In this indicative case study presented in Figure 3, three successive plasma-ion surges are directed earthward

(bottom panel) tightly associated with three distinct dipolarizations labeled A, B and C (in fourth panel) and marked with red arrows along the B_z trace. We use the magnetic field datasets (in nanoteslas) provided by the Geotail/MGF instrument [20] with 62-ms resolution, while the plasma data are generated by taking the moments over 12-s intervals of ion distribution functions measured by the low energy particle (LEP) instrument [21] . The event was observed at (X, Y, Z)_GSM = (−28.1, 1.3, 4.4) R_E, and we may conclude that the magnetosphere is repetitively transiting from a stretched geometry to a more dipole-like configuration every ~40 s. That is, the source is clearly activated three times ejecting earthward three ion jets. In a more detailed look, we would like to comment and organize our observations in a chronological sequence, as they probably took place during this event:

1) First, we suppose that the plasma flows signify separate source activations; at the same time, we assume that each dipolarization process was initiated just earthward of the IDR, although it will arrive (at S/C) a little time later. The latter is evident by the apparent S/C motion; it moves inward (i.e., toward the neutral sheet) three times. Obviously, we interpret the decreases of B_x as plasma sheet expansions which were initiated before the arrival of stepwise increases of B_z; the tail’s rearrangement begins at the same time with the source activation.

2) Each increase of B_y is assumed as due to an earthward flowing FAC; thus, within the north plasma sheet, the B_y is switched from a slightly negative to a large positive value three times, every time the S/C crosses the FAC while moving inward. The current is caused by the ion jet directed earthward from the source; the ions ride on the magnetic field lines just out of the IDR. We consider that the ions flow faster along the magnetic field lines, and consequently arrive earlier than the dipolarization front itself. Thus, the B_y increases before each front arrival. And each reduction of B_z, before each front, is due to the plasma diamagnetic effect, plus an ongoing inflation of tail.

3) Three successive dipolarization fronts arrive at S/C while the B_z increases as much as an order of magnitude. The crossing of each front coincides with a rapidly reducing plasma velocity until the next activation takes place.

The above processes can be understood in the context of Figure 2. According to our model, the source is activated three times and ejects (earthward) three distinct ion jets. The latter initiates three successive increases of FACs which, in turn, are manifested by the B_y deflections. And actually, for the presented event, the earthward flowing FACs produce positive deflections of B_y in north plasma sheet, where Geotail was located.

It seems that the FAC is particularly intense for the first plasma surge; this is the reason for the extreme enhancement of B_y from about zero to 12 nT (i.e., a temporal effect). For the dipolarizations B and C, the current crossing changes the sign of B_y to −3 nT (i.e., a spatial effect). Only for the first dipolarization the peaked B_y value coincides to aB_z transition to more positive values. Thus, one could classify this structure as a typical MFR one. Nevertheless, the whole phenomenon (of Figure 3) probably carries the information that this MFR-like structure is not different from the next two. The whole phenomenology (in our view) is the result of an ion FAC and the depolarization front which is propagating. The satellite is first influenced by the nearby flowing FAC because the B_y deflection increases almost simultaneously with the initiation of an ion flow. Conversely, the B_z increase delays since it is depended on the slower process that piles up magnetic flux close to the Earth. Consequently, the structures A-C are essentially interpreted here as (mainly) due to temporal rather than spatial effects, although both effects interrelate. Therefore, we conclude that the supposed rope-like “structure A” has neither a B_y core, nor a helical structure. And we conclude, from this indicative phenomenon, that the peaked values of B_y and B_z may coincide or not; this is not the critical factor for a deeper understanding of the ultimate physical process. We think that the same underlying mechanism produces all the three structures in Figure 3; it would be misleading to categorize only the “structure A” as belonging to a particular class.

During the second ion jet, the B_z abruptly increases ~20 nT; that is, more than twice as compared to the pre-disturbed B_total. Such a striking reconfiguration remind us the process taking place over the dayside magnetopause boundary and this view is described in the preceded section.

There are cases for which the B_y deflections are essentially decoupled from any bipolar signature visible along the B_z trace or any peaked value observable along the total magnetic field trace; accordingly, these events are obviously incompatible with any real MFR structure. In Figure 4, we present six events with positive or negative B_y deflections occurring with very intense earthward plasma-ion flows, given that the V_x velocity increases up to 1400 - 3000 km∙s⁻¹. They are mostly observed in combination with (or during) the dipolarization phase of

the magnetic field (characterized by the high and positive values of B_z, fourth panels) simultaneously occurring with strong B_y deflections (third panels). In particular, we pay attention to the B_y polarity which is positive in the north plasma sheet, where the B_x is clearly greater than zero, as shown by the events a-c; and negative in the south plasma sheet with B_x less than zero, as shown by the events d-f. In the context of the already presented model, the FACs generated by the earthward ejected ions will cause B_y deflections (outside the IDR) with the appropriate polarity in the north and south plasma sheet. It is worth noticing that the event “b” shows three successive and impulsive plasma flows accompanied by three distinct positive B_y deflections. Most importantly, the second B_y deflection clearly precedes (~2 min) the second intense local dipolarization. The latter lines up with the concept that the ion flow first affects the B_y component (via the FACs); while the B_z is later affected by the flux pile-up. In general, the B_y deflection increases stronger with the plasma velocity than any B_z variation. In seldom cases we can observe (i.e., with regard to the situation of event “f”) that the |B_y| deflection increases even before the detection of plasma flow; additionally, no clear dipolarization is detected at all for this event.

Figure 5 shows four events with an MFR-like structure occurring with a) an earthward plasma-ion flow (i.e., positive values for the velocity V_x, fifth panels); b) a distinct dipolarization (i.e., negative-then-positive B_z variation) evident along each B_z trace (fourth panels); c) maximized B_y values (marked with dashed-vertical lines) occurring simultaneously with the dipolarizations; and d) peaked values of the total magnetic field when the B_z changes sign. Consequently, based on morphological criteria, one can plausibly argue that we observe four typical MFR structures; that is, the bipolar B_z signature may correspond to the helical structure and the B_y deflection is potentially the rope’s core. However, we suggest a different interpretation: The sign for the B_y deflection is determined by the satellite placement in north (positive) or south (negative) plasma sheet. Actually the events “a” and “b” are characterized by negative B_y deflections in south plasma sheet (B_x negative) and the events “c” and “d” are related with positive B_y deflections in north plasma sheet (B_x positive), in agreement with the exhibited model in subsection 2. Therefore, in our framework, the shown structures are actually pseudo- MFRs; neither cores nor helical structures do really exist. The duration of each event varies from 1 to 2 minutes and the time in between the peaked |B_y| and |B_z| values varies from 10 to 60 s. If the reader thinks that the B_x trace, for the event “b”, is not sufficiently representative (case corresponding to south PS), then he/she could consult two similar examples studied in the past a) by Snekvik et al., [22] with CLUSTER, or b) by Sormakov and Sergeev [23] , in their Figure 2; on both cases the negative B_y deflections clearly occur in south PS with earthward plasma flows.

The fundamental principle, in our model, is that the FACs are developed and diagnosed before the arrival of the dipole-like structure (i.e., a temporal effect). First, a source activation occurs producing an earthward ion jet, and inflating the PS as a whole. In turn, the ion jet sets up the FACs (associated with the B_y deflection), while the PS expansion is organized as a dipolarization front. And the reader must pay attention to the fact that the B_y deflection, at times, essentially occurs simultaneously with the negative B_z deflection (and not when the B_z changes sign, as conventionally expected with respect to a rope structure). The latter is absolutely clear for the “a” and “b” events of Figure 5 if we focus our attention to the really dominant |B_y| values. They clearly occur before the abrupt increase of B_z from negative to positive values; something that is evident in Figure 6 showing the ratios of B_z/B_total (blue line) and B_y/B_total (red line). When the B_z is very small or negative, the (by far) dominant component is the B_y. We obviously identify not a symmetric loop-like rope structure, but a mere boundary crossing, a simple passage from one region to another.

In summary, we conclude that the observed B_y deflections essentially form a certain pattern; they are positive in the north and negative in the south plasma sheet. A more systematic-statistical approach would further support this finding; the already described selected events are considered as representative.

In a recently published work, Sarafopoulos [9] emphasized that in a thin plasma sheet, after an explosion process initiated recurring activations of substorm, the bulk tailward motion of demagnetized ions (forming an “ion channel” inside the IDR) is catalytically affected by the local magnetotail’s motions. He suggested that even an ion vortex is formed from a tailward ion jet affected by local upward or downward tail's motions. The type of motion eventually determines the clockwise or counterclockwise direction of the ion cyclic motion (over the XZ plane) and, consequently, the sign for the B_y deflection due to this vortex-current. The upward (downward) motion is associated with a positive (negative) B_y deflection. In a similar way, in this work we initially stress that the earthward ejected ions stream along an “ion channel” (inside the IDR) which eventually splits into two

branches forming the “downward FACs” and causing the appropriate B_y deflections. Furthermore, it is more realistic to assume that the two current branches are probably asymmetric. This concept is schematically illustrated in Figure 7(a), where the northern branch of FAC is much more intense than the southern one. And this magnetosphere response could result by two combined factors: First, a local downward motion of magnetotail (earthward of the source) and, second, the ballistic character of orbits of demagnetized ions within the IDR. Thus, the earthward ejected ions essentially ride on the lines of northern hemisphere PS. Consequently, a satellite (S/C) crossing the whole plasma sheet will (almost exclusively) detect positive B_y deflections; only adjacent to the south branch of FAC, a negative and weak B_y deflection might be detected. Parenthetically, it would be instructive to highlight that the same type of asymmetry could be detected for a tailward ion jet associated with an upward tail’s motion: The satellite has to record, throughout the whole plasma sheet, a positive B_y deflection caused by an ion surface current, as it is shown in Figure 7(b). That is, each tail’s motion does not necessarily lead to a vortex-like current, as one could speculate.

In supporting the claim of asymmetric currents (introduced in the preceded paragraph), we present in Figure 8 an indicative event occurred on June 22, 2002, around 06:35 UT. In this episode the satellite, while receiving an earthward plasma flow with high velocities, moves from the southern to northern PS and detects prevailing positive B_y values. Only the more negative excursions of B_x are associated with negative B_y deflections; in parallel, the positive B_y deflection is indeed particularly intense in north PS. Additionally, we show the scatter-plot of B_x and B_y (at the bottom panel), for the period 06:33:55 to 06:36 UT; actually, the positive B_y values dominate irrespectively of the B_x sign. The thick-red line constitutes the (third order) polynomial fitting to the data. Needless to say that this result is obviously in disharmony with any north-south symmetric FAC or even the possibility of a symmetric guide field generated by Hall currents in a typical X-line, if the latter was the case. We schematically know that the guide field must change sign around a neutral sheet crossing. Consequently, the diagnosed asymmetry may introduce a radically different perspective from what was emphatically pointed out by Runov et al., [4] ; they presented scatter-plots between B_x and B_y showing a symmetry (of variations) with respect to the neutral sheet. Certainly, in the above mentioned work, the authors assume that the satellite was inside the IDR. In any case, it remains a question how the exhibited asymmetry (with respect to the neutral sheet) could be interpreted and explained in the context of an X-line “reconnection model”.

It was already pointed out that we deal with structures related to earthward plasma flows; however, since the issue of “asymmetric scatter plots” around the neutral sheet like the one of Figure 8, is particularly important,

either with earthward or tailward flows, it turns out to be helpful to know that the exhibited asymmetry is commonly detected at both sides of the source. That is, we shall extend our discussion by two more cases occurring when the satellite has just crossed the source when moving tailward. The latter will further deepen our horizon and we shall better understand a few new key factors related to the magnetotail’s response, given that this asymmetry was not emphasized in the past. Thus, we show in Figure 9 two cases for which the Geotail, after passing by the source, and the earthward flows are switched to tailward, encounters asymmetric (B_x, B_y) scatter-plots. The asymmetry is profound; the same polarity of B_y is detectable almost throughout the whole PS. The events occurred a) on 11/05/2006, at ~09:22 UT with a downward motion of the magnetotail and a negative B_y deflection extended from B_x = −15 to + 15 nT, and b) on 17/03/2003, at ~14:04 UT, with an upward tail’s motion and positive B_y deflection within a region confined in between ±15 nT. Negative (positive) B_y deflection is observed even with negative (positive) B_x values in case 9a (b). Scatter-plots for these events (shown in Figure 9(c) and Figure 9(d) correspond to the periods from 09:19 to 09:24 UT (for the first event), and from 14:02:20 to 14:07 UT (for the second event). The guide field B_y is almost monopolar, negative (for the left panel) and positive (for the right), irrespectively of the sign of B_x. Obviously, this result is not in agreement with the typically symmetric “B_y-guide field” produced by Hall currents in the context of a collisionless X-line. In our view, the shown B_y deflections are not due to Hall currents; they are modulated by the joint action of a tailward directed ion jet plus an upward (downward) motion of the tail. The ions stream within the IDR and finally ride on the already curved magnetic field lines of the southern (northern) plasma sheet and move along the lines causing positive (negative) B_y deflections within almost the whole plasma sheet. The deflections are dictated by the surface ion current and Ampere´s law; an indicative interpretation scheme is shown in Figure 7(b) where a tailward ion jet encounters an upward curved magnetic field structure (i.e., lines like a convex function with monotonically increasing slope, at the plasma sheet/lobe interface). Conversely, if the tailward ion jet encounters downward curved magnetic field lines (i.e., lines like a concave function with monotonically decreasing slope, at the plasma sheet/lobe interface), then a negative B_y deflection will be caused across the whole plasma sheet. It is a great challenge for the X-line reconnection model to give a convincing interpretation for the coexisting values of B_y @ 20 nT and B_x @ B_z @ 0 nT, at ~14:04 UT, in Figure 9(b).

Within the “ion channel” (being inside the IDR) the demagnetized ions cross the magnetic field lines and do not convect with them; they do not obey to the “frozen-in condition”. Most importantly, the earthward ejected ions finally build up a termination boundary with B_z @ B_total, even at the very CPS, as it was proposed in Section 2. Already from 1992, Sergeev et al. [24] , noted “that the amplitude of the main B_z impulse is systematically larger at the satellite which is closer to the centre of the PS, both in the tailward and in the earthward streaming MFRs”. The incident ions (over the boundary) give rise to a downward surface current, which progressively grows more intense. This local dipolarization is of fundamental importance because the ion populations behind the front will produce the FACs causing the B_y deflections with different polarities in north and south PS. And according to this scenario, given that it is impossible to have an MFR entity with two oppositely directed “B_y cores”, the usually observed (earthward of the source) structures are “pseudo-MFRs”; very close to the Earth a genuine MFR is obviously not formed. According to Sarafopoulos [9] a rope embedded in the tailward plasma flow is often formed when a local motion (upward or downward) of the tail affects the “ion channel” of the IDR. Something similar, very close to the Earth, is rather unreasonable to occur because the source is relatively close to the “dipolarization front-surface”. In contrast, if the source is activated far away in the tail, then an MFR could be formed embedded in the earthward plasma flow, too.

Furthermore, the abrupt increase of B_z in CPS is associated (for a satellite crossing the whole structure) with a south-then-north signature (of B_z). The latter is schematically shown in Figure 2 via a satellite (S/C) moving tailward, although, actually, the whole structure moves earthward. This geometry is the same with that proposed by Sergeev et al. [24] and Sormakov and Sergeev [23] . They suggested that “a reconnected magnetic tube moving earthward can crumple the background field by creating a compressed B_y on its front and strongly increasing the B_z values in the positive direction; a short-time negative variation in B_z occurs due to the inclination of the compressed background field”. In our view, such a compression actually takes place, an impulsive B_z boundary is really built in CPS and the bipolar B_z signature is of chief importance; however, we add in this scenario the key role resulting by the ion surface current over the B_z front. The fundamental difference (expressed in a sim-

plified manner) is that in the X-line model the source ejects plasma, whereas in our model the source ejects ions (outside the IDR). The ions build up the impulsive B_z boundary (outside of the IDR) and produce the FACs associated with the B_y deflection pattern. In all the presented MFR-like events the −B_z/+B_z signature is highly asymmetric, the gradients of B_z are extremely high and, most importantly, the B_y deflection frequently occurs simultaneously with the negative B_z values, as we have pointed out in subsection 3.3. And the background field does not always determine the B_y polarity: For the event “4a”, although the background B_y is essentially zero, the B_y deflection is more than −20 nT, while for the events “4c and d” the background field is slightly negative and the B_y positive!

If the “B_z boundary” is built, it is supposed to play an additional role. From the already suggested production mechanism, energetic ions highly varying in energy (i.e., 30 - 1000 keV) must be incident on the boundary from their tailward source; in our approach, the B_z front is not built after an elapsed time of convection. Thus, we would logically infer that this distinct boundary associated with the dipolarization front is probably the so- called “injection boundary” (as defined by McIlwain [25] or Lopez et al., [26] ). The latter is considered to be a region tailward of which all electrons and protons are energized together at the same time, at substorm onset; and the particle signatures observed by satellites therefore show the normal dispersion due to the different energy depended drift speeds of the particles from the injection boundary to the satellite [27] . Furthermore, it should be underlined that the location or shape of such an injection boundary was a matter of conjecture [28] . In our approach, the dispersionless injection signature would be obtained if indeed the satellite was within or very near to the injection boundary region, but the particles are not locally accelerated. The particles are accelerated at the source far away tailward, and in turn, propagate as an ion jet (carrying plasma ions and energetic populations at the same time) that modulates the dipolarization front, which is finally detected by the satellite as a dispersionless injection. Again, in our view, we do not need a near-Earth acceleration mechanism being at work adjacent to the injection boundary; the particles are not energized by a local plasma instability. Certainly, the idea about particles moving inward, perhaps from several R_E away, is an old one: In the “convection surge” model [29] - [31] particles are energized by the dipolarization process, and subsequently convected earthward by the inductive electric field. Even long drift paths could be possible in this perspective, if the magnetic field change propagated with the particles and cancels the normal radial magnetic field gradient that otherwise might separate particles of different energies. Simulations supporting the latter concept are performed by Sarris and Li [32] . The concept underlying this work may be distinguished from the others in that the populations are essentially energized at the source site by a specific mechanism, for instance, the explosive phase of the twin Double Layer (DL) structure, as we already suggested [11] . According to the Double Layer model, the resulting ion particles are going to build the B_z dipolarization front. We basically decouple the dipolarization front from any reconnection process taking place via a supposed X-line; the produced energy cannot be considered as the result of any reconnection, but as the result of an electrostatic discharge potentially occurring within the twin-DL structure.

This work, along with the recently published one by Sarafopoulos [9] , could lead to a unified scheme concerning the pattern of B_y deflections at both sides of the source, on fluid scale lengths. Taking into account that we study plasma flows or magnetic field deflections lasting (typically) 3 min and having velocities (on average) 500 k∙ms⁻¹, then the involved scale length is about 15 R_E; i.e., this work focuses on the so-called fluid scale.

Furthermore, in order to achieve an overall pattern of B_y deflections, we apply the fluid-scale results (that line up with our observations) to the microscale response of a typical X-line. We try to obtain the best fit of our results to the mainstream X-type reconnection model, although major inconsistencies finally remain. Thus, in the next subsection, while raising additional objections against the X-line model, we finally support a radically different opinion.

At presently, we consider a simplified, symmetric topology, for the region of the source, although we have (to a certain extent) scrutinized asymmetric (with respect to the neutral sheet) B_y patterns. We are particularly interested in the area located earthward of the source; accordingly, the earthward side of the outflowing source is sketched in Figure 10. The reader can recognize basic features for all the three scales: The smallest scale lengths (and the shortest timescales as well), corresponding to an electron accumulation area, are confined inside the EDR (green line rectangular), the ion scale lengths corresponding to the region inside the IDR (which is en-

closed by the blue line), and eventually the fluid scale lengths (that include the space outside the IDR). At present, it is assumed that the narrow region of the so-called “electron currents” (in Figure 10) corresponds to the “Hall current of collisionless reconnection”. In general, in order to achieve coupling of different scales, measurements of at least two satellites are required; however, in this work we merely use single satellite datasets plus the knowledge gained from previous studies.

The electron Hall current, within the IDR, results from a fundamental difference between ions and electrons in collisionless plasma [16] : The ions are unmagnetized, whereas the electrons remain magnetized and move strictly transverse to the electric and magnetic fields; our Figure 10 is a restricted two-dimensional version over the XZ plane. In this context, the electron populations are carried outward by the newly reconnected (within the EDR) magnetic field lines, whereas the ions will ride on the reconnected lines later; that is, the ions will become adiabatic when they arrive at the boundary of the IDR. Therefore, earthward of the IDR, two branches of the field-aligned ion current would be apparently formed connecting the source to the ionosphere. Furthermore, these FACs will produce the appropriate B_y magnetic field variations; as a result, a magnetic field domain is developed in the neighborhood of an activated source (on fluid scale lengths). Interestingly, the B_y pattern outside the IDR is the same with that related to the Hall currents inside the IDR. A satellite (S/C) moving in trajectories like those shown in Figure 10 and crossing either the Hall current (within the IDR) or the two branches of the ion current (outside the IDR) will identify (inside the PS) the same transition of deflection from −B_y to +B_y. In this perspective, the early results of Runov et al. [4] concerning the pattern of B_y deflections, within the IDR, could be supplementary to our B_y related results on larger scales. However, a major objection will be raised in the next subsection concerning this item.

In addition, in our framework, asymmetric B_y patterns (at both sides of the source) could be easily interpreted. For instance, if the tail earthward of the source locally moves downwards, then only the north hemisphere plasma sheet surface ion current would be essentially developed and exclusively associated with a prevailing positive B_y deflection, like the event of Figure 8. On the other hand, if the tail, tailward of the source, is locally moving upward (downward), then only “one-sided surface ion current” would be mainly formed. The latter corresponds to the event shown in Figure 9(a) and Figure 9(b), where actually the tail locally moves downward (upward) and the dominant B_y deflection is negative (positive).

In this subsection we are going to express our skepticism about the alleged existence of Hall currents originating from an X-line collisionless reconnection; that is, we refer to the electron currents supposed to be outward directed for the inflow regions and inward directed for the outflow ones, inside the IDR. First, we deal with the inflow region Hall currents: Axiomatically, the EDR is (in the context of an X-line) highly populated by the incoming electrons; consequently, an amphipolar electric field E_z pointing inward is formed (within the IDR). Schematics showing such an electric field can be found even in textbooks like the one of Heikilla [33] . In this regard, we would like to make three comments as a preparation for the forthcoming discussions: First, if the E_z were greater than E_y, then a downward drift will prevail for the inflowing electrons and the Hall current J_z would be totally blocked off. Second, if the reconnection geometry were actually X-shaped, then the electrons following a line convecting inward will probably escape along the line by repulsive electrostatic forces, and the Hall currents would be insignificant, too. Third, if the reconnection process actually took place in the tail and the electrons indeed generated a Hall current due to the drift E_y x B_x, then the Hall current J_z flowing against the E_z would initiate a dynamo process (given that E_z・J_z < 0); conversion of magnetic energy is probably not required.

In a second place, we scrutinize the supposed Hall currents flowing along the “outflow regions” of an X-line: It is commonly accepted that the just reconnected magnetic field lines, a process taking place inside the EDR, will move outward (in a narrow region along the X-axis) carrying the “frozen-in” field electrons. The resulting electric field will be inward directed with respect to the X-line, while the electrons are propelled by the “sling effect”. However, we assume that it is highly unrealistic to assume an electron drift taking place inside the IDR. The electron convection velocity is much less than “the ballistic” electron velocity (due to the source) along the magnetic field lines. For only 500 eV electrons the ratio of convection to ballistic velocities will be ~1/7. In our view, a more pragmatic approach would be as follows: First, the electrons and ions are accelerated in the same manner at the source region (S) simultaneously by a mechanism associated with the “twin-Double Layer structure” in its explosive phase, as proposed by Sarafopoulos [11] . Then, both electrons and ions move outward from the source, with the difference that the electrons are apparently going to become adiabatic just outside the EDR (green rectangle in Figure 10), whereas the ions will become adiabatic just outside the IDR (blue rectangle for the same figure). The electrons, when leaving the EDR, ride on the magnetic field lines and produce a bifurcated electron current (inside the IDR), an electron FAC as it is shown in Figure 10. Therefore, the outward streaming electrons produce homo-parallel and mutually attracted currents plus the associated B_y guide fields. In this context the guide fields are indeed caused by an electron current, but not a Hall current; Hall currents do not exist at all. Moreover, the attracted electron currents tend to preserve the thin current sheet structure and, in this way, sustain a lengthened “ion channel” within the IDR. The ions outside the IDR will produce a bifurcated ion current (dashed red line in Figure 10), which inside the PS creates a B_y pattern similar to that of Hall current, too. However, in our situation these B_y deflections are related to the fluid scale (and not the IDR).

In summary, the earthward streaming electrons of the source (along the lines) obviously produce an electron FAC (and not a Hall current, as it was hypothesized in past works to occur very close to an X-line). A satellite (S/C) crossing (brown-dashed line trajectory within the IDR) the electron FAC current will detect the same B_y pattern as in the erroneously assumed situation of Hall current, which, according to its definition, flows as a convection current perpendicular to the plane of the B_z and the E_y-field.

In a similar way, the ions jetting tailward of the source (marked by S in Figure 11) will sustain the electron FACs flowing along the magnetic field lines enveloping the jet; therefore, the B_y deflections recorded by a satel- lite (inside the PS) will be due to the electron FACs and not the electron Hall current, although the pattern is the same. Therefore, based on the latter inference, we can argue that, in past research efforts, the B_y deflections

due to ion or electron FACs could have been erroneously credited to Hall electron currents (e.g., as in the statistical work of Ueno et al. [34] and Eastwood et al., [35] ).

Obviously, reconnection as magnetic field reconfiguration actually takes place; the major question is whether this process converts magnetic energy into mechanical one and produces energetic populations. In addition to what was presented in the preceded paragraph, the fundamental idea of transport of magnetic flux, via convecting magnetic field “frozen-in” plasma, may be very misleading in many cases, too. Although the schematic picture showing the circulating magnetic flux, in the Earth’s magnetosphere, is very attractive, however, it does not mean that such a circulation must be really based on the frozen-in condition or even the reconnection concept. The convection is probably not the dominant process when explosive events (during substorms) are underway; instead, we are of the opinion that a major driving mechanism that largely modulates the magnetosphere’s topology is the jetting ions out of the source and within an extended IDR. Thus, the dipolarization front may be built up by an earthward ion jet, and large scale MFRs propagating tailward (of the source) may originate from tailward ion jets [9] . At times with low geomagnetic activity, the ions and electrons alike will drift earthward performing E_y × B_z motions. At substorm-times, an “ion jet” (i.e., an exodus “ion channel”) streaming outwards from the acceleration source will readily cross the magnetic field lines. That is, in this way the “ion channel” essentially determines the extent (along the X-axis) of the IDR measuring, at times, a few R_E. The criterion for the IDR may be (a) that V_x @ V_x^ (where the V_x^ is the earthward directed component of the plasma-ion velocity being perpendicular to the magnetic field in very CPS) and (b) that the B_total @ B_z (i.e., the field is almost vertical to the current sheet plane). The extent of the “ion channel” is determined according to the same criterion, while the acceleration source will be much more localized. In general, the magnetic field reconfigures as the current systems progressively dictate; however, the plasma does not carry the magnetic field lines. Naturally, the positive (negative) B_z values are related to earthward (tailward) E_y × B_z ion drift motions; nevertheless, within the IDR we can even observe tailward flows with positive B_z values (as commonly were observed; for instance look at the work of Ohtani et al., [36] ).

It may be a profound truth that the concept of “merging or reconnection” is a gigantic pseudo-science (Alfvén, [37] [38] ); what is actually happening in magnetotail, may be far from any frozen-in condition or the reconnection process converting magnetic energy to mechanical one. Obviously, as reconnection we do not mean here a simple magnetic field reconfiguration; the main point constantly remains the mechanism accelerating particles up to ~1 MeV [39] - [41] . In a persistently thinned PS, where all the phenomena take place within the IDR, the FACs (and not the Hall currents) will be the ultimate source producing guide fields. If an IDR is locally established in tail, then the ions will cease drifting earthward within this region; consequently, a net earthward electron flow will be established. Furthermore, the IDR will become a cavity with increased charge densities (for both charge carriers); and such a cavity potentially leads to the “twin-DL structure” with repetitive activations as suggested by Sarafopoulos [11] . And the DLs would convert the magnetic energy into mechanical one and readily produce energetic particles.

The work of Runov et al. [4] was frequently cited by many authors; therefore, it would be constructive to support further our different explanation scheme for the presented observations. A scheme unrelated to the X-type reconnection model in which a few unanswered issues intrude, too.

The quadrapole magnetic field structure around the source was indeed observed; the question focuses on their ultimate driving mechanism. We have clearly suggested in this work that the B_y pattern was accomplished by FACs established within and outside the IDR. And we further discuss three worth noticing points in respect to this work:

1) They study B_y deflections caused by Hall currents obviously within the IDR. Hence, they observe proton bulk velocities up to ~1000 km∙s⁻¹ within the IDR. Naturally, a question is raised: Which is the mechanism, in the context of the X-line model, forcing the ions to be accelerated within the IDR at these high velocities? The “moving magnetic field lines” does not offer such a mechanism; probably the exact opposite happens. A highly localized source (as we have proposed, for instance, the twin-DL structure) accelerates both electrons and ions simultaneously and, in turn, eject them outward within the IDR.

2) They do not show plasma-electron bulk velocities perpendicular to the local magnetic field having higher velocities than those observed for protons (e.g., look at the simulations of Pritchett, [42] ). Thus, they do not actually observe Hall currents; they just make a conjecture. The presence and operation of FACs can therefore be expected with high probability.

3) The flow reverse of plasma from tailward to earthward occurred in the range of ~1 min; and this is the involved time unit for the analyzed quadrapole field. However, we have very clear results that such a reverse of plasma flow, and most importantly, a simultaneous reverse for the ~300 keV ions, happens even within 10 s [15] . Consequently, the studied region has wider extent than the source; a region in which the source populations sustain the related FACs.

In a locally thinned current sheet the demagnetized ions form the cross-tail bifurcated J_y-currents (Figure 12), which, along with the existing magnetic field, produce the inward directed magnetic forces F = J_y × B_x. Moreover, these ion currents and the homoparallel electron J_y-current (of akis) are mutually attracted as well. Thus, the PS locally “pinches” and the curvature radius of akis further decreases leading to even higher demagnetization rates. The resulting collapse of the negatively charged akis, being an integral part of the twin-DL structure, generates an electrostatic discharge within the space of DLs. The transient electric fields may accelerate energetic particles up to ~1 MeV, while the PS undergoes a local dipolarization. Therefore, around the “pinched” region, the cross-tail current is disrupted generating ±ΔB_z deflections. And the deflections (along with the produced ion jets within the IDR) are probably developed in two propagating fronts just outside the IDR: A strong front characterized with negative B_z values that moves tailward, and a positive B_z-front that moves earthward. The latter is usually described by the term “X-type reconnection”. If the negative ΔB_z deflection is weak then one would observe tailward plasma flows at the same time with the prevailing positive B_z values; a situation often verified by observations.