrons which have higher probability to
998: be formed outside the nucleus. 
999: As a consequence, reinteractions will decrease and so will the slow proton rates.
1000: This $E_{HAD}$ and $Q^2$ dependence is also consistent with the decrease of the 
1001: average values of $E_{HAD}$ and $Q^2$ in events with identified \bp as shown 
1002:  in Table~\ref{table:kine_table}.
1003: In the \bpim case the dependence of the yield on $Q^2$ and on $E_{HAD}$ is 
1004: less pronounced. This can be understood since the \bpim rates are a less 
1005: sensitive probe of nuclear effects because \bpim production on a stationary 
1006: nucleon is kinematically allowed for momenta up to about half the nucleon mass.  
1007: Furthermore \bpim can be produced in the decay of forward going resonances 
1008: or unstable particles. 
1009: \begin{figure}[htb]
1010: \vskip -0.5 cm
1011: \begin{center}
1012:    \epsfig{file=fig/bpy_ehad_new.eps,width=120mm}
1013: \end{center}
1014: \vskip -.3cm
1015: \caption{{\it The average number of \bp ($ 370<P<700 MeV/c $) 
1016: per event as a function of the hadronic energy $E_{HAD}$ (left) and of $Q^2$ (right).}}
1017: \label{bpy_ehad}
1018: \end{figure}
1019: \begin{figure}[htb]
1020: \begin{center}
1021:    \epsfig{file=fig/bpimy_ehad_new.eps,width=120mm}
1022: \end{center}
1023: \vskip -.2cm
1024: \caption{{\it The average number of \bpim ($ 350<P<800 MeV/c $)  
1025: per event as a function of the hadronic energy $E_{HAD}$ (left) and of $Q^2$ (right).}}
1026: \label{bpimy_ehad}
1027: \end{figure}
1028: 
1029: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1030: \subsection{Multiplicities of slow particles}
1031: \label{slow}
1032: \noindent Since mostly slow hadrons are produced inside the nucleus,
1033: they are predominantly the ones that can scatter off nucleons, producing 
1034: both slow protons and backward particles. Therefore, in the framework of 
1035: the rescattering model, a correlation between the number of emitted slow 
1036: protons and the multiplicity of mesons is expected  \cite{ishii}.\\
1037: The increase in the rate of events with slow protons as a function of
1038: hadron multiplicities was first observed by
1039: the E745 collaboration \cite{kitagaki} in a $\nu$-Freon
1040: experiment. This correlation was used \cite{ishii} to
1041: extract information on the formation zone of secondary hadrons.
1042: This effect was also observed by the BEBC collaboration\cite{guy}.\\
1043: \noindent We have observed this correlation by studying the 
1044: fraction of events with at least one \bp as a function of the 
1045: multiplicity of low momentum ($P<700 MeV/c$) positively or negatively
1046: charged hadrons (Fig.~\ref{slow_mult}). 
1047: In both cases, this fraction of events increases with
1048: increasing multiplicity of low momentum tracks.
1049: However, the positive hadronic multiplicity is biased by the presence
1050: of protons from rescattering itself and therefore it is less representative
1051: of the original multiplicity of positive slow hadrons from the neutrino 
1052: interaction. The true correlation can therefore be better studied as a function
1053: of the multiplicity of negative slow hadrons.
1054: \begin{figure}[htb]
1055: \begin{center}
1056:    \epsfig{file=fig/slow_mult_new.eps,width=120mm}
1057: \end{center}
1058: \vskip -.5cm
1059: \caption{{\it Fraction $F_{Bp}$ of events with at least one \bp
1060: as a function of the multiplicity of low--momentum ($P<700 MeV/c$)
1061: positively (left) or negatively (right) charged hadrons.}}
1062: \label{slow_mult}
1063: \end{figure}
1064: 
1065: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1066: \subsection{Backward $\pi^-$ spectra and the Fermi momentum tail}
1067: 
1068: As already pointed out in the Introduction, in the
1069: short range correlation models the spectrum of fast backward going
1070: hadrons reflects the tail of the Fermi momentum distribution.
1071: For this study we used \bpim because, as opposed to $Bp$, they do not 
1072: need to range out in the target thus yielding data up to larger momenta.\\
1073: To estimate qualitatively the contribution of this tail we have compared
1074: our \bpim data to the predictions of two different Fermi momentum 
1075: distributions \cite{bodek,benhar}. 
1076: \begin{figure}[htb]
1077: \begin{center}
1078:    \epsfig{file=fig/mcpion.eps,width=90mm}
1079: \end{center}
1080: \vskip -.5cm
1081: \caption{{\it 
1082: Invariant spectrum for \bpim in MC (open circles) and data (full circles). 
1083: The solid line is the curve of Ref.~\cite{benhar} superimposed with an 
1084: arbitrary normalization. }}
1085: \label{mcpion}
1086: \end{figure}
1087: The distribution of Ref.~\cite{bodek} has a long momentum tail up to $4~GeV/c$.
1088: In Ref.~\cite{benhar} the spectrum of backward going hadrons in lepton--nucleon 
1089: scattering is derived from the spectral function $p({\bf k},E)$, 
1090: which is the probability of finding a nucleon with momentum
1091: ${\bf k}$ and removal energy $E$. The function, whose integral over $E$ is the 
1092: Fermi distribution, is obtained from non relativistic many body 
1093: theory and a specific procedure has been developed to extrapolate it to 
1094: large values of ${\bf k}$. In this case the Fermi momentum tail is shorter and
1095: ends at about $1~GeV/c$ \cite{benhar_priv}.\\
1096: In Fig.~\ref{mcpion} the invariant spectrum of MC $B\pi^{-}$, simulated 
1097: using the Fermi distribution of Ref.~\cite{bodek}, is shown 
1098: together with the data.
1099: A clear disagreement is visible in the tail of the distributions,
1100: the MC  being much larger than the data above $\simeq 0.5~(GeV/c)^{2}$.
1101: Our results therefore do not support a Fermi momentum distribution
1102: with a long tail as proposed in Ref.~\cite{bodek} while they agree
1103: with the dependence predicted by Ref.~\cite{benhar}, as also 
1104: shown in Fig.~\ref{mcpion}.\\
1105: 
1106: 
1107: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1108: \subsection{Effects of short range correlations in \bp production}
1109: 
1110: According to the picture proposed in Ref.~\cite{fs1} \bp production 
1111: is explained as the result of a neutrino interaction within a correlated 
1112: cluster of two or more nucleons. 
1113: This cluster is formed, for a short time, when two nucleons in their motion 
1114: inside the nucleus approach each other so closely as to come under the effect
1115: of the short range component of the nuclear force ($r_c = 0.5 \div 0.7 fm$). 
1116: As a result, the high relative momenta of the correlated nucleon pair
1117: manifest themselves when the backward moving spectator is released in 
1118: the interaction of the incoming virtual $W$ with the forward going nucleon. 
1119: In this model, if the effects of reinteractions are neglected, the released 
1120: backward nucleon can leave the nucleus keeping its original momentum.
1121: These correlated pairs have recently been observed in the 
1122: reaction $e+^{16}O~\rightarrow~e^{'} p p~^{14}C $ at low values of 
1123: the energy transfer $(180\div240 MeV)$ \cite{starink}.\\
1124: To study these correlations it is customary to use the variable 
1125: $\alpha$ defined as:
1126: \be
1127: \alpha=(E~-P_l)/ M
1128: \label{eq_alpha}
1129: \ee
1130:  \noindent where $E$ and $P_l$ are respectively
1131: the energy and  longitudinal momentum of the \bp and 
1132: $M$ is the nucleon mass. For $Bp$, $\alpha > 1$ since $P_l$ is negative.
1133: In this model, due to the target motion a correlation between 
1134: the Bjorken scaling variable $x$ and the variable $\alpha$ is expected. 
1135: In particular the average $x$ for events with a \bp is expected to be smaller
1136: than in events where no \bp is present, as indeed observed in our data 
1137: (see Table~\ref{table:kine_table}). The same correlation is expected to 
1138: hold if the variable $v = xy$ is used, where $y = E_{HAD}/E_{vis}$.
1139: This variable is related to the muon kinematics and  can be written
1140: as : 
1141: \be
1142: v=(E_{\mu}-P_\mu^{l})/M
1143: \label{eq_v}
1144: \ee
1145: \noindent where $E_{\mu}$ and $P_\mu^{l}$ are the muon energy and longitudinal 
1146: momentum. Since, generally, in neutrino experiments the muon kinematic variables 
1147: are well measured, $v$ is better suited than $x$ for this purpose.\\
1148: \noindent The $(v,\alpha)$ correlations were searched for in the data
1149: by calculating for each event $\alpha$ and $v$ as defined in Eq.~\ref{eq_alpha} and
1150: \ref{eq_v}. For each $\alpha$ bin we plot the variable $<v_N>$ defined as:
1151: \be
1152: <v_N>~=~{<v>_{Bp} \over <v>_{no~Bp}}
1153: \ee
1154: \noindent where $<v>_{no~Bp}$ is obtained from the full sample of events 
1155: without a $Bp$. According to Ref.~\cite{fs1} the average values of 
1156: $v$ in events with a $Bp$, $<v>_{Bp}$, is related to the average value of $v$ in events 
1157: where no \bp is present, $<v>_{no~Bp}$, by:
1158: \be
1159: <v>_{Bp}~=~<v>_{no~Bp}~(2-\alpha)
1160: \ee
1161: 
1162: \noindent More generally for a cluster composed of $\xi$ nucleons
1163: the relation is \cite{multiq}:
1164: 
1165: \be
1166: <v>_{Bp}~=~<v>_{no~Bp} \left(1 - {\alpha \over \xi} \right)~ {\xi \over {\xi-1}}
1167: \label{v_cluster}
1168: \ee
1169: 
1170: \noindent Fig.~\ref{alpha}a shows  $<v_N>$ as a function of $\alpha$.
1171: The data indicate a slope of $-0.22\pm0.06$ with a $\chi^{2}/ndf=5.3/6$ 
1172: (the $\chi^{2}/ndf$ in the hypothesis of no dependence on $\alpha$ is 19.1/7). 
1173: The two--nucleon correlation mechanism ($\xi=2$, the most probable case when
1174: considering the overlapping probabilities of $\xi$ nucleons in the nucleus)
1175: fails to describe our data. Either higher order structures are playing
1176: a leading role \cite{multiq} or the observed low level of correlation is due
1177: to the presence of reinteraction processes. In this case part of the 
1178: \bp are emitted as a result of reinteractions in the nucleus and are not
1179: related to the target nucleon. The presence of intranuclear cascade processes
1180: could therefore dilute the existing correlation to the observed level.
1181: \begin{figure}[htb]
1182: \begin{center}
1183:    \epsfig{file=fig/alpha.eps,width=120mm}
1184: \end{center}
1185: \vskip -.5cm
1186: \caption{{\it The variable $<v_N>$ plotted as a function of $\alpha$. 
1187: The lines represent the predicted correlation (Eq.~\ref{v_cluster}) 
1188: for a number $\xi$ of nucleons in the cluster equal to 2, 3 and 4. 
1189: In (a) all the \bp events were used; 
1190: in (b) only events having a \bp with $cos{\theta_{j}} < 0$, $\theta_{j}$ 
1191: being the angle of the \bp with respect to the hadronic jet direction.}}
1192: \label{alpha}
1193: \end{figure}
1194: \noindent To test this hypothesis we tried to reduce the component 
1195: due to rescattering in the selected \bp sample. Having observed the 
1196: correlation existing between the multiplicities of slow tracks
1197: and rescattered protons (Sec.~\ref{slow}) we applied increasingly 
1198: tighter cuts on the number of slow tracks ($P<700~MeV/c$). 
1199: As a consequence of these cuts the degree of correlation between 
1200: $\alpha$ and $v$ increases. 
1201: The fit values of the slopes are reported in Table~\ref{table:alpha}
1202: together with the definitions of the cuts applied.\\
1203: We have also observed a strong correlation between the presence of protons travelling
1204: backward in the lab but forward with respect to the hadronic jet direction, and the 
1205: concentration of events  at small $Q^2$ values and large angles with respect to the beam. 
1206: Since also a small $Q^2$ indicates the presence of rescattering,
1207: the exclusion of these events should highlight the expected correlation.
1208: The resulting slope is $-0.39\pm0.07$ with a $\chi^{2}/ndf=4.9/6$
1209: (see Fig.~\ref{alpha}b).\\
1210: The observed behaviour is consistent with the hypothesis of the correlations
1211: effects being to some degree hidden by the presence of rescattering. 
1212: Reducing the rescattering component these correlations seem to become stronger.
1213: 
1214: \begin{table}[ht]
1215: \centering
1216: \begin{tabular}{cclc}
1217: \hline 
1218:  ~~$n^+$   & ~~$n^-$    & ~~~~~~~slope & $\chi^{2}/ndf$ \\ \hline \smallskip
1219:  $\leq~3$  & $\leq~2$ & $-0.23\pm0.06$ & 5.7/6  \\ 
1220:  $\leq~2$  & $\leq~1$ & $-0.25\pm0.07$ & 3.3/6 \\ 
1221:  $\leq~2$  &       0  & $-0.30\pm0.07$ & 3.7/6 \\
1222:        1   &       0  & $-0.37\pm0.10$ & 3.9/6 \\ \hline
1223: \end{tabular}
1224: \caption{\it { The fitted value of the ($\alpha, v$) slope,
1225: and the corresponding $\chi^{2}/ndf$, for \bp
1226: selected from events with various numbers of positive ($n^+$)
1227: and negative ($n^-$) low  momentum ($P < 700 MeV/c$) particles.}}
1228: \label{table:alpha}
1229: \end{table}
1230: 
1231: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1232: \section{Conclusions}
1233: 
1234: We have observed backward proton and $\pi^-$ production in $\nu$-nucleus 
1235: interactions in the NOMAD detector.\\
1236: The slope parameter $B$ of the invariant cross section,
1237: parametrized as $C~e^{-BP^{2}}$, has been measured and found to be consistent
1238: with previous $\nu$-nucleus and hadron-nucleus experiments.
1239: We found that $B$ does not depend on $E_{HAD}$ and $Q^{2}$ over a wide range
1240: of values. This is in agreement with the ``nuclear scaling'' previously
1241: observed in hadronic experiments.
1242: The observed invariant spectrum is not consistent with the existence of 
1243: a ``long'' Fermi momentum tail as the one proposed in Ref.~\cite{bodek} 
1244: but agrees with the prediction of Ref.~\cite{benhar}.\\
1245: The \bp rate in the NOMAD target (mainly carbon) has been measured 
1246: and compared with the values obtained on different nuclei. 
1247: While the $A$ dependence for neutrino scattering on heavy nuclei  
1248: is consistent with that of hadron experiments, and can be parametrized
1249: as $<N_{BP}> \propto A^{\alpha}$ with $\alpha = 0.68$ \cite{dayon}, the NOMAD result
1250: does not fit this dependence. The $A$ dependence of \bpim
1251: has been found to be steeper than that of $Bp$.\\
1252: The backward proton data have been compared with the predictions of
1253: reinteractions and short range models. The observed energy dependence
1254: is consistent with the ``formation zone'' mechanism. The correlation
1255: between the multiplicitiy of slow ($P<700 MeV/c$) tracks and \bp
1256: events indicates the effects of reinteractions. However when appropriate cuts
1257: are applied in order to reduce the intranuclear cascade contributions,
1258: the correlation between the \bp and the muon scaling variable $v$, predicted
1259: by the short range models, becomes stronger.
1260: 
1261: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1262: \begin{ack}
1263: We thank the management and staff of CERN and of all
1264: participating institutions for their vigorous support of the experiment.
1265: Particular thanks are due to the CERN accelerator and beam-line staff
1266: for the magnificent performance of the neutrino beam. The following
1267: funding agencies have contributed to this experiment:
1268: Australian Research Council (ARC) and Department of Industry, Science, and
1269: Resources (DISR), Australia;
1270: Institut National de Physique Nucl\'eaire et Physique des Particules (IN2P3), 
1271: Commissariat \`a l'Energie Atomique (CEA), Minist\`ere de l'Education 
1272: Nationale, de l'Enseignement Sup\'erieur et de la Recherche, France;
1273: Bundesministerium f\"ur Bildung und Forschung (BMBF, contract 05 6DO52), 
1274: Germany; 
1275: Istituto Nazionale di Fisica Nucleare (INFN), Italy;
1276: Russian Foundation for Basic Research, % (grant 96-02-18562), 
1277: Institute for Nuclear Research of the Russian Academy of Sciences, Russia; 
1278: Fonds National Suisse de la Recherche Scientifique, Switzerland;
1279: Department of Energy, National Science Foundation (grant PHY-9526278), 
1280: the Sloan and the Cottrell Foundations, USA. F.J.P. Soler is
1281: supported by a TMR Fellowship from the European Commission.
1282: We are grateful to O. Benhar, S. Fantoni and G. Lykasov for
1283: stimulating discussions on the subject of this paper.
1284: \end{ack}
1285: 
1286: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1287: \begin{thebibliography}{99}
1288: \bibitem{baldin}
1289: A. Baldin et al., Sov. J. Nucl. Phys. {\bf 18} (1973) 41.
1290: \bibitem{kopel}
1291: \noindent V. B. Kopeliovich, Phys. Rep. {\bf 139} (1986) 51.
1292: \bibitem{ranft}
1293: J. Ranft, Zeit. Phys. {\bf C 43} (1989) 439;\\
1294: \noindent H. J. M\"orig and J. Ranft, Zeit. Phys. {\bf C 52} (1991) 643.
1295: \bibitem{ferrari-ranft}
1296: \noindent A. Ferrari et al., Z. Phys. {\bf C 70} (1996) 413.
1297: \bibitem{stodolsky}
1298: L. Stodolsky, Proc. $V^{th}$ Int. Coll. on Multiparticle Reactions,
1299: pag. 577, Oxford 1975.
1300: \bibitem{niko}
1301: G. V. Davidenko and N. N. Nikolaev, Nucl. Phys {\bf B 135} (1978) 333.
1302: \bibitem{eliseev}
1303: A. El Naghy and S. M. Eliseev, J. Phys. G Nucl. Part. Phys. {\bf 16} (1990) 39;\\
1304: \noindent S. M. Eliseev and B. S. Yuldashev, Sov. J. Nucl. Phys. {\bf 40} (1984) 601.
1305: \bibitem{arneodo}
1306: \noindent M. Arneodo, Phys. Rep. {\bf 240} (1994) 301.
1307: \bibitem{burov}
1308: \noindent V. V. Burov et al., Phys. Lett. {\bf B 67} (1977) 46.
1309: \bibitem{fs1}
1310: L. L. Frankfurt and M. Strikmann, Phys. Lett {\bf B 69} (1977) 93; \\
1311:  Phys. Rep. {\bf 76} (1981) 215.
1312: \bibitem{fujita1}
1313: T. Fujita, Phys Rev. Lett. {\bf 39} (1977) 174.
1314: \bibitem{fujita2}
1315: \noindent T. Fujita and J. H\"ufner, Nucl. Phys {\bf A 314} (1979) 317.
1316: \bibitem{carlson}
1317: C. E. Carlson, K. E. Lassila and U. P. Sukhatme, Phys. Lett. {\bf B 263} (1991) 277\\
1318:  \noindent C. E. Carlson, J. Hanlon and K. E. Lassila, hep--ph/9902281.
1319: \bibitem{multiq}
1320: \noindent L. A. Kondratyuk and M. Zh. Shmatikov, Z. Phys. {\bf A 321} (1985) 301.
1321: \bibitem{bosveld}
1322: G. D. Bosveld, A. E. L. Dieperink and A. G. Tenner, Phys. Rev. {\bf C 49} (1994) 2379.
1323: \bibitem{fredriksson}
1324: S. Fredriksson et al, Phys. Rep. {\bf 144} (1987) 187.
1325: \bibitem{berge}
1326: J. P. Berge et al., Phys. Rev. {\bf D 18} (1978) 1367\\
1327: \noindent V. I. Efremenko et al. Phys. Rev. {\bf D 22} (1980) 2581.
1328: \bibitem{skat}
1329: A. A. Ivanilov et al., JETP Lett. {\bf 30} (1979) 362.
1330: \bibitem{ammosov}
1331: V. V. Ammosov et al., Sov. J. Nucl. Phys. {\bf 43} (1986) 759.
1332: \bibitem{matsinos}
1333: E. Matsinos et al. Z. Phys. {\bf C 44} (1989) 79.
1334: \bibitem{dayon}
1335: M. Dayon et al. Z. Phys. {\bf C 56} (1992) 391;\\
1336: \noindent R. G. Ammar et al., JETP Lett. {\bf 49} (1989) 219.
1337: \bibitem{nom1}
1338: J. Altegoer et al. NOMAD Coll., Nucl. Instr. and Meth. {\bf A 404} (1998) 96.
1339: \bibitem{nom2}
1340: J. Altegoer et al. NOMAD Coll., Phys. Lett. {\bf B 431} (1998) 219.
1341: \bibitem{nomad_dc}
1342: M. Anfreville et al., {\em ``The drift chambers of the NOMAD experiment''}, 
1343: hep-ex/0104012, submitted for publication in Nucl. Instrum. Methods. 
1344: \bibitem{trd}
1345: G. Bassompierre et al., Nucl. Instr. and Meth. {\bf A 403} (1998) 363;\\
1346: G. Bassompierre et al., Nucl. Instr. and Meth. {\bf A 411} (1998) 63.
1347: \bibitem{ecal}
1348: D. Autiero et al., Nucl. Instr. and Meth. {\bf A 372} (1996) 556;\\
1349: D. Autiero et al., Nucl. Instr. and Meth. {\bf A 373} (1996) 358;\\
1350: D. Autiero et al., Nucl. Instr. and Meth. {\bf A 387} (1997) 352;\\
1351: D. Autiero et al., Nucl. Instr. and Meth. {\bf A 411} (1998) 285.
1352: \bibitem{trigger}
1353: J. Altegoer et al., Nucl. Instr. and Meth. {\bf A 428} (1999) 299.
1354: \bibitem{gianmaria}
1355: G. Collazuol et al., Nucl. Instr. and Meth. {\bf A 449} (2000) 609.
1356: \bibitem{star}
1357: G. Barichello et al., Nucl. Instr. and Meth. {\bf A 419} (1998) 1.
1358: \bibitem{lepto}
1359: G.~Ingelman, LEPTO 6.1, in Proc. of Physics at HERA,
1360: Edited by W. Buchmueller, G. Ingelman, DESY, Hamburg (1992) 1366.
1361: \bibitem{jetset}
1362: T.~Sj\"ostrand, Computer Phys. Commun. {\bf 39} (1986) 347.
1363: \bibitem{geant}
1364: GEANT, CERN Program Library Long Writeup W5013.
1365: \bibitem{bodek}
1366: A. Bodek and J. L. Ritchie, Phys. Rev. {\bf D 23} (1981) 1070.
1367: \bibitem{alanakyan}
1368: K. V. Alanakyan et al., Sov. J. Nucl Phys {\bf 26} (1977) 539.
1369: \bibitem{bayukov}
1370: Y. D. Bayukov et al., Phys Rev {\bf C 20} (1979) 764.
1371: \bibitem{komarov}
1372: V. I. Komarov et al., Nucl Phys {\bf A 326} (1979) 297.
1373: \bibitem{yulda}
1374: B.S. Yuldashev et al., Phys Rev {\bf D 46} (1992) 45.
1375: \bibitem{d2}
1376: D. Allasia et al., Zeit. Phys. {\bf C 24} (1984) 119;\\
1377: \noindent A. G. Tenner, NIKHEF--H/86--7, NIKHEF Report.
1378: \bibitem{na22}
1379: N. M. Agababyan et al, EHS/NA22 Coll., Z. Phys. {\bf C 66} (1995) 385.
1380: \bibitem{ishii}
1381: C. Ishii, K. Saito and F. Takagi, Prog. of Theor. Phys. {\bf 86} (1991) 205.
1382: \bibitem{kitagaki}
1383: T. Kitagaki et al. E745 Coll., Phys. Lett. {\bf B 214} (1988) 281.
1384: \bibitem{guy}
1385: J. Guy et al.,  Phys. Lett. {\bf B 229} (1989) 421.
1386: \bibitem{benhar}
1387:  O. Benhar, S. Fantoni and G. Lykasov, Eur. Phys. J. {\bf A 7} (2000) 3,415;\\  
1388: \noindent O. Benhar et al., Phys. Rev. {\bf C 55} (1997) 244.
1389: \bibitem{benhar_priv}
1390:  O. Benhar, private communication.
1391: \bibitem{starink}
1392: R. Starink et al.,  Phys. Lett. {\bf B 474} (2000) 33.
1393: 
1394: \end{thebibliography}
1395: 
1396: \end{document}
1397: