MOND is a paradigm that contends to account for the mass discrepancies in the Universe without invoking `dark' components, such as `dark matter' and `dark energy'. It does so by supplanting Newtonian dynamics and General Relativity, departing from them at very low accelerations. Having in mind readers who are historians and philosophers of science, as well as physicists and astronomers, I describe in this review the main aspects of MOND -- its statement, its basic tenets, its main predictions, and the tests of these predictions -- contrasting it with the dark-matter paradigm. I then discuss possible wider ramifications of MOND, for example the potential significance of the MOND constant, $a_0$, with possible implications for the roots of MOND in cosmology. Along the way I point to parallels with several historical instances of nascent paradigms. In particular, with the emergence of the Copernican world picture, that of quantum physics, and that of relativity, as regards their initial advent, their development, their schematic structure, and their ramifications. For example, the interplay between theories and their corollary laws, and the centrality of a new constant with converging values as deduced from seemingly unrelated manifestations of these laws. I demonstrate how MOND has already unearthed a number of unsuspected laws of galactic dynamics (to which, indeed, $a_0$ is central) predicting them a priori, and leading to their subsequent verification. I parallel the struggle of the new with the old paradigms, and the appearance of hybrid paradigms at such times of struggle. I also try to identify in the history of those established paradigms a stage that can be likened to that of MOND today.

Ogle et al. have fallaciously argued recently that fast-rotating disc galaxies break with the predictions of MOND: the 6 fastest rotators of the 23 galaxies in their sample appear to have higher rotational speeds than is consistent with the MOND relation between the baryonic mass of a galaxy, $M$, and its `rotational speed', $V$. They interpret this departure as a break in the observed $M-V$ relation from a logarithmic slope near the MOND-predicted $4$, to a shallow slope of $\approx 0$. However, Ogle et al. use the MAXIMAL rotational speed of the galaxies, $V_{max}$, not the ASYMPTOTIC one, $V_\infty$, which appears in the MOND prediction, $V_\infty^4=MGa_0$. Plotting their $M$ vs. $V_{max}$ pairs on an $M$ vs. $V_\infty$ plot from Lelli et al. (2016), they arrive erroneously at the above tension with MOND. The $H_\alpha$ rotation curves used by Ogle et al. are far too short reaching to probe the asymptotic regime, and determine $V_\infty$. However, it is well documented for fast rotators with observed, extended, HI rotation curves, that they can have $V_{max}$ considerably larger than the MOND-relevant $V_\infty$ [Noordermeer and Verheijen (NV) (2007) and others]. E.g., the fastest rotator in the NV sample has $V_{max}\approx 490{\rm ~km/s}$, but $V_\infty\approx 250{\rm ~km/s}$. NV also show that in a (MOND-irrelevant) $M-V_{max}$ plot the high-speed galaxies fall off the power-law line defined by the lower-speed ones, creating a break in the $M$ vs. $V$ relation, in just the way claimed by Ogle et al. But, when plotting the MOND-relevant $M$ vs. $V_\infty$ all galaxies fall near the same power-law relation, without a break.

I limelight and review a potentially crucial aspect of MOND: The near equality of the MOND acceleration constant, $a_0$ -- as deduced from local, galactic phenomena -- and cosmological parameters. To wit, $a_0\sim c H_0\sim c^2\Lambda^{1/2}\sim c^2/\ell_U$, where $H_0$ is the present value of the Hubble-Lema\^{i}tre constant, $\Lambda$ is the `cosmological constant', and $\ell_U$ is a cosmological characteristic length; e.g., the Hubble distance, or the de Sitter radius associated with $\Lambda$. In itself, this near equality has some important phenomenological consequences, such as the impossibility of black holes, and of cosmological strong lensing, in the MOND regime. More importantly perhaps, this `coincidence' may be a pointer to the `FUNDAMOND' -- the more basic theory underlying MOND phenomenology. The manners in which such a relation emerges in existing, underlying scheme of MOND are also reviewed, interlaced with examples of similar relations in other physical systems, between apparently-fundamental velocity, length, and acceleration constants. Such analogies may point the way to explanation of the MOND `coincidence'.

MOND has limelighted the fact that Newtonian dynamics (ND) and general relativity (GR) have not been verified at accelerations below MOND's $a_0$. In particular, we do not know that all the principles underlying ND or GR apply below $a_0$. I discuss possible breakdown of general covariance (GC) in this limit. This resonates well with MOND, which hinges on accelerations. Relaxing GC affords more freedom in constructing MOND theories. I exemplify this with a simplified theory whose gravitational Lagrangian is $\mathcal{L}_M\propto \ell_M^{-2}\mathcal{F}(\ell_M^{2}\mathcal{R})$, where $\mathcal{R}= g^{\mu\nu} (\Gamma^\gamma_{\mu\nu}\Gamma^\lambda_{\lambda\gamma}-\Gamma^\gamma_{\mu\lambda} \Gamma^\lambda_{\nu\gamma})/2$. $\Gamma^\gamma_{\mu\nu}$ is the Levi-Civita connection of a metric, $g_{\mu\nu}$, and $\ell_M=c^2/a_0$ is the MOND length. Requiring $\mathcal{F}(z)\rightarrow z+\zeta$, for $z\gg 1$ gives GR with a cosmological constant $\zeta c^{-4}a_0^2$ for high accelerations. In the MOND limit $\mathcal{F}'(z\ll 1)\propto z^{1/2}$. In the nonrelativistic limit the metric is of the form $g_{\mu\nu}\approx \eta_{\mu\nu}-2\phi\delta_{\mu\nu}$, as in GR, but the potential $\phi$ solves a MOND, nonlinear Poisson analog. This form of $g_{\mu\nu}$ also produces gravitational lensing as in GR only with the MOND potential. I show that this theory is a fixed-gauge expression of BIMOND, with the auxiliary metric constrained to be flat. The latter theory is thus a covariantized version of the former a-la St\"{u}ckelberg. This theory is also a special case of so-called $f(\mathcal{Q})$ theories -- aquadratic generalizations of `symmetric, teleparallel GR', which are, in turn, also equivalent to constrained BIMOND-type theories. (Abridged.)

I describe a heuristic model where MOND dynamics emerge in a universe viewed as a nearly spherical brane embedded in a higher-dimensional flat space. The brane, described by $\xi(\Omega)$, is of density $\sigma$ ($\xi$ and $\Omega$ are the radial and angular coordinates in the embedding space). The brane and matter -- confined to the brane and of density $\rho(\Omega)\ll\sigma$ -- are coupled to a potential $\varepsilon(\xi)$. I restrict myself to shallow perturbations, $\xi(\Omega)=\ell_0+\zeta(\Omega)$, $|\zeta|\ll\ell_0$. A balanced brane implies $\hat a_0\equiv\varepsilon'(\ell_0)\sim T/\sigma\ell_0$, $T$ is the brane tension, yielding for the velocity of small brane perturbations $c^2\sim T/\sigma\sim \ell_0\hat a_0$. But, $\hat a_0$ plays the role of the MOND acceleration constant in local gravitational dynamics; so $\hat a_0\sim c^2/\ell_0$. What we, in the brane, perceive as the gravitational potential is $\phi\equiv\varepsilon[\xi(\Omega)]\approx \phi_0+\hat a_0\zeta$. Aspects of MOND that may emerge naturally as geometrical properties are: a. The special role of acceleration in MOND, and why it is an acceleration, $a_0$, that marks the transition from the standard dynamics much above $a_0$ to scale-invariant dynamics much below $a_0$. b. The intriguing connection of $a_0$ with cosmology. c. The Newtonian limit corresponds to local departure $|\zeta|\ll\ell_0$; i.e., $\phi-\phi_0\sim a_0\zeta\ll a_0\ell_0\sim c^2$ - whereas relativity enters when $|\zeta|\not\ll\ell_0$. The model also opens new vistas for extension, e.g., it points to possible dependence of $a_0$ on $\phi$, and to $a_0$ losing its status and meaning altogether in the relativistic regime. The required global balance of the brane might solve the `old' cosmological-constant problem. I discuss possible connections with the nearly-de-Sitter nature of our Universe. (Abridged.)

Intermediate-richness galaxy groups are an important test ground for MOND. First, they constitute a distinct type of galactic systems, with their own evolution histories and underlying physical processes; secondly, they probe little-chartered regions of parameter space, as they have baryonic masses similar to massive galaxies, and similar velocity dispersions, but much larger sizes -- similar to cluster cores (or even to clusters), but much lower dispersions. Importantly in the context of MOND, they have the lowest internal accelerations reachable inside galactic systems. I analyze a sample of 56 medium-richness groups having a large number ($\ge 15$) of members with measured velocities. The groups obey the deep-MOND, baryonic-mass-velocity-dispersion relation, $M_MGa_0=(81/4)\sigma^4$, with individual, MOND $M_M/L_K$ ratios of order $1$ solar unit, with $(M_M/L_K)_{median}=0.7$ s.u. compared with the much larger Newtonian $M_d/L_K$ -- several tens s.u., and $(M_d/L_K)_{median}=37$ s.u. The same MOND relation describes dwarf spheroidals -- 2-3 orders smaller in size, and 7-8 orders lower in mass. The groups conformation to the MOND relation is equivalent to their lying on the deep-MOND branch of the `acceleration-discrepancy relation', $g\approx (g_N a_0)^{1/2}$, for $g$ as low as a few percents of $a_0$ ($g_N$ is the baryonic, Newtonian, gravitational acceleration, and $g$ the actual one). This argues against a breakdown of MOND at extremely low accelerations. This conformation also argues against the hypothesis that the remaining MOND conundrum in cluster cores bespeaks a breakdown of MOND on large-distance scales; our groups are as large as cluster cores, but do not show obvious disagreement with MOND. I also discuss the possible presence of the idiosyncratic, MOND external-field effect.

We consider disk stability in the quasi-linear formulation of MOND (QUMOND), the basis for some $N$-body integrators. We generalize the Toomre criterion for the stability of disks to tightly wound, axisymmetric perturbations. We apply this to a family of thin exponential disks with different central surface densities. By numerically calculating their QUMOND rotation curves, we obtain the minimum radial velocity dispersion required for stability against local self-gravitating collapse. MOND correctly predicts much higher rotation speeds in low surface brightness galaxies (LSBs) than does Newtonian dynamics without dark matter. Newtonian models thus require putative very massive halos, whose inert nature implies they would strongly stabilize the disk. MOND also increases the stability of galactic disks, but in contradistinction to Newtonian gravity, this extra stability is limited to a factor of 2. MOND is thus rather more conducive to the formation of bars and spiral arms. Therefore, observation of such features in LSBs could be problematic for Newtonian galaxy models. This could constitute a crucial discriminating test. We quantitatively account for these facts in QUMOND. We also compare numerical QUMOND rotation curves of thin exponential disks to those predicted by two algebraic expressions commonly used to calculate MOND rotation curves. For the choice that best approximates QUMOND, we find the circular velocities agree to within 1.5% beyond $\approx 0.5$ disk scale lengths, regardless of the central surface density. The other expression can underestimate the rotational speed by up to 12.5% at one scale length, though rather less so at larger radii.

Galaxy groups, which have hardly been looked at in MOND, afford probing the acceleration discrepancies in regions of system-parameter space that are not accessible in well-studied galactic systems, such as galaxies, galaxy clusters, and dwarf-spheroidal satellites of galaxies. Groups are typically the size of galaxy-cluster cores, but have masses typically only a few times that of a single galaxy. Accelerations in groups get far below those in galaxies, and far below the MOND acceleration. So much so, that many groups might be affected by the external-field effect, which is unique to MOND, due to background accelerations. Here, I analyze the MOND dynamics of 53 galaxy groups, recently catalogued in 3 lists. Their Newtonian, K-band, dynamical $M/L$ ratios are a few tens to several hundreds solar units, with $\langle{M_d/L_K}\rangle= (56, 25, 30)~M_{\odot}/L_{\odot}$, respectively for the 3 lists; thus evincing very large acceleration discrepancies. I find here that MOND requires dynamical $M_M/L_K$ values of order $1~M_{\odot}/L_{\odot}$, with $\langle{M_M/L_K}\rangle=(0.8, 0.56, 1.0) ~M_{\odot}/L_{\odot}$, for the 3 lists, which are in good agreement with population-synthesis stellar values, and with those found in individual galaxies. MOND thus accounts for the observed dynamics in those groups with baryons alone, and no need for dark matter -- an important extension of MOND analysis from galaxies to galactic systems, which, to boot, have characteristic sizes of several hundred kiloparsecs, and accelerations much lower than probed before -- only a few percent of MOND's $a_0$. The acceleration discrepancies evinced by these groups thus conform to the deep-MOND prediction: $g\approx (g_Na_0)^{1/2}$, down to these very low accelerations ($g$ is the measured, and $g_N$ the baryonic, Newtonian acceleration).

The dwarf galaxy NGC1052-DF2 has recently been identified as potentially lacking dark matter. If correct, this could be a challenge for MOND, which predicts that low surface brightness galaxies should evince large mass discrepancies. However, the correct prediction of MOND depends on both the internal field of the dwarf and the external field caused by its proximity to the giant elliptical NGC1052. Taking both into consideration under plausible assumptions, we find $\sigma_{\rm MOND} = 13.4^{+4.8}_{-3.7}\;\mathrm{km}\,\mathrm{s}^{-1}$. This is only marginally higher than the claimed 90\% upper limit on the velocity dispersion ($\sigma < 10.5\;\mathrm{km}\,\mathrm{s}^{-1}$), and compares well with the observed root mean square velocity dispersion ($\sigma = 14.3\;\mathrm{km}\,\mathrm{s}^{-1}$). We also discuss a few caveats on both the observational and theoretical side. On the theory side, the internal virialization time in this dwarf may be longer that the time scale of variation of the external field. On the observational side, the paucity of data and their large uncertainties call for further analysis of the velocity dispersion of NGC1052-DF2, to check whether it poses a challenge to MOND or is a success thereof.

Aniyan et al. (2018) have recently published direct measurements of the baryonic mass distribution and the rotation curve of the almost-face-on disc galaxy NGC 628. While its very low inclination makes this galaxy anything but ideal for rotation-curve analysis, these new results, taken at face value, have interesting ramifications for MOND. The methods employed afford a direct determination of the stellar mass in the disc, which, in turn, affords a parameter-free MOND prediction of the rotation curve, which I show. In comparison, the dark-matter fits that Aniyan et al. present have two free parameters. To boot, these results further negate an earlier claim deleterious to MOND. It is that stellar $M/L$ ratios deduced from vertical velocity dispersions in disc galaxies are rather lower than what is required by MOND fits to rotation curves. Specifically, it was claimed that even high-surface-density discs are, by and large, sub maximal; viz., that they show substantial mass discrepancies near their center. This is contrary to the prediction of MOND that in such high-acceleration regions only small discrepancies should appear, if any to speak of. Such claims of low $M/L$ values have been rebutted before, and the fallacy that may have led to them pointed out. The new results strongly buttress these rebuttals.

Genzel et al. have recently published the rotation curves of six high-redshift disc galaxies ($z\sim 0.9-2.4$), which they find to be `baryon dominated' within the studies radii. While not up to the standard afforded by data available for analysis in the nearby Universe, these data are valuable in constraining cosmological evolution of either DM scenarios, or -- as I discuss here -- $z$-dependence of MOND. Indeed, these results, if taken at face value, teach us useful lessons in connection with MOND. a. The dynamical accelerations at the half-light radii, found by Genzel et al., are rather high compared with the MOND acceleration constant, as measured in the nearby Universe: $g(R_{1/2})= (3-11)a_0$. MOND then predicts fractions of `phantom matter' at $R_{1/2}$ of at most a few tens of percents, which, galaxy by galaxy, agree well with what Genzel et al. find. b. The asymptotic rotational speeds predicted by MOND from the baryonic-mass estimates of Genzel et al. are substantially lower ($0.55-0.75$) than the maximal speeds of the RCs. MOND thus predicts a substantial decline of the RCs beyond the maximum. This too is in line with what Genzel et al. find. c. Arguably, the most important lesson is that the findings of Genzel et al. cast very meaningful constraints on possible variation of $a_0$ with cosmic time. For example, they all but exclude a value of the MOND constant of $\sim 4a_0$ at $z\sim 2$, excluding, e.g., $a_0\propto (1+z)^{3/2}$.

We highlight phenomenological aspects of Verlinde's recent proposal to account for the mass anomalies in galactic systems without dark matter -- in particular in their relation to MOND. Welcome addition to the MOND lore as it is, this approach have reproduced, so far, only a small fraction of MOND phenomenology, and is still rather tentative, both in its theoretical foundations and in its phenomenology. What Verlinde has extracted from this approach, so far, is a formula -- of rather limited applicability, and with no road to generalization in sight -- for the effective gravitational field of a spherical, isolated, static baryonic system. This formula cannot be used to calculate the gravitational field of disk galaxies, with their rich MOND phenomenology. Notably, it cannot predict their rotation curves, except asymptotically. It does not apply to the few-, or many-body problem; so, it cannot give, e.g., the two-body force between two galaxies, or be used to conduct N-body calculations of galaxy formation, evolution, and interactions. The formula cannot be applied to the internal dynamics of a system embedded in an external field, where MOND predicts important consequences. etc. MOND is backed by full-fledged, Lagrangian theories that can be, and are, routinely applied to all the above phenomena, and more. Verlinde's formula, as it now stands, strongly conflicts with solar-system and possibly earth-surface constraints, and cannot fully account for the mass anomalies in the cores of galaxy clusters (a standing conundrum in MOND). The recent weak-lensing test of the formula is, in fact, testing a cornerstone prediction of MOND, one that the formula does reproduce, and which has been tested before in the very same way.

Keller and Wadsley (2016) have smugly suggested, recently, that the end of MOND may be in view. This is based on their claim that their highly-restricted sample of $\Lambda$CDM-simulated galaxies are "consistent" with the observed MOND mass-discrepancy-acceleration relation (MDAR), in particular, with its recent update by McGaugh et al. (2016), based on the SPARC sample. From this they extrapolate to "$\Lambda$CDM is fully consistent" with the MDAR. I explain why these simulated galaxies do not show that $\Lambda$CDM accounts for the MDAR. a. Their sample of simulated galaxies contains only 18 high-mass galaxies, within a narrow range of one order of magnitude in baryonic mass, at the very high end of the observed, SPARC sample, which spans 4.5 orders of magnitude in mass. More importantly, the simulated sample has none of the low-mass, low-acceleration galaxies -- abundant in SPARC -- which encapsulate the crux and the nontrivial aspects of the predicted and observed MDAR. The low-acceleration part of the simulated MDAR is achieved, rather trivially, from the flattish-velocity-curve regions of the simulated high-mass galaxies. b. Half of the simulated galaxies have "wrong" rotation curves that differ greatly from any observed ones. This, does not prevent these wrong galaxies from lying on the observed MDAR (for trivial reasons, again). They, in fact, define the high-acceleration branch of the simulated MDAR. c. To boot, even if $\Lambda$CDM were made "consistent" with the MDAR through the elaborate adjustments that the simulations allow, this would not obviate MOND, which predicts much more than the MDAR.

McGaugh et al. (2016) have used their extensive SPARC sample to update the well-known mass-discrepancy-acceleration relation (MDAR), which is one of the major predicted "MOND laws". This is not a newly discovered relation. Rather, it improves on the many previous studies of it, with more and better data. Like its precedents, it bears crucial ramifications for the observed dynamical anomalies in disc galaxies, and, in particular, on their resolution by the MOND paradigm. Their result, indeed, constitute a triumph for MOND. However, unlike previous analyses of the MDAR, McGaugh et al. have chosen to obfuscate the MOND roots of their analysis, and its connection with, and implications for, this paradigm. For example, the fitting formula they use, seemingly as a result of some unexplained inspiration, follows in its salient properties from the basic tenets of MOND, and has already been used in the past in several MOND analyses. No other possible origin for such a function is known. Given that this formula had already been shown to reproduce correctly the observed rotation curves from the baryon distribution (as a MOND effect), it must have been clear, a priory, that it should describe correctly the MDAR, which is but a summary of rotation curves. The present paper corrects these oversights -- bringing to light the deep connections with MOND, suppressed by McGaugh et al. It also gives due credit to previous works, and discusses some new, important, but less known, aspects of this MOND relation. (Substantially abridged.)

I derive a new MOND relation for pure-disc galaxies: The `dynamical' central surface density, $\Sigma^0_D$, deduced from the measured velocities, is a universal function of only the true, `baryonic' central surface density, $\Sigma^0_B$: $\Sigma^0_D=\Sigma_M \mathcal{S}(\Sigma^0_B/\Sigma_M)$, where $\Sigma_M\equiv a_0/2\pi G$ is the MOND surface density constant. This surprising result is shown to hold in both existing, nonrelativistic MOND theories. $\mathcal{S}(y)$ is derived: $\mathcal{S}(y)=\int_0^y\nu(y')dy'$, with $\nu(y)$ the interpolating function of the theory. The relation aymptotes to $\Sigma^0_D=\Sigma^0_B$ for $\Sigma^0_B\gg\Sigma_M$, and to $\Sigma^0_D=(4\Sigma_M\Sigma^0_B)^{1/2}$ for $\Sigma^0_B\ll\Sigma_M$. This study was prompted by the recent finding of a correlation between related attributes of disc galaxies by Lelli et al. (2016). The MOND central-surface-densities relation agrees very well with these results.

Galactic systems, and the Universe at large, exhibit large dynamical anomalies: The observed matter in them falls very short of providing enough gravity to account for their dynamics. The mainstream response to this conundrum is to invoke large quantities of `dark matter' -- which purportedly supplies the needed extra gravity -- and also of `dark energy', to account for further anomalies in cosmology, such as the observed, accelerated expansion. The MOND paradigm offers a different solution: a breakdown of standard dynamics (gravity and/or inertia) in the limit of low accelerations -- below some acceleration $a_0$. In this limit, dynamics become space-time scale invariant, and is controlled by a gravitational constant $\mathcal{A}_0\equiv Ga_0$, which replaces Newton's $G$. With the new dynamics, the various detailed manifestations of the anomalies in galaxies are predicted with no need for dark matter. The cosmological anomalies could, but need not have to do with small accelerations. For example, the need for dark matter in accounting for the expansion history of the Universe is eliminated if the relevant gravitational constant is $\approx 2\pi G$. Such a `renormalization' of $G$ could be a dimensionless parameter of a MOND theory. The constant $a_0$ turns out to carry cosmological connotations, in that $2\pi a_0\approx cH_0\approx c^2(\Lambda/3)^{1/2}$, where $H_0$ is the present expansion rate of the Universe, and $\Lambda$ the measured `cosmological constant'. There are MOND theories in which this `coincidence' is natural. I draw on enlightening historical and conceptual analogies from quantum theory to limelight aspects of MOND. I also explain how MOND may have strong connections with effects of the quantum vacuum on local dynamics.

Angus et al. (2015) have recently faulted MOND as follows: Studying thirty disc galaxies from the DiskMass survey, they derive the profiles of velocity dispersion perpendicular to the discs as predicted by MOND, $\sigma_M(r)$, and compare them with $\sigma(r)$, measured as part of the DiskMass project. This is a new test of MOND, different from rotation-curve analysis. A nontrivial accomplishment of MOND -- not discussed by Angus et al. -- is that the shapes of $\sigma_M$ and $\sigma$ agree very well, i.e, $\eta(r)\equiv\sigma_M(r)/\sigma(r)$ is well consistent with being $r$-independent (while $\sigma$ and $\sigma_M$ are strongly $r$ dependent). The fault found with MOND was that $\eta$ is systematically above 1 (with an average of about 1.3). I have suggested to Angus et al. that the fault may lie with the DiskMass dispersions, which may well be too low for the purpose at hand: Being based on population-integrated line profiles, they may be overweighed by younger populations, known to have much smaller dispersions, and scale heights, than the older populations, which weigh more heavily on the light distributions. I discuss independent evidence that supports this view, and show, besides, that if the DiskMas dispersions are underestimates by only 25 %, on average, the MOND predictions are in full agreement with the data, in shape and magnitude. Now, Aniyan et al. (2015) have questioned the DiskMass $\sigma$ on the same basis. They show for the solar column in the Milky Way that: " Combining the (single) measured velocity dispersion of the total young + old disc population... with the scale height estimated for the older population would underestimate the disc surface density by a factor of $\sim 2.$" If this mismatch found for the Milky Way is typical, correcting for it would bring the measured DiskMass $\sigma(r)$ to a remarkable agreement with the predicted MOND $\sigma_M(r)$.

MOND reduces greatly the mass discrepancy in clusters of galaxies, but does leave a consistent global discrepancy of about a factor of two. It has been proposed, within the minimalist and purist MOND, that clusters harbor some indigenous, yet-undetected, cluster baryonic (dark) matter (CBDM). Its total amount has to be comparable with that of the observed hot gas. Following an initial discovery by van Dokkum & al. (2015a), Koda & al. (2015) have recently identified more than a thousand ultra-diffuse galaxy-like objects (UDGs) in the Coma cluster. Robustness of the UDGs to tidal disruption seems to require, within Newtonian dynamics, that they are much more massive than their observed stellar component. Here, I propound that a considerable fraction of the CBDM is internal to UDGs, which endows them with robustness. The rest of the CBDM objects formed in now-disrupted kin of the UDGs, and is dispersed in the intracluster medium. While the discovery of cluster UDGs is not in itself a resolution of the MOND cluster conundrum, it lends greater qualitative plausibility to CBDM as its resolution, for reasons I discuss. Alternatively, if the UDGs are only now falling into Coma, their large size and very low surface brightness could result from the adiabatic inflation due to the MOND external-field effect, as described in Brada & Milgrom (2000). I also consider briefly solutions to the conundrum that invoke more elaborate extensions of purist MOND, e.g., that in clusters, the MOND constant takes up larger-than-canonical values of the MOND constant.

Accepting that galactic mass discrepancies are due to modified dynamics, I show why it is specifically the MOND paradigm that is pointed to cogently. MOND is thus discussed here as a special case of a larger class of modified dynamics theories whereby galactic systems with large mass discrepancies are described by scale-invariant dynamics. This is a novel presentation that uses more recent, after-the-fact insights and data (largely predicted beforehand by MOND). Starting from a purist set of tenets, I follow the path that leads specifically to the MOND basic tenets. The main signposts are: (i) Space-time scale invariance underlies the dynamics of systems with large mass discrepancies. (ii) In these dynamics, $G$ must be replaced by a single "scale-invariant" gravitational constant, Q0 (in MOND, Q0=A0=Ga0, where a0 is MOND's acceleration constant). (iii) Universality of free fall points to the constant q0=Q0/G as the boundary between the G-controlled, standard dynamics, and the Q0-controlled, scale-invariant dynamics (in MOND, q0=a0). (iv) Data clinches the case for q0 being an acceleration (MOND).

The proximity of the MOND acceleration constant with cosmological accelerations -- for example, a0 cH0/2pi -- points to its possibly decreasing with cosmic time. I begin to consider the secular changes induced in galactic systems by such presumed variations, assumed adiabatic. It is important to understand these effects, in isolation from other evolutionary influences, in order to identify or constrain a0 variations by detection of induced effects, or lack thereof. I find that as long as the system is fully in the deep-MOND regime -- as applies to many galactic systems -- the adiabatic response of the system obeys simple scaling laws. For example, in a system that would be stationary for fixed a0, the system expands homologously as a0^{-1/4}, while internal velocities decrease uniformly as a0^{1/4}. If a0 is proportional to cH at all relevant times, this change amounts to a factor of ~ 2.5 since redshift 10. For a system stationary in a rotating frame, the angular frequency decreases as a0^{1/2}. The accelerations increase relative to a0 as a0^{-1/4}, pushing the system towards the Newtonian regime. All this follows from the appearance of a0 in MOND and the scale invariance of the deep-MOND limit -- two basic tenets of MOND. More complicated evolution ensues when parts of the system become Newtonian, or are so from inception. For example, these parts may become unstable, not being protected by MOND's stabilizing effects. The existence of such regions also modifies the MONDian regime, since they affect the potential everywhere, and since constituents might migrate between the Newtonian and MONDian regimes. Studying these last effects would require detailed numerical calculations.

A general account of MOND theory is given. I start with the basic tenets of MOND, which posit departure from standard dynamics in the limit of low acceleration -- below an acceleration constant a0 -- where dynamics become scale invariant. I list some of the salient predictions of these tenets. The special role of a0 and its significance are then discussed. In particular, I stress its coincidence with cosmologically relevant accelerations. The deep-MOND limit and the consequences of its scale invariance are considered in some detail. General aspects of MOND theories are then described, after which I list briefly presently known theories, both nonrelativistic and relativistic. Most full-fledged theories modify the gravitational action, hinge on a0, introduce an interpolating function between the low and high accelerations, and obey MOND requirements in the opposite limits. These theories have much heuristic value as proofs of various concepts (e.g., that covariant MOND theories can be written with correct gravitational lensing). But, probably, they are, at best, effective theories of limited applicability. I then outline several other promising approaches to constructing MOND theories that strive to obtain MOND as an effective theory from deeper concepts, for example, by modifying inertia and/or gravity as a result of interactions with some omnipresent agent. Some theories do enjoy a natural appearance of a cosmological-constant-like contribution that, furthermore, exhibits the observed connection with a0. However, none were shown to address fully the mass discrepancies in cosmology and structure formation that are otherwise explained by cosmological dark matter. We have no clues as to whether and how MOND aspects enter non-gravitational phenomena, but I discuss briefly some possibilities.

An important and useful relation is known to hold in two specific MOND theories. It pertains to low-acceleration, isolated systems of pointlike masses, m_p, at positions r_p, subject to gravitational forces F_p. It reads sum_p r_p.F_p=-(2/3)(Ga0)^{1/2}[(\sum_p m_p)^{3/2}-\sum_p m_p^{3/2}]; a0 is the MOND acceleration constant. Here I show that this relation holds in the nonrelativistic limit of any modified-gravity MOND theory. It follows from only the basic tenets of MOND (as applied to such theories): departure from standard dynamics at accelerations below a0, and space-time scale invariance in the nonrelativistic, low-acceleration limit. This implies space-dilatation invariance of the static, gravitational-field equations, which, in turn, leads to the above point-mass virial relation. Thus, the various MOND predictions and tests based on this relation hold in any modified-gravity MOND theory. Since we do not know that any of the existing MOND theories point in the right direction, it is important to identify such predictions that hold in a much larger class of theories. Among these predictions are the MOND two-body force for arbitrary masses, and a general mass-velocity-dispersion relation of the form sigma^2=(2/3)(MGa0)^{1/2}[1-\sum_p (m_p/M)^{3/2}], where M is the total mass.

I consider the weak field limit (WFL) of bimetric MOND (BIMOND): the lowest order in the small departures Hmn and H'mn from double Minkowski space-time. In particular, I look at propagating solutions, for a favorite subclass of BIMOND. The WFL splits into two sectors for two linear combinations, Hmn+ and Hmn-, of Hmn and H'mn. The Hmn+ sector is equivalent to the WFL of general relativity (GR), with its gauge freedom, and has the same vacuum gravitational waves. The Hmn- sector is fully nonlinear even for the weakest Hmn-, and inherits none of the coordinate gauge freedom. The equations of motion are scale invariant in the deep-MOND limit of purely gravitational systems. In these last two regards, the BIMOND WFL is greatly different from that of other bimetric theories studied to date. Despite the strong nonlinearity, an arbitrary pair of harmonic GR wave packets of Hmn and H'mn moving in the same direction, is a solution of the (vacuum) BIMOND WFL.

MOND predicts a number of laws that galactic systems should obey irrespective of their complicated, haphazard, and mostly unknowable histories -- as Kepler's laws are obeyed by planetary systems. The main purpose of this work is to show how, and to what extent, these MOND laws follow from only the paradigm's basic tenets: departure from standard dynamics at accelerations a<~a0, and space-time scale invariance in the limit a<<a0. Such predictions will be shared by all MOND theories that embody these premises. This is important because we do not know which of the existing MOND theories, if any, is a step in the right direction. In the Newtonian-dynamics-plus-dark-matter paradigm, the validity of such clear-cut laws -- which tightly constrain baryons, `dark matter', and their mutual relations -- is contrary to expectations.

I first give a succinct account of the MOND paradigm--emphasizing the centrality of scale invariance in the nonrelativistic, deep-MOND limit--and describing rudiments of its phenomenology. I then present my credo, and some generalities, concerning existing MOND theories. Then I concentrate on one relativistic formulation of MOND in the form of a bimetric theory (BIMOND). I describe its various limits: the weak field, with application to gravitational waves, the nonrelativistic limit, and their further deep-MOND (low acceleration) limits, which are scale invariant. Other aspects of BIMOND that have been explored are aspects of cosmology, matter fluctuations in cosmology, and matter-twin-matter interactions. BIMOND remains largely unexplored, despite its promise in several regards: It tends to GR for a0 goes to 0 (a0 is the MOND constant); it has a simple nonrelativistic limit; it describes gravitational lensing correctly; and, it has a generic appearance of a cosmological-constant term that is of order a0^2/c^4, as observed.

We employ recently published measurements of the velocity dispersions in the newly discovered dwarf satellite galaxies of Andromeda to test our previously published predictions of this quantity. The data are in good agreement with our specific predictions for each dwarf made a priori with MOND, with reasonable stellar mass-to-light ratios, and no dark matter, while Newtonian dynamics point to quite large mass discrepancies in these systems. MOND distinguishes between regimes where the internal field of the dwarf, or the external field of the host, dominates. The data appear to recognize this distinction, which is a unique feature of MOND not explicable in LCDM.

MOND predicts that the asymptotic gravitational potential of an isolated, bounded (baryonic) mass, M, is phi(r)=(MGa0)^{1/2}ln(r); a0 is the MOND constant. Relativistic MOND theories predict that the lensing effects of M are dictated by phi(r) as general-relativity lensing is dictated by the Newtonian potential. Thus, MOND predicts that the asymptotic Newtonian potential deduced from galaxy-galaxy gravitational lensing will have: (1) a logarithmic r dependence, and (2) a normalization (parametrized standardly as 2s^2) that depends only on M: s=(MGa0/4)^{1/4}. I compare these predictions with recent results of galaxy-galaxy lensing, and find agreement on all counts. For the "blue"-lenses subsample ("spiral" galaxies) MOND reproduces the observations well with an r'-band M/L of 1-3 solar units, and for "red" lenses ("elliptical" galaxies) with M/L of 3-6 solar units, both consistent with baryons only. In contradistinction, Newtonian analysis requires, typically, M/L values of about 130 solar units, bespeaking a mass discrepancy of a factor of about 40. Compared with the staple, rotation-curve tests, MOND is here tested in a wider population of galaxies, through a different phenomenon, using relativistic test objects, and is probed to several-times-lower accelerations--as low as a few percent of a0.

We compare the recently published velocity dispersions for seventeen Andromeda dwarf spheroidals with estimates of the MOND predictions, based on the luminosities of these dwarfs, with reasonable stellar M/L values, and no dark matter. We find that the two are consistent within the uncertainties. We further predict the velocity dispersions of another ten dwarfs for which only photometric data are currently available.

Cosmological models that invoke warm or cold dark matter can not explain observed regularities in the properties of dwarf galaxies, their highly anisotropic spatial distributions, nor the correlation between observed mass discrepancies and acceleration. These problems with the standard model of cosmology have deep implications, in particular in combination with the observation that the data are excellently described by Modified Newtonian Dynamics (MOND). MOND is a classical dynamics theory which explains the mass discrepancies in galactic systems, and in the universe at large, without invoking dark entities. MOND introduces a new universal constant of nature with the dimensions of acceleration, a0, such that the pre-MONDian dynamics is valid for accelerations a >> a0, and the deep MONDian regime is obtained for a << a0, where space-time scale invariance is invoked. Remaining challenges for MOND are (i) explaining fully the observed mass discrepancies in galaxy clusters, and (ii) the development of a relativistic theory of MOND that will satisfactorily account for cosmology. The universal constant a0 turns out to have an intriguing connection with cosmology: \bar a0 == 2 pi a0 \approx c H_0 \approx c^2(Lambda/3)^{1/2}. This may point to a deep connection between cosmology and internal dynamics of local systems.

Different formulations of MOND predict somewhat different rotation curves for the same mass distribution. Here I consider a global attribute of the rotation curve that might provide a convenient discriminant between theories when applied to isolated, pure-disk galaxies that are everywhere deep in the MOND regime. This parameter is Q=<V^2>/V0^2, where <V^2> is the mean squared rotational speed of the galaxy, and V0 is the asymptotic (constant) rotational speed. The comparison between the observed and predicted values of Q is oblivious to the distance, the inclination, the mass, and the size of the disk, and to the form of the interpolating function. For the known modified-gravity theories Q is predicted to be a universal constant (independent of the mass distribution in the disk): Q=2/3. The predicted Q value for modified-inertia theories does depend on the mass distribution. However, surprisingly, I find here that it varies only little among a very wide range of mass distributions, Q=0.73+-0.01. While the difference between the theories amounts to only about 5 percent in the predicted RMS velocity, a good enough sample of galaxies may provide the first discerning test between the two classes of theories.

The gravitational fields of two isolated ellipticals, NGC 720 and NGC 1521, have been recently measured to very large galactic radii (~100 and ~200 kpc), assuming hydrostatic balance of the hot gas enshrouding them. They afford, for the first time to my knowledge, testing MOND in ellipticals with force and quality that, arguably, approach those of rotation-curve tests in disk galaxies. In the context of MOND, it is noteworthy that the measured accelerations span a wide range, from more than 10a0 to about a0/10, unprecedented in individual ellipticals. I find that MOND predicts correctly the measured dynamical mass runs (apart from a possible minor tension in the inner few kpc of NGC 720, which might be due to departure from hydrostatic equilibrium): The predicted mass discrepancy increases outward from none, near the center, to ~10 at the outermost radii. The implications for the MOND-vs-dark-matter controversy go far beyond the simple fact of two more galaxies conforming to MOND.

I discuss a novel MOND effect that entails a correction to the dynamics of isolated mass systems even when they are deep in the Newtonian regime: systems whose extent R<<Rm, where Rm=sqrt(GM/a0) is the MOND radius, and M the total mass. Interestingly, even if the MOND equations approach Newtonian dynamics arbitrarily fast at high accelerations, this correction decreases only as a power of R/Rm. The effect appears in formulations of MOND as modified gravity, governed by generalizations of the Poisson equation. The MOND correction to the potential is a quadrupole field G*Pij*Ri*Rj, where R is the radius from the center of mass. In QUMOND (quasilinear MOND), Pij=-qQij/Rm^5, where Qij is the quadrupole moment of the system, and q>0 is a numerical factor that depends on the interpolating function. For example, the correction to the Newtonian force between two masses, m and M, a distance L apart (L<<Rm), is Fa=2q(L/Rm)^3(mM)^2(M+m)^{-3}a0 (attractive). For generic MOND theories, which approach Newtonian dynamics quickly for accelerations beyond a0, the predicted strength of the effect in the Solar system is rather much below present testing capabilities. In MOND theories that become Newtonian only beyond k*a0, the effect is enhanced by k^2.

I discuss theories that describe fully nonlinear physics, while being practically linear (PL), in that they require solving only linear differential equations. These theories may be interesting in themselves as manageable nonlinear theories. But, they can also be chosen to emulate genuinely nonlinear theories of special interest, for which they can serve as approximations. The idea can be applied to a large class of nonlinear theories, exemplified here with a PL analogs of scalar theories, and of Born-Infeld (BI) electrodynamics. The general class of such PL theories of electromagnetism are governed by a Lagrangian L=-(1/2)F_mnQ^mn+ S(Q_mn), where the electromagnetic field couples to currents in the standard way, while Qmn is an auxiliary field, derived from a vector potential that does not couple directly to currents. By picking a special form of S(Q_mn), we can make such a theory similar in some regards to a given fully nonlinear theory, governed by the Lagrangian -U(F_mn). A particularly felicitous choice is to take S as the Legendre transform of U. For the BI theory, this Legendre transform has the same form as the BI Lagrangian itself. Various matter-of-principle questions remain to be answered regarding such theories. As a specific example, I discuss BI electrostatics in more detail. As an aside, for BI, I derive an exact expression for the short-distance force between two arbitrary point charges of the same sign, in any dimension.

Translation of a popular article on MOND vs. dark matter that appeared, in Hebrew, in the Israeli magazine "Odyssea", dedicated to "thoughts and ideas in the forefront of science and philosophy".

After a succinct review of the MOND paradigm--with its phenomenology, and its various underlying theories--I concentrate on so called modified inertia (MI) formulations of MOND, which have so far received only little attention. These share with all MOND theories the salient MOND predictions, such as asymptotically flat rotation curves, and the universal mass-asymptotic-speed relation. My emphasis here is, however, on the fact that MI theories can differ substantially from their "modified-gravity" (MG) kin in predicting other phenomena. Because MI theories are non local in time, MOND effects depend on the full trajectory of a system, not only on its instantaneous state, as in MG theories. This may lead to rather different predictions for, e.g., the external-field effect (EFE): A subsystem, such as a globular cluster or a dwarf galaxy, moving in the field of a mother galaxy, or a galaxy in a cluster, may be subject to an EFE that depends on the accelerations all along its orbit, not only on the instantaneous value. And, it is even possible to construct MI theories with practically no EFE. Other predictions that may differ are also discussed. Since we do not yet have a full fledged, modified-inertia formulation, simple, heuristic models have been used to demonstrate these points.

MOND predictions are compared with the mass discrepancy, Gamma (the dynamical-to-baryon mass ratio) deduced from the recently measured rotation curve, for the gas-rich, dwarf galaxy KK246, "the only galaxy observed in the local void". KK246 is special in at least two regards: a. It is, to my knowledge, the record holder for the largest mass discrepancy deduced from a rotation curve, Gamma= 15. b. It is very isolated, residing in a large, very empty void. I also discuss another extreme case: Andromeda IV, a dwarf considered here for the first time in light of MOND, with a very large mass discrepancy, Gamma =12, also conforming accurately to the MOND prediction. In both cases, MOND predicts Gamma, or the total dynamical mass at the last observed radius, from only the knowledge of the small mass of baryons. If MOND is accepted as the root of the mass discrepancy, these are just two more expected, albeit reassuring, conformities. However, in the framework of the dark-matter paradigm--where the mass discrepancy is strongly dependent on the buildup history of a galaxy--every new such conformity with a tight law is another difficult-to-understand surprise, and does carry a new import: What, in the LCDM paradigm, would prevent such galactic baryons from residing in a halo of half, or twice, the observed rotational velocities, instead of selecting exactly the velocities predicted by MOND? This conundrum is especially poignant for KK246, whose great isolation points to a relatively unique buildup history. This note underscores the individual importance of each galaxy as a new test, as opposed to the view of them all as a statistical ensemble.

Survival of high-energy cosmic rays (HECRs) against gravitational Cherenkov losses is shown not to cast strong constraints on MOND theories that are compatible with general relativity (GR): theories that coincide with GR in the high-acceleration limit. The energy-loss rate, L, is shown to be many orders smaller than those derived in the literature for theories with no extra scale. The gravitational acceleration produced by a HECR in its vicinity is much higher than the MOND acceleration a0. So, modification to GR, which underlies L, enters only beyond the MOND radius of the particle, within which GR holds sway: r_M=sqrt(Gp/c a0). The spectral cutoff, which enters L quadratically, is thus 1/r_M, not the particle's, much larger, de Broglie wavenumber: k_{dB}= p/hbar. Thus, L is smaller than published rates, which use k_{dB}, by a factor (r_M k_{dB})^2~10^{39}(cp/3.10^{11}Gev)^3. With 1/r_M as cutoff, the distance a HECR can travel without major losses is q l_M, where l_M=c^2/a0 is the MOND length, and q is a dimensionless function of parameters of the problem. Since l_M is ~2 pi times the Hubble distance, survival of HECRs does not strongly constrain GR-compatible, MOND theories. Such theories also easily satisfy existing preferred-frame limits, inasmuch as these limits are gotten in high-acceleration systems. I exemplify the results with MOND adaptations of Einstein-Aether theories.

The MOND paradigm posits a departure from standard Newtonian dynamics, and from General Relativity, in the limit of small accelerations. The resulting modified dynamics aim to account for the mass discrepancies in the universe without non-baryonic dark matter. I briefly review this paradigm with its basic tenets, and its underlying theories--nonrelativistic and relativistic--including a novel, bimetric MOND gravity theory. I also comment on MOND's possible connection to, and origin in, the cosmological state of the universe at large. Some of its main predictions, achievements, and remaining desiderata are listed. I then succinctly pit MOND against the competing paradigm of standard dynamics with cold, dark matter. (Abridged)

I describe the MOND paradigm, which posits a departure from standard physics below a certain acceleration scale. This acceleration as deduced from the dynamics in galaxies is found mysteriously to agree with the cosmic acceleration scales defined by the present day expansion rate and by the density of `dark energy'. I put special emphasis on phenomenology and on critical comparison with the competing paradigm based on classical dynamics plus cold dark matter. I also describe briefly nonrelativistic and relativistic MOND theories.

I look at the growth of weak density inhomogeneities of nonrelativistic matter, in bimetric-MOND (BIMOND) cosmology. I concentrate on matter-twin-matter-symmetric versions of BIMOND, and assume that, on average, the universe is symmetrically populated in the two sectors. MOND effects are absent in an exactly symmetric universe, apart from the appearance of a cosmological constant, Lambda~(a0/c)^2. MOND effects-local and cosmological-do enter when density inhomogeneities that differ in the two sectors appear and develop. MOND later takes its standard form in systems that are islands dominated by pure matter. I derive the nonrelativistic equations governing small-scale fluctuation growth. The equations split into two uncoupled systems, one for the sum, the other for the difference, of the fluctuations in the two sectors. The former is governed strictly by Newtonian dynamics. The latter is governed by MOND dynamics, which entails stronger gravity, and nonlinearity even for the smallest of perturbations. These cause the difference to grow faster than the sum, conducing to matter-twin-matter segregation. The nonlinearity also causes interaction between nested perturbations on different scales. Because matter and twin matter (TM) repel each other in the MOND regime, matter inhomogeneities grow not only by their own self gravity, but also through shepherding by flanking TM overdensitie. The relative importance of gravity and pressure in the MOND system depends also on the strength of the perturbation. The development of structure in the universe, in either sector, thus depends crucially on two initial fluctuation spectra: that of matter alone and that of the matter-TM difference. I also discuss the back reaction on cosmology of BIMOND effects that appear as "phantom matter" resulting from inhomogeneity differences between the two sectors.

Bimetric MOND (BIMOND) theories, propounded recently, predict peculiar gravitational interactions between matter and twin matter (TM). Twin matter is hypothetical matter that might couple directly only to the second metric of the theory. Considerations of cosmology in the BIMOND framework suggest that such TM might exist and copy matter in its attributes. Here I investigate the indirect interactions that BIMOND theories predict between nonrelativistic masses of matter and TM. The most salient result is that in the deep-MOND regime of the matter-TM-symmetric theories, TM behaves as if it has a negative gravitational mass. To wit, interaction within each sector is attractive MOND gravity, but between matter and TM it is repulsive MOND gravity. In the high-acceleration regime, the interaction depends on a parameter, beta. For the favored value beta=1, matter and TM do not interact in this regime; for beta<1 they attract; and for beta>1 they repel each other. Such interactions may have substantial ramifications for all aspects of structure formation, such as matter distribution, peculiar velocities, and effects on the CMB. The repulsive interactions probably lead to segregation of matter and TM structures, leading, in turn, to intermeshing of the respective cosmic webs, with high-density nodes of one sector residing in the voids of the other (possibly conducing to efficient evacuation of the voids). Weak lensing by TM seems the best way to constrain its attributes. In the MOND regime a TM body acts on matter photons as a diverging lens. Strong lensing occurs in the high acceleration regime, and thus depends on the value of beta. For beta=1 no strong lensing effects of TM are expected. I also discuss briefly asymmetric theories (abridged).

A new formulation of MOND as a modified-potential theory of gravity is propounded. In effect, the theory dictates that the MOND potential phi produced by a mass distribution rho is a solution of the Poisson equation for the modified source density rho*=-(1/4 pi G)divergence(g), where g=nu(|gN|/a0)gN, and gN is the Newtonian acceleration field of rho. This makes phi simply the scalar potential of the algebraic acceleration field g. The theory thus involves solving only linear differential equations, with one nonlinear, algebraic step. It is derivable from an action, satisfies all the usual conservation laws, and gives the correct center-of-mass acceleration to composite bodies. The theory is akin in some respects to the nonlinear Poisson formulation of Bekenstein and Milgrom, but it is different from it, and is obviously easier to apply. The two theories are shown to emerge as natural modifications of a Palatini-type formulation of Newtonian gravity, and are members in a larger class of bi-potential theories.

A new relativistic formulation of MOND is advanced, involving two metrics as independent degrees of freedom: the MOND metric g_mn, to which alone matter couples, and an auxiliary metric g*_mn. The main idea hinges on the fact that we can form tensors from the difference, C^a_bc, of the Levi-Civita connections of the two metrics, and these act like gravitational accelerations. In the context of MOND we can form dimensionless `acceleration' scalars, and functions thereof, from contractions of C^a_bc/a0. I look at a class of bimetric MOND theories governed by an action with the gravitational Lagrangian density b sqrt(g)R+a sqrt(g*) R* -2(gg*)^{1/4}f(k)a0^2M(U/a0^2), and with matter actions I(g_mn,psi)+I*(g*_mn,chi), with U a scalar quadratic in the C^a_bc, k=(g/g*)^{1/4}, and allowing for the existence of twin matter, chi, that couples to g*_mn alone. In particular, I concentrate on one interesting and simple choice of the scalar U. This theory introduces only one new constant, a0; it tends simply to General Relativity in the limit a0 goes to 0, and to a phenomenologically valid MOND theory in the nonrelativistic limit. The theory naturally gives MOND and "dark energy" effects from the same term in the action, both controlled by the MOND constant a0. As regards gravitational lensing by nonrelativistic systems--a holy grail for relativistic MOND theories--the theory predicts that the same potential that controls massive-particle motion also dictates lensing in the same way as in GR. This last result can be modified with other choices of U, but lensing is still enhanced and MOND-like, with an effective logarithmic potential.

Prompted by the recent claim, by Donato et al., of a quasi-universal central surface density of galaxy dark matter halos, I look at what MOND has to say on the subject. MOND, indeed, predicts a quasi-universal value of this quantity for objects of all masses and of any internal structure, provided they are mostly in the Newtonian regime; i.e., that their mean acceleration is at or above a0. The predicted value is qSm, with Sm= a0/2 pi G= 138 solar masses per square parsec for the nominal value of a0, and q a constant of order 1 that depends only on the form of the MOND interpolating function. This gives in the above units log(Sm)=2.14, which is consistent with that found by Doanato et al. of 2.15+-0.2. MOND predicts, on the other hand, that this quasi-universal value is not shared by objects with much lower mean accelerations. It permits halo central surface densities that are arbitrarily small, if the mean acceleration inside the object is small enough. However, for such low-surface-density objects, MOND predicts a halo surface density that scales as the square root of the baryonic one, and so the range of the former is much compressed relative to the latter. This explains, in part, the finding of Donato et al. that the universal value applies to low acceleration systems as well. Looking at literature results for a number of the lowest surface-density disk galaxies with rotation-curve analysis, I find that, indeed, their halo surface densities are systematically lower then the above "universal" value. The prediction of Sm as an upper limit, and accumulation value, of halo central surface densities, pertains, unlike most other MOND predictions, to a pure "halo" property, not to a relation between baryonic and "dark matter" properties.

This is a semi-popular account of the MOND paradigm and its comparison with the competing Newtonian-dynamics-plus-CDM paradigm. It was published recently in the online magazine of the Israel Physical Society in the IYA 2009 issue.

I pinpoint a previously unrecognized MOND effect that may act in the inner solar system, and is due to the galactic acceleration, g_g=eta*a0: a byproduct of the MOND external-field effect. Predictions of the effect are not generic to the MOND paradigm, but depend on the particular MOND formulation at hand. However, the modified-Poisson formulation, on which I concentrate, uniquely predicts a subtle anomaly that may be detected in planetary and spacecraft motions (and perhaps in other precision systems, such as binary pulsars), despite their very high accelerations, and even if the MOND interpolating function is arbitrarily close to unity at high accelerations. Near the sun, this anomaly appears as a quadrupole field, with the acceleration at position u from the sun being g_i(u)=-q_{ij}u^j, with q_{ij} diagonal, axisymmetric, and traceless: -2q_{xx}=-2q_{yy}=q_{zz}=q(eta)*(a0/R_M), where R_M=(MG/a0)^{1/2}=8*10^3 au is the MOND radius of the sun. The anomaly is described and analyzed as the Newtonian field of the fictitious cloud of ``phantom matter'' that hovers around the sun. I find, for the relevant range of eta values, and for a range of interpolating functions, mu(x), values of 10^{-2}<-q< 0.3, which turn out to be sensitive to the form of mu(x) around the MOND-to-Newtonian transition. This range verges on the present bounds from solar system measurements. There might thus exist favorable prospects for either measuring the effect, or constraining the theory and the relevant parameters.

The MOND limit is shown to follow from a requirement of space-time scale invariance of the equations of motion for nonrelativistic, purely gravitational systems; i.e., invariance of the equations of motion under (t,r) goes to (qt,qr), in the limit a0 goes to infinity. It is suggested that this should replace the definition of the MOND limit based on the low-acceleration behavior of a Newtonian-MOND interpolating function. In this way, the salient, deep-MOND results--asymptotically flat rotation curves, the mass-rotational-speed relation (baryonic Tully-Fisher relation), the Faber-Jackson relation, etc.--follow from a symmetry principle. For example, asymptotic flatness of rotation curves reflects the fact that radii change under scaling, while velocities do not. I then comment on the interpretation of the deep-MOND limit as one of "zero mass": Rest masses, whose presence obstructs scaling symmetry, become negligible compared to the "phantom", dynamical masses--those that some would attribute to dark matter. Unlike the former masses, the latter transform in a way that is consistent with the symmetry. Finally, I discuss the putative MOND-cosmology connection in light of another, previously known symmetry of the deep-MOND limit. In particular, it is suggested that MOND is related to the asymptotic de Sitter geometry of our universe. It is conjectured, for example, that in an exact de Sitter cosmos, deep-MOND physics would exactly apply to local systems. I also point out, in this connection, the possible relevance of a de Sitter-conformal-field-theory (dS/CFT) duality.

MOND predicts that a mass, M, contained within its transition radius r_t=(MG/a0)^{1/2}, may exhibit a feature at about that radius in the form of a shell, or projected ring, in the deduced distribution of its phantom dark matter. This occurs despite the absence of any underlying feature in the true (``baryon'') source distribution itself. The phenomenon is similar to the appearance of an event horizon and other unusual physics ``in the middle of nothing'' near the transition radius of General Relativity $MG/c^2$. We consider the possibility that this pure MOND phenomenon is in the basis of the recent finding of such a ring in the galaxy cluster Cl 0024+17 by Jee et al. We find that the parameters of the observed ring can be naturally explained in this way; this feature may therefore turn out to be a direct evidence for MOND. We study this phenomenon in simple, axisymmetric configurations aligned with the line of sight: spherical masses, a dumbbell of spherical masses, and an elongated, thin structure. The properties of the apparent ring: its radius, surface density, and contrast, depend on the form of the MOND interpolating function and on the exact three dimensional distribution of the sources (the thin-lens approximation is quite invalid in MOND). We also comment on the possible appearance of orphan features, marking the Newtonian-to-MOND transition, in high surface brightness galaxies. In particular, we find that previously unexplained structure in the rotation curves of some galaxies may be evidence for such features.

I review briefly different aspects of the MOND paradigm, with emphasis on phenomenology, epitomized here by many MOND laws of galactic motion--analogous to Kepler's laws of planetary motion. I then comment on the possible roots of MOND in cosmology, possibly the deepest and most far reaching aspect of MOND. This is followed by a succinct account of existing underlying theories. I also reflect on the implications of MOND's successes for the dark matter (DM) paradigm: MOND predictions imply that baryons alone accurately determine the full field of each and every individual galactic object. This conflicts with the expectations in the DM paradigm because of the haphazard formation and evolution of galactic objects and the very different influences that baryons and DM are subject to during the evolution, as evidenced, e.g., by the very small baryon-to-DM fraction in galaxies (compared with the cosmic value). All this should disabuse DM advocates of the thought that DM will someday be able to reproduce MOND: It is inconceivable that the modicum of baryons left over in galaxies can be made to determine everything if a much heavier DM component is present.

I start with a brief introduction to MOND phenomenology and its possible roots in cosmology--a notion that may turn out to be the most far reaching aspect of MOND. Next I discuss the implications of MOND for the dark matter (DM) doctrine: MOND's successes imply that baryons determine everything. For DM this would mean that the puny tail of leftover baryons in galaxies wags the hefty DM dog. This has to occur in many intricate ways, and despite the haphazard construction history of galaxies--a very tall order. I then concentrate on galaxy clusters in light of MOND, which still requires some yet undetected cluster dark matter, presumably in some baryonic form (CBDM). This CBDM might contribute to the heating of the x-ray emitting gas and thus alleviate the cooling-flow puzzle. MOND, qua theory of dynamics, does not directly enter the microphysics of the gas; however, it does force a new outlook on the role of DM in shaping the cluster gasdynamics: MOND tells us that the cluster DM is not cold dark matter, is not so abundant, and is not expected in galaxies; it is thus not subject to constraints on baryonic DM in galaxies. The mass in CBDM required in a whole cluster is, typically, similar to that in hot gas, but is rather more centrally concentrated, totally dominating the core. The CBDM contribution to the baryon budget in the universe is thus small. Its properties, deduced for isolated clusters, are consistent with the observations of the ``bullet cluster''. Its kinetic-energy reservoir is much larger than that of the hot gas in the core, and would suffice to keep the gas hot for many cooling times. Heating can be effected in various ways depending on the exact nature of the CBDM, from very massive black holes to cool, compact gas clouds.

I consider in light of MOND the three debris galaxies discussed recently by Bournaud et al.. These exhibit mass discrepancies of a factor of a few within several scale lengths of the visible galaxy, which, arguably, flies in the face of the cold dark matter paradigm. I show here that the rotational velocities predicted by MOND agree well with the observed velocities for each of the three galaxies, with only the observed baryonic matter as the source of gravity. There is thus no need to invoke a new form of baryonic, yet-undetected matter that dominates the disc of spiral galaxies, as advocated by Bournaud et al. I argue on other grounds that the presence of such ubiquitous disc dark matter, in addition to cold dark matter, is not likely.

We consider two-level detectors coupled to a scalar field and moving on arbitrary trajectories in Minkowski space-time. We first derive a generic expression for the response function using a (novel) regularization procedure based on the Feynmann prescription that is explicitly causal, and we compare it to other expressions used in the literature. We then use this expression to study, analytically and numerically, the time dependence of the response function in various non-stationarity situations. We show that, generically, the response function decreases like a power in the detector's level spacing, $E$, for high $E$. It is only for stationary world-lines that the response function decays faster than any power-law, in keeping with the known exponential behavior for some stationary cases. Under some conditions the (time dependent) response function for a non-stationary world-line is well approximated by the value of the response function for a stationary world-line having the same instantaneous acceleration, torsion, and hyper-torsion. While we cannot offer general conditions for this to apply, we discuss special cases; in particular, the low energy limit for linear space trajectories.

We present the results of numerical simulations of stationary, spherically outflowing, electron-positron pair winds, with total luminosities in the range 10^{34}--10^{42} ergs/s. In the concrete example described here, the wind injection source is a hot, bare, strange star, predicted to be a powerful source of pairs created by the Coulomb barrier at the quark surface. We find that photons dominate in the emerging emission, and the emerging photon spectrum is rather hard and differs substantially from the thermal spectrum expected from a neutron star with the same luminosity. This might help distinguish the putative bare strange stars from neutron stars.

We present MOND analysis for several of the lowest mass disc galaxies currently amenable to such analysis--with (baryonic) masses below 4x10^8 solar masses. The agreement is good, extending the validity of MOND and its predicted mass velocity relation, to such low masses.

We present the results of numerical simulations of stationary, spherically outflowing, electron-positron pair winds, with total luminosities in the range 10^{34}- 10^{42} ergs/s. In the concrete example described here, the wind injection source is a hot, bare, strange star, predicted to be a powerful source of electron-positron pairs created by the Coulomb barrier at the quark surface. We find that photons dominate in the emerging emission, and the emerging photon spectrum is rather hard and differs substantially from the thermal spectrum expected from a neutron star with the same luminosity. This might help distinguish the putative bare strange stars from neutron stars.

I show that massive-particle dynamics can be simulated by a weak, spherical, external perturbation on a potential flow in an ideal fluid. The effective Lagrangian is of the form mc^2L(U^2/c^2), where U is the velocity of the particle relative to the fluid and c the speed of sound. This can serve as a model for emergent relativistic inertia a la Mach's principle with m playing the role of inertial mass, and also of analog gravity where it is also the passive gravitational mass. m depends on the particle type and intrinsic structure, while L is universal: For D dimensional particles L is proportional to the hypergeometric function F(1,1/2;D/2;U^2/c^2). Particles fall in the same way in the analog gravitational field independent of their internal structure, thus satisfying the weak equivalence principle. For D less or equal 5 they all have a relativistic limit with the acquired energy and momentum diverging as U approaches c. For D less or equal 7 the null geodesics of the standard acoustic metric solve our equation of motion. Interestingly, for D=4 the dynamics is very nearly Lorentzian. The particles can be said to follow the geodesics of a generalized acoustic metric of a Finslerian type that shares the null geodesics with the standard acoustic metric. In vortex geometries, the ergosphere is automatically the static limit. As in the real world, in ``black hole'' geometries circular orbits do not exist below a certain radius that occurs outside the horizon. There is a natural definition of antiparticles; and I describe a mock particle vacuum in whose context one can discuss, e.g., particle Hawking radiation near event horizons.

I briefly highlight the salient properties of modified-inertia formulations of MOND, contrasting them with those of modified-gravity formulations, which describe practically all theories propounded to date. Future data (e.g. the establishment of the Pioneer anomaly as a new physics phenomenon) may prefer one of these broad classes of theories over the other. I also outline some possible starting ideas for modified inertia.

We consider a time dependent, spherically outflowing wind, in Schwarzschild space-time, consisting of electron-positron pairs and photons . Without assuming thermal equilibrium, we account for the microphysics, including two-body processes and their radiative three-body variants. We present a finite-difference scheme for solving the general relativistic kinetic Boltzmann equations for electron-positron pairs and photons. We apply this to the concrete example of a wind from a hot, bare, strange star, predicted to be a powerful source of hard X-ray photons and pairs created by the Coulomb barrier at the quark surface. We study the kinetics of the wind particles and the emerging emission in photons and pairs for stationary winds with total luminosities in the range 10^{34}-10^{39} ergs/s, for different values of the injected photon-to-pair ratio. The wind parameters--such as the mean optical depth for photons, the rates of particle number and energy outflows, bulk velocity, and number density of the pair plasma--are presented as functions of the distance from the stellar surface, as well as characteristics of the emergent radiation. We find that photons dominate in the emerging emission, and the emerging photon spectrum is rather hard and differs substantially from the thermal spectrum expected from a neutron star with the same luminosity. This might help distinguish the putative bare strange stars from neutron stars.

We examine two corollaries of MOND pertaining to the properties of the equivalent dark-matter halo. MOND predicts for pure exponential discs a tight relation involving the halo and disc scale lengths and the mean acceleration in the disc, which we find to test favorably against the Verheijen sample of Ursa Major galaxies. A correlation between halo and disc length scales is also apparent when the "maximum disc" contribution is assumed, but we demonstrate that this follows from the more general MOND prediction. The second MOND prediction involves the existence of a maximum halo acceleration, which also tests favorably against the Ursa Major sample for different assumptions on the disc contribution.

We consider a stationary, spherically outflowing wind consisting of electron-positron pairs and photons. We do not assume thermal equilibrium, and include the two-body processes that occur in such a wind: Moller and Bhaba scattering of pairs, Compton scattering, two-photon pair annihilation, and two-photon pair production, together with their radiative three-body variants: bremsstrahlung, double Compton scattering, and three-photon pair annihilation, with their inverse processes. In the concrete example described here, the wind injection source is a hot, bare, strange star. Such stars are thought to be powerful sources of pairs created by the Coulomb barrier at the quark surface. We present a new, finite-difference scheme for solving the relativistic kinetic Boltzmann equations for pairs and photons. Using this method we study the kinetics of the wind particles and the emerging emission for total luminosities of L=10^{34}-10^{42} ergs/s. We find the rates of particle number and energy outflows, outflow velocities, number densities, energy spectra, and other parameters for both photons and pairs as functions of the distance. We find that for L>2x10^{35} ergs/s, photons dominate the emerging emission. As L increases from ~ 10^{34} to 10^{42} ergs/s, the mean energy of emergent photons decreases from ~400-500 keV to 40 keV, as the spectrum changes in shape from that of a wide annihilation line to nearly a blackbody spectrum with a high energy (> 100 keV) tail. These results are pertinent to the deduction of the outside appearance of hot bare strange stars, which might help discern them from neutron stars.

The recent findings of Romanowsky et al., of an ``unexpectedly'' small mass discrepancy within 5 effective radii in several elliptical galaxies, are not surprising in the context of MOND. As we show here, they are, in fact, in full concordance with its predictions. One is dealing with high-surface-density galaxies with mean accelerations rather larger than the acceleration constant of MOND. These findings continue, and are now the extreme examples of, the trend predicted by MOND: the mass discrepancy sets in at larger and larger scaled radii in galaxies with larger and larger mean surface densities; or, equivalently, mean accelerations.

We present the results of numerical simulations of stationary, spherically outflowing, pair winds, with total luminosities of L=10^{35}- 10^{42} ergs/s. These results have direct relevance to the emission from hot, bare, strange stars, which are thought to be powerful sources of electron-positron pairs created by the Coulomb barrier at the quark surface. The spectra of emergent photons and pairs are calculated. For L > 2x10^{35} erg/s, photons dominate the emerging emission. As L increases from 10^{35} to 10^{42} ergs/s, the mean photon energy decreases from ~ 400-500 keV to 40 keV, while the spectrum changes in shape from a wide annihilation line to being nearly blackbody with a high energy (> 100 keV) tail. Such a correlation of the photon spectrum with the luminosity, together with the fact that super-Eddington luminosities can be achieved, might be a good observational signature of hot, bare, strange stars.

I estimate MOND M/L values for nine galaxy groups that were recently studied by Tully et al.. Instead of the large M/L values that they find with Newtonian dynamics (up to 1200 solar units) the MOND estimates cluster around 1 solar unit. Tully et al. find a systematic and significant difference between the M/L values of groups that do not contain luminous galaxies and those that do: Dwarfs-only groups have larger M/L values (by a factor of about 5). The MOND M/L values do not show this trend; the Newtonian disparity is traced back to the dwarfs-only groups having systematically smaller intrinsic accelerations (similar sizes, but rather smaller velocity dispersions).

I discuss open theoretical questions pertaining to the modified dynamics (MOND)--a proposed alternative to dark matter, which posits a breakdown of Newtonian dynamics in the limit of small accelerations. In particular, I point the reasons for thinking that MOND is an effective theory--perhaps, despite appearance, not even in conflict with GR. I then contrast the two interpretations of MOND as modified gravity and as modified inertia. I describe two mechanical models that are described by potential theories similar to (non-relativistic) MOND: a potential-flow model, and a membrane model. These might shed some light on a possible origin of MOND. The possible involvement of vacuum effects is also speculated on.

I investigate the properties of forces on bodies in theories governed by the generalized Poisson equation div[mu(abs(grad_phi))grad_phi]=G rho, for the potential phi produced by a distribution of sources rho. This equation describes, inter alia, media with a response coefficient, mu, that depends on the field strength, such as in nonlinear, dielectric, or diamagnetic, media; nonlinear transport problems with field-strength dependent conductivity or diffusion coefficient; nonlinear electrostatics, as in the Born-Infeld theory; certain stationary potential flows in compressible fluids, in which case the forces act on sources or obstacles in the flow. The expressions for the force on a point charge is derived exactly for the limits of very low and very high charge. The force on an arbitrary body in an external field of asymptotically constant gradient, E, is shown to be F=QE, where Q is the total effective charge of the body. The corollary Q=0 implies F=0 is a generalization of d'Aembert's paradox. I show that for G>0 (as in Newtonian gravity) two point charges of the same (opposite) sign still attract (repel) each other. The opposite is true for G<0. I discuss the generalization of this to extended bodies, and derive virial relations.

An account is given of the development, and the status, of the modified dynamics (MOND)--a proposed alternative to dark matter, which posits a breakdown of Newtonian dynamics in the limit of small accelerations.

In a recent paper, Kaplinghat and Turner (2001) (KT) advertise that MOND can be derived naturally in the CDM paradigm. They actually proceed to produce a more limited result: Every galaxy should have a transition radius, $r_t$, below which baryons dominate, and above which dark matter (DM) takes over; the acceleration at $r_t$ is nearly the same for all galaxies; and due to a coincidences this is of order $a_0\sim cH_0$. This follows from their tacit, intermediate result, whereby CDM halos of galaxies have a very nearly universal acceleration profile $a(r)\approx v^2(r)/r\approx A\hat a(r/\ell)$, where A is universal, and only the scale $\ell$ varies from halo to halo. (This remains so when baryons are added because they assume a universal baryon-collapse factor.) The KT scenario is phenomenologically wrong--observed galaxies are simply not like that. For example, it precludes altogether the existence of LSB galaxies, in which the acceleration is everywhere smaller than $a_0$. The phenomenologically sound outcome--i.e., the role of $a_0$ as a transition acceleration in high-surface-brightness galaxies--pertains to only a small part of the statement of MOND. There are several other, independent roles that $a_0\sim cH_0$ plays in MOND phenomenology, and other predictions of MOND, not related to the value of $a_0$, that are not explainable in the KT scenario. The results of KT also disagree with those of CDM simulations, which, as they now stand, do not reproduce any aspect of MOND phenomenology.

Analyses of halo shapes for disc galaxies are said to give incongruous results. I point out that the modified dynamics (MOND) predicts for disc galaxies a distribution of fictitious dark matter that comprises two components: a pure disc and a rounder halo. The former dominates the true disc in regions of small accelerations, where it controls the z-dynamics in the disc (disc flaring etc.); it has a finite total mass. It also dominates the round component near the centre where the geometry is nearly planar. The second component controls motions far from the plane, has a total enclosed mass that diverges linearly with radius, and determines the rotation curve at large radii. Its ellipticity may be appreciable at small radii but vanishes asymptotically. This prediction of MOND differs from what one expects from galaxy-formation scenarios with dark matter. Analyses to date, which, as they do, assume one component--usually with a constant ellipticity, perforce give conflicting results for the best-value ellipticity, depending on whether they probe the disc or the sphere, small radii or large ones.

In a significant recent paper A. Edery undertakes a new study of light deflection in generalizations of general relativity (GR). He claims to prove that any metric-based gravitational theory that proposes to explain the flat rotation curves of disk galaxies without postulating dark matter halos must conflict with observations of gravitational lensing by galaxies and clusters of galaxies because any such theory inevitably make a negative contribution to light deflection. Here we show that some of the basic steps of Edery's argument are invalid, and no such general result obtains.

In the modified dynamics (MOND) the inner workings of dwarf satellites can be greatly affected by their mother galaxy-over and beyond its tidal effects. Because of MOND's nonlinearity a system's internal dynamics can be altered by an external field in which it is immersed (even when this field, by itself, is constant in space). As a result, the size and velocity dispersion of the satellite vary as the external field varies along its orbit. A notable outcome of this is a substantial increase in the dwarf's vulnerability to eventual tidal disruption-rather higher than Newtonian dynamics (with a dark-matter halo) would lead us to expect for a satellite with given observed parameters.

We propose that a gamma-ray burst in one member of a binary may induce a supernova-like explosion of a close, white-dwarf companion. Such an explosion might be brought about in rather light companions, which cannot undergo the standard accretion-induced explosion. This would give some GRB-associated supernova an appearance rather unlike that of the typical Type I. GRB 980425, if indeed associated with SN 1998bw, is too weak to have produced it through our proposed mechanism.

There is an effect in the modified dynamics (MOND) that is conducive to formation of warps. Because of the nonlinearity of the theory the internal dynamics of a galaxy is affected by a perturber over and above possible tidal effects. For example, a relatively distant and light companion or the mean influence of a parent cluster, with negligible tidal effects, could still produce a significant warp in the outer part of a galactic disk. We present results of numerical calculations for simplified models that show, for instance, that a satellite with the (baryonic) mass and distance of the Magellanic clouds can distort the axisymmetric field of the Milky Way enough to produce a warp of the magnitude (and position) observed. Details of the warp geometry remain to be explained: we use a static configuration that can produce only warps with a straight line of nodes. In more realistic simulations one must reckon with the motion of the perturbing body, which sometimes occurs on time scales not much longer than the response time of the disk.

General analytic arguments lead us to expect that in the modified dynamics (MOND) self-gravitating disks are more stable than their like in Newtonian dynamics. We study this question numerically, using a particle-mesh code based on a multi-grid solver for the (nonlinear) MOND field equation. We start with equilibrium distribution functions for MOND disk models having a smoothly truncated, exponential surface-density profiles and a constant Toomre $Q$ parameter. We find that, indeed, disks of a given ``temperature'' are locally more stable in MOND than in Newtonian dynamics. As regards global instability to bar formation, we find that as the mean acceleration in the disk is lowered, the stability of the disk is increased as we cross from the Newtonian to the MOND regime. The degree of stability levels off deep in the MOND regime, as expected from scaling laws in MOND. For the disk model we use, this maximum degree of stability is similar to the one imparted to a Newtonian disk by a halo three times as massive at five disk scale lengths.

Why does there appear in the modified dynamics (MOND) an acceleration constant, a0, of cosmological significance? An intriguing possibility is that MOND, indeed inertia itself--as embodied in the actions of free particles and fields, is due to effects of the vacuum. Either cosmology enters local dynamics by affecting the vacuum, and inertia in turn, through a0; or, the same vacuum effect enters both MOND (through a0) and cosmology (e.g. through a cosmological constant). For the vacuum to serve as substratum for inertia a body must be able to read in it its non-inertial motion; this indeed it can, by detecting Unruh-type radiation. A manifestation of the vacuum is also seen, even by inertial observers, in a non-trivial universe (marked, e.g., by curvature or expansion). A non-inertial observer in a nontrivial universe will see the combined effect. An observer on a constant-acceleration (a) trajectory in a de Sitter universe with cosmological constant L sees Unruh radiation of temperature T\propto [a^2+a0^2]^{1/2}, with a0=(\L/3)^{1/2}. The temperature excess over what an inertial observer sees, T(a)-T(0), turns out to depend on a in the same way that MOND inertia does. An actual inertia-from-vacuum mechanism is still a far cry off.

We have recently discovered that the modified dynamics (MOND) implies some universal upper bound on the acceleration that can be contributed by a `dark halo'--assumed in a Newtonian analysis to account for the effects of MOND. Not surprisingly, the limit is of the order of the acceleration constant of the theory. This can be contrasted directly with the results of structure-formation simulations. The new limit is substantial and different from earlier MOND acceleration limits (discussed in connection with the MOND explanation of the Freeman law for galaxy disks, and the Fish law for ellipticals): It pertains to the `halo', and not to the observed galaxy; it is absolute, and independent of further physical assumptions on the nature of the galactic system; and it applies at all radii, whereas the other limits apply only to the mean acceleration in the system.

The Modified dynamics (MOND) has been propounded as an alternative to Dark matter. It imputes the mass discrepancy in galaxy systems to failure of standard dynamics in the limit of small accelerations. After a brief description of the MOND tenets, I discuss how its predictions now compare with the data. I put special emphasis on rotation-curve analysis--whence comes the most clear-cut support for MOND, and on the cores of rich x-ray clusters, where MOND does not yet explain away the mass discrepancy. I then outline the MOND program, especially work still left to do. This is followed by general comments on cosmology and structure formation in MOND. I conclude with some incomplete thoughts on the possible origin of MOND (as an effective theory); in particular on the possibility that it comes about as a vacuum effect.

I estimate Modified-Dynamics (MOND), median M/L values for recently published catalogues of galaxy groups. While the median, Newtonian M/L values quoted for these catalogues are 110-200 solar units, the corresponding values for MOND are less than 10 solar units.

We add to the lore of spherical, stellar-system models a two-parameter family with an anisotropic velocity dispersion, and a central point mass (``black hole''). The ratio of the tangential to radial dispersions, is constant--and constitutes the first parameter--while each decreases with radius as r^{-1/2}. The second parameter is the ratio of the central point mass to the total mass. The Jeans equation is solved to give the density law in closed form: rho\propto (r/r0)^{-c}/[1+(r/r0)^{3-c}]^2, where r0 is an arbitrary scale factor. The two parameters enter the density law only through their combination c. At the suggestion of Tremaine, we also explore models with only the root-sum-square of the velocities having a Keplerian run, but with a variable anisotropy ratio. This gives rise to a more versatile class of models, with analytic expressions for the density law and the dispersion runs, which contain more than one radius-scale parameter.

Eisenstein Loeb and Turner (ELT) have recently proposed a method for estimating the dynamical masses of large-scale filaments, whereby the filament is modeled by an axisymmetric, isothermal cylinder, for which ELT derive a global relation between the (constant) velocity dispersion and the total line density. We first show that the model assumptions of ELT can be relaxed materially: an exact relation between the velocity and line density is derived for any cylinder (not necessarily axisymmetric), with an arbitrary constituent distribution function (so isothermality need not be assumed). We then consider the same problem in the context of the modified dynamics (MOND). After a brief comparison between scaling properties in the two theories, we study idealized MOND model filaments. A preliminary application to the segment of the Perseus-Pisces filament treated by ELT, gives MOND M/L estimates of order 10 s.u., compared with the Newtonian value of about 450, which ELT find. In spite of the large uncertainties still besetting the analysis, this instance of MOND application is of particular interest because: 1. Objects of this geometry have not been dealt with before. 2. It pertains to large-scale structure. 3. The typical accelerations involved are the lowest so far encountered in a semi-virialized system.

Very-low-surface-density galactic systems have very low mean accelerations. They thus provide quintessential tests of the Modified Dynamics (MOND), which predicts an increasing mass discrepancy with decreasing acceleration. We describe succinctly the results pertinent to several classes of such objects: Low-luminosity (dwarf) spirals, irregular dwarf spirals, normal-lumonosity-but -low-surface-density spirals, and dwarf-spheroidal satellites of the Milky Way.

The little we do know of the physical conditions in gamma-ray bursters makes them conducive to the acceleration of high-energy cosmic rays, especially if they are at cosmological distances. We find that, with the observed statistics and fluxes of gamma-ray bursts, cosmological bursters may be an important source of cosmic rays in two regions of the observed spectrum: 1. At the very-high-energy end (E>10^{19} eV), where cosmic rays must be of extragalactic origin. 2. Around and above the spectral feature that has been described as a bump and/or a knee, which occurs around 10^{15} eV, and starts at about 10^{14} eV. The occasional bursters that occur inside the Galaxy--about once in a few hundred thousand years if burst emission is isotropic; more often, if it is beamed--could maintain the density of galactic cosmic rays at the observed level in this range. These two energy ranges might correspond to two typical energy scales expected from bursters: one pertinent to acceleration due to interaction of a magnetized-fireball front with an ambient medium; the other to acceleration in the fireball itself (e.g. shock acceleration).

We point out that each of the error boxes of the two highest-energy cosmic-ray shower events known, overlaps with that of a strong gamma-ray burst (GRB). The GRBs precede the cosmic rays by 5.5, and 11 months respectively. In one case the strongest known cosmic ray is paired with the strongest gamma-ray burst in the BATSE catalogue. The probability of this to have occurred by chance seems rather small. Without building on post-factum statistics, we think the above is remarkable enough to suggest that the cosmic ray and gamma-ray burst were produced by the same outburst. A time delay (and a small positional disparity) is expected, since the trajectory of a charged cosmic-ray particle is wriggled by intervening magnetic fields. We estimate that the Galaxy's field alone may produce a delay of the order observed. We discuss some of the implications that follow if such an association is confirmed. For example, the upper limit on the distance to the cosmic-ray source, combined with a much-better-determined position of the gamma-ray burst source, narrows greatly the volume in which to look for an optical counterpart. There is also useful information in the time delay regarding, e.g., intergalactic magnetic fields.

Gerhard had recently analyzed the data on seven dwarf spheroidals, and concluded that these disagree with the predictions of MOND. We contend that this conclusion is anything but correct. With new data for three of the dwarfs the observations of all dwarfs are in compelling agreement with the predictions of MOND. Gerhard found MOND M/L values that fall around a few solar units, as expected if MOND is a valid alternative to dark matter. His sole cause for complaint was that some of his MOND M/L values were still outside the range of ``reasonable'' stellar values. This, we say, was easily attributable to uncertainties in the data, such as in the velocity dispersions, luminosities, core radii, etc., and in the assumptions that underlie the analysis, such as isotropic velocity distribution, light-distribution fits, etc.. There are now new, much improved determinations of the velocity dispersions for three of the dwarfs, in particular for the two dwarfs for which Gerhard found high MOND M/L values. In fact, within just the quoted errors on the velocity dispersions and the luminosities, the MOND M/L values for all seven dwarfs are now perfectly consistent with stellar values, with no need for dark matter.

We describe a family of circular, and elliptical, finite disks with a disk potential that is a power of the radius. These are all flattened ellipsoids, obtained by squashing finite spheres with a power-law density distribution, and cutoff at some radius Ro. First we discuss circular disks whose circular rotation speed v is proportional to r^alpha, with any alpha> -1/2. The surface-density of the disks is expressed in terms of hypergeometric functions of 1-(Ro/r)^2. We give closed expressions for the full 3-D potentials in terms of hypergeometric functions of two variables. We express the potential and acceleration in the plane at r>Ro, and along the rotation axis, in terms of simple hypergeometric functions. All the multipoles of the disk are given. We then generalize to non-axisymmetric disks. The potential in the midplane is given in terms of the hypergeometric function of two variables. For integer values of 2 alpha the above quantities are given in more elementary terms. All these results follow straightforwardly from formulae we derive for the general, cutoff, power-law, triaxial ellipsoid.

We Consider models of thin disks (with and without bulges) in the Bekenstein-Milgrom formulation of MOND as a modification of Newtonian gravity. Analytic solutions are found for the full gravitational fields of Kuzmin disks, and of disk-plus-bulge generalizations of them. For all these models a simple algebraic relation between the MOND potential field and the Newtonian potential holds everywhere outside the disk. We give exact expressions for the rotation curves for these models. We also find that the algebraic relation is a very good approximation for exponential disks. The algebraic relation outside the disk is then extended into the disk to derive an improved approximation for the MOND rotation curve of disk galaxies that requires only knowledge of the Newtonian curve and the surface density.

We describe a simple derivation of virial relations (VR) for arbitrary action-governed systems. These follow directly from the action with no need to go via the equations of motion. When some of the degrees of freedom are of the same type a tensor virial theorem presents itself. Further generalizations are discussed. Symmetries of the action may lead to identities involving the VR. Beside pointing to a unified provenance of the VR, and affording general systematics of them, our method is a simple prescription for deriving such relations. It is particularly useful for treating high-derivative and non-local theories. We demonstarte the procedure with several examples.

Lo, Sargent, and Young (1993) have recently concluded that the masses of some dwarf galaxies, as deduced by the modified dynamics (MOND) from the observed velocity dispersions, are systematically smaller than the observed masses, by a factor of ten or more. We show here that the MOND mass estimator used by Lo et al. is smaller than the proper expression, by a factor of about twenty. We derive the proper mass estimator as an exact virial-like relation between the 3-D rms velocity, Vs, and the total mass, M, of an arbitrary, self-gravitating system, made of light constituents, that is everywhere in the very-low-acceleration regime of MOND. This reads M=(9/4)Vs^2/G a0. (For a system that is not stationary, Vs involves also an average over time.) We further generalize this relation to cases with constituent masses that are not small compared with that of the whole system. We discuss various applications of the M-V relation. With the correct estimator the predictions of MOND are, by and large, in good agreement with the total observed masses (the observed gas mass plus a stellar mass corresponding to an M/L of order one solar unit).

We investigate particle laws of motion derived from nonstandard kinetic actions of a special form. We are guided by a phenomenological scheme--the modified dynamics (MOND)--that imputes the mass discrepancy observed in galactic systems to a departure from Newtonian dynamics below a certain acceleration scale a0. In the limit a0 goes to 0 the theory goes to Newtonian dynamics. In the opposite limit the action becomes proportional to 1/a0. Galilei-invariant such theories must be strongly non-local; this is a blessing, as such theories need not suffer from the illnesses that are endemic to higher-derivative theories. We discuss the possibility that such a modified law of motion is an effective theory resulting from the elimination of degrees of freedom pertaining to the universe at large (the near equality a0=cH0 being a trace of that connection). A virial relation for bounded trajectories is derived. Exact solutions are obtained for circular orbits, which pertain to rotation curves of disk galaxies. We also explore, in passing, theories that depart from the conventional Newtonian dynamics for very low frequencies. The present formulation contrasts, on important accounts, with earlier, modified-gravity formulations.