The energetic electrons and ions embedded in Earthward-moving plasmoid structures have been observed. These plasmoids are associated with a rotational local Bz component (bi-polar) signature. Energetic electrons are found to be confined in a smaller spatial region than ions inside the plasmoid. Energetic ions and electrons seem to be a good indicator for the structure boundary. The fleet of Cluster spacecraft cross the plasmoid structure in a “first entry, last out” order (Note: when spacecraft cross a planar discontinuity, e.g. magnetopause, they will be in “first entry, first out” order). This documents the fact that the plasmoid has a non-planar nested structure. The large separation distance (around 1 RE) of the Cluster satellites in October 2002 is an advantage to provide constraints on the size and shape of the plasmoid structure of interest. In addition, the plasmoid (with closed field lines) should preserve the ion composition information where it is formed. The ion composition observed in the plasmoid shows significantly lower O and He than in the ambient plasma. This implies few heavy ions are involved in the reconnection process where the plasmoid is formed. Multiple flux ropes/plasmoids observation presented in this paper can be interpreted as strong evidence for multiple X-lines.
 Plasmoids and/or Flux ropes are important feature in various kinds of eruptive processes in astrophysical plasmas, notably in the occurrence of magnetospheric substorm and solar Coronal Mass Ejection (CME). The typical observational signatures of a plasmoid have been defined in two dimensions by a bipolar perturbation in Bz, accompanied by high-speed tailward flowing energetic ions and electrons beams. Such a tailward-moving large loop-like magnetic structure is generated as the result of X-type reconnection in the near magnetotail [Hones, 1979]. Many plasmoids may have a non-zero By [Moldwin and Hughes, 1991, 1993; Zong et al., 1997, 1998; Baker et al., 2002; Slavin et al., 2003a, 2003b]. Hughes and Sibeck  examined the three dimensional topology and morphology of a plasmoid of a finite dawn-dusk component. Magnetic structures in the magnetotail are rather complicated configurations as shown in Figure 1. From the left the four schematics show (a) the original 2-dimensional Hones concept, (b) a helical flux rope embedded in an outer closed loop magnetic field, (c) a helical flux rope, (d) remote sensing of a plasmoid by means of its moving magnetic imprint in the lobes (travelling compression region or TCR). If a plasmoid passes down tail in the central plasma sheet, the magnetic flux in the tail lobes will be temporarily compressed as the cross-sectional area of the lobe is reduced. The structure with these signatures - a transient increase in the lobe field strength and a deflection of the magnetic field vector, first northward and then southward is called a TCR [e.g., Slavin et al., 2003a, 2003b, 2003c]. Recently, the flux ropes in tail have been further referred to as “BBF” flux ropes (earthward flux rope with a bursty bulk flow) and “plasmoid” flux ropes (tailward flux rope) [Slavin et al., 2003a]. The “BBF” flux ropes are associated with quasi-sinusoidal ΔBz perturbations and Earthward plasma flow whereas the “plasmoid” flux ropes are associated with ±ΔBz perturbations and tailward plasma flow.
 The energetic particle data presented in this paper were obtained by the RAPID/Cluster instrument [Wilken et al., 1997]. Furthermore, we use magnetometer measurements from the fluxgate magnetometer (FGM) on board Cluster, which makes high resolution vector field measurements [Balogh et al., 1997] and plasma data from the Cluster Ion Spectrometer (CIS) experiment [Reme et al., 1997].
 On Oct. 28, 2002, the Cluster spacecraft were travelling in the tail plasma sheet at -11 RE in the northern part of the tail. Figure 2 gives an overview of CLUSTER measurement between 19:30 and 20:00 UT on Oct. 28, 2002. The Cluster Tetrahedron in GSE X-Z plane is attached in this figure. As we can see, S/C 1 and 4 are in the tail side of S/C 2 and 3 while S/C 3 and 4 are more closer to the Z = 0 plane. The separation of Cluster spacecraft in GSE X direction is up to 1 RE. The tail plasma sheet is characterized as a region with plasma density between 0.3 and 0.5 ions/cm3.
 The plasmoid event lasted from 1942 to 1952 UT – P1 can be identified by the bipolar signature in magnetic field Z-component. From 1935 to 1942 UT, as a precursor of this plasmoid, the total magnetic field intensity was compressed from 20 to 23 nT (the Bx, By and Bz components showed an increase as well) prior to the arrival of the plasmoid at 1947 UT. This may be caused by a compression of the lobe magnetic field and lobe plasma density in front of the Earthward moving plasmoid. In the plasmoid, the Bz component shows (−/+/−) “W” signature. The δ Bz of the peak-to-peak signature at 1947 UT was 8.5 nT. There is no indication of a very strong core field at the time of the inflection point (where Bz changes its sign). This kind of plasmoid ((−/+) signatures in the Z-component) is consistent with typical plasmoid moving Earthward as illustrated in the Figure 1. The Bz “W” shape magnetic field variation signature may be explained as this structure has a two-loop – the trailing half of one flux rope followed by a second one [Zong et al., 1997, 2003].
 Inspection of Figure 2 shows that during this event, the peak of the energetic particle fluxes for both ions and electrons occurred at the reflection point while the total magnetic field was around the local minimum. In particular, the magnetic field of this event is consistent with a “closed loop” plasmoid with a symmetry axis roughly in the Y plane. Figure 2 shows the appearance of a strong energetic proton and electron burst associated with this event (from 1944 to 1955 UT). The particle burst followed the profile of the disturbed magnetic field. Both ion and electron spikes started about 2 min after the magnetic field signature was observed. It should be mentioned that the (−/+) bipolar signature passed over the spacecraft essentially during the interval for which the By component was negative. Between 1952 and 1958 UT the Bz component observed by Cluster S/C 1 was essentially negative, while the Z-component fluctuated with small amplitudes. Between 1952 and 2000 UT, three TCRs were observed by Cluster S/C 1 which was the spacecraft farthest from to the central plasma sheet. During the same time period, a smaller plasmoid labeled “P2” and two TCRs were observed consequently by S/C 4 which was the spacecraft closer to the central plasma sheet.
 An important point to be noted is that TCR signatures were observed by both Cluster S/C 1 and S/C 4 between 1952 and 2000 UT. This is additional evidence for (1) continuing existence of a neutral line tailward of Cluster spacecraft, (2) the direct connection of the post-plasmoid configuration to an active neutral line, and (3) curved lobe field lines caused by the huge size of the plasmoid in the x–z plane.
 RAPID ion composition and electron observations from spacecraft 1 to 4 are shown in Figure 3. It is evident that all of the four spacecraft recorded similar features for the plasmoid P1. However, the following small structure P2 was only observed by S/C3 and 4. The most clear signature is the sharp enhancement of the flux of electrons. The separation of Cluster spacecraft is up to 1 RE in this case. The two panels display the ratio of the flux of helium ions (70 to 1500 keV) to that of protons (30 to 1500 keV) and the flux of CNO ions (140 to 1500 keV) to that of protons, respectively. It is interesting to note that just within the plasmoid, these two ratios decrease significantly to a value around 0.2 and 0.03, respectively. The ratios on either side of the plasmoid encountered were found to be between 0.6 and 0.1. This indicates that the ion composition in the plasmoid is significantly different compared to the ambient plasma. The much lower heavy ions density has been “trapped” in the Earthward-moving plasmoid which was formed in the distant tail region. The plasmoid should preserve the ion composition information where it was formed. In contrast, a high abundance of heavier ions is often observed in the tailward moving plasmoids [Zong et al., 1997, 1998]. It should be noted here that O/H, He/H Flux ratios obtained by Cluster C1 and C4 for the P2 plasmoid period are almost the same as the ambient plasma whereas O/H, He/H Flux ratios obtained by Cluster C2 and C3 are still smaller than the ambient plasma. The reason is due to the P2 plasmoid was directly observed by C2 and C4, whereas C1 and C3 were outside the plasmoid, only TCR signatures were observed. This is consistent with magnetic field and energetic electron observations shown in Figure 2.
 As we can see from Figure 3, the onset times of the proton and electron flux in the plasmoid are quite different (about 1 min) from proton enhancement in advance. The thickness of ion and electron separation is estimated as 0.84 RE by using the speed of the plasmoid from their time of arrival at the different spacecraft. This is 4 times larger than the proton gyro-radius (0.21 RE), thus, it cannot be simply explained as the gyro-radius effect, the reason needs to be further studied. The scale size of this plasmoid is calculated as 5.04 RE.
3. Interpretation and Discussion
 The spacecraft in this event cross the electron and ion burst embedded in the plasmoid in a “first entry, last out” order (Note: when spacecraft cross a planar discontinuity, e.g. the magnetopause, they will be in “first entry, first out” order). The plasmoid studied here was first encountered by the spacecraft C1 and then followed by the spacecraft C4, C3 and C2, consecutively (see Figure 3). On the trailing edge the spacecraft leaving the structure were in the order of C2, C3, C4 and C1. The maximum duration time is obtained by C1 (Samba, the first satellite to encounter both the electron and proton bursts). This might be explained by nested structure for the plasmoid as illustrated at Figure 4. Energetic electrons (e.g. 50 keV) travel along field lines with a high speed of around 20 REs−1. These swift electrons trace out field lines in the magnetosphere in a few seconds, therefore, could provide nearly instantaneous information about the changes in the open and closed magnetic field line configuration in the geospace regions. Thus, the high flux electron embedded in this plasmoid suggests it is a closed field line structure.
 A sketch of the inferred orientation and the motion of the observed plasmoid with respect to the four Cluster tetrahedral configuration is also shown in of Figure 4. The XZ plane extended plasmoid with a significant duskward speed will lead the Cluster spacecraft cross the plasmoid structure in the order of C1 − > C4 − > C3 − >C2. This is consistent with the energetic particle observations.
 It has been proposed to modify a single-plasmoid model [e.g., Hones, 1979] in order to explain the multiple-plasmoid formation in the course of substorm event [Zong et al., 1997; Slavin et al., 2003a, 2003b]. The sequence of plasmoids at about 1954UT and TCRs is an indication of the repeated development of NENLs. The plasmoids can occur intermittently and repeatedly [e.g., Lee et al., 1985]. The repetition rate may depend on the conditions of the solar wind and the IMF. The significance of multiple plasmoids/flux ropes in the plasma sheet is that their formation can be understood in terms of simultaneous reconnection at multiple X-lines (MRX) in the magnetotail [Slavin et al., 2003a; Deng et al., 2004]. Multiple x-lines due to tearing mode reconnection could produce magnetic islands. The resulting overall field configuration is that of a magnetic island without a core field in the center. If the tail field lines are not precisely antiparallel, the islands will contain a field component along the island's axis to form a flux rope [Hughes and Sibeck, 1987].
 The observations made in this paper also raise some interesting questions regarding the “fate” of Earthward moving plasmoids in the plasma sheet. The Earthward-moving plasmoids pushed up against the geomagnetic field will probably dissipate quickly since the orientation of their magnetic fields is favourable for reconnection with the geomagnetic closed field lines in the near Earth as suggested by Slavin et al. [2003a]. In particular, multiple Earthward-moving plasmoid/flux ropes could be possible triggers for the substorm injections observed in the geostationary orbit. The first injection in the geostationary orbit occurred at 2008 UT, Oct. 28, 2002 which is about 14 min later than the first plasmoid signature observed. These issues need to be further studied.
 The main results of this paper can be summarized as follows: (1) The energetic electrons and ions embedded in an Earthward-moving plasmoid structure have been observed. This plasmoid is associated with a rotational local Bz component (bi-polar). Energetic electrons are a good indicator of the structure boundary. (2) Ions and electrons co-exist in the plasmoid. Electrons are found to be confined in a smaller spatial volume than ions. (3) Multiple flux ropes/plasmoids observed in the tail plasma sheet can be interpreted as strong evidence for multiple X-lines, observed S-N TCR signatures are caused by the Earthward-moving flux ropes/plasmoids. (4) The ion species ratios– He/H and O/H in the Earthward moving plasmoid are found to be significantly lower than the ambient plasma. The plasmoid should preserve the ion composition information where it is formed. In contrast, a high abundance of heavier ions are often observed in the tailward motion plasmoid [Zong et al., 1997, 1998].