ReadingTimeMachine/rtm-sgt-ocr-v1

1source,target2. This prescription aims to model the balance between the dissociation of molecules by Lyman-Werner band photons. and the formation of molecules on dust grains.," This prescription aims to model the balance between the dissociation of molecules by Lyman-Werner band photons, and the formation of molecules on dust grains."3 We refer the readers to the aforementioned papers for the full derivation. and simply repeat the numerical prescription here.," We refer the readers to the aforementioned papers for the full derivation, and simply repeat the numerical prescription here."4 The molecular fraction is given by: for s«2 and fg»= Ofors= 2., The molecular fraction is given by: for $s<2$ and $f_{\rm H2} = 0$ for $s\geq 2$ .5"5=In(1|0.6\0.014 2)/(0.67.). where y=0.76(1|3.12779), and τι=0.066Xa(Μιpe7) Z'."," $s = {\rm ln}6(1+0.6\chi + 0.01\chi^2)/(0.6\tau_{\rm c})$ , where $\chi =70.76(1+3.1Z'^{0.365})$, and $\tau_{\rm c} = 0.066\Sigma_{\rm8 cloud}/(\msun {\rm pc^{-2}})\times Z'$ ."9 Z'is the metallicity divided by the solar metallicity., $Z'$is the metallicity divided by the solar metallicity.10 This formalism for deriving fg» assumes chemical equilibrium., This formalism for deriving $f_{\rm H2}$ assumes chemical equilibrium.11 It is worth a quick note that there are numerous prescriptions for determining the bbalance in the ISM of simulations. some of which include time-dependent chemistry.," It is worth a quick note that there are numerous prescriptions for determining the balance in the ISM of simulations, some of which include time-dependent chemistry."12 developed an empirical pressure-based methodology for calculating the ffraction in the neutral ISM. based on observations of local galaxies.," developed an empirical pressure-based methodology for calculating the fraction in the neutral ISM, based on observations of local galaxies."13 Similarly. both semi-analytic models2009).. as well as full numerical solutions exist which model the effect of dissociating photons through models of galaxies2010).," Similarly, both semi-analytic models, as well as full numerical solutions exist which model the effect of dissociating photons through models of galaxies."14. We motivate our usage of the analytic prescription of for two reasons., We motivate our usage of the analytic prescription of for two reasons.15 First. some observational evidence suggests that on small scales (<<LOO pe). Equation + may fare better than pressure-based prescriptions in describing the state of the neutral ISM in low-metallicity dwarf galaxies2010).," First, some observational evidence suggests that on small scales $<100$ pc), Equation \ref{eq:kmt} may fare better than pressure-based prescriptions in describing the state of the neutral ISM in low-metallicity dwarf galaxies."16. Second. a comparison between Equation 4. and a numerical treatment of time-dependent chemical reaction network and radiative transfer in galaxies suggests that the analytic approximation is reasonable at metallicities above 0.01 Z.1)..," Second, a comparison between Equation \ref{eq:kmt} and a numerical treatment of time-dependent chemical reaction network and radiative transfer in galaxies suggests that the analytic approximation is reasonable at metallicities above 0.01 $Z_{\odot}$."17 Because we aim to model actively star-forming systems in this work. we find that the mass-weighted metallicity of our model clouds is always higher than this fiducial value and expect that the analytic approximation is therefore reasonable.," Because we aim to model actively star-forming systems in this work, we find that the mass-weighted metallicity of our model clouds is always higher than this fiducial value and expect that the analytic approximation is therefore reasonable."18 With Novas and Algo detined. the radius of the cloud is Known.," With $\Sigma_{\rm cloud}$ and ${M}_{\rm H2}$ defined, the radius of the cloud is known."19 In order to account for the turbulent compression of gas. we scale the volumetric densities of the GMCs by a factor e7»7 where numerical simulations show where Ap ids the | dimensional Mach of: the turbulence3JostOl.," In order to account for the turbulent compression of gas, we scale the volumetric densities of the GMCs by a factor $e^{\sigma_\rho^2/2}$ where numerical simulations show where $M_{\rm 1D}$ is the 1 dimensional Mach of the turbulence."20padOQ2. Because the temperature calculation is dependent on the density of the GMC (see below). solving for the density and temperature simultaneously is a computationally lengthy process for the multi-million-cell grids that concern us.," Because the temperature calculation is dependent on the density of the GMC (see below), solving for the density and temperature simultaneously is a computationally lengthy process for the multi-million-cell grids that concern us."21 Thus. to calculate the turbulence-driven density enhancement. we assume the temperature of the GMC is 10 K. which as we shall show. is a good approximation for the bulk of the GMCs in these simulations.," Thus, to calculate the turbulence-driven density enhancement, we assume the temperature of the GMC is 10 K, which as we shall show, is a good approximation for the bulk of the GMCs in these simulations."22 We calculate the [D velocity dispersion in the cloud: where Goo is the mean square sum of the subgrid turbulen velocity dispersion within the GMC and the resolved nontherma velocity dispersion., We calculate the 1D velocity dispersion in the cloud: where $\sigma_{\rm cell}$ is the mean square sum of the subgrid turbulent velocity dispersion within the GMC and the resolved nonthermal velocity dispersion.23 The subgrid turbulent velocity dispersion is calculated from the external pressure from the hot ISM using στ=P/py though we impose a ceiling of IO which comes from average values found in turbulen feedback simulations2011)., The subgrid turbulent velocity dispersion is calculated from the external pressure from the hot ISM using $\sigma^2=P/\rho_{cell}$ though we impose a ceiling of 10 which comes from average values found in turbulent feedback simulations.24.. The resolved nonthermal componen is calculated by finding the turbulent velocity dispersion of the nearest neighbouring cells in the simulation., The resolved nonthermal component is calculated by finding the turbulent velocity dispersion of the nearest neighbouring cells in the simulation.25 In detail. we calculate the standard deviation of the velocities of the nearest neighbour cells in the ο. 7 and 5 directions. and define the nonthermal velocity dispersion as the mean of these.," In detail, we calculate the standard deviation of the velocities of the nearest neighbour cells in the $\hat{x}$, $\hat{y}$ and $\hat{z}$ directions, and define the nonthermal velocity dispersion as the mean of these."26 In cases where the GMC is unresolved. a floor ai I8 set by assuming the cloud is in virial balance with a virial parameter aj;=1. for agbaudt (COM). so that for Mu=100M;pe7? where A is the mass of the cloud.," In cases where the GMC is unresolved, a floor $\sigma_{\rm27 vir}$ is set by assuming the cloud is in virial balance with a virial parameter $\alpha_{\rm vir} = 1$ , for $\alpha_{\rm vir} \equiv 5 \sigma_{\rm vir}^2 R/(GM)$ , so that for $\Sigma_{\rm cloud}=100 \ \msun {\rm pc}^{-2}$ where $M$ is the mass of the cloud."28 Finally. we calculate the temperature of the model GMCs.," Finally, we calculate the temperature of the model GMCs."29 The model is based on that developed by1).. and we describe the relevant details here as it is an important aspect of our model.," The model is based on that developed by, and we describe the relevant details here as it is an important aspect of our model."30 The temperature of the molecular ISM is determined by a balance of heating and cooling processes in the gas. heating and cooling of the dust. and a dust-gas thermal exchange.," The temperature of the molecular ISM is determined by a balance of heating and cooling processes in the gas, heating and cooling of the dust, and a dust-gas thermal exchange."31 For the gas. we consider grain photoelectric heating at à rate per H nucleus Le. cosmic ray heating at a rate Pour. and cooling via either CIT or CO line cooling at a rate αμ.," For the gas, we consider grain photoelectric heating at a rate per H nucleus $\Gamma_{\rm pe}$ , cosmic ray heating at a rate $\Gamma_{\rm32 CR}$, and cooling via either CII or CO line cooling at a rate $\Lambda_{\rm line}$."33 The dust can be heated by the background infrared radiation field at à rate Poni. and cool via thermal emission at a rate Adis.," The dust can be heated by the background infrared radiation field at a rate $\Gamma_{\rm dust}$, and cool via thermal emission at a rate $\Lambda_{\rm dust}$."34 Finally. there is an energy exchange between dust and gas at a rate μι where V. is positive if the dust is hotter than the gas.," Finally, there is an energy exchange between dust and gas at a rate $\Psi_{\rm gd}$ where $\Psi_{\rm gd}$ is positive if the dust is hotter than the gas."35 If the gas and dust are in thermal balance. then we have the following equations: The equation is solved by simultaneously iterating on the temperatures of the gas and'..," If the gas and dust are in thermal balance, then we have the following equations: The equation is solved by simultaneously iterating on the temperatures of the gas and."36 The grain photoelectric heating rate is assumed to be attenuated by half the mean extinction of the cloud (asthe heating rate is expected to decrease toward the cloud interiors) and is given by:, The grain photoelectric heating rate is assumed to be attenuated by half the mean extinction of the cloud (asthe heating rate is expected to decrease toward the cloud interiors) and is given by:37Classical novae are interacting binary systems and the most energetic tvpe of cataclysmic variable.,Classical novae are interacting binary systems and the most energetic type of cataclysmic variable.38 The central svstem consists of a white dwarf primary and a main sequence secondary which fills its Roche lobe., The central system consists of a white dwarf primary and a main sequence secondary which fills its Roche lobe.39 Lvdrogen-rich matter is accreted from the secondary. onto the white cwarl via an accretion disc., Hydrogen-rich matter is accreted from the secondary onto the white dwarf via an accretion disc.40 Once the pressure at the base of the accreted envelope. reaches a critical level. thermonuclear reactions begin under degenerate conditions leading to an explosion (Starrfield 1989).," Once the pressure at the base of the accreted envelope reaches a critical level, thermonuclear reactions begin under degenerate conditions leading to an explosion (Starrfield 1989)."41 Xs a direct result the bolometric luminosity of the svstem increases by at least three orders of magnitude over a timescale of a few cavs., As a direct result the bolometric luminosity of the system increases by at least three orders of magnitude over a timescale of a few days.42 Observational estimates of the mass of material ejected are typically 107 10* M. with velocities in the range a few hundred to several thousand knis.|., Observational estimates of the mass of material ejected are typically $10^{-5}$ – $10^{-4}$ $_{\odot}$ with velocities in the range a few hundred to several thousand $^{-1}$.43 Correlations exist between the rate at which the visual light declines from. maximum ancl the peak absolute magnitude and the ejection velocity., Correlations exist between the rate at which the visual light declines from maximum and the peak absolute magnitude and the ejection velocity.44" These are in the sense that the novae. which fade most. rapicdly (""last novae"") are intrinsically brightest and eject material at the highest speeds.", These are in the sense that the novae which fade most rapidly (“fast novae”) are intrinsically brightest and eject material at the highest speeds.45 τοῦ Cas (Nova Cassiopeiae. 1995) was discovered. at 1 292magon 1995 Xugust 24 by M. Yamamoto (LHirosawa et al 1995) which for the purposes of this paper will be defined. as dav zero., V723 Cas (Nova Cassiopeiae 1995) was discovered at $V=9.2$ mag on 1995 August 24 by M. Yamamoto (Hirosawa et al 1995) which for the purposes of this paper will be defined as day zero.46 Phe nova reached visual maximum of \V=7.1 mag 115 days later on December 17., The nova reached visual maximum of $V$ =7.1 mag 115 days later on December 17.47 The light curve characteristics of V723 Cas display very slow evolution with a long pre-maximunm halt (Hachisu Ixato 2004) and have ed it to be classified as a very slow nova alongside LIH Del and RR Pie (Chochol Pribulla 1997)., The light curve characteristics of V723 Cas display very slow evolution with a long pre-maximum halt (Hachisu Kato 2004) and have led it to be classified as a very slow nova alongside HR Del and RR Pic (Chochol Pribulla 1997).48 Oscillations in the light curve are evident post-maximum. a feature also seen in the evolution of LR. Del. Optical spectroscopy (lijima. hosino Della Valle 1998. Munari ct al 1996) wior to maximum light revealed [αν narrow emission components (full width half maximum PWIALY 90 1j with P Cvgni absorptions blueshifted ~100 + relative o the emission peak.," Oscillations in the light curve are evident post-maximum, a feature also seen in the evolution of HR Del. Optical spectroscopy (Iijima, Rosino Della Valle 1998, Munari et al 1996) prior to maximum light revealed fairly narrow emission components (full width half maximum $\sim$ 90 $^{-1}$ ) with P Cygni absorptions blueshifted $\sim 100$ $^{-1}$ relative to the emission peak."49 Munari et al (1996) report that. post-maximun the emission lines increase bv a factor of several imes in intensity and. broaden to ENIIM-A550 kim s whilst the absorption components remain at a similar width to pre-maximum., Munari et al (1996) report that post-maximum the emission lines increase by a factor of several times in intensity and broaden to $\sim550$ km $^{-1}$ whilst the absorption components remain at a similar width to pre-maximum.50 Grouncl ancl space-based infrared observations (evans et al 2003) show line profiles with δις330 km and indicate an cjected mass. of 2.6«107 M. from the Dra line and 4.3«101 M. from the frec-[ree emission (assuming a distance of 4 kpc)., Ground and space-based infrared observations (Evans et al 2003) show line profiles with $\sim 330$ km $^{-1}$ and indicate an ejected mass of $2.6\times 10^{-5}$ $_\odot$ from the $\alpha$ line and $4.3\times 10^{-4}$ $_\odot$ from the free-free emission (assuming a distance of 4 kpc).51 Radio emission in novae typically arises due to free- emission from gas at temperatures of approximately 104 ly (Seacquist. 1989)., Radio emission in novae typically arises due to free-free emission from gas at temperatures of approximately $^4$ K (Seaquist 1989).52 Non-thermal components have been detected. in a few novae. the most notable example being," Non-thermal components have been detected in a few novae, the most notable example being"53We interpret the observations iu the coutext of the Starburst99 model (8B99.Leithereretal.1999)... which includes predictions for how the EWs of Ibv«droseu recombination line evolve over time.,"We interpret the observations in the context of the Starburst99 model \citep[SB99,][]{leitherer99}, which includes predictions for how the EWs of Hydrogen recombination line evolve over time."54 Therefore. our first task is to estimate IT2 line streneths from the cata.," Therefore, our first task is to estimate $\beta$ line strengths from the data."55 We attribute the excess light iu the J band. compared to the continu light measured in the I aud I bauds. to combined effect of enmission lues.," We attribute the excess light in the J band, compared to the continuum light measured in the I and H bands, to combined effect of emission lines."56 Therefore. we can compute the combined equivalent width as follows: where Wy=2815À is the effective width of the J-filter response curve. ;=1.7 to correct the observed EW to the rest frame. and f is the fux density f£ iu the respective filters.," Therefore, we can compute the combined equivalent width as follows: where $W_{J}=2845\rm{\AA}$ is the effective width of the $J$ -filter response curve, $z=1.7$ to correct the observed EW to the rest frame, and $f$ is the flux density $f_\nu$ in the respective filters."57 The relative contributions of the various emission Ies are constrained by fitting Gaussian coniponeuts to the 3 ocnuüsionu lines seen in the exis spectra shown in Figure L. keeping the ratio between the two |OIIT| colmpoucuts fixed at 3.," The relative contributions of the various emission lines are constrained by fitting Gaussian components to the 3 emission lines seen in the grism spectra shown in Figure \ref{spec}, keeping the ratio between the two [OIII] components fixed at 3."58 We only use the 3 UDS spectra as I> is only mareially detected in the GSD spectrin., We only use the 3 UDS spectra as $\beta$ is only marginally detected in the GSD spectrum.59 The enussion line ratios are remarkably similar for all 3 objects: IL) contributes 1/8 to the combined line huninosity. sugeesting a very low ietallicity (sec.e.g...Salzeretal.2005:Amorin2010.for conrparisons)..," The emission line ratios are remarkably similar for all 3 objects: $\beta$ contributes 1/8 to the combined line luminosity, suggesting a very low metallicity \citep[see, e.g.,][for60comparisons]{salzer05, amorin10}."61 Because the flux is dominated by the lino aud is therefore more directly related to our |OII]s007observations. we show the inferred συν EWs in Figure 5. (also sce Table 13).," Because the flux is dominated by the $_{5007}$ line and is therefore more directly related to our observations, we show the inferred $_{5007}$ EWs in Figure \ref{mv_ew} (also see Table \ref{tab}) )."62 However. we model the observations by fitting the inferred IL) EWs to the SB99 predictions.," However, we model the observations by fitting the inferred $\beta$ EWs to the SB99 predictions."63 These are assumed to be always l/sth of the combined EW., These are assumed to be always 1/8th of the combined EW.64 The unavoidable iutrinsic scatter in this couversion is, The unavoidable intrinsic scatter in this conversion is65Empirical mass loss formulas are pivotal lor the construction of empirical and semiempirical stellar atmosphere and wind models. stellar evolution computations and studies of the interstellar medium. among other (topics.,"Empirical mass loss formulas are pivotal for the construction of empirical and semiempirical stellar atmosphere and wind models, stellar evolution computations and studies of the interstellar medium, among other topics."66" Historically. the mass loss rate M ol late-tvpe elants and supergiants has been described by “Reimers” law’. given as M=jrDE 1971).. with L.. R,. M, as stellar luminosity. raclius. and mass. respectively. given in solar units. and 7 is a fitting parameter."," Historically, the mass loss rate $\dot{M}$ of late-type giants and supergiants has been described by “Reimers' law"", given as $\dot{M} = \eta \cdot \frac{L_* R_*}{M_*}$ \citep{rei75,rei77}, , with $L_*$, $R_*$, $M_*$ as stellar luminosity, radius, and mass, respectively, given in solar units, and $\eta$ is a fitting parameter."67 Other empirical mass loss formulas have been presented by Lamers(1981).. deJager.Nieuwenhuijzen.&vanderIlucht (1983).. and Nieuwenhuijzen&deJager(1990).. but they do not distinguish between the strong. ancl now well-described dust-driven winds (e.g.Wachterοἱal.2002).. ancl the physically. very dillerent case of nondust-driven winds.," Other empirical mass loss formulas have been presented by \cite{lam81}, \cite*{dej88}, and \cite{nie90}, but they do not distinguish between the strong, and now well-described dust-driven winds \cite[e.g.,][]{wac02}, and the physically very different case of nondust-driven winds."68 Despite its wide-ranging success. (he mass loss formula by Reimers sulfers [rom two important deficiencies.," Despite its wide-ranging success, the mass loss formula by \citeauthor{rei75} suffers from two important deficiencies."69 First. it is solely based. on dimensional scaling arguments without anv physical interpretation.," First, it is solely based on dimensional scaling arguments without any physical interpretation."70 In. particular. the appearance of (he stellar luminosity in the formula is awkward noting that for cool star winds. with the exception of molecule-driven and cdust-diven winds. the huninosity of the star is not expected to be relevant 1985).," In particular, the appearance of the stellar luminosity in the formula is awkward noting that for cool star winds, with the exception of molecule-driven and dust-driven winds, the luminosity of the star is not expected to be relevant \cite[e.g.,][]{hol85}."71". In fact. the Reimers law seems (o suggest that a certain fraction of the stellar luminosity L, is utilized to lift the wind material from (he photosphere."," In fact, the Reimers law seems to suggest that a certain fraction of the stellar luminosity $L_*$ is utilized to lift the wind material from the photosphere."72 The second deficiency. consists in the necessity of adjustimg the fitting parameter η. as different 7 values are required ad hoe (o match observed mass-Ioss rates from different tvpes of giants and supergiants.," The second deficiency consists in the necessity of adjusting the fitting parameter $\eta$, as different $\eta$ values are required ad hoc to match observed mass-loss rates from different types of giants and supergiants."73 The same is (rue. if reasonable mass loss vields and final masses are {ο be achieved through stellar evolution models with prescribed mass loss.," The same is true, if reasonable mass loss yields and final masses are to be achieved through stellar evolution models with prescribed mass loss."74 For more evolved AGB egiants. the Reimers relation is better replaced by. e.g.. deJagerοἱal.(1983).. which sugeests up to three times as much mass-loss for the tip-AGB (Schroder&Sedlmavr2001).," For more evolved AGB giants, the Reimers relation is better replaced by, e.g., \cite{dej88}, which suggests up to three times as much mass-loss for the tip-AGB \citep{schr01}."75. More recently. the Reimers relation also fails to describe revised mass loss rates [rom Ix ancl AL eiant stars. based on updated Ca II ionization balances which consider photoionization raciation deduced from FUSE spectra (larperetal.2004).," More recently, the Reimers relation also fails to describe revised mass loss rates from K and M giant stars, based on updated Ca II ionization balances which consider photoionization radiation deduced from FUSE spectra \citep{har04}."76. For further updated information on mass-loss mechanisms see. e.g.. the review by Willson(2000).," For further updated information on mass-loss mechanisms see, e.g., the review by \cite{wil00}."77. In the present work. we overcome these deficiencies bv adopting a more physical picture.," In the present work, we overcome these deficiencies by adopting a more physical picture."78 In our approach. the non-radiative energy input into the wind is assumed to be given by the turbulent energy. density. within the chromosphere or underneath. possibly related to the manifestation of (magneto-)acoustic waves.," In our approach, the non-radiative energy input into the wind is assumed to be given by the turbulent energy density, within the chromosphere or underneath, possibly related to the manifestation of (magneto-)acoustic waves."79 This approach appears to be consistent with the major conclusion by Judge&Stencel(1991).. who presented a detailed empirical analysis of the global thermodyvnanmical properties of the outer atmosphlieres and. winds of a set of well-studied cool giant and supergiant stars.," This approach appears to be consistent with the major conclusion by \cite{jud91}, who presented a detailed empirical analysis of the global thermodynamical properties of the outer atmospheres and winds of a set of well-studied cool giant and supergiant stars."80 They concluded that7[...] mass loss rates are not strongly dependent on the actual physical processes driving the winds [suggesting] that, They concluded that“[...] mass loss rates are not strongly dependent on the actual physical processes driving the winds [suggesting] that81The preseuce of absorbing material alone the iue of sight is generally believed. to be the only difference between Type 2 and a Type 1 Active Calactic Nuclei (AGN).,The presence of absorbing material along the line of sight is generally believed to be the only difference between Type 2 and a Type 1 Active Galactic Nuclei (AGN).82 This material obscures oth the cussion lines from) the Broad Line Reeion (BLR) aud the N-rav spectrum. being he main ineredient5 of the so-called Unificatiou Model.," This material obscures both the emission lines from the Broad Line Region (BLR) and the X-ray spectrum, being the main ingredient of the so-called Unification Model."83 It is usually euvisaged as a compact ‘tors’. ocated at a peo scale distance from the nucleus (e.g.2)..," It is usually envisaged as a compact `torus', located at a pc scale distance from the nucleus \citep[e.g.][]{antonucci93}."84" This distance is basically confirmed both wo dudiect techniques. such απ considerations owed on photoionization codes (e.g.?77).. aud direct imaeiueoe of the torus itself (οι,°?).. IToor therewo"," This distance is basically confirmed both by indirect techniques, such as considerations based on photoionization codes \citep[e.g.][]{bmi01,mass06}, and direct `imaging' of the torus itself \citep[e.g.][]{jaffe04}."85 oqueds expeselenceoidcne that. thisla simplc«d5 scenario NOVO.mav not hold for all objects.," However, there is evidence that this simple scenario may not hold for all objects."86" The a. Compton-thick -torus ⋅aud∖↽↴∖⊳∖∢ a ""mECompton cii 2,≓⊳↽⋅⋈⋠thin material. on a nmch larger scale. seenis to better account extendedfor the observed pheuoumnenoloss (ee.2h."," The co-existence of a Compton-thick torus and a Compton-thin material, extended on a much larger scale, seems to better account for the observed phenomenology \citep[e.g.][]{matt00b}."87" The latter absorber mav be naturalh associated to dust-lanes (οιο,» 2).. or to molecular eas in. the ealactic. disks. (7).."," The latter absorber may be naturally associated to dust-lanes \citep[e.g.][]{mvr98}, , or to molecular gas in the galactic disks \citep{lamastra06}."88 The presence of. obscuring1. matterHt on 1laree AM--(pe-kpe)] scales1 iuxi detached from the unclear torus is also supported . ↴⋝↖↽∣≲∣⇂∣↑−↙↙↕∣⊽↴∖↴↑⋯∐↸∖↴∖↴∪↕⋀∖∐↕↴↕∏∐∐∐∪∏↴∖↴≺∣∏⋜↧↴∖↴⋜∐⋅↴∖↴⋜↧↑; ⋅ hnieh⋅⋅ z which⋅ are very likely. hosted bv dusty ealaxies− (c.g.⋅⋅2?) Mor ο”... Jtat a anuabo: of Sevtert 2s presents showsignificantd variabilitylarge of the absorbing cobluun deusitv (Ng) on timescales as low as months. thus sugeesting that the absorbing material. should be ich closer to the uucleus than . ⋅ ⋅ pl ⋅⊔∪↥⋅⋯∐↖↽⇀−∙↖↖↽↸∖∐⋅≺⊔∪↴∖↴↸∖≼⋪↧↴∖↴↥⋅⋪⊔," The presence of obscuring matter on large (pc-kpc) scales and detached from the nuclear torus is also supported by studies of MIR luminous quasars at high z which are very likely hosted by dusty galaxies \citep[e.g.][]{poll08,ms06}89 Moreover, \citet{risa02b} showed that a large number of Seyfert 2s presents significant variability of the absorbing column density $\mathrm{N_H}$ ) on timescales as low as months, thus suggesting that the absorbing material should be much closer to the nucleus than assumed for the torus, possibly in the BLR itself."90∖∩⊾↖↽ picture ↴⋅⋅secs the ouly tenable for the. objects. .o ⋅⇁ ⋅⋅ list↥⋅∪⋯↓↖∏∖↸∖↨↘↽↑∪⋜∏⋝∪∏↑∩⋯∪∐↑∐∖↴∙⋜↧∐∪↖↖↽∐↓∶↴⋁∏↴∖↴ within davs- or even hours: NGC 1388NN ]o (C2). NGC 1365 (2) and NOC 1151 (2)...," This picture seems the only tenable for the objects, which present the most rapid $\mathrm{N_H}$ variations ever observed, within days or even hours: NGC 4388 \citep{elvis04}, NGC 1365 \citep{ris05} and NGC 4151 \citep{puc07}."91 Is this the cud of the torus paradie1i?, Is this the end of the torus paradigm?92 Or is it only au exception on a laudful of peculiar objects?, Or is it only an exception on a handful of peculiar objects?93 NGC“cic are7582 (z=0.0053).ra being ∙∙included iu. the ? catalog. has targeted by most Acay Clescopes: been (%).. (2). CH). Einstei," NGC 7582 (z=0.0053), being included in the \citet{pic82} catalog, has been targeted by most X-ray telescopes: \citep{mp81}, \citep{tp89}, \citep{war93}, \citep{schac98,xue98}."94nThe (?).. emerged Cringefron hese studies4 was that of a pictureflat thatS-ray spectrum dominated by heavy obscuration., The picture that emerged from these studies was that of a flat X-ray spectrum dominated by heavy obscuration.95 Thauks to the -BeppoSAN⋟∏∙DEM broad‘ baudpass.ALL: .? reported] for. he first time∙ the detection ofj a more couples geometryeconmoetrv of the absorbing material. likely constituteconstituted w two differentof components. one of which This scenario was confirmed by a combined inaeiug analysis performed with andHST.. which suggested. that the Conptou-thick orus coexists with a large-scale Comptou-tlin naterial associated with the dust lane aud cireumnelear eas is photoionized by the ACN alone torus-free LBies of sight (?)..," Thanks to the BeppoSAX broad bandpass, \citet{turn00} reported for the first time the detection of a more complex geometry of the absorbing material, likely constituted by two different components, one of which This scenario was confirmed by a combined imaging analysis performed with and, which suggested that the Compton-thick torus coexists with a large-scale Compton-thin material associated with the dust lane and circumnuclear gas is photoionized by the AGN along torus-free lines of sight \citep{bianchi07b}."96 The most interesting results came frou the alysis of the two observations. aken [ wears apart. in 2001 aud 2005 (7).," The most interesting results came from the analysis of the two observations, taken 4 years apart, in 2001 and 2005 \citep{pico07}."97. Both clearly show a completely different spectral and Hux state with respect to the 1998 BeppoSAX 6sorvation., Both clearly show a completely different spectral and flux state with respect to the 1998 BeppoSAX observation.98 The spectrum can ve well described by a imodel consisting of a conibination. of. a heavily. absorbed (Ng-~--107! chiz7) power law aud a pure reflection. component. soth obseured. by a column density of few ο1022 ," The spectrum can be well described by a model consisting of a combination of a heavily absorbed $\mathrm{N_H}\sim10^{24}$ $^{-2}$ ) power law and a pure reflection component, both obscured by a column density of few $\times10^{22}$ $^{-2}$."99"Notably.of ? detect a significant iucrease oa factor ~2 in the cohuun density of the inner, thicker absorber covering the primary ray source, between 2001 and 2005."," Notably, \citet{pico07} detect a significant increase by a factor $\sim2$ in the column density of the inner, thicker absorber covering the primary X-ray source, between 2001 and 2005."100 Iu this paper. we preseut a monitorug canrpaign. of DAP.NGC Lr7582. which.. together with. a now observation.. confinus the ↖↽⋜∐⋅↕⋜∏⋝↕∐↑⋅↖↽∪↕↑∐↸∖⋯↕⋯⊔∐≼∐∖∐↴∖↴↕↑⋅↖↽∪↕↑∐↸∖∐∐∐∖↥⋅⋅⋅⋅ . . . . absorber. but down to timescales⋅ smaller thia clay.," In this paper, we present a monitoring campaign of NGC 7582, which, together with a new observation, confirms the variability of the column density of the inner absorber, but down to timescales smaller than a day."101 During⋅⋝ the second Aunouncecmieut of. 0BODENtunity ⋜↧↴∖↴↴∖↴⋯⊔↸∖≼⊔∪↥⋅∐↸∖↑∪↥⋅∏↴∖↴∙↻∪↴∖↴↴∖↴↕↴⋝↕∙↖⇁∐↕↑∐↸∖↕≧∫⇀↕⊰↕↑↴∖↴↸∖∐∙(A02) Ni .. beat observe NGC 7582 prepat different timescales. This . . ↖↖⇁∐↕↸⊳↕⊓∐⋅↸∖↴∖↴↸∖∐↑↑∐∖⋯∪↴∖↴↑↥⋅⋜∏⋯↧⋀∖∐↖↽⋜∐⋅↕⋜↕⊓∪∐↴∖↴↸∖↖⇁↸∖↥⋅ . . probe ‘distances as1| :close asthe ‘the BLEBLR iau observed.⋅⋅ almost as far as the traditional torus.," During the second Announcement of Opportunity (AO2), we proposed a strategy to observe NGC 7582 at different timescales, from 1 week to about 6 months, allowing us to probe distances as close as the BLR and almost as far as the traditional torus."102 Moreover. his campaign complemented the scales of the order of vears already. tested withNewton.," Moreover, this campaign complemented the scales of the order of years already tested with."103 Therefore. NGC 7582 was observed four tics v du 2007 (PE AL Cliaberee): on Max Ist aud 28th. aud November 9th aud 16th.," Therefore, NGC 7582 was observed four times by in 2007 (PI: M. Chiaberge): on May 1st and 28th, and November 9th and 16th."104" rav huagine Spectrometer (NIS) aud IbLbud X-rav Detector (IND) event files were reprocessed with the latest calibration files available (2008- release). using 6.5 and Suzaku software.VersionEP 9. adopting"" standard filtering:∙⋡"," X-ray Imaging Spectrometer (XIS) and Hard X-ray Detector (HXD) event files were reprocessed with the latest calibration files available (2008-07-09 release), using 6.5 and Suzaku softwareVersion 9, adopting standard filtering procedures."105 Source⋅ ⋅↴↴⋅and background spectran for procedures. all the three NIS“Te detectors were extracted frou. circular regions of 2.9 arcnmün radius. avoiding," Source and background spectra for all the three XIS detectors were extracted from circular regions of 2.9 arcmin radius, avoiding"106Tn ow sample 59% of the X-ray detected Sevfert 1 ealaxies show significant N-rvav variability during the ROSAT AlbSky Survey aud ROSAT pointed observations.,In our sample $59\%$ of the X-ray detected Seyfert 1 galaxies show significant X-ray variability during the ROSAT All-Sky Survey and ROSAT pointed observations.107 The corresponding X-rav light curves are shown in Appcuclix AppendixC:.., The corresponding X-ray light curves are shown in Appendix \ref{tables_light}.108" Iu Πο,", In fig.109 7 we compare the ROSAT All-Sky Survey count rate with the count rate measured iu ROSAT PSPC pointed observations., \ref{fig_varias1} we compare the ROSAT All-Sky Survey count rate with the count rate measured in ROSAT PSPC pointed observations.110 The most extreme factor of variability is found for NCC 3516 (a factor of about 33 ou a timescale of 718 days)., The most extreme factor of variability is found for NGC 3516 (a factor of about 33 on a timescale of 718 days).111 For interacting and isolated Sevtert 2 ealaxies no indication for significant X-ray variability on timescales above 0.5 vears is found by comparing the ROSAT AII-Sky Survey and ROSAT PSPC pointed observations (Fie. S))., For interacting and isolated Seyfert 2 galaxies no indication for significant X-ray variability on timescales above 0.5 years is found by comparing the ROSAT All-Sky Survey and ROSAT PSPC pointed observations (Fig. \ref{fig_varias2}) ).112 The galaxy NGC 5506 is classified by Lipovetski ct al. (, The galaxy NGC 5506 is classified by Lipovetski et al. (1131987) as Sevtert type 2.,1987) as Seyfert type 2.114 This source exhibits the largest factor of variability of about 2.7 on a timescale of 375 clays., This source exhibits the largest factor of variability of about 2.7 on a timescale of 375 days.115 ILowever. for three out of the 36 Sevfert 2 ealaxies. NGC 10685. NGC. [388 and ΠΑΡ FOI175-07I0. indications for X-ray variability are) found iu ROSAT pointed observations.," However, for three out of the 36 Seyfert 2 galaxies, NGC 1068, NGC 4388 and IRAS F01475-0740, indications for X-ray variability are found in ROSAT pointed observations."116 In fig. (op) the pointed observation light curve o [the Sevfert 2 galaxy NCC LOGS is shown., In \ref{fig_NGC1068} ) the pointed observation light curve of the Seyfert 2 galaxy NGC 1068 is shown.117 An Increase in count rate from 1.823 to 2.080 countss |. correspouding to a factor of 1.11 or Acps=0.256. within 2cays is detected.," An increase in count rate from 1.823 to 2.080 $\rm counts\;s^{-1}$ , corresponding to a factor of 1.14 or $\Delta\rm cps = 0.256$, within 2days is detected."118 A coustant model fit using the 4? test cau be rejected with a probability of 99.83%. corresponding to 30.," A constant model fit using the $\chi^2$ test can be rejected with a probability of $99.83\%$, corresponding to $\sigma$."119 Tndicatious for N-rav variability in NGC 1068 are also found in other pointed observations (cf., Indications for X-ray variability in NGC 1068 are also found in other pointed observations (cf.120 fig. Bj)., fig. \ref{fig_NGC1068b}) ).121 The ROSAT PSPC lieht curve for the Sevfert 2 ealaxy IRAS 01175-0718 is shown in fig., The ROSAT PSPC light curve for the Seyfert 2 galaxy IRAS 01475-0748 is shown in fig.122 9. Griddle)., \ref{fig_NGC1068} ).123 A decrease m the count rate from 1.061 to 0.021 countss+ within 12.9 hours is detected., A decrease in the count rate from 0.064 to 0.021 $\rm counts\;s^{-1}$ within 12.9 hours is detected.124 This variability correspouds to a factor of 3 aud to a chanee iu the count rate of Acps=0.013., This variability corresponds to a factor of 3 and to a change in the count rate of $\Delta\rm cps = 0.043$.125 A constant model fit eives a probability of (1o)., A constant model fit gives a probability of $\%$ $\sigma$ ).126 Iu fie., In fig.127 9 (bottom) the ταν light curve of NGC [388 is shown., \ref{fig_NGC1068} ) the X-ray light curve of NGC 4388 is shown.128 The count rate decreases from 0.0586 to 0.0322 countss! corresponding to a factor of variabilitv of about 1.8 aud a chauge in the count rate of Acps=0.026 within 21 days., The count rate decreases from 0.0586 to 0.0322 $\rm counts\ s^{-1}$ corresponding to a factor of variability of about 1.8 and a change in the count rate of $\Delta \rm cps = 0.026$ within 21 days.129" A constant model fit can be rejected with a probability of 97.5964. corresponding to 2a,"," A constant model fit can be rejected with a probability of $\%$, corresponding to $\sigma$."130 Recently. Ceorgantopoulos Papadakis (2000) found evidence for spectral (and timing) variability for four Sevtert 2 ealaxics in RATE observations.," Recently, Georgantopoulos Papadakis (2000) found evidence for spectral (and timing) variability for four Seyfert 2 galaxies in RXTE observations."131the Spruit-Tayler dyanamo.,the Spruit-Tayler dyanamo.132 The binary system initially consists of a 18Meo star and a 17Mo star in a 4 day orbit., The binary system initially consists of a $18~\mathrm{M_\odot}$ star and a $17~\mathrm{M_\odot}$ star in a 4 day orbit.133" Mass transfer starts at t=8.09x10°yr, when the helium mass fraction in the hydrogen burning core has increased to 0.94."," Mass transfer starts at $t= 8.09134\times 10^{6}~\mathrm{yr}$, when the helium mass fraction in the hydrogen burning core has increased to 0.94."135" The mass transfer rate rises up to 8x10-4 Moyr-!, which roughly corresponds to M1/tTKH1 where M; and tTxKH,1 denote the mass and the Kelvin-Helmoltz time scale of the primary star, respectively."," The mass transfer rate rises up to $8\times10^{-4}~\mathrm{M_\odot yr^{-1}}$ , which roughly corresponds to $M_\mathrm{1}/\tau_\mathrm{KH,1}$ where $M_\mathrm{1}$ and $\tau_\mathrm{KH,1}$ denote the mass and the Kelvin-Helmoltz time scale of the primary star, respectively."136 The primary mass decreases to 7.5Mo by the end of the Case A transfer (see Fig. 4))., The primary mass decreases to $7.5~\mathrm{M_\odot}$ by the end of the Case A transfer (see Fig. \ref{fig:chem}) ).137 The second Roche-lobe overflow begins at t=8.513x10°yr when the envelope of the primary star expands due to hydrogen shell burning during the helium core contraction phase (Case AB mass transfer)., The second Roche-lobe overflow begins at $t= 8.513\times 10^{6}~\mathrm{yr}$ when the envelope of the primary star expands due to hydrogen shell burning during the helium core contraction phase (Case AB mass transfer).138" The primary star loses most of the hydrogen envelope as a result, exposing its helium core of 3.95Mo having a small amount of hydrogen (My=0.04 Μο)) in the outermost layers, as shown in the third panel of Fig. 4.."," The primary star loses most of the hydrogen envelope as a result, exposing its helium core of $3.95~\mathrm{M_\odot}$ having a small amount of hydrogen $M_\mathrm{H} =1390.04$ ) in the outermost layers, as shown in the third panel of Fig. \ref{fig:chem}."140" Although the star remains compact (R«0.9 Ro) during core helium burning, helium shell burning activated after core helium exhaustion leads to the expansion of the envelope up to ~12Ro (see Fig. 3))"," Although the star remains compact $R < 0.9~\mathrm{R_\odot}$ ) during core helium burning, helium shell burning activated after core helium exhaustion leads to the expansion of the envelope up to $\sim 12~\mathrm{R_\odot}$ (see Fig. \ref{fig:hrseq9}) )"141 during core carbon burning., during core carbon burning.142" A Case ABB mass transfer does notoccur, however, due to the large orbital separation (A=~121 Ro) at this stage, while it does"," A Case ABB mass transfer does notoccur, however, due to the large orbital separation $A = \sim143121~\mathrm{R_\odot}$ ) at this stage, while it does"144(6)) and (7)). but using the full 7-parameter Kepler formalism. not just the three Ixepler parameters displaved in equations (1)) aud (5)).,"\ref{eqn:bij}) ) and \ref{eqn:fi}) ), but using the full 7-parameter Kepler formalism, not just the three Kepler parameters displayed in equations \ref{eqn:basicform}) ) and \ref{eqn:trueform}) )."145 That is. even though the adopted orbits are circular ancl edge-on. I allow for [vee fits to the eccentricity and inclination. ancl so [or correlations between (hese parameters (as well as the (wo remaining Kepler parameters) and (he parameters of interest (aunplitude and period).," That is, even though the adopted orbits are circular and edge-on, I allow for free fits to the eccentricity and inclination, and so for correlations between these parameters (as well as the two remaining Kepler parameters) and the parameters of interest (amplitude and period)."146 The effect of allowing for these covariances is lo increase (he errors in semi-unplitude ancl period bv modest amounts relative to what would be obtained using equation (5))., The effect of allowing for these covariances is to increase the errors in semi-amplitude and period by modest amounts relative to what would be obtained using equation \ref{eqn:trueform}) ).147 Panels(a) and (b) of Figure 1. show the ratios of the true errors for the amplitude aud period. respectively. relative to the naive equations (8)) aud (9)). for both the RV (green) and astrometric (red) cases.," Panels(a) and (b) of Figure \ref{fig:apm} show the ratios of the true errors for the amplitude and period, respectively, relative to the naive equations \ref{eqn:massfrac}) ) and \ref{eqn:periodfrac}) ), for both the RV ) and astrometric ) cases."148 In fact. as Ε will discuss in 4.. for P?=T. the errors depends on phase as well as period.," In fact, as I will discuss in \ref{sec:phase}, for $P\ga T$, the errors depends on phase as well as period."149 Figure 1l therefore shows the root-mean-square of the errors. averaged over all phases.," Figure \ref{fig:apm} therefore shows the root-mean-square of the errors, averaged over all phases."150 Note that the RV amplitude errors follow the naive form until P/T~LA and then deteriorate relatively gracefullv., Note that the RV amplitude errors follow the naive form until $P/T\sim 1.1$ and then deteriorate relatively gracefully.151 By contrast. the astrometric errors beein deviaüng al P/T~0.75 and then deteriorate much more quickly," By contrast, the astrometric errors begin deviating at $P/T\sim 0.75$ and then deteriorate much more quickly."152 For the period errors. deterioration begins al P/1~0.85 for RV and 2/7~0.65 for astrometry. but the overall pattern is qualitatively similar.," For the period errors, deterioration begins at $P/T\sim 0.85$ for RV and $P/T\sim 0.65$ for astrometry, but the overall pattern is qualitatively similar."153 The mass estimates for astrometry and RY depend on dillerent combinations of amplitude and period. Ilence the fractional error in the mass (or misin/ in the case of BV) is related to the errors in (he [it parameters bv In the limit P«&T. e(m)/m—ofay)/ay. but for P2 T. the period error and the correlations become important.," The mass estimates for astrometry and RV depend on different combinations of amplitude and period, Hence the fractional error in the mass (or $m\sin i$ in the case of RV) is related to the errors in the fit parameters by In the limit $P\ll T$, $\sigma(m)/m\rightarrow \sigma(a_1)/a_1$, but for $P\ga T$ , the period error and the correlations become important."154 Figure lec shows the results of caleulations that apply equation (11))., Figure \ref{fig:apm}c c shows the results of calculations that apply equation \ref{eqn:masscovar}) ).155 Ol course. (he star may have more (han one companion (planetary or otherwise). and one may imagine arbitrarily complicated configurations.," Of course, the star may have more than one companion (planetary or otherwise), and one may imagine arbitrarily complicated configurations."156 Here I restrict myself to the next level of complication. a second companion that is sufficiently [ar away Chat its effect on the star may be (treated as uniform acceleration.," Here I restrict myself to the next level of complication, a second companion that is sufficiently far away that its effect on the star may be treated as uniform acceleration."157 Even if no such acceleration is identified. one might decide to fit lor it on the grounds that there aay be such a companion that has," Even if no such acceleration is identified, one might decide to fit for it on the grounds that there be such a companion that has"158low mass stars. the material does not form stars but remains as cooled gas. or the cooling flow model is incorrect.,"low mass stars, the material does not form stars but remains as cooled gas, or the cooling flow model is incorrect."159 Consequently. (here was considerable excitement when X-ray observations Claimed (o discover large amounts of cooled gas in galaxy. clusters wilh approximately (he masses expected from a long-lived cooling flow (Whiteetal.1991). (hereafter WFJMA).," Consequently, there was considerable excitement when X-ray observations claimed to discover large amounts of cooled gas in galaxy clusters with approximately the masses expected from a long-lived cooling flow \citep{wfjma} (hereafter WFJMA)."160 Thev used Einstein SSS data [or 21 clusters. corrected for a (üme-dependent ice build-up. and their spectral fits vielded an absorption column which they compared to the Galactic value obtained from the large-beam Bell Labs survey (Starketal.1992)..," They used Einstein SSS data for 21 clusters, corrected for a time-dependent ice build-up, and their spectral fits yielded an absorption column which they compared to the Galactic value obtained from the large-beam Bell Labs survey \citep{sgwblhh}."161 About half of the clusters (12/21) had. N-ray. absorption columns in excess of (he Galactic HI column bv al least 3o. and the excess was correlated with the deduced rate of cooling gas.," About half of the clusters (12/21) had X-ray absorption columns in excess of the Galactic HI column by at least $3\sigma$, and the excess was correlated with the deduced rate of cooling gas."162 The mass of absorbing gas within the cluster was determined to be 3xLOM—10.AL... which is approximately the amount of cooled gas that would be produced by a cooling flow over its lifetime.," The mass of absorbing gas within the cluster was determined to be $3\tenup{11}-10^{12}$, which is approximately the amount of cooled gas that would be produced by a cooling flow over its lifetime."163 The WFJMA study led to searches at other wavelengths for cold gas in cooling flow clusters. since LOM—107 oo! III or wwould be easily detected. if ils properties were simular to Galactic eas.," The WFJMA study led to searches at other wavelengths for cold gas in cooling flow clusters, since $10^{11}-10^{12}$ of HI or would be easily detected if its properties were similar to Galactic gas."164 Observational searches for HI usually vielded upper limits (Jalfe1987.1991:Dwarakanath.vanGorkom&Owen1994:ODea.Gallimore&Bani 1995).. and when HI was detected. it was (vpically two orders of magnitude lower than the expected HI mass 1995)..," Observational searches for HI usually yielded upper limits \citep{jaf87,jaf91,dvo,ogb}, and when HI was detected, it was typically two orders of magnitude lower than the expected HI mass \citep{jaf90,mob,ngjh,hjn}."165 One concern was that the ILE mieht have a velocity dispersion similar to the cluster. making it difficult to detect in narrow bandwidth studies.," One concern was that the HI might have a velocity dispersion similar to the cluster, making it difficult to detect in narrow bandwidth studies."166 ILowever. a recent wide bandwidth search for III rules out such emission. (wpically at a level of 5xLO? citepopk..," However, a recent wide bandwidth search for HI rules out such emission, typically at a level of $5\tenup{9}$ \\citep{opk}."167 searches [ον molecular hydrogen have often focused on emission or absorption Irom CO millimeter lines. which have led to stringent upper limits (AleNamara&Jaffe1994:ODeaetal.1994:Braine&DuprazDraine 1995).," Searches for molecular hydrogen have often focused on emission or absorption from CO millimeter lines, which have led to stringent upper limits \citep{mj,obmts,bd,bwrhl}."168. Recently. searches have emploved the infrared lines. usually the ((1-0)9C1) line. and emission has been detected in a lew cases (Jaffe&Bremer1997:Falcke 1998).," Recently, searches have employed the infrared lines, usually the (1-0)S(1) line, and emission has been detected in a few cases \citep{jb,frrsw}."169. In their analysis of the detections. Jalle&Bremer(1997). deduce masses that are about. 107.10AL... still inadequate by two orders of magnitude to be in agreement with the X-rav observations.," In their analysis of the detections, \citet{jb} deduce masses that are about $10^{10}$, still inadequate by two orders of magnitude to be in agreement with the X-ray observations."170 Given the limits on HI andII5.. theoretical investigations have examined whether the eas could be hidden in a form that would be clilficult to detect.," Given the limits on HI and, theoretical investigations have examined whether the gas could be hidden in a form that would be difficult to detect."171 The work of, The work of172considering the collision events of each ταν will the source surface. the quantities of interest may be read off.,"considering the collision events of each ray with the source surface, the quantities of interest may be read off."173 In the stellar-source case. however. i( is necessary to solve for the initial direction of a photon so that it will reach the observer.," In the stellar-source case, however, it is necessary to solve for the initial direction of a photon so that it will reach the observer."174 Hence we have a boundary-value problem., Hence we have a boundary-value problem.175 Figure | is à schematic illustration of our method., Figure \ref{example} is a schematic illustration of our method.176 Photons are emitted from nearby spacetime points on the stellar orbits aud travel to the observer., Photons are emitted from nearby spacetime points on the stellar orbits and travel to the observer.177 In the picture. the star emits “Minkowski photons” which feel no space curvature ancl travel in straight lines: redshift depends only on the velocity and time dilation οἱ the star.," In the left-hand picture, the star emits “Minkowski photons” which feel no space curvature and travel in straight lines; redshift depends only on the velocity and time dilation of the star."178 This is in effect the approximation used in previous work., This is in effect the approximation used in previous work.179 In (he middle picture. (he star emits “Schwarzschild photons” which [eel space curvature.," In the middle picture, the star emits “Schwarzschild photons” which feel space curvature."180" In the right-hand picture. (he star emits “frame drageine photons"" which feel spin as well as space curvalure."," In the right-hand picture, the star emits “frame dragging photons” which feel spin as well as space curvature."181 Below. ?? details the problem to be solved and the method used [ον caleulatioΕν of the redshilt.," Below, \ref{Algorithm} details the problem to be solved and the method used for calculation of the redshift."182 The Matlab scripts implementing our algorithm are available as an online supplement., The Matlab scripts implementing our algorithm are available as an online supplement.183 Then 2.2. presents the black-hole model ancl associated metrics which we use in our approach. and ?? «derives how (he various effects scale with orbit size.," Then \ref{theMetric} presents the black-hole model and associated metrics which we use in our approach, and \ref{scalesec} derives how the various effects scale with orbit size."184 We apply our algorithm to the star $2. and detail the results in Section ??..," We apply our algorithm to the star S2, and detail the results in Section \ref{Results}."185 In order to calculate the redshilt of à moving star as observed by a fixed observer. we need to solve the geodesic equations for both the star and for photons.," In order to calculate the redshift of a moving star as observed by a fixed observer, we need to solve the geodesic equations for both the star and for photons."186 Geoclesic equations are commonlv expressed in terms of (he Lagrangian. with cols denoting derivatives with respect to the affine parameter.," Geodesic equations are commonly expressed in terms of the Lagrangian, with dots denoting derivatives with respect to the affine parameter."187 Bul an equivalent formulation exists in Cerms of a Hamiltonian, But an equivalent formulation exists in terms of a Hamiltonian188the velocity at small 0p the spectral index is —1.76+0.02 using the global mean field compared to —1.9740.02 using the local mean field.,the velocity at small $\theta_B$ the spectral index is $-1.76 \pm 0.02$ using the global mean field compared to $-1.97 \pm 0.02$ using the local mean field.189 This is because the magnetic field fluctuations are large enough that the local mean field direction seen by an eddy is not the same as the global mean field direction., This is because the magnetic field fluctuations are large enough that the local mean field direction seen by an eddy is not the same as the global mean field direction.190" If the fluctuations are in critical balance, the angle between the local and global mean fields is B/Bo©kj/k ."," If the fluctuations are in critical balance, the angle between the local and global mean fields is $\delta\mathbf{B}_\perp/B_0\approx k_\para/k_\perp$."191" This suggests that when using the global mean field, the parallel scaling cannot be correctly distinguished from the perpendicular scaling, even for small óB,/Bo, because the angle of measurement to the local mean field needs to be less than ky/k,."," This suggests that when using the global mean field, the parallel scaling cannot be correctly distinguished from the perpendicular scaling, even for small $\delta\mathbf{B}_\perp/B_0$, because the angle of measurement to the local mean field needs to be less than $k_\para/k_\perp$."192 This interpretation is in agreement with previous solar wind studies that have used local and global mean field methods., This interpretation is in agreement with previous solar wind studies that have used local and global mean field methods.193" Those that use the global mean field method do not detect spectral index anisotropy (Sari&Valley1976;Tesseinetal.2009) and those that use a local mean field method do detect it (Horburyetal.2008;Podesta2009;Luo&Wu2010;Wicksetal.2010, 2011)."," Those that use the global mean field method do not detect spectral index anisotropy \citep{sari76,tessein09} and those that use a local mean field method do detect it \citep{horbury08,podesta09a,luo10,wicks10a,wicks11}."194". A similar situation is also seen in simulations, where scaling anisotropy is detected when a local mean field is used (Cho&Vishniac2000;Maron&Goldreich2001) but not when a global mean field is used (Grappin&Müller2010)."," A similar situation is also seen in simulations, where scaling anisotropy is detected when a local mean field is used \citep{cho00,maron01} but not when a global mean field is used \citep{grappin10}."195". Here, we have shown that when keeping all other parameters constant, it is indeed the use of the global or local mean field that determines whether the anisotropic scaling is measured."," Here, we have shown that when keeping all other parameters constant, it is indeed the use of the global or local mean field that determines whether the anisotropic scaling is measured."196" It seems, therefore, that the ffluctuations, both in solar wind turbulence and forced RMHD turbulence simulations, are more sensitive to the local mean field at the scale of the fluctuations than the global large scale field."," It seems, therefore, that the fluctuations, both in solar wind turbulence and forced RMHD turbulence simulations, are more sensitive to the local mean field at the scale of the fluctuations than the global large scale field."197" In the decaying simulation (not shown in Fig. 8)),"," In the decaying simulation (not shown in Fig. \ref{fig:localvsglobal}) ),"198" the local and global mean field methods are much more similar, with the parallel scaling being steeper than —2 in all cases."," the local and global mean field methods are much more similar, with the parallel scaling being steeper than $-2$ in all cases."199" One possible reason for this is that the scale separation between the global mean field and the fluctuations is not large, meaning that the global and local mean fields are similar."," One possible reason for this is that the scale separation between the global mean field and the fluctuations is not large, meaning that the global and local mean fields are similar."200" This, combined with the smaller fluctuation amplitudes in the decaying simulation, could account for the observed behaviour."," This, combined with the smaller fluctuation amplitudes in the decaying simulation, could account for the observed behaviour."201 This could be tested by performing a decaying simulation with a larger inertial range., This could be tested by performing a decaying simulation with a larger inertial range.202" In this paper, we measure the power and spectral index anisotropy of tturbulence in the solar wind and RMHD simulations using second-order structure functions."," In this paper, we measure the power and spectral index anisotropy of turbulence in the solar wind and RMHD simulations using second-order structure functions."203" The analysis technique is essentially the same for both, allowing us to make a direct comparison."," The analysis technique is essentially the same for both, allowing us to make a direct comparison."204" In the slow solar wind, we find that the magnetic field power and spectral index are anisotropic with respect to the local magnetic field direction."," In the slow solar wind, we find that the magnetic field power and spectral index are anisotropic with respect to the local magnetic field direction."205 This anisotropy has now been seen by several different methods in both fast and slow wind., This anisotropy has now been seen by several different methods in both fast and slow wind.206 In both forced and decaying simulations we also find that the power and spectral index are anisotropic in both the velocity and magnetic field., In both forced and decaying simulations we also find that the power and spectral index are anisotropic in both the velocity and magnetic field.207" In the solar wind, the perpendicular spectral index of the magnetic field is close to —5/3, in agreement with the theory of Goldreich&Sridhar(1995)."," In the solar wind, the perpendicular spectral index of the magnetic field is close to $-5/3$, in agreement with the theory of \citet{goldreich95}."208". In the forced simulation, the perpendicular spectral indices are close to —5/3 for velocity and —3/2 for the magnetic field."," In the forced simulation, the perpendicular spectral indices are close to $-5/3$ for velocity and $-3/2$ for the magnetic field."209" We are not aware of any theory that can account for this difference, although it may be caused by the velocity forcing."," We are not aware of any theory that can account for this difference, although it may be caused by the velocity forcing."210" In the decaying simulation, the perpendicular spectral index is close to —5/3 for both the velocity and magnetic field."," In the decaying simulation, the perpendicular spectral index is close to $-5/3$ for both the velocity and magnetic field."211" In all cases, the spectral index steepens at small angles to the magnetic field."," In all cases, the spectral index steepens at small angles to the magnetic field."212" The parallel scaling obtained in the solar wind and forced simulations is close to —2, which agrees with the theories based on critical balance of both Goldreich&Sridhar(1995) and Boldyrev(2006)."," The parallel scaling obtained in the solar wind and forced simulations is close to $-2$, which agrees with the theories based on critical balance of both \citet{goldreich95} and \citet{boldyrev06}."213". The parallel spectral indices in the decaying simulation are —2.33+0.03 for the velocity and —2.30+0.03 for the magnetic field, which are steeper than the critical balance predictions."," The parallel spectral indices in the decaying simulation are $-2.33 \pm 0.03$ for the velocity and $-2.30 \pm 0.03$ for the magnetic field, which are steeper than the critical balance predictions."214We based our initial classification on the ratios of diagnostic cussion lines.,We based our initial classification on the ratios of diagnostic emission lines.215 Previous studies used iustead a conibinatioun of line ratios and equivalent widths (see the Iutroduction) that can be more affected Gn particular when only upper hits can be derived) by the quality of the data and bv the contrast with the continu -τσLl.sion., Previous studies used instead a combination of line ratios and equivalent widths (see the Introduction) that can be more affected (in particular when only upper limits can be derived) by the quality of the data and by the contrast with the continuum emission.2164. Similarly. we only used ratios of lines with small wavelength separation. not affected by the possible effects oe-ernal reddening. that cau be particularly severe when considering e.g. the 10 ΠΙΛΟΤΟΥ line.," Similarly, we only used ratios of lines with small wavelength separation, not affected by the possible effects of internal reddening, that can be particularly severe when considering e.g. the [O $\lambda$ 3727 line."217 Thus our procedure is expected to produce a rather robust method of spectral identification., Thus our procedure is expected to produce a rather robust method of spectral identification.218 Nonetheless. our classificatious are overall in good agreement with those found in the iterature ou a object by object basis.," Nonetheless, our classifications are overall in good agreement with those found in the literature on a object by object basis."219 For example. comparing our results with those of Willottetal.(1999) for the 3CRR sources we fouud 52 objects in common.," For example, comparing our results with those of \citet{willott99} for the 3CRR sources we found 52 objects in common."220 Leaving aside 2 objects of the newly iutroduced. class of ELEC. aud 3 objects that we consider as unclassitied. (2 reported as LEC. namely 3€ 035 aud 3€ 319. 1 as TEC. 3C 138) the identification i the various classes coiucides with oulv 3 exceptions for the remaining £7 radio-galaxies.," Leaving aside 2 objects of the newly introduced class of ELEG, and 3 objects that we consider as unclassified (2 reported as LEG, namely 3C 035 and 3C 319, 1 as HEG, 3C 438) the identification in the various classes coincides with only 3 exceptions for the remaining 47 radio-galaxies."221 These are: 3€ 388. à LEG from our analysis (with an excitation index of E.L20.62) against the previous TEC identification. and two galaxies. 3C 079 aud 3C 223. where we do not see a broad linecoumponcut?.. contrasting with their sugeestedOO membership in the class of Weak Quasars.," These are: 3C 388, a LEG from our analysis (with an excitation index of E.I.=0.62) against the previous HEG identification, and two galaxies, 3C 079 and 3C 223, where we do not see a broad line, contrasting with their suggested membership in the class of Weak Quasars."222 Ledlow&Owen(1996) conrpared the optical Ro band magnitude of the host galaxies with the total radio cluission at 1.[ GIIz., \citet{ledlow96} compared the optical R band magnitude of the host galaxies with the total radio emission at 1.4 GHz.223 They found that sources locate in different areas of the plot depending ou their radio morphology: as already. known frou the pioneering study of Εαπατο&Riley(1971) FR II sources have higher radio powers than FR I sources and they separate at a huninosity of ~2«4107? W |! at 178 MIIEz., They found that sources locate in different areas of the plot depending on their radio morphology: as already known from the pioneering study of \citet{fanaroff74} FR II sources have higher radio powers than FR I sources and they separate at a luminosity of $\sim 2\times10^{25}$ W $^{-1}$ at 178 MHz.224 The novel result of Ledlow&Owen(1996) cousists in the fact that the FR I/II division shows a depeudeuce on Mj., The novel result of \citet{ledlow96} consists in the fact that the FR I/II division shows a dependence on $M_{\rm host}$.225 FR I sources hosted bv the nore lunünous galaxies cau have radio powers higher than the average ER I/FR II separation., FR I sources hosted by the more luminous galaxies can have radio powers higher than the average FR I/FR II separation.226 The separation between FR I and FR IT is rather sharp over the whole range of radio power., The separation between FR I and FR II is rather sharp over the whole range of radio power.227 Iu Fig., In Fig.228 10. (left panel) we plotted the 3CR sources in the plane radio huninosity (at 178 MIIz) versus the naenitude in IT head ofthe ost galaxy (reported in Table 2))., \ref{ledlow} (left panel) we plotted the 3CR sources in the plane radio luminosity (at 178 MHz) versus the magnitude in H band of the host galaxy (reported in Table \ref{speclas}) ).229 We selected the IT band since it provides the most complete coverage (~ ) for the 3€CR sample bv using neasurclcuts from the 2ALASS (Skrutskieetal.2006) or. When this is not available. from IST images (Donzellietal. 2007).," We selected the H band since it provides the most complete coverage $\sim$ ) for the 3CR sample by using measurements from the 2MASS \citep{skrutskie06} or, when this is not available, from HST images \citep{donzelli07}."230. For the BLO we also corrected the host iuninositv for the coutribution of their brigh IR unelei. ucasured by Baldietal.(2009).," For the BLO we also corrected the host luminosity for the contribution of their bright IR nuclei, measured by \citet{baldi09}."231. In order to compare our results with those of Ledlow&Owen(1996) we used the color correction from Manuuncecietal.(2001).. R- II 2 2.5. and scaled the 1.1 CUz data to 178 MITz adopting a radio spectriun in the form FyXvU.T," In order to compare our results with those of \citet{ledlow96} we used the color correction from \citet{mannucci01}, R - H = 2.5, and scaled the 1.4 GHz data to 178 MHz adopting a radio spectrum in the form $F_{\nu} \propto \nu^{-0.7}$."232 The relative scarcity of FR Tsources in the 3CR sample prevents us from exploring in detail the host maenitude- separation between the FR classes., The relative scarcity of FR I sources in the 3CR sample prevents us from exploring in detail the host magnitude-dependent separation between the FR classes.233 However. the FR location for our sample is consistent with the separation introduced bv Ledlow&Owen(1996).," However, the FR location for our sample is consistent with the separation introduced by \citet{ledlow96}."234. We also checked that this result holds using radio ποσάπως at 1.1 1) as well as host magnitude in other bands (ic. V xd)., We also checked that this result holds using radio luminosities at 1.4 GHz as well as host magnitude in other bands (i.e. V band).235 Du line with their results we fud a few exceptions. associated with FR Isources of extremely high radio powers m vorv massive hosts.," In line with their results we find a few exceptions, associated with FR I sources of extremely high radio powers in very massive hosts."236 Iun the rieht xuel woe iutroduced the optical spectroscopic classification. separating the 3C'R sources iuto WEG and LEC.," In the right panel we introduced the optical spectroscopic classification, separating the 3CR sources into HEG and LEG."237 For the LEG class we further consider the FR type., For the LEG class we further consider the FR type.238 WEG and LEG/FR II sources are well mixed above the FR I/FR II separation. having the same median in terius of radio power. and oulv a small offset iu the," HEG and LEG/FR II sources are well mixed above the FR I/FR II separation, having the same median in terms of radio power, and only a small offset in the"239and Fedunetal.(2011). have indicated how vortex motions generate a significant amount of Povnting Ilux directed outwards from the photosphere. and. as a result. may be (he source of various observed. MIID wave moces.,"and \citet{fedun11} have indicated how vortex motions generate a significant amount of Poynting flux directed outwards from the photosphere, and, as a result, may be the source of various observed MHD wave modes."240 In this paper. we use high spatial and temporal resolution observations. in addition to numerical simulations. to determine (he velocity distribution of a large sample of ΔΙ)».," In this paper, we use high spatial and temporal resolution observations, in addition to numerical simulations, to determine the velocity distribution of a large sample of MBPs."241 The observations and munerical simulations are described in 2.. while the methodology used. and the values obtained for the velocities of AIBP structures. are detailed in 3..," The observations and numerical simulations are described in \ref{obs}, while the methodology used, and the values obtained for the velocities of MBP structures, are detailed in \ref{analy}."242 As our tracking algorithm can detect and monitor bright point chains. as well as isolated brghtenings and merger events. we believe (hat (his is a unique study of the. dynamics of AIBPs in the solar photosphere.," As our tracking algorithm can detect and monitor bright point chains, as well as isolated brightenings and merger events, we believe that this is a unique study of the dynamics of MBPs in the solar photosphere."243 Differences between (he velocity characteristics of MDPs. and (those that undergo mergers wilh other bright points. are discussed in 3..," Differences between the velocity characteristics of non-merging MBPs, and those that undergo mergers with other bright points, are discussed in \ref{analy}."244 Finally. our concluding remarks are given in 1..," Finally, our concluding remarks are given in \ref{conc}."245 The data emploved in this study. were obtained using the Rapid Oscillations in the Solar Atmosphere (ROSA:Jessetal.2010) instrument. which is installed as a user facility at the 76 em Dunn Solar Telescope (DST). in New Mexico. USA.," The data employed in this study were obtained using the Rapid Oscillations in the Solar Atmosphere \citep[ROSA;][]{Jess10} instrument, which is installed as a common-user facility at the 76 cm Dunn Solar Telescope (DST), in New Mexico, USA."246 Observations were obtained during a period of excellent seeing on 2009 May. 28. using a 9.2 wide filter centred al 43805 (G-band).," Observations were obtained during a period of excellent seeing on 2009 May 28, using a $9.2$ wide filter centred at $4305$ (G-band)."247" We observed a 70x10"" quiet Sun region at disk centre for ~50 minutes. achieving diffraction-limited imaging with 0"".069 !."," We observed a $70'' \times 70''$ quiet Sun region at disk centre for $\sim$ 50 minutes, achieving diffraction-limited imaging with $0''.069$ $^{-1}$ ."248" The images were reconstructed using Speckle algorithms (Wógeretal.2003).. while image ce-stretching was performed using a 40x grid (equatingtoaz1"".7separationbetweenspatialsamples:etal.2007. 2008)."," The images were reconstructed using Speckle algorithms \citep{Wog08}, while image de-stretching was performed using a $40 \times 40$ grid \citep[equating to a $\approx1''.7$ separation between 249spatial samples;][]{Jess07, Jess08}."250. These processes were implemented to remove (he effects of atmospheric seeing from the dataset., These processes were implemented to remove the effects of atmospheric seeing from the dataset.251 G-band images were taken at a raw cadence of 0.033 s. while alter speckle reconstruction the cadence was reduced to 0.528 s. Reconstructed images were then binned into consecutive groups of four to improve the signal-to-noise anc reduce (he overall volume of the dataset. providing a final image cadence of 2.1 s. simulated G-band images were produced using the detailed radiative transport technique described by Shelvagetal.(2004)... with the solar photospheric magneto-convection models for the radiative transport calculations provided by the MUBRAM radiative MIID code 2005).," G-band images were taken at a raw cadence of 0.033 s, while after speckle reconstruction the cadence was reduced to 0.528 s. Reconstructed images were then binned into consecutive groups of four to improve the signal-to-noise and reduce the overall volume of the dataset, providing a final image cadence of 2.1 s. Simulated G-band images were produced using the detailed radiative transport technique described by \citet{Shel04}, with the solar photospheric magneto-convection models for the radiative transport calculations provided by the MURaM radiative MHD code \citep{Vog05}."252.. A computational domain of size 12x1.4 Mm. was emploved [or the simulations. resolved by 480x100 grid cells. providing a horizontal (wo-pixel resolution of 50 km.," A computational domain of size $12\times12\times1.4$ $^3$, was employed for the simulations, resolved by $480\times480\times100$ grid cells, providing a horizontal two-pixel resolution of 50 km."253 The level corresponding to the visible solar surface is located approximately GOO kin below the upper boundary of thedomain., The level corresponding to the visible solar surface is located approximately 600 km below the upper boundary of thedomain.254 Side boundariesof the domain are periodic. while the upper boundary is closed [or vertical ancl stress-free horizontal plasma motions and," Side boundariesof the domain are periodic, while the upper boundary is closed for vertical and stress-free horizontal plasma motions and"255 (o: M.. aw (e.g..Johnson1981.1987:1993.2003;Stelzer2004).," $\alpha\omega$ $M_\odot$ $\alpha\omega$ \citep[e.g.,][]{johnson81,johnson87,tagliaferri90,drake96,fleming00,rutledge00,fleming93,fleming03,stelzer04}."256 (a7 (e.g..Durneyetal.," $\alpha^2$ \citep[e.g.,][]{durney93}."2571993).. Ha (Gizisetal.2000:Mohanty&Basri2003).. (e.g..Fleming," $\alpha$ \citep{gizis00,mohanty03}. \citep[e.g.,][]{fleming93,fleming03}."258etal.1993.2003).. Berger(2002) Putnam2005).," \citet{berger02} \citep{berger05,burgasser05}."259. regime with the Australia Telescope Compact Array (ATCA)., regime with the Australia Telescope Compact Array (ATCA).260 Our target. ε Ind Bab. is the closest known brown dwarf and is composed of two objects with spectral types ΤΙ and Τό separated by 0773 (2.65 AU at a distance of 3.626 pe; Scholz2004)).," Our target, $\epsilon$ Ind Bab, is the closest known brown dwarf and is composed of two objects with spectral types T1 and T6 separated by $0\farcs 73$ $2.65$ AU at a distance of $3.626$ pc; \citealt{scholz03,smith03,volk03,mccaughrean04}) )."261 We list the binary’s main characteristics in Table 1I.., We list the binary's main characteristics in Table \ref{tab:epsind}.262 Blank(2005) failed to detect e Ind Bab with ATCA. but our radio observation goes deeper.," \citet{blank05} failed to detect $\epsilon$ Ind Bab with ATCA, but our radio observation goes deeper."263 Despite our best effort to coordinate the and ATCA observations of the e Ind Bab binary. the X-ray observations were delayed by a few days due to satellite safety reasons.," Despite our best effort to coordinate the and ATCA observations of the $\epsilon$ Ind Bab binary, the X-ray observations were delayed by a few days due to satellite safety reasons."264 A log of the observations is given in Table 2.., A log of the observations is given in Table \ref{tab:log}.265 The ATCA was in a long-baseline configuration (6D); we used 4.8 GHz and 8.64 GHz receivers with bandwidths of 128 MHz., The ATCA was in a long-baseline configuration (6D); we used 4.8 GHz and 8.64 GHz receivers with bandwidths of 128 MHz.266" Observing scans ranged from 10 min to 20 min on source. depending on the weather conditions. whereas we used 3 min scans for the phase calibrator,"," Observing scans ranged from 10 min to 20 min on source, depending on the weather conditions, whereas we used 3 min scans for the phase calibrator."267" About 5 minutes at the start of each observing round were spent on the flux calibrator,", About 5 minutes at the start of each observing round were spent on the flux calibrator.268 The observing conditions on the first day were average with cloud coverage: however. the last three hours of the first day of observation were essentially useless due to strong winds and a thunderstorm.," The observing conditions on the first day were average with cloud coverage; however, the last three hours of the first day of observation were essentially useless due to strong winds and a thunderstorm."269 In contrast. the weather conditions were much better during the second day with generally a cloud-free sky.," In contrast, the weather conditions were much better during the second day with generally a cloud-free sky."270 We combined both observing rounds and reduced the ATCA data using the MIRIAD software (Saultetal.1995)., We combined both observing rounds and reduced the ATCA data using the MIRIAD software \citep{sault95}.271 We detected 9 sources in the 4.8 GHz map: we used boxes of about width centered on these sources and applied a CLEAN algorithm using uniform. weighting., We detected 9 sources in the 4.8 GHz map; we used boxes of about width centered on these sources and applied a CLEAN algorithm using uniform weighting.272 About 5.4 mJy were thus removed in 202 iterations (we stopped when a negative value was encountered)., About $5.4$ mJy were thus removed in 202 iterations (we stopped when a negative value was encountered).273 Although no sources were visible in the dirty map at 8.64 GHz. we used the boxes around the first five brightest sources detected at longer wavelengths and performed a CLEAN algorithm.," Although no sources were visible in the dirty map at 8.64 GHz, we used the boxes around the first five brightest sources detected at longer wavelengths and performed a CLEAN algorithm."274 About 0.5 mJy were removed after 38 iterations., About $0.5$ mJy were removed after 38 iterations.275 The rms noise levelinthecleaned mapsreached 26.4 and 37.3 at 4.8 and 8.64 GHz. respectively.," The rms noise levelinthecleaned mapsreached $26.4$ and $37.3$ at 4.8 and 8.64 GHz, respectively."276 The values are close to thetheoreticalvalues (25.1 and 35.6 jiJy))., The values are close to thetheoreticalvalues $25.1$ and $35.6$ ).277 No source was, No source was278Let us now discuss the conditions required for the strong-cooling model developed in the preceding sections to be valid.,Let us now discuss the conditions required for the strong-cooling model developed in the preceding sections to be valid.279 First. an obvious necessary. coudiGon for the strong-cooling. strong-compression regime is (hat equation (2.4)) has a large-;1 solution. Ac1.," First, an obvious necessary condition for the strong-cooling, strong-compression regime is that equation \ref{eq-entropy-A}) ) has a $A$ solution, $A\gg 1$."280 However. the actual situation is somewhat more subtle.," However, the actual situation is somewhat more subtle."281 The condition sls>1 is just the condition for the of a stationary stronglv-cooled. state of the reconnection laver., The condition $A\gg 1$ is just the condition for the of a stationary strongly-cooled state of the reconnection layer.282 In addition. however. we must impose an extra evolutionary condition for the svstem to be able to reach this state.," In addition, however, we must impose an extra evolutionary condition for the system to be able to reach this state."283 As we shall see below. this will result in a certain requirement for the radiative cooling function.," As we shall see below, this will result in a certain requirement for the radiative cooling function."284 The picture thal we have in mind here is the following., The picture that we have in mind here is the following.285 The ambient plasma upstream of the reconnection laver is rather tenuous: when it just enters (he laver. it becomes subject to ohmic heating ancl its temperature rises. whereas ils density does not change appreciably ab first.," The ambient plasma upstream of the reconnection layer is rather tenuous; when it just enters the layer, it becomes subject to ohmic heating and its temperature rises, whereas its density does not change appreciably at first."286 If radiative cooling can be neglected. one always gets the classical SweetParker laver solution. with relatively low density η2ny and relatively high temperature TT4 (corresponding to “lt7 1).," If radiative cooling can be neglected, one always gets the classical Sweet–Parker layer solution, with relatively low density $n\simeq n_0$ and relatively high temperature $T\simeq T_{\rm eq}$ (corresponding to $A\simeq 1$ )."287 The transition to the strong-cooling. strong-compression 12» reeime described in the previous sections happens only if that <Ac1 SweetParker laver becomes unsustainable in (he presence of radiative cooling. ie.. if it is able to cool and collapse towards the d21 solution sulliciently rapidly.," The transition to the strong-cooling, strong-compression $A\gg 1$ regime described in the previous sections happens only if that $A\simeq 1$ Sweet–Parker layer becomes unsustainable in the presence of radiative cooling, i.e., if it is able to cool and collapse towards the $A\gg 1$ solution sufficiently rapidly."288 For this to happen. we must require that the radiative cooling of the corresponding “lL1 SweetParker solution be stronger that the corresponding Οπής heating. i.e.: Quanto. (n0]] > η. (1)]y. where we used equation (2.4)) in the last step.," For this to happen, we must require that the radiative cooling of the corresponding $A\simeq 1$ Sweet–Parker solution be stronger that the corresponding Ohmic heating, i.e.: [n_0, (n_0)] > [n_0, (n_0)], where we used equation \ref{eq-Q_ohm-2}) ) in the last step."289 Now. assuming that a stationary strong-cooling solution with slc1 does exist. it is convenient to make use of the corresponding heating-cooling balance equation," Now, assuming that a stationary strong-cooling solution with $A\gg 1$ does exist, it is convenient to make use of the corresponding heating-cooling balance equation"290Results for (&) obtained by dividing the inner galaxy into radial bins of width LR. are shown on figure 9.,Results for $\left< \kappa \right>$ obtained by dividing the inner galaxy into radial bins of width $\frac{1}{30} R_{\odot}$ are shown on figure 9.291 The histogram shows an arithmetic average over (lie various lines ol sight through each annulus., The histogram shows an arithmetic average over the various lines of sight through each annulus.292 Also shown on figure 9 as small circles are the individual channel values of 7 in (he absorption spectra. converted to & by mulliplving by the velocity eradient. a given by the rotation curve (see Burton. 1055. [or a review of the significance ol the velocity graclient).," Also shown on figure 9 as small circles are the individual channel values of $\tau$ in the absorption spectra, converted to $\kappa$ by multiplying by the velocity gradient, $\frac{dv}{ds}$, given by the rotation curve (see Burton, 1988, for a review of the significance of the velocity gradient)."293 The curves on figure 9 indicate observational selection based on our noise level in 7 for the case of G328.4240.22. one of our brighter sources.," The curves on figure 9 indicate observational selection based on our noise level in $\tau$ for the case of G328.42+0.22, one of our brighter sources."294 The lower curve shows the opacity which would result [rom optical depth equal to one sigma. as defined bv the emission fluctuations discussed above.," The lower curve shows the opacity which would result from optical depth equal to one sigma, as defined by the emission fluctuations discussed above."295 Weaker absorption (han (his is not detectable. so multiplving by the velocity eradient eives (he corresponding lower limit for detectable opacity.," Weaker absorption than this is not detectable, so multiplying by the velocity gradient gives the corresponding lower limit for detectable opacity."296" Spectra toward Lainter sources (hat give higher o, will have higher lower limits.", Spectra toward fainter sources that give higher $\sigma_{\tau}$ will have higher lower limits.297 Points near or below this curve on ligure 9 are upper limits on &., Points near or below this curve on figure 9 are upper limits on $\kappa$.298" On the high side. when the absorption lines are very deep. the noise prevents us from distinguishing between optical depths greater than 7,4,=—ln(20;}."," On the high side, when the absorption lines are very deep, the noise prevents us from distinguishing between optical depths greater than $\tau_{max}=-\ln \left( 2 \sigma_{\tau} \right)$."299" The line saturates al this point (typically 7 3). ancl we set the optical depth to this value ibe*«2o,."," The line saturates at this point (typically $\tau \sim 3$ ), and we set the optical depth to this value if $e^{-\tau}< 2 \sigma_{\tau}$."300 Multiplied by the velocity gradient. (his maximum detectable 7 gives an upper limit on the measurable &. shown for the case of G328.42+0.22 by the upper curve.," Multiplied by the velocity gradient, this maximum detectable $\tau$ gives an upper limit on the measurable $\kappa$, shown for the case of G328.42+0.22 by the upper curve."301 Again. for spectra wilh higher noise level (he upper limit is lower. as evidenced by chains of points al various levels corresponding (o saturated lines.," Again, for spectra with higher noise level the upper limit is lower, as evidenced by chains of points at various levels corresponding to saturated lines."302 Points near the upper curve on figure 9 are (hus lower limits on &., Points near the upper curve on figure 9 are thus lower limits on $\kappa$.303 The velocity integrals of (he absorption spectra are given on table 1. columns 6 and 7.," The velocity integrals of the absorption spectra are given on table 1, columns 6 and 7."304 Column 6 gives (he integral over the entire negative velocily range corresponding to the inner galaxy., Column 6 gives the integral over the entire negative velocity range corresponding to the inner galaxy.305 Column 7 gives the integral over therestricted velocity range corresponding to the points on figure 9. i.e. [rom zero km ! to the recombination line velocity.," Column 7 gives the integral over therestricted velocity range corresponding to the points on figure 9, i.e. from zero km $^{-1}$ to the recombination line velocity."306 Whether the region is at the near or [ar distance. we can be confident that at least over (his velocity range onlv (he near-side gas can contribute to the absorption.," Whether the region is at the near or far distance, we can be confident that at least over this velocity range only the near-side gas can contribute to the absorption."307 For the regions we stop 3km ! short of the recombination line velocity. (o avoid the deep absorption usually seen just bevond the region's velocity.," For the regions we stop 3 km $^{-1}$ short of the recombination line velocity, to avoid the deep absorption usually seen just beyond the region's velocity."308 The SNR. G328.42+0.22. is known to be bevond the solar circle on Che far side of the Galaxy (Gaensler. Dickel. and Green. 2000): in this case we carry the velocity integration to the terminal velocity at (his longitude. (," The SNR G328.42+0.22, is known to be beyond the solar circle on the far side of the Galaxy (Gaensler, Dickel, and Green, 2000); in this case we carry the velocity integration to the terminal velocity at this longitude. ("309Varving the 3 km | offset to 10 or even 15 km ! has minimal effect on the opacities on figure 9. although it reduces the number of lines of sight contributing to some of the annuli.),"Varying the 3 km $^{-1}$ offset to 10 or even 15 km $^{-1}$ has minimal effect on the opacities on figure 9, although it reduces the number of lines of sight contributing to some of the annuli.)"310 Figure 10 shows the geometrv of the Galactic plane in the fourth. cquacdrant (from MeClue-GQrffiths et al., Figure 10 shows the geometry of the Galactic plane in the fourth quadrant (from McClure-Griffiths et al.311 2001)., 2001).312 The sun's location is assumed to be al (roy)=(0.8.5). and we plot lines of sight at longitude ancl3337... the boundaries of the test region studied in (his paper.," The sun's location is assumed to be at $(x,y)=(0,8.5)$, and we plot lines of sight at longitude and, the boundaries of the test region studied in this paper."313 Annuli ave plotted with radii from 0.4 to 1.0 Ro. and the major spiral features," Annuli are plotted with radii from 0.4 to 1.0 $R_{\odot}$ , and the major spiral features"314"N-rav- endssion associated with radio jets in extragalactic sources now cones ina large variety of diverse characteristics,",X-ray emission associated with radio jets in extragalactic sources now comes in a large variety of diverse characteristics.315 When this sort of Cluission was first isolated (c.g. the radio ealaxy AST with the EINSTEIN OBSERVATORY. Schreier. Gorenstein. Feigelsou 1982) aud there were oulv a few examples. it was tempting to work on the assumption that the ΟΡΙΟ process was definable aud would apply to all examples.," When this sort of emission was first isolated (e.g. the radio galaxy M87 with the EINSTEIN OBSERVATORY, Schreier, Gorenstein, Feigelson 1982) and there were only a few examples, it was tempting to work on the assumption that the emission process was definable and would apply to all examples."316 This notion persisted into the ROSAT era until a couvincing case was iade that the terminal hotspots of Cyenus A represented svuchrotrou selt-Conmptou (SSC) emission. (Warris. Carilli. Perley. 1991).," This notion persisted into the ROSAT era until a convincing case was made that the terminal hotspots of Cygnus A represented synchrotron self-Compton (SSC) emission (Harris, Carilli, Perley, 1994)."317 With the advent of the CUANDRA OBSERVATORY. he uuuber of sources has alinost tripled (from 7 to at least 19) ye there remain substantial problems in deternünius the cussion process responsible or the N-rays in some sources.," With the advent of the CHANDRA OBSERVATORY, the number of sources has almost tripled (from 7 to at least 19) yet there remain substantial problems in determining the emission process responsible for the X-rays in some sources."318" Althoueli broken oower laws connecting the radio. optical. aud. X-rav data are still viable spectral models for a few sources, a uunmboer of other sources are observed to rave such a low flux deusitv iu the optical that the indicated cutoff iu the spectrum would preclude a simple conection to the N-ray data."," Although broken power laws connecting the radio, optical, and X-ray data are still viable spectral models for a few sources, a number of other sources are observed to have such a low flux density in the optical that the indicated cutoff in the spectrum would preclude a simple connection to the X-ray data."319 The introduction of the beaming model (Tavecchio et al., The introduction of the 'beaming model' (Tavecchio et al.320 2000. Celotti et al.," 2000, Celotti et al."321 2001) in the particular case of PISSOG37 was presented as an escape from this dilemuna., 2001) in the particular case of PKS0637 was presented as an escape from this dilemma.322 Iu this model. the euhancement of," In this model, the enhancement of"323»ulse observed in high-energv channel is narrower than in ower-energv. channel. which is manifested by the fact that he FAWHIAIs of the pulses in separated energy. channels decrease with the energy in à power-law form.,"pulse observed in high-energy channel is narrower than in lower-energy channel, which is manifested by the fact that the FWHMs of the pulses in separated energy channels decrease with the energy in a power-law form."324 However. he pulse narrowing we obtained is less prominent than hat observed in real GRBs for which pulse width decay »ower-law. index -0.4: while the pulse width decay index we obtained. for instance in the case of Figure 2. (one-sided exponential decay emission profile). is -0.13.," However, the pulse narrowing we obtained is less prominent than that observed in real GRBs for which pulse width decay power-law index $\sim$ -0.4; while the pulse width decay index we obtained, for instance in the case of Figure \ref{expfig} (one-sided exponential decay emission profile), is -0.13."325" Other than the ""Band"" spectrum. we also used. an alternative function - a low-energy power [aw plus the . . ⋅∕⋅ ↓⊔⋏∙≟↓↥⊣⋅⊔⋖⊾↓⋅⋏∙≟∙∖⇁⋖⋅⇀∖↓≻∪⊔⋖⋅⊔⇂⋯↓≼↛⋯−∪∐⋜⊔∆↗∣−⇂∪↓⋅↿↓⊔⋅↓⋅∢⋅⊳∖⇂−∐⋅⋜⋯↓⋖⊾⋅ spectrum."," Other than the “Band” spectrum, we also used an alternative function - a low-energy power law plus the high-energy exponential cut-off at $E_p'$ - for the rest-frame spectrum."326 For the typical values of the parameters we have used. this changing of spectrum only narrows the Channel 4 pulse by ~6%.. hence hardly changes the slope of the pulse width versus the energy.," For the typical values of the parameters we have used, this changing of spectrum only narrows the Channel 4 pulse by $\sim$, hence hardly changes the slope of the pulse width versus the energy."327 The spectral lag is an important observational property of je pulse in GRBs in that it may be usec to derive the 'osmological distribution of (ας (Norris 2002) ancl to iscriminate the internal shock signature and the external shock signature in the pulses (e.g.. Llakkila Ciblin 2004).," The spectral lag is an important observational property of the pulse in GRBs in that it may be used to derive the cosmological distribution of GRBs (Norris 2002) and to discriminate the internal shock signature and the external shock signature in the pulses (e.g., Hakkila Giblin 2004)."328 This motivates us to probe the dependences of the peak lags oin other physical parameters of the simple model., This motivates us to probe the dependences of the peak lags on other physical parameters of the simple model.329 We choose le symmetric Gaussian. profile as the intrinsic emission profile. which includes an intrinsic rising phase.," We choose the symmetric Gaussian profile as the intrinsic emission profile, which includes an intrinsic rising phase."330" We alter the Lorentz factor E. the spectral parameters of the rest frame emission (a. 3 and £7,7 ). the radius of the radiation surface H. and the rest-frame duration ἐς of the intrinsic radiation. respectively. and see how the Channel 1/3 and the Channel 1/4 peak lags vary. with these changes."," We alter the Lorentz factor $\Gamma$, the spectral parameters of the rest frame emission $\alpha$, $\beta$ and $E_p'$ ), the radius of the radiation surface $R$, and the rest-frame duration $t_d'$ of the intrinsic radiation, respectively, and see how the Channel 1/3 and the Channel 1/4 peak lags vary with these changes."331 Figure 4 shows that the lag decreases with the Lorentz factor following Lag xLot., Figure \ref{lag_gamma} shows that the lag decreases with the Lorentz factor following Lag $\propto \Gamma^{-1}$.332 We think this is a natural outcome of the relativistic boosting of the time structure., We think this is a natural outcome of the relativistic boosting of the time structure.333 ‘Those pulses whose lags are larger may come from colliding shells with low Lorentz factors. according to the current standard models (e.g. Piran 1999).," Those pulses whose lags are larger may come from colliding shells with low Lorentz factors, according to the current standard models (e.g., Piran 1999)."334bv an equivalent deviation An of the differential antenna direction.,by an equivalent deviation $\Delta\bf{n}$ of the differential antenna direction.335 The deviation induced by {he pseudo-dipole signal in a released WMAP map can be easily modeled., The deviation induced by the pseudo-dipole signal in a released WMAP map can be easily modeled.336 For each measured temperature difference in the TOD used to produce the map. we substitute Cae pseudo-dipole signal caleulated by Eq.," For each measured temperature difference in the TOD used to produce the map, we substitute the pseudo-dipole signal calculated by Eq."337 2 for an assumed An (o produce a new., 2 for an assumed $\Delta\bf{n}$ to produce a new.338 The temperature map produced by map [rom ihe new TOD can be used as a model of the pseudo-dipole signal in the released map., The temperature map produced by map from the new TOD can be used as a model of the pseudo-dipole signal in the released map.339 In calculation we use the spacecraft coordinate svstem (N.Y.Z). where the X. axis is parallel to plane of radiators. —Z is the anti-sun direction of the spin axis. and Y is perpendicular to both.," In calculation we use the spacecraft coordinate system $(X, Y, Z)$ , where the $X$ axis is parallel to plane of radiators, $-Z$ is the anti-sun direction of the spin axis, and $Y$ is perpendicular to both."340 The WMAP spacecraft scans the skv with a livbrid motion mode consists of rotation and precessing., The WMAP spacecraft scans the sky with a hybrid motion mode consists of rotation and precessing.341 In spacecralt coordinates. (he LOS unit vectors of its (wo antennas are close lo Cr.y.z)=(0.0.94.20.33) aud Ce.y.2)=(0.—0.94. —0.33). and the spacecraft rotation is around the Z-axis.," In spacecraft coordinates, the LOS unit vectors of its two antennas are close to $(x,y,z)=(0, 0.94, -0.33)$ and $(x,y,z)=(0, -0.94, -0.33)$ , and the spacecraft rotation is around the $Z$ -axis."342" Suppose the overall LOS error An from all such effects is made up of three small vectors (0.0.0). (0.0.0) and (0.0.0) with ὁ= 0.01. each alone on the final map induces its own full-sky distribution of deviation 90/,. 9/, and 9/.. respectively,"," Suppose the overall LOS error $\Delta\bf{n}$ from all such effects is made up of three small vectors $(\delta, 0, 0)$, $(0,\delta, 0)$ and $(0, 0, \delta)$ with $\delta=0.01$ , each alone on the final map induces its own full-sky distribution of deviation $\delta t_x$ , $\delta t_y$ and $\delta t_z$, respectively."343" The induced deviation upon the released WAITAPT vear-1 Ql1-band map. 0/,. 0/,. and of. are shown in Fig."," The induced deviation upon the released WMAP7 year-1 Q1-band map, $\delta t_x$, $\delta t_y$, and $\delta t_z$ are shown in Fig."344 1., 1.345 Results lor other vears and other bands are similar., Results for other years and other bands are similar.346 It is easy to see [rom Fig., It is easy to see from Fig.347" 1 that 9/, and of. are highly correlated.", 1 that $\delta t_y$ and $\delta t_z$ are highly correlated.348" In. order (o avoid degeneracy issue we only use οἱ, and οἱ, in model fitting.", In order to avoid degeneracy issue we only use $\delta t_x$ and $\delta t_y$ in model fitting.349" The clean full skv temperature map where // is the corresponding WAIAP CAMB temperature map and the coellicients ο, and e, can be determined by minimizing /7,,,,.", The clean full sky temperature map where $t'$ is the corresponding WMAP CMB temperature map and the coefficients $c_x$ and $c_y$ can be determined by minimizing $t_{clean}^2$.350" Usinge the standard IDL programex “reeress”e we gel e,=—0.35. e,=—0.18 for the WMAPT vear-1 QlI-band map."," Using the standard IDL program ""regress"" we get $c_x=-0.35$, $c_y=-0.78$ for the WMAP7 year-1 Q1-band map."351 We also model and remove the pseudo-dipole signal from the released WAIAP7T vear-1 to vear-7 maps of QI. Q2. Vl. V2. WI. W2. W3 and W4 bands.separatelv.," We also model and remove the pseudo-dipole signal from the released WMAP7 year-1 to year-7 maps of Q1, Q2, V1, V2, W1, W2, W3 and W4 bands,."352. From the clean maps we calculate (heir power spectraand residual quadruples., From the clean maps we calculate their power spectraand residual quadruples.353 Table 1lists the obtained residual cquacdrupoles ofdifferent bands., Table 1lists the obtained residual quadrupoles ofdifferent bands.354 The overallaverage clean quadiupole power for all bands is found to be, The overallaverage clean quadrupole power for all bands is found to be355Iu any case. due to the degeneracy of the material iu the density aud temperature ranges considered heroe. the computed ignition times (or the fits) rarely differ between the two cases more than,"In any case, due to the degeneracy of the material in the density and temperature ranges considered here, the computed ignition times (or the fits) rarely differ between the two cases more than."35650.. niaiu-sequence star comes frou the CNO aud °°FeimmeleiMostoftheinitialmctallicity inherited frou its ambient interstellar iuediuni at birth., Most of the initial metallicity of main-sequence star comes from the CNO and } nuclei inherited from its ambient interstellar medium at birth.357 The slowest step in the CNO cvele is , The slowest step in the hydrogen-burning CNO cycle is proton capture onto }.358"ThisresultsinalltheCNO catalysts piling up into hwdrogeu-buruimg1ENwhenlydrogenprotoncaptureoutoHN. burning ou παπάς, all of the themainsequenceiscompleted.", This results in all the CNO catalysts piling up into } when hydrogen burning on the main sequence is completed.359Divinghelimm !Nisconvertediuto??No.," During helium burning, all of the } is converted into }."360 Asa proxy for investigating the effects of metallicity in our a-chain based reaction network. we consider ignition while increasing the fraction of Νο(audthus this surrogate by usiug," Asa proxy for investigating the effects of metallicity in our $\alpha$ -chain based reaction network, we consider the ignition of a $\cfrac = 0.5$ constant-pressure ignition while increasing the fraction of } (and thus decreasing the abundance of oxygen)."361 Neinlaveernetworks.decreasingtheabuudauceofoxvecu).We'llverify, We'll verify this surrogate by using } in larger networks.362 The effects of increasingBS. from 0 to 0.05. and then further to 0.1 and 0.2. is shown in Fie. 5..," The effects of increasing from 0 to 0.05, and then further to 0.1 and 0.2, is shown in Fig. \ref{fig:nediff}. ."363 Addition of even fairly modest amounts of neon can siguificautly 201) reduce the ignition times for, Addition of even fairly modest amounts of neon can significantly ) reduce the ignition times for364 It is now fairly well-established that WofRavet (WR) stars av display two distinct (but pr'bably nou mutually exclusive) spectroscopic p)otterust of variability: (a) small-scale enuüission features servations:1ically moving from the line ceuter to the line on| an ⋅hourly Igitimescale le(c.g.?d .of Lépine1998)): (5) ιατ]ςnoetric‘ally larger line-profile deformations operating ou mich contimmun. onegcr basis ( davs. e.g.. Sunith&Willis199," It is now fairly well-established that Wolf-Rayet (WR) stars may display two distinct (but probably non mutually exclusive) spectroscopic patterns of variability: (a) small-scale emission features systematically moving from the line center to the line wings on an hourly timescale (e.g., \cite{Lepinephd}) ); (b) dramatically larger line-profile deformations operating on a much longer basis $\sim$ days, e.g., \cite{Smithwillis}) )."365" 1)). i the first phenomenon. observed ii most. (Gf not all) WR stars. is believed to be the consequence of the ↕⋟↥⋅⋜↧∶↴∙⊾⋯↸∖∐↸∖≼↧∙↻∪↴∖∷∖↴∏⋝↕↖⇁↑↿∐⋅↴⋝∏↕↸∖∐↑∐⋜↧⊓∐⋅↸∖∪↕≯↑∐↸∖⋯↕↑∏∪↖↖⇁∙↑↕↓↸∖ origi⋅⋅o of the ⋡↾↕↕⇈‘latter 4tvpeY of, variabiliUtiili vods still very much elusive."," Although the first phenomenon, observed in most (if not all) WR stars, is believed to be the consequence of the fragmented, possibly turbulent nature of the outflow, the origin of the latter type of variability is still very much elusive."3661 Remarkable; in. this. respect. is. the existence of a wel-estalished (although strougv epoch-cdepeudcut) large-scale. pattern of varjabilitv iu tjio two apparcutly single WR stars (P = 2Miwaτουpan’ + 0.002 d: Firmianiὁal. 1980)) and (P 227 + O01 d MeCandlissetal.199111 Moreletal. 1999)).," Remarkable in this respect, is the existence of a well-established (although strongly epoch-dependent) large-scale, pattern of variability in the two apparently single WR stars $\cal P$ = 3.763 $\pm$ 0.002 d; \cite{Firmani}) ) and $\cal P$ = 2.27 $\pm$ 0.04 d; \cite{McCandliss94}; \cite{Morel98b}) )."367" A majx observaional effort has been cirected on cstablising the true nature of these peculiar ohjects. Ίνα, whether this evelical variability is imduced by an orbiting uuseen (colapsed?)"," A major observational effort has been directed on establishing the true nature of these peculiar objects, i.e., whether this cyclical variability is induced by an orbiting unseen (collapsed?)"368 companion or bv the LOational modulation of large-scale wind structures (6.8... Vreuxctal.1992:: Morel1 908: Ihuriesetal. 1999... aud reerences therein).," companion or by the rotational modulation of large-scale wind structures (e.g., \cite{Vreux}; \cite{Morel98phd}; \cite{Harries}, and references therein)."369 Αλλοιeh the exact nature of these stews has vet to be muahicuouslv. settled. these studies reveal that rotatioial iioc.lation constitutes an attractive alernative to the|iuuv livothesis. especiaIv considering the recent recoenilon tha sole © stars (ti6 progenitors of WR stars) might possess such aziniutliaIv structured 0!tilows (Fullerto1etal.!997: Ikapereta. 1997)).," Although the exact nature of these stars has yet to be unambiguously settled, these studies reveal that rotational modulation constitutes an attractive alternative to the binary hypothesis, especially considering the recent recognition that some O stars (the progenitors of WR stars) might possess such azimuthally structured outflows \cite{Fullerton}; \cite{Kaper}) )."370" A prime target for furt1er Investigations is the seldom, apparentv single WN 5 star (ID lool) that has receutlv been shown to prese ioa spectral"," A prime target for further investigations is the seldom-studied, apparently single WN 5 star (HD 4004) that has recently been shown to present a spectral"371be possible only on high-mass acereting WDs (Mj&1 AL. ).,be possible only on high-mass accreting WDs $M_{1} \gtappeq 1$ $_{\odot}$ ).372 Moreover. Ne are the only class of novae in which the WD is expected to gain more mass between eruptions than it loses during them.," Moreover, RNe are the only class of novae in which the WD is expected to gain more mass between eruptions than it loses during them."373 This would make E Pyx a strong candidate Type la supernova progenitor., This would make T Pyx a strong candidate Type Ia supernova progenitor.374 However. the recent study of the system by Schaefer.Pagnotta&Shara(2000). (see also Selvellietal. 2003)) suggests. first. that P. Pyx.js. in fact. in a transient evolutionary state. and. second. that. integrated over many nova eruptions. its WD does lose more mass than it. gains.," However, the recent study of the system by \cite{b18} (see also \citealp{b20}) ) suggests, first, that T Pyx, in fact, in a transient evolutionary state, and, second, that, integrated over many nova eruptions, its WD does lose more mass than it gains."375 Alore specifically. Schaefer.Pagnotta&Shara(2009) suggest that T Pyx was an ordinary cataclysmic variable until it erupted as a nova in. 1866.," More specifically, \cite{b18} suggest that T Pyx was an ordinary cataclysmic variable until it erupted as a nova in 1866."376" This eruption triggered a winel-clriven supersoft: X-ray phase (as. first. suggested bv Ἱνηίσσο,Wine&Patterson 2000)). resulting in an unusually high. luminosity and. acerction rate."," This eruption triggered a wind-driven supersoft X-ray phase (as first suggested by \citealp{b7}) ), resulting in an unusually high luminosity and accretion rate."377" However. unlike in the original scenario proposed by WKnigee.Wine&Patterson (2000). Che supersoft phase is not self-sustaining. so that the aceretion rate has. been declining ever. since the 1866 nova eruption. Fron. AL~lo"" M. vero to 10 M."," However, unlike in the original scenario proposed by \cite{b7}, the supersoft phase is not self-sustaining, so that the accretion rate has been declining ever since the 1866 nova eruption from $\dot{M} \sim 10^{-7}$ $_{\odot}$ $^{-1}$ to $10^{-8} $ $_{\odot}$ $^{-1}$."378 As a result. T Pyx has faded: by almost 2 magnitudes since the nova eruption (Schaefer.Pagnotta&Shara 2009)).," As a result, T Pyx has faded by almost 2 magnitudes since the nova eruption \citealp{b18}) )."379 Dased on this. and the fact that T Pvx has already passed. its mean recurrence time by more than 20 vears. Schaefer.Pagnotta&Shara(2009) argue that Po Pyx might no longer even. be a recurrent. nova.," Based on this, and the fact that T Pyx has already passed its mean recurrence time by more than 20 years, \cite{b18} argue that T Pyx might no longer even be a recurrent nova."380 Lf these ideas are correct. T Pyx is not a viable SN. la progenitor. ancl its remaining Lifetime can be substantially longer than a few million vears.," If these ideas are correct, T Pyx is not a viable SN Ia progenitor, and its remaining lifetime can be substantially longer than a few million years."381 However. if all its ordinary nova eruptions are followed by. relatively long-lived ( 100. vrs) intervals of winel-clriven evolution at. high AL. its secular evolution may nevertheless be strongly affected. with significant implications for CV evolution more generally (sce also Section 5)).," However, if all its ordinary nova eruptions are followed by relatively long-lived (> 100 yrs) intervals of wind-driven evolution at high $\dot{M}$, its secular evolution may nevertheless be strongly affected, with significant implications for CV evolution more generally (see also Section \ref{SD}) )."382 A kev assumption in virtually all of these arguments is that the photometric period measured by Pattersonοἱ(1998). is. in fact. the orbital period of the system.," A key assumption in virtually all of these arguments is that the photometric period measured by \cite{b13} is, in fact, the orbital period of the system."383 So far. there has only been one attempt to obtain a spectroscopic period for P Pyx. by. Vogt.ctal.(1990)... who reporte a spectroscopic modulation with P=344 hours.," So far, there has only been one attempt to obtain a spectroscopic period for T Pyx, by \cite{b26}, who reported a spectroscopic modulation with $P=3.44$ hours."384 Such a long orbital period above the CV period gap would. be much more consistent with the high accretion rate four in T Pyx., Such a long orbital period above the CV period gap would be much more consistent with the high accretion rate found in T Pyx.385 In this study. we present the first. definitive spectroscopic determination of the orbital period of T Pyx. showing that it ijs. in fact. consistent with Patterson e al," In this study, we present the first definitive spectroscopic determination of the orbital period of T Pyx, showing that it is, in fact, consistent with Patterson et al."386s photometric period.,'s photometric period.387" We also use our. time-resolvec spectroscopy to estimate the main system parameters; such as the velocity semi-amplitude of the white cwarl (A1). the mass-ratio (q). the masses (M, anc AM») and the orbita inclination. (7)."," We also use our time-resolved spectroscopy to estimate the main system parameters, such as the velocity semi-amplitude of the white dwarf $K1$ ), the mass-ratio (q), the masses $M_{1}$ and $M_{2}$ ) and the orbital inclination $i$ )."388 Finally. we discuss the implications of our results for the evolution of P Pyx and related svstenis.," Finally, we discuss the implications of our results for the evolution of T Pyx and related systems."389 Alulti-Bibre Spectroscopy. of T. Pyx was obtained. during five nights in 2004 and 2005 with the GLIILAEFISFLAALES instrument mounted on the Unit Telescope 2 of the VET at ESO Paranal. Chile.," Multi-fibre Spectroscopy of T Pyx was obtained during five nights in 2004 and 2005 with the GIRAFFE/FLAMES instrument mounted on the Unit Telescope 2 of the VLT at ESO Paranal, Chile."390 Phe data were taken in the integrated lieldeunit mode., The data were taken in the integrated field-unit mode.391" “Phe total field. of view in this mode is about 11.5"" 7.3"" and thus covers most of T Pyx's 10"" diameter nove shell (Williams1982))."," The total field of view in this mode is about 11.5"" $\times$ 7.3"" and thus covers most of T Pyx's 10"" diameter nova shell \citealp{b29}) )."392. We used. the fibre system ARGUS. which consists of BIT fibres. cüstributed across the field. of which 5 are pointing to a calibration unit ancl 15 are pointing to sky.," We used the fibre system ARGUS, which consists of 317 fibres distributed across the field, of which 5 are pointing to a calibration unit and 15 are pointing to sky."393 Ehe grating order was 4. which gives à resolution of Ro12000.," The grating order was 4, which gives a resolution of R=12000."394 Phe wavelength range was chosen between 4501 to 5078AL. so that the emission-line speetrum would include the Bowen blend at 4645 4650 and the Hell at 4686A.," The wavelength range was chosen between 4501 to 5078, so that the emission-line spectrum would include the Bowen blend at 4645 – 4650 and the HeII at 4686."395 With this setup. the dispersion is 0.2 pix. corresponding to about 12.5 pix.," With this setup, the dispersion is 0.2 /pix, corresponding to about 12.5 $^{-1}$ /pix."396 The full widths at. half-maximum (EWILMsS) of a few spectral lines obtained simultaneously with the science spectra from libres pointing to the calibration unit. indicate. that. the spectral resolution is about 0.4A., The full widths at half-maximum (FWHMs) of a few spectral lines obtained simultaneously with the science spectra from fibres pointing to the calibration unit indicate that the spectral resolution is about 0.4.397 A log of the observations can be found in Table L.., A log of the observations can be found in Table \ref{tab:obs}.398 The initial steps in the data reduction were performed using the ESO pipeline for CEILAEFEIS., The initial steps in the data reduction were performed using the ESO pipeline for GIRAFFE.399 Ehe pipeline is based on the reduction software BLDRS from the Observatory of Geneva., The pipeline is based on the reduction software BLDRS from the Observatory of Geneva.400 Phe basie functions. of the pipeline. are to oovide master calibration data and. dispersion. solutions., The basic functions of the pipeline are to provide master calibration data and dispersion solutions.401 The pipeline also provides an image of the reconstructed icld of view. which can be used to associate a specific fibre o à given object.," The pipeline also provides an image of the reconstructed field of view, which can be used to associate a specific fibre to a given object."402 In order to extract the spectrum. from he desired. fibre and to correct. for the contribution from he sky. background: and. cosmic ravs. the output. from the pipeline was processed. further in LRAF.," In order to extract the spectrum from the desired fibre and to correct for the contribution from the sky background and cosmic rays, the output from the pipeline was processed further in IRAF."403 The PSE of our arget. T Pyx. is spread out over several fibres and 6 single ibre spectra containing significant target [Tux were extracted and median combined after weighting cach spectrum by the mean Fux in the region 4660 - 4710 A(covering Holl at 4686 A)).," The PSF of our target, T Pyx, is spread out over several fibres and 6 single fibre spectra containing significant target flux were extracted and median combined after weighting each spectrum by the mean flux in the region 4660 - 4710 (covering HeII at 4686 )."404 Cosmic rays were removed by first binning the spectra over 7 pixels (the cosmic ravs have a typical width of 2 6 pixels)., Cosmic rays were removed by first binning the spectra over 7 pixels (the cosmic rays have a typical width of 2 $-$ 6 pixels).405 Phe smoothed spectra were then subtracted from the corresponding fibre spectra to only. leave the residuals and the cosmic rays., The smoothed spectra were then subtracted from the corresponding fibre spectra to only leave the residuals and the cosmic rays.406 Cosmic ravs were then removed and the residuals added back to the target spectra., Cosmic rays were then removed and the residuals added back to the target spectra.407 Fibres containing the sky were extracted and combined to ereate a master-sky spectrum., Fibres containing the sky were extracted and combined to create a master-sky spectrum.408 Finally. the master-sky. was subtracted from the combined science spectrum.," Finally, the master-sky was subtracted from the combined science spectrum."409 T Pyx was observed again during four nights in March 2008. this time on the 6.5 meter Baacle telescope Magellan Lat Las Campanas. Chile (see Table L for a log of the observations).," T Pyx was observed again during four nights in March 2008, this time on the 6.5 meter Baade telescope Magellan I at Las Campanas, Chile (see Table \ref{tab:obs} for a log of the observations)."410" Long-«lit. spectroscopy. was obtained with the instrument INLACS using the Cra-L200-17.45 grating with a 0.9"" slit. resulting in a clispersion of 0.386. —pix. corresponcling o about 25 pix."," Long-slit spectroscopy was obtained with the instrument IMACS using the Gra-1200-17.45 grating with a 0.9"" slit, resulting in a dispersion of 0.386 /pix, corresponding to about 25 $^{-1}$ /pix."411 We estimate that the spectral resolution of these observations is about 1.5A. as measured rom the EWIIMSs of a few spectral lines in the arc-Iamp spectra.," We estimate that the spectral resolution of these observations is about 1.5, as measured from the FWHMs of a few spectral lines in the arc-lamp spectra."412 The overall spectra span over four CCDs and cover a total wavelength range of 4000 500A., The overall spectra span over four CCDs and cover a total wavelength range of 4000 – 4800.413. ATE frames from he four CCDs were treated separately during both the 2D and LD reduction steps., All frames from the four CCDs were treated separately during both the 2D and 1D reduction steps.414 The speetra were reduced. in. LAF using. standard packages., The spectra were reduced in IRAF using standard packages.415 A master bias produced. from combining all bias Tames obtained during the four nights was subtracted from he science frames and the overscan region was used. to remove the residual bias., A master bias produced from combining all bias frames obtained during the four nights was subtracted from the science frames and the overscan region was used to remove the residual bias.416 A master fat-Hield. corrected. for illumination effects was produced.," A master flat-field, corrected for illumination effects was produced."417 The spectra were then lat-Giclel corrected. ancl extracted., The spectra were then flat-field corrected and extracted.418 Line. iclentifications of, Line identifications of419xocessius.,processing.420" However. the existeuce of Πο would have Oo be taken iuto account. because old svstenis with donors that have undergone significant mass loss could wave Ny;=2.105 and would thus have high nova rates, regardless of the accreted moetallicitv."," However, the existence of $^3$ He would have to be taken into account, because old systems with donors that have undergone significant mass loss could have $X_3\gtrsim2\E{-3}$ and would thus have high nova rates, regardless of the accreted metallicity."421. A proper wecdiction of the effect of donor composition requires a population-svuthesis calculation that includes further colplications such as binary aud donor evolutiou., A proper prediction of the effect of donor composition requires a population-synthesis calculation that includes further complications such as binary and donor evolution.422 We cave this exercise for future work., We leave this exercise for future work.423 To date. most observations only report ealactically-averaged ova rates. although some ΑΟ studies (Ciarcdulloetal.1987:Capaccioli1989:Daruleyal.2006) have found that NM31*s bulee produces more πονας per stellar huuinositv than its disk by a factor of ~5.," To date, most observations only report galactically-averaged nova rates, although some M31 studies \citep{ciar87,cap89,darn06} have found that M31's bulge produces more novae per stellar luminosity than its disk by a factor of $\sim 5$."424 On the other haud. ealactically-averaged uova rates do not see any imnorphologv depeudeuce. finding instead that the huninosity-specific nova rate (LSNR) is roughly across all galaxy types at a value of δν| constantaoeLoa) (Williams&Shatter2001).. where Log is the A-—band solar huninositv.," On the other hand, galactically-averaged nova rates do not see any morphology dependence, finding instead that the luminosity-specific nova rate (LSNR) is roughly constant across all galaxy types at a value of $2 \pm 1$ $^{-1}$ $(10^{10} \ L_{\odot, \, K})^{-1}$ \citep{ws04}, where $L_{\odot, \, K}$ is the -band solar luminosity."425 The LAIC and SAIC (and possibly Vireo chwart elliptical ealaxies: Neill&Shara 2005)). which have LSNRs higher by a factor of 3. are exceptions.," The LMC and SMC (and possibly Virgo dwarf elliptical galaxies; \citealt{ns05}) ), which have LSNRs higher by a factor of 3, are exceptions."426 These mneasurenmieuts have large error bars due to μπα πο statistics and issues of completion caused by. both extinction and iufrequeut observations., These measurements have large error bars due to small number statistics and issues of completion caused by both extinction and infrequent observations.427 A better measurement of nova rates will conie with new deep optical surveys with high cadeuces such as Pan-STARRS-1. Pau-STARRBS-1. iud the Large Synoptic Survey Telescope. which will see thousauds of novae every vear.," A better measurement of nova rates will come with new deep optical surveys with high cadences such as Pan-STARRS-1, Pan-STARRS-4, and the Large Synoptic Survey Telescope, which will see thousands of novae every year."428 These will reduce the nova rate error bars and also possibly allow us to measure rates in different populations within other galaxies besides M21, These will reduce the nova rate error bars and also possibly allow us to measure rates in different populations within other galaxies besides M31.429 Tn addition to this population-averaged observable. the composition could also have a detectable effect on individual systems.," In addition to this population-averaged observable, the composition could also have a detectable effect on individual systems."430 Ta particular. the depletion of fuel can significantly increase the surface Ηχος] above the baseline set by the entropy released ching compression of the accreted laver (see Figs.," In particular, the depletion of fuel can significantly increase the surface luminosity above the baseline set by the entropy released during compression of the accreted layer (see Figs."431 10 and 11))., \ref{fig:lumz} and \ref{fig:lumx3}) ).432 While this increase is still well below the accretiou lIuimunositv associated with eravitational energy release. it would be visible while the svstemi was du disk quiescence.," While this increase is still well below the accretion luminosity associated with gravitational energy release, it would be visible while the system was in disk quiescence."433 Recurrent novae. in particular. would be ideal systenis in which to observe this brightening in quiesceut DIunünositv due to their short recurrence times of ~30 vr.," Recurrent novae, in particular, would be ideal systems in which to observe this brightening in quiescent luminosity due to their short recurrence times of $\sim 30 $ yr."434 These observables are depeudent on the assuiptio- that convection is not initiated in C-cuhauced material., These observables are dependent on the assumption that convection is not initiated in C-enhanced material.435 If instead U-vich euvelope material penetrates into C-rich material aud triggers convection there. the accreted composition will have little effect.," If instead H-rich envelope material penetrates into C-rich material and triggers convection there, the accreted composition will have little effect."436 Thus. if these colmpositiou-dependcut effects are observed. they will provide evidence that convection for imu novae is initiated iu C-poor material aud that CNO euricluneut for these novae is due to convective shear mixiug or overshoot.," Thus, if these composition-dependent effects are observed, they will provide evidence that convection for many novae is initiated in C-poor material, and that CNO enrichment for these novae is due to convective shear mixing or overshoot."437 We thank D. Towuslev for discussions. D. Paxton aud F. Tiuuues for valuable assistance with MESA. aud J. Steiufadt for nova rate calculations.," We thank D. Townsley for discussions, B. Paxton and F. Timmes for valuable assistance with MESA, and J. Steinfadt for nova rate calculations."438 This work was supported bv the National Science Foundation uuder evauts PITY 05-51161 aud AST 07-07633, This work was supported by the National Science Foundation under grants PHY 05-51164 and AST 07-07633.439This suggests that. bv studying the backgrounds behind dwarl stars. we can predict the limes of future events and plan frequent. sensitive. high-resolution observations to test [or evidence of the planet's effect as a lens (2)..,"This suggests that, by studying the backgrounds behind dwarf stars, we can predict the times of future events and plan frequent, sensitive, high-resolution observations to test for evidence of the planet's effect as a lens \citep{DiStefano2005}."440 What can we learn [rom (targeted observations?, What can we learn from targeted observations?441 We can determine if (here is a planet in the habitable zone: we will either discover evidence of it. or place a quantifiable limit on the existence of one.," We can determine if there is a planet in the habitable zone: we will either discover evidence of it, or place a quantifiable limit on the existence of one."442 If there is a planet. we can. under ideal circumstances. measure ils mass and its projected distance [rom the star.," If there is a planet, we can, under ideal circumstances, measure its mass and its projected distance from the star."443 To plan such observations. we must identilv a large reservoir of potential lenses ling in [ront of dense backerounds.," To plan such observations, we must identify a large reservoir of potential lenses lying in front of dense backgrounds."444" In. front of the Magellanie Clouds. e.g.. we expect there to be approximately 200 (1600) dwarf stars with D,«50 pe (CD,<200 pe)."," In front of the Magellanic Clouds, e.g., we expect there to be approximately $200$ $1600$ ) dwarf stars with $D_L < 50$ pc $D_L < 200$ pc)."445 Equation (4) indicates that. if observations with small values of fy and/or ον can be conducted. roughly 1056 of these stars. will produce detectable events within a decade.," Equation (4) indicates that, if observations with small values of $f_T$ and/or $\theta_{mon}$ can be conducted, roughly $10\%$ of these stars will produce detectable events within a decade."446 The first step. therefore. is to study the backgrounds behind the known dwarf stars Iviug in front of the chosen background field. to determine which are the ones most likely to produce events in the near future. and (o predict the likely event ünmes so (hat appropriate monitoring observations can be planned for the duration of the predicted event.," The first step, therefore, is to study the backgrounds behind the known dwarf stars lying in front of the chosen background field, to determine which are the ones most likely to produce events in the near future, and to predict the likely event times so that appropriate monitoring observations can be planned for the duration of the predicted event."447 For dwarl stars that do not lie in front of dense fields. serendipitous events are nevertheless possible: one such event has been observed (?7?)..," For dwarf stars that do not lie in front of dense fields, serendipitous events are nevertheless possible; one such event has been observed \citep{Gaudi2007, Fukui2007}."448 between the positions of nearby dwarf stars and the poistions of more distant sources of light may identify future positional coincidences (hat can be predicted with hieh accuracy (?).., Cross-correlation between the positions of nearby dwarf stars and the poistions of more distant sources of light may identify future positional coincidences that can be predicted with high accuracy \citep{SalimGould}. .449 Targeted lensing observations have been suggested by a variety of authors (??7??).. but have not vet been carried out.," Targeted lensing observations have been suggested by a variety of authors \citep{Feibelman1986, PaczynskiNearby, SalimGould,450DiStefano2005}, but have not yet been carried out."451 Instead. astronomers have conducted large observing prograis in which (μον have monitored tens of millions of stars per night. (???)..," Instead, astronomers have conducted large observing programs in which they have monitored tens of millions of stars per night \citep{MACHO5.7,452OGLEIIIcatalog, EROSII3year}."453 More than 3.500 microlensing event candidates have been discovered to date.," More than $3,500$ microlensing event candidates have been discovered to date."454 The microlenses tend to be located al distances greater (han several kiloparsecs. and their presence is revealed (hrough (heir action as lenses.," The microlenses tend to be located at distances greater than several kiloparsecs, and their presence is revealed through their action as lenses."455 While theoretical work has explored the discovery of nearby lenses by these programs (?).. several examples of nearby lenses suggest that dwarf stars constitute as many as LO-20% of the lenses producing detectable events (????)..," While theoretical work has explored the discovery of nearby lenses by these programs \citep{DiStefano2007}, several examples of nearby lenses suggest that dwarf stars constitute as many as $10-20\%$ of the lenses producing detectable events \citep{Nguyen2004, Kallivayalil2006,456Gaudi2007, Fukui2007}."457 For monitoring programs. the expected rate of events caused by nearby dwarfs. with constant spatial density. Ny. in front of a sourcefield of area Q. is," For monitoring programs, the expected rate of events caused by nearby dwarfs, with constant spatial density $N_L,$ in front of a sourcefield of area $\Omega,$ is"458to orbital phase over more than one rotation period iu cach season (e.g. 1996.. Busaetal. 19993).,"to orbital phase over more than one rotation period in each season (e.g. \citealt{1996ApJ...470.1172D}, \citealt{1999A&A...350..571B}) )."459 Iu this work we analyse the short aud long-term claomospheric activity of the nou-eclipsiug RS CVni stars most observed by IUE aud we compare the chromospheric and photospheric patterus of variability., In this work we analyse the short and long-term chromospheric activity of the non-eclipsing RS CVn stars most observed by IUE and we compare the chromospheric and photospheric patterns of variability.460 Iu particular. in Section refsec.calibbaja.. we derive a relation to determine the Mouut Wilson iudex from the Me line-core fluxes measured on IUE low-resolution spectra.," In particular, in Section \\ref{sec.calibbaja}, we derive a relation to determine the Mount Wilson index from the Mg line-core fluxes measured on IUE low-resolution spectra."461 As an application of this calibration. iu Section refsec.rscvn owe studv the Mount Wilson indices we derived from both IVE hieh aud low-resolution spectra of the RS CVn stars ΠΟ 22168 (V711 Tau. UR 1099). TD 21212 (UN Ari) and ΠΟ 221085 (II Peg).," As an application of this calibration, in Section \\ref{sec.rscvn} we study the Mount Wilson indices we derived from both IUE high and low-resolution spectra of the RS CVn stars HD 22468 (V711 Tau, HR 1099), HD 21242 (UX Ari) and HD 224085 (II Peg)."462 Iu Paper πο iute-calibrated the iudex 5 aud Mg Hhxes using quasi-multaueous IUE high-resolution observations for a set of dwarf stars with spectral types F to I. Since in the present work we inteud to use this calibration to study the activity of RS CV svstems. which have sub-egiauts as their primary star. we check the accuracy of applying this calibration to cool sub-eiauts.," In Paper I we inter-calibrated the index $S$ and Mg fluxes using quasi-simultaneous IUE high-resolution observations for a set of dwarf stars with spectral types F to K. Since in the present work we intend to use this calibration to study the activity of RS CVn systems, which have sub-giants as their primary star, we check the accuracy of applying this calibration to cool sub-giants."463 To do so. we obtained the IVE Ale fiuxes for those cool sub-giaut stars also observed at Alot Wilson Observatory.," To do so, we obtained the IUE Mg fluxes for those cool sub-giant stars also observed at Mount Wilson Observatory."464 In Fig., In Fig.465 1 we show the calibration of Paper L aud we include these sub-eiauts.," \ref{fig.calibS_alta} we show the calibration of Paper I, and we include these sub-giants."466 It cau be seen that these stars follow this calibration within the statistical OCYYOIS., It can be seen that these stars follow this calibration within the statistical errors.467 Since the IUE database also provides a large number of low-resolution spectra of stars. in particular RS CVu stars. it is inportaut to incorporate to the svsteniatic studies of magnetic activity.," Since the IUE database also provides a large number of low-resolution spectra of late-type stars, in particular RS CVn stars, it is important to incorporate to the systematic studies of magnetic activity."468 To this cud. in what follows we analyse the relation between the Me line-core fixes derivedfrom IUE low-resolution spectra aud the Mouut Wilson iudex.," To this end, in what follows we analyse the relation between the Mg line-core fluxes derivedfrom IUE low-resolution spectra and the Mount Wilson index."469 Iu Fig., In Fig.470 2. we preseut some examples of IVE low-resolution spectra of three F. € aud IK iain sequence stars.," \ref{fig.esp_b} we present some examples of IUE low-resolution spectra of three F, G and K main sequence stars."471 These spectra present a resolution of R=100 at 2700A.. they are available from the IUE public library (at /sdc.laeff.esa.es/cgi-ines/IUEdbsMY)). aud have been calibrated using the NEWSIPS (New Spectral nage Processing System) algorithin (Carhartetal. 1997)..," These spectra present a resolution of $R=400$ at 2700, they are available from the IUE public library (at ), and have been calibrated using the NEWSIPS (New Spectral Image Processing System) algorithm \citep{1997IUENN..57....1G}. ."472 The iuterual accuracy of the low-resolution flux calibration is (Massa&Fitzpatrick 2000).., The internal accuracy of the low-resolution flux calibration is \citep{2000ApJS..126..517M}. .473where D is the distance to the svstem.,where $D$ is the distance to the system.474 The quantities on the rhs of (his equation can all be measured. astrometicallv., The quantities on the rhs of this equation can all be measured astrometically.475 We will assume that M. the mass of the more huninous component. can be estimated photometrically or spectroscopicallv.," We will assume that $M$, the mass of the more luminous component, can be estimated photometrically or spectroscopically."476 And we will focus on the case in which the companion is known to be dark (or at least extremely. dim compared to the primary). /«L.," And we will focus on the case in which the companion is known to be dark (or at least extremely dim compared to the primary), $l\ll L$."477 Under (hese assumptions. it is straightforward to determine i. the mass of the dark companion. Irom the astrometric observations.," Under these assumptions, it is straightforward to determine $m$, the mass of the dark companion, from the astrometric observations."478 In general. a can be measured with approximately (he same precision as (he parallax. π.," In general, $\alpha$ can be measured with approximately the same precision as the parallax, $\pi$."479 OL course this does not hold exactly., Of course this does not hold exactly.480 Even for circular binary orbits. the inclination of the orbit will not match exactly the ecliptic latitude (1.e.. the inclination of the parallactic circle). so there will be either more or less information about the binary orbit than about (he reflex motion of (he Earth's orbit (parallax).," Even for circular binary orbits, the inclination of the orbit will not match exactly the ecliptic latitude (i.e., the inclination of the parallactic circle), so there will be either more or less information about the binary orbit than about the reflex motion of the Earth's orbit (parallax)."481 Moreover. for certain binary orbits. notably edee-on highly eccentric orbits that “point” in our direction. the errors in a will be much larger (han the parallax errors because (he binary will show almost no astrometric motion.," Moreover, for certain binary orbits, notably edge-on highly eccentric orbits that “point” in our direction, the errors in $\alpha$ will be much larger than the parallax errors because the binary will show almost no astrometric motion."482" Nevertheless. [rom the standpoint of making an estimate of the errors lor a random ensemble of binaries. setting σι,~8, is a good approximation."," Nevertheless, from the standpoint of making an estimate of the errors for a random ensemble of binaries, setting $\sigma_\alpha \sim \sigma_\pi$ is a good approximation."483" This is confirmed by Figure 1.. where we plot a,/σι lor astrometric binaries will orbital solutions (i.e.. binaries of (wpe Ὁ} in the llipparcos catalog (ESA1997.Vol.10).."," This is confirmed by Figure \ref{fig:one}, where we plot $\sigma_\alpha/\sigma_\pi$ for astrometric binaries with orbital solutions (i.e., binaries of type `O') in the Hipparcos catalog \citep[Vol.\ 10]{hip}."484 While these fits macle use of some auxiliary ground- spectroscopic information (uainly to establish the period). or constrained orbits to be circular. this should not have a major impact on the errors in a for periods P<3.3vi. the duration of the mission.," While these fits made use of some auxilliary ground-based spectroscopic information (mainly to establish the period), or constrained orbits to be circular, this should not have a major impact on the errors in $\alpha$ for periods $P\la 3.3\,\yr$, the duration of the mission."485 While the figure shows some scatter. the two errors are roughly equal o1 average.," While the figure shows some scatter, the two errors are roughly equal on average."486 Figure 2. shows (he sensilivily (50 detection) of Hipparcos to dark companions as a function of stellar (wpe. i.e.. the number ofLipparcos stars that can be probed lor companions ol a given mass.," Figure \ref{fig:two} shows the sensitivity $5\,\sigma$ detection) of Hipparcos to dark companions as a function of stellar type, i.e., the number of stars that can be probed for companions of a given mass."487 These (vpes were assigned based on position in the color-amagnitude diagram when the parallaxes were sufficiently accurate. and on position in (he reduced proper-notion diagram otherwise.," These types were assigned based on position in the color-magnitude diagram when the parallaxes were sufficiently accurate, and on position in the reduced proper-motion diagram otherwise."488 In the latter case. cistances were assigned based on stellar (ype and color and magnitude.," In the latter case, distances were assigned based on stellar type and color and magnitude."489" The figure shows that white dwarf (WD). NS. and DII companions of mass 0.6. 1.4 and 7 M... are respectively detectable among39%...G8Y%.. and of allHipparcos stars (Nii,— 118.000)."," The figure shows that white dwarf (WD), NS, and BH companions of mass 0.6, 1.4 and 7 $M_\odot$, are respectively detectable among, and of all stars $N_{Hip}=118,000$ )."490 For periods of P=1.5vr. these fractions fall to21%.," For periods of $P=1.5\,\yr$, these fractions fall to,."491..47%...52%... At P~dva.sensitivity is seriously compromised by parallax aliasing and at shorter periods the sensititv. falls off rapidly.," At $P\sim 1\,\yr$,sensitivity is seriously compromised by parallax aliasing and at shorter periods the sensitity falls off rapidly."492 On the other hand. for P23.8vr. orbital solutions become rapidly unstable.," On the other hand, for $P\ga 3.3\,\yr$, orbital solutions become rapidly unstable."493 Hence. the sensitivities peak fairly sharply at ?~3.3ντ.," Hence, the sensitivities peak fairly sharply at $P\sim 3.3\,\yr$."494 The overwhelming majority of theseMipparcos stars ave F ancl G cdwarfs. or giant stars whose progenilors are overwhelmingly F and G dwarls.," The overwhelming majority of these stars are F and G dwarfs, or giant stars whose progenitors are overwhelmingly F and G dwarfs."495 The Ireequency of companions per log period for P~3.3vr among such stars is dfy/dlogP~17 (Duquennov&Mavor 1991)).," The frequency of companions per log period for $P\sim 3.3\,\yr$ among such stars is $df_b/d\log P\sim 7\%$ \citealt{DM91}) )."496 From the previous paragraph. is sensitive to companions over about hall a dex," From the previous paragraph, is sensitive to companions over about half a dex"497We apply the same offset technique to metallicities as a function of mass.,We apply the same offset technique to metallicities as a function of mass.498 The total stellar mass-metallicity relation of unbarred galaxies is fit with a second order polynomial and A O/H is defined as the otfset between the metallicity of a barred galaxy and the value predicted for its total stellar mass: The O/H residuals Gin bins of stellar mass) of the unbarred control sample have median values around. zero and scatter ypically less than 0.02 dex., The total stellar mass-metallicity relation of unbarred galaxies is fit with a second order polynomial and $\Delta$ O/H is defined as the offset between the metallicity of a barred galaxy and the value predicted for its total stellar mass: The O/H residuals (in bins of stellar mass) of the unbarred control sample have median values around zero and scatter typically less than 0.02 dex.499 Note that. unlike the SPRs. no attempt is made to correct he abundance value to a total metallicity.," Note that, unlike the SFRs, no attempt is made to correct the abundance value to a total metallicity."500 The metallicity values used throughout this paper are fibre values., The metallicity values used throughout this paper are fibre values.501 Kewlev et al. (, Kewley et al. (5022005) ound that covering fractions of >20% should yield abundances hat are representative of integrated light spectra over the entire galaxy.,2005) found that covering fractions of $>$ should yield abundances that are representative of integrated light spectra over the entire galaxy.503 The majority of our galaxies have tibre covering fractions «20t.. so that imposing such a cut would leave a sample hampered by small number statistics.," The majority of our galaxies have fibre covering fractions $<$, so that imposing such a cut would leave a sample hampered by small number statistics."504 However. considering the fibre stellar mass-metallicity relation circumvents the issue of aperture bias. under the assumption that the unbarred galaxies have a similar distribution of radial coverage to the barred sample.," However, considering the fibre stellar mass-metallicity relation circumvents the issue of aperture bias, under the assumption that the unbarred galaxies have a similar distribution of radial coverage to the barred sample."505 The consistent distributions of mass and redshift between the barred and unbarrec galaxies indicate that this is a reasonable assumption., The consistent distributions of mass and redshift between the barred and unbarred galaxies indicate that this is a reasonable assumption.506 The upper panel of Figure 4. shows AO/H as a function of total stellar mass., The upper panel of Figure \ref{delta_oh} shows $\Delta \rm{O/H}$ as a function of total stellar mass.507" The metallicities of barred galaxies are higher than the unbarred sample by «0.06 dex when the total mass of the galaxy M, > 10! M..", The metallicities of barred galaxies are higher than the unbarred sample by $\sim$ 0.06 dex when the total mass of the galaxy $_{\star}$ $>$ $^{10}$ $_{\odot}$.508 This is a similar transition mass to the enhanced star formation rates in Figure 3.., This is a similar transition mass to the enhanced star formation rates in Figure \ref{delta_sfr}.509 The lower panel of the same figure shows the offset from the mass-metallicity relation., The lower panel of the same figure shows the offset from the mass-metallicity relation.510 The metallicities of barred galaxies are now higher by ~ 0.06 dex across the entire mass range., The metallicities of barred galaxies are now higher by $\sim$ 0.06 dex across the entire mass range.511 The appearance of enhanced metallicities at low masses when tibre quantities are considered can be understood by plotting the fraction of mass in the fibre as a function of total stellar mass (Figure 51)., The appearance of enhanced metallicities at low masses when fibre quantities are considered can be understood by plotting the fraction of mass in the fibre as a function of total stellar mass (Figure \ref{delta_mass}) ).512" Since barred galaxies with M, « 10! M. are dominated by late-types (Nair Abraham 2010b) whose bulges and bars are small compared to their disks. the fibre contains a smaller fraction of the total galaxy mass (see also Figure 1)."," Since barred galaxies with $_{\star}$ $<$ $^{10}$ $_{\odot}$ are dominated by late-types (Nair Abraham 2010b) whose bulges and bars are small compared to their disks, the fibre contains a smaller fraction of the total galaxy mass (see also Figure \ref{bar_images}) )."513" Above M, > 10! M.. there is a rapid transition to a population dominated by early-type spirals. whose bulge fraction is much higher."," Above $_{\star}$ $>$ $^{10}$ $_{\odot}$, there is a rapid transition to a population dominated by early-type spirals, whose bulge fraction is much higher."514 The low fraction of mass covered by the tibrefor low stellar mass galaxies results in a more pronounced ditference in the fibre mass-metallicity oftsets in the lower panel of Figure 4.., The low fraction of mass covered by the fibrefor low stellar mass galaxies results in a more pronounced difference in the fibre mass-metallicity offsets in the lower panel of Figure \ref{delta_oh}. .515We have searched for C» and CN emission bands.,We have searched for $_2$ and CN emission bands.516 Figure 8. presents the observed spectra obtained in the 3700-4000 rrange (CN band). and Fig.," Figure \ref{f:cn} presents the observed spectra obtained in the 3700-4000 range (CN band), and Fig."517 9. presents the 4800-5200 rregion (C+ bands) after subtraction of the solar spectrum., \ref{f:c2} presents the 4800-5200 region $_2$ bands) after subtraction of the solar spectrum.518 A theoretical spectrum of both CN and C» emission bands has been superimposed on these spectra., A theoretical spectrum of both CN and $_2$ emission bands has been superimposed on these spectra.519 We computed the CN spectrum by using the model described in ?.., We computed the CN spectrum by using the model described in \cite{zucconi:1985}.520" We computed the C spectrum with the model described in ? with transition moments |[D,-yF=[D107? aatomie unit.", We computed the $_2$ spectrum with the model described in \cite{rousselot:2000} with transition moments $|D_{a-X}|^2=|D_{c-X}|^2=3.5\times 10^{-6}$ atomic unit.521 Both spectra were computed for similar heliocentric distance and velocity km.s! ) and convolved with an instrument response function similar to that of FORS | in the mode used during our observations oof FWHM)., Both spectra were computed for similar heliocentric distance and velocity (-3.097 $^{-1}$ ) and convolved with an instrument response function similar to that of FORS 1 in the mode used during our observations of FWHM).522 As we've seen. no CN nor C» emission lines are apparent.," As we've seen, no CN nor $_2$ emission lines are apparent."523 It is only possible to derive an upper limit for both of these species., It is only possible to derive an upper limit for both of these species.524" In the case of CN the brightest possible CN emission band can be estimated to have an intensity equal to about 2.0x107"" 7.s eerg.cmA! (corresponding to a 4-sigma detection", In the case of CN the brightest possible CN emission band can be estimated to have an intensity equal to about $2.0\times 10^{-17}$ $^{-2}$ $^{-1}.$ $^{-1}$ (corresponding to a 4-sigma detection525 , 526fits the data well.,fits the data well.527 The temperature of the seed photons is again fixed at 0.3 keV. As for the global spectrum. we remark that the absorption column returned from the fit with this model is slightly higher than the value obtained with CPL.," The temperature of the seed photons is again fixed at 0.3 keV. As for the global spectrum, we remark that the absorption column returned from the fit with this model is slightly higher than the value obtained with CPL."528 We note a significant evolution of the absorption column density and of the power law photon index between the two intervals., We note a significant evolution of the absorption column density and of the power law photon index between the two intervals.529 In order to check whether the evolution of both was real. we re-performed the fits freezing Ny to its mean value (Table 6)).," In order to check whether the evolution of both was real, we re-performed the fits freezing $N_{\mathrm H}$ to its mean value (Table \ref{tab:rxtefit}) )."530 The spectral parameters obtained for both fits are compatible with those found leaving all parameters free to vary. except the power law photon index which tends to a softer value in interval 1 (EF.=2.11+ 0.05). and a to harder one for interval 2 (PV=1.38+ 0.03).," The spectral parameters obtained for both fits are compatible with those found leaving all parameters free to vary, except the power law photon index which tends to a softer value in interval 1 $\Gamma=2.11\pm 0.05$ ), and a to harder one for interval 2 $\Gamma=1.38\pm0.03$ )."531 Since Ny and E are tightly correlated. we also re-performed the fit freezing Γ to its mean value. and allowing Ny to vary.," Since $N_{\mathrm H}$ and $\Gamma$ are tightly correlated, we also re-performed the fit freezing $\Gamma$ to its mean value, and allowing $N_{\mathrm H}$ to vary."532 While for interval 2 the spectral parameters obtained in this case are close to the ones obtained when everything is free to vary. this method yields a poor fit for interval 1.," While for interval 2 the spectral parameters obtained in this case are close to the ones obtained when everything is free to vary, this method yields a poor fit for interval 1."533 We take these results as evidence that both F and Ny vary between both intervals., We take these results as evidence that both $\Gamma$ and $N_{\mathrm H}$ vary between both intervals.534 Note that this likely variation of the absorption is reinforced by the variations of Ny we observe between Obs., Note that this likely variation of the absorption is reinforced by the variations of $N_{\mathrm H}$ we observe between Obs.535 |. 2 and 3 (Table 6)).," 1, 2 and 3 (Table \ref{tab:rxtefit}) )."536 As mentioned previously in all the and spectra. an iron Ka fluorescence line i5 required in the spectral fits.," As mentioned previously in all the and spectra, an iron $\alpha$ fluorescence line is required in the spectral fits."537 The parameters of the line obtained from the spectral fit to each observation are reported in Table 8.., The parameters of the line obtained from the spectral fit to each observation are reported in Table \ref{tab:line}.538 Note that these are obtained from the fits with the phenomenological models. but no significant differences are found in the spectra where a model is used.," Note that these are obtained from the fits with the phenomenological models, but no significant differences are found in the spectra where a model is used."539 One could wonder whether the line is intrinsic to itself. or whether it could originate from the Galactic background.," One could wonder whether the line is intrinsic to itself, or whether it could originate from the Galactic background."540 The main argument that points towards an origin intrinsic to the system is that if the line was due to the Galactic ridge. we would expect its flux to be roughly constant.," The main argument that points towards an origin intrinsic to the system is that if the line was due to the Galactic ridge, we would expect its flux to be roughly constant."541 This is obviously not the case here., This is obviously not the case here.542" It is interesting to note that in almost all cases. (except in the “Bright” and ""Faint states). the parameters inferred for the line could be indicative of a narrow line. rather than a broad line."," It is interesting to note that in almost all cases, (except in the “Bright"" and ""Faint"" states), the parameters inferred for the line could be indicative of a narrow line, rather than a broad line."543 In fact for both instruments the upper limit on the line width indicates that we are limited by the instrumental spectral resolution., In fact for both instruments the upper limit on the line width indicates that we are limited by the instrumental spectral resolution.544 The case of the faint and bright states seem different since our fits indicate a broad line (Table 8))., The case of the faint and bright states seem different since our fits indicate a broad line (Table \ref{tab:line}) ).545 Our spectral fits to the data (Sec. 3.2.1)), Our spectral fits to the data (Sec. \ref{sec:integspec}) )546" 1dicate that the ""Bright"" state is spectrally intermediate between the “Faint” state and the ""Ultra-bright one. as we will discuss further below."," indicate that the “Bright” state is spectrally intermediate between the “Faint” state and the “Ultra-bright” one, as we will discuss further below."547 In particular in the soft X-rays (4-8 keV). a black body component could be present in the spectra of the “Bright” state. and represents the data well for the faint state.," In particular in the soft X-rays (4–8 keV), a black body component could be present in the spectra of the “Bright” state, and represents the data well for the faint state."548 In both cases. a fit to the data with a black body and a Gaussian (besides the power law) does not converge on sensitive. parameters for either of the components.," In both cases, a fit to the data with a black body and a Gaussian (besides the power law) does not converge on sensitive parameters for either of the components."549" The broad line we found instead could be indicative of a ""mixture"" of faint black body emission (poorly constrained given the 4 keV lower boundary of our fits) and a Gaussian line.", The broad line we found instead could be indicative of a “mixture” of faint black body emission (poorly constrained given the 4 keV lower boundary of our fits) and a Gaussian line.550 This possibility is compatible with the evolution between the three “states”. as clearly seen of Fig. 2..," This possibility is compatible with the evolution between the three “states”, as clearly seen of Fig. \ref{fig:integspec},"551" where black body emission dominates the soft X-ray in the “Faint state"" (when either no line is needed or a very broad one). to the ""Ultra Bright"" state. where no black body is detected. and with a good constraint on the We performed a thorough spectral analysis of the source using a well-sampled high energy monitoring with in 2003 March-May. and adding 3RXTE observations performed at different epochs."," where black body emission dominates the soft X-ray in the “Faint state” (when either no line is needed or a very broad one), to the “Ultra Bright” state, where no black body is detected, and with a good constraint on the We performed a thorough spectral analysis of the source using a well-sampled high energy monitoring with in 2003 March–May, and adding 3 observations performed at different epochs."552 As already observed (Paper 1). is highly variable on timescales from months down to hours. and it can show variations on shorter timescales as seen during observation 3 (Fig. 5)).," As already observed (Paper 1), is highly variable on timescales from months down to hours, and it can show variations on shorter timescales as seen during observation 3 (Fig. \ref{fig:PCAHXT}) )."553 This behaviour is reminiscent of Galactic X-ray binaries (XRB). and our deep analysis further confirms the Galactic nature of IGR 71914040951. already proposed in other publications (Paper|. Corbet et al.," This behaviour is reminiscent of Galactic X-ray binaries (XRB), and our deep analysis further confirms the Galactic nature of IGR J19140+0951, already proposed in other publications (Paper1, Corbet et al."554 When observed withRXTE. the source was dim. with a |- keV (unabsorbed) luminosity of ~3.4x 10°°x(D/10 Κρο)” erg/s (Obs.3). and a spectrum typical of Comptonisation of soft photons by a low temperature plasma (AT~5 keV) with a relatively high optical depth (7~ 5).," When observed with, the source was dim, with a 1-200 keV (unabsorbed) luminosity of $\sim 3.4 \times 10^{36}\times$ (D/10 $^2$ erg/s (Obs.3), and a spectrum typical of Comptonisation of soft photons by a low temperature plasma $kT\sim 5$ keV) with a relatively high optical depth $\tau \sim 5$ )."555" This could Correspone to the ""ultra faint state” which seems to be the state in which the source spends most of its time as indicated by our monitoring.", This could correspond to the “ultra faint state” which seems to be the state in which the source spends most of its time as indicated by our monitoring.556 During the observations. the luminosity is up to about 10 times higher. with a maximum of ~3.7x10 «D/IO kpey erg/s. Here significant spectral evolution is observed since in one case a," During the observations, the luminosity is up to about 10 times higher, with a maximum of $\sim 3.7 \times 10^{37}\times$ (D/10 $^2$ erg/s. Here significant spectral evolution is observed since in one case a"557Once loggg is set. Chere is a direct relation between the eravitw of the belt loggii and ils velocity Viu.,"Once $\log{g_{wd}}$ is set, there is a direct relation between the gravity of the belt $\log{g_{belt}}$ and its velocity $V_{belt}$."558" For the accretion belt plus WD composite models. we created a grid in WD temperature Tip from 16.000IX. to 35.000Ix. in steps of 1000Ilx. and in accretion belt temperature {νι from 25.000IX (to 55.000IN. in steps of 100019. However. the results are nol verv sensitive to the value of loggpa as long as logi4;<7.5. corresponding to Visinizz3.000—4.000 km !. where the lower limit corresponds to loggj.)=8.37 (a 0.8687. WD) and the upper limit corresponds to logguy=8.54 (a 0.96AM,, WD). ancl we have assumed 7=60 degrees."," For the accretion belt plus WD composite models, we created a grid in WD temperature $T_{eff}$ from 16,000K to 35,000K in steps of 1000K, and in accretion belt temperature $T_{belt}$ from 25,000K to 55,000K in steps of 1000K. However, the results are not very sensitive to the value of $\log{g_{belt}}$ as long as $\log{g_{belt}} < 7.5$, corresponding to $V_{belt} \sin{i} \approx 3,000-4,000$ km $^{-1}$, where the lower limit corresponds to $\log{g_{wd}}=8.37$ (a $0.86 M_{\odot}$ WD) and the upper limit corresponds to $\log{g_{wd}}=8.54$ (a $0.96 M_{wd}$ WD), and we have assumed $i=60$ degrees."559 In the case of the WD plus belt the distance d is given in a way similar to the WD only case with where we sel d=65 pe., In the case of the WD plus belt the distance $d$ is given in a way similar to the WD only case with where we set d=65 pc.560" The best-litting white dwarf plus accretion belt fit with fixed distance (d=65pc) and a free radius A; vielded 7,,;—23. 000K. Si abundance = 2.0 x solar. C abundance = 0.2 x solar. V,sini= 4O0knm J|. with a radius Ry=0.0087... corresponding (o a mass Moyo=0.96.U. or logg=8.54."," The best-fitting white dwarf plus accretion belt fit with fixed distance (d=65pc) and a free radius $R_{wd}$ yielded $T_{wd} = 23,000$ K, Si abundance = 2.0 $\times$ solar, C abundance = 0.2 $\times$ solar, $V_{rot}\sin{i}= 400$ km $^{-1}$ , with a radius $R_{wd}=0.0087R_{\odot}$, corresponding to a mass $M_{wd}=0.96 M_{\odot}$ or $\log{g}=8.54$."561 The belt temperature was {γι=48. 000Ix. with a velocity of Vigsin£24. 000km !.," The belt temperature was $T_{belt} = 48,000$ K, with a velocity of $V_{belt}\sin{i} \approx 4,000$ km $^{-1}$."562 In this model. the white dwarf contributed of the FUV flux and the accretion belt of the FUV flus. and the fractional area of the accretion belt was24%.," In this model, the white dwarf contributed of the FUV flux and the accretion belt of the FUV flux, and the fractional area of the accretion belt was."563.. The 42 value of this fit was 7.06., The $\chi^2_{\nu}$ value of this fit was 7.06.564 This best-fitting white dwarf plus accretion belt composite model is displaved in figure 5 aud in Table 2 (model 12)., This best-fitting white dwarf plus accretion belt composite model is displayed in figure 5 and in Table 2 (model 12).565 We find this composite fit to be clearly superior to the single temperature and accretion disk - only fits., We find this composite fit to be clearly superior to the single temperature and accretion disk - only fits.566 The remarkable lowering of the reduced 42 value for the accretion belt fit is additional confirmation of the findings by Sionetal.(1996.1997.2001). and Gansicke&Benermann(1996) that VW llis white dwarl has an inhomogeneous temperature distribution and that the most likely explanation is that of an accretion belt of higher temperatureat its equatorial latitudes.," The remarkable lowering of the reduced $\chi^2_{\nu}$ value for the accretion belt fit is additional confirmation of the findings by \citet{sio96,sio97,sio01} and \citet{gan96} that VW Hyi's white dwarf has an inhomogeneous temperature distribution and that the most likely explanation is that of an accretion belt of higher temperatureat its equatorial latitudes."567 On the other hand. the best-litting combination WD plus accretion belt model with a fixed WD radius and distance(model 13 in Table 2. M=0.86... ). had a 472 = 10.9.," On the other hand, the best-fitting combination WD plus accretion belt model with a fixed WD radius and distance(model 13 in Table 2, $M=0.86 M_{\odot}$ ), had a $\chi^2_{\nu}$ = 10.9."568 The stellar and belt parameters corresponding to (hiis best-fit are as follows., The stellar and belt parameters corresponding to this best-fit are as follows.569" The WD has an average surface temperature Typp = 22.000. V;,;sin= 400knm b with solar abundances."," The WD has an average surface temperature $T_{eff}$ = 22,000K, $V_{rot}\sin{i}= 400$ km $^{-1}$, with solar abundances."570" The accretion belt has the parameters 77,5; = 50.000. and τιsin;= 3.000km !."," The accretion belt has the parameters $T_{belt}$ = 50,000K, and $V_{belt} \sin{i}= 3,000$ km $^{-1}$."571 The bell area is of the WD surface and the belt contributes of the total flux while the WD contributes of flux., The belt area is of the WD surface and the belt contributes of the total flux while the WD contributes of flux.572 This composite two-lemperature fit is shown in figure 6., This composite two-temperature fit is shown in figure 6.573 It is interesting to note that irrespective of whether the radius (and consequently the mass) of the white dwarf was kept fixed or not in the models. (he best fit models with a lower AZ andwith parameters (Mu. d. /) consistent with the values assessed for the system," It is interesting to note that irrespective of whether the radius (and consequently the mass) of the white dwarf was kept fixed or not in the models, the best fit models with a lower $\chi^2_{\nu}$ andwith parameters $M_{wd}$ , $d$ , $i$ ) consistent with the values assessed for the system"574of the ages have been given in that paper: occasionally thev max be as large as a factor of 2 to 3: errors that large will uot detract frou our main conclusions.,of the ages have been given in that paper; occasionally they may be as large as a factor of 2 to 3; errors that large will not detract from our main conclusions.575 Our sample has been selected from the catalogue of stars within pc from the Sun by Woolleyctal.(1970): this catalogue is definitely incomplete and so iust be our saluple., Our sample has been selected from the catalogue of stars within pc from the Sun by \citet{wool:70}; this catalogue is definitely incomplete and so must be our sample.576 Even within the distance Bits eiven in Table l1 stars will exist that we could have inchided but did uot., Even within the distance limits given in Table \ref{tab:distlimits} stars will exist that we could have included but did not.577 This incompleteness does not. however. iutroduce a statistical bias: we have checked that for a given spectral type the distribution of the stellar distances is the same for stars with a disk as for stars without a disk: this is illustrated by the average distances in Table 7..," This incompleteness does not, however, introduce a statistical bias: we have checked that for a given spectral type the distribution of the stellar distances is the same for stars with a disk as for stars without a disk; this is illustrated by the average distances in Table \ref{tab:distances}."578 Fig., Fig.579 6 presets eraphically the fraction of the (visual) stellar light recmutted im the infrared bv the disk as a function of the stellar age., \ref{fig:tau-of-age} presents graphically the fraction of the (visual) stellar light reemitted in the infrared by the disk as a function of the stellar age.580 Similar diagrams based mainly on IRAS results. have been published before- sec. for example. Tollandetal.(1998).," Similar diagrams based mainly on IRAS results, have been published before- see, for example, \citet{holl:98}."581". À eeneral. continuous correlation appears: disks around PMS-stars (οιο, Herbie AeDoe) are more massive than disks around stars like 3 Pic and Vega. and the disk around the Sun is still less massive,"," A general, continuous correlation appears: disks around PMS-stars (e.g. Herbig AeBe) are more massive than disks around stars like $\beta$ Pic and Vega, and the disk around the Sun is still less massive."582 These earlier diagrams have almost no data on the age range shown in Fig., These earlier diagrams have almost no data on the age range shown in Fig.583 6 aud the new ISO data fill in an important hole., \ref{fig:tau-of-age} and the new ISO data fill in an important hole.584" Table & suniuanizes the detections at GO. separatelv for sars of different agen and of ciffereut spectral type together with the same nunibers for stars with a disk: iu the column marked “tot” the total umber of stars (disks plus no-disks) is shown and uuder the heading ""disk the iuuber of stars with a disk.", Table \ref{tab:detstat} summarizes the detections at 60 separately for stars of different age and of different spectral type together with the same numbers for stars with a disk; in the column marked “tot” the total number of stars (disks plus no-disks) is shown and under the heading “disk” the number of stars with a disk.585 The total count is Sl mscad of 81 because for three of our target stars (two A-stars iud one IK-star) the age could not be estimated ina satisfying manner., The total count is 81 instead of 84 because for three of our target stars (two A-stars and one K-star) the age could not be estimated in a satisfying manner.586 Table & shows that the stars with a etected disk are systematically vounger than the stars without disk: out of the 15 stars vouuger than LOO Ayr uine (6054)) have a disk: out of the 66 older stars ouly five have a disk (1)., Table \ref{tab:detstat} shows that the stars with a detected disk are systematically younger than the stars without disk: out of the 15 stars younger than 400 Myr nine ) have a disk; out of the 66 older stars only five have a disk ).587 Eurthermore. there exists a more or less sharply defined age aove Which a star has no longer a disk.," Furthermore, there exists a more or less sharply defined age above which a star has no longer a disk."588 This is best demonstrated by the A-stars., This is best demonstrated by the A-stars.589 Six A-stars have a disk: the stellar ages are 220. 2LO. 280. 350. 360. 380 Myr.," Six A-stars have a disk; the stellar ages are 220, 240, 280, 350, 360, 380 Myr."590 For the A-stars without disk the corresponding ages are 300. 320. 350. 380. 120. 150. 5LO. 890. 1230 My: 350 to LOO Ner is a well-defined transition region.," For the A-stars without disk the corresponding ages are 300, 320, 350, 380, 420, 480, 540, 890, 1230 Myr: 350 to 400 Myr is a well-defined transition region."591 We couclude that the A stars in genera arive on the Dualn-sequeuce with a disk. but that hey loose the disk within 50 Myr when they are about 350.3vr old.," We conclude that the A stars in general arrive on the main-sequence with a disk, but that they loose the disk within 50 Myr when they are about 350 Myr old."592 Is what is true for he A-stars also valid or the stars of other spectral types?, Is what is true for the A-stars also valid for the stars of other spectral types?593" Our answer is ""probably ves”: of the five EF. C. and Is stars vounecr than LOO Ny. three (GO%)) have a disk."," Our answer is “probably yes”: of the five F, G, and K stars younger than 400 Myr three ) have a disk."594 Of the GI. EF. €. aud K stars older than LOO Myr five iive a disk (one in twelve or )).," Of the 61 F, G, and K stars older than 400 Myr five have a disk (one in twelve or )."595 The percentages are he same as for the A-stars out the for voung Cc and Is-sars Is based on only three detections., The percentages are the same as for the A-stars but the for young G- and K-stars is based on only three detections.596 Tt secs that the disks around EF. €. aud I& stars decay in a sinularly short time after arrival on the main sequence.," It seems that the disks around F, G, and K stars decay in a similarly short time after arrival on the main sequence."597 An inuuediate question is: do stars arrive at the nai sequence with a disk?, An immediate question is: do stars arrive at the main sequence with a disk?598 Studies of EC-MWUL-SCquchce stars show that disks are common. mt whether they always exist is unknown.," Studies of pre-main-sequence stars show that disks are common, but whether they always exist is unknown."599 The sequence| of ages of the A-stars shows that thie three voungest A-stars have a disk., The sequence of ages of the A-stars shows that the three youngest A-stars have a disk.600 This sueeestsOO that all stars arrive on he main secpuede? with a disk. but the sugecstionOO is based on simall-uimuber," This suggests that all stars arrive on the main sequence with a disk, but the suggestion is based on small-number"601This is the same form as used in Yang et al. (,This is the same form as used in Yang et al. (602"2008), with scaling parameters: logM;= 9.8, logΜι,=10.7, a=0.6 and 8=2.9.","2008), with scaling parameters: $log~M_s = 9.8$ , $log~M_h = 10.7$ , $\alpha = 0.6$ and $\beta = 2.9$."603 Halo masses for the spiral galaxies in our sample were estimated using the spiral galaxy relation of Mandelbaum et al. (, Halo masses for the spiral galaxies in our sample were estimated using the spiral galaxy relation of Mandelbaum et al. (604"2006), also shown in Fig. 1..","2006), also shown in Fig. \ref{fig1}."605 Note the data in Fig., Note the data in Fig.606" 2 show a similar form to the relationship adopted between galaxy stellar and halo mass (Fig. 1)),"," \ref{fig2} show a similar form to the relationship adopted between galaxy stellar and halo mass (Fig. \ref{fig1}) ),"607 with only an offset in the X-axis values., with only an offset in the X-axis values.608 Galaxy stellar masses are taken from Spitler et al. (, Galaxy stellar masses are taken from Spitler et al. (6092008) and Peng et al. (,2008) and Peng et al. (6102008).,2008).611 For the Spitler et al. (, For the Spitler et al. (612"2008) estimates, the Chabrier (2003) initial stellar mass function is used to match the Mandelbaum et al. (","2008) estimates, the Chabrier (2003) initial stellar mass function is used to match the Mandelbaum et al. ("6132006) relation.,2006) relation.614 There is a small (0.07 dex in logMstetiar) systematic offset between the masses derived using the Spitler et al. (, There is a small (0.07 dex in $log~M_{stellar}$ ) systematic offset between the masses derived using the Spitler et al. (6152008) technique and those published in Peng et al. (,2008) technique and those published in Peng et al. (6162008).,2008).617 This offset is removed from the Peng et al. (, This offset is removed from the Peng et al. (6182008) masses before analysis.,2008) masses before analysis.619 The GC system numbers in Spitler et al. (, The GC system numbers in Spitler et al. (6202008) were converted to GC system total masses by multiplying the numbers by the average GC mass of 4x10° Mo.,2008) were converted to GC system total masses by multiplying the numbers by the average GC mass of $4\times10^5$ $_{\odot}$.621 Peng et al. (, Peng et al. (6222008) summed the total stellar mass of all GCs in each galaxy.,2008) summed the total stellar mass of all GCs in each galaxy.623" For galaxies in common, the Spitler et al. ("," For galaxies in common, the Spitler et al. ("6242008) GC masses were used because they come from wide-field imaging where the entire spatial coverage of the GC system was observed.,2008) GC masses were used because they come from wide-field imaging where the entire spatial coverage of the GC system was observed.625 The NGC 3311 GC system number estimate is from Wehner et al. (, The NGC 3311 GC system number estimate is from Wehner et al. (6262008).,2008).627" For reference, a table is available online with relevant properties of the main sample."," For reference, a table is available online with relevant properties of the main sample."628" The GC system mass of Local Group (LG) dwarf galaxies is estimated by summing the individual GC stellar masses inferred from V-band photometry (Harris 1996; Webbink 1985; Da Costa Mould 1988) and applying a mass-to-light ratio of 2.2 for an old, metal-poor stellar population (Bruzual Charlot 2003)."," The GC system mass of Local Group (LG) dwarf galaxies is estimated by summing the individual GC stellar masses inferred from V-band photometry (Harris 1996; Webbink 1985; Da Costa Mould 1988) and applying a mass-to-light ratio of 2.2 for an old, metal-poor stellar population (Bruzual Charlot 2003)."629 LG dwarf galaxy stellar masses are from V-band absolute magnitudes (Lotz et al., LG dwarf galaxy stellar masses are from V-band absolute magnitudes (Lotz et al.630 2004) with appropriate mass-to-light ratios from Bruzual Charlot (2003) for the age and metallicity of the stellar populations (Lotz et al., 2004) with appropriate mass-to-light ratios from Bruzual Charlot (2003) for the age and metallicity of the stellar populations (Lotz et al.631 2004)., 2004).632" Total masses (Miotalος06, where σο is the central velocity dispersion) and distances to these galaxies are from Mateo (1998)."," Total masses $M_{total}\propto\sigma_0^2$, where $\sigma_0$ is the central velocity dispersion) and distances to these galaxies are from Mateo (1998)."633 The analysis includes five galaxy clusters selected because their central galaxy has a reliable GC system number measurement available., The analysis includes five galaxy clusters selected because their central galaxy has a reliable GC system number measurement available.634" GCs associated with galaxy clusters will reside in the central cluster galaxy, around satellite galaxies and in the intracluster medium (see refresults))."," GCs associated with galaxy clusters will reside in the central cluster galaxy, around satellite galaxies and in the intracluster medium (see \\ref{results}) )."635" The total mass of GCs associated with satellite galaxies was approximated by integrating the observed cluster galaxy mass functions (Sandage, Bingegeli, Tammann et al."," The total mass of GCs associated with satellite galaxies was approximated by integrating the observed cluster galaxy mass functions (Sandage, Bingegeli, Tammann et al."636 1985; Ferguson Sandage 1991; Yagi et al., 1985; Ferguson Sandage 1991; Yagi et al.637" 2002; Trentham, Tully Mahdavi 2006) after convolving them with a quadratic fit to data in Fig. 2.."," 2002; Trentham, Tully Mahdavi 2006) after convolving them with a quadratic fit to data in Fig. \ref{fig2}."638 No global constraint on an intracluster GC population exists., No global constraint on an intracluster GC population exists.639" We therefore use a prediction from computer simulations of galaxy clusters (Bekki Yahagi 2006) that intracluster GCs make up (with RMS = 5%)) of the total cluster GC mass, independent of the clusters total mass."," We therefore use a prediction from computer simulations of galaxy clusters (Bekki Yahagi 2006) that intracluster GCs make up (with RMS = ) of the total cluster GC mass, independent of the clusters total mass."640" Because the study of Bekki Yahagi (2006) was limited to a rudimentary GC formation prescription, formal uncertainties on these total cluster GC masses are taken to be40%."," Because the study of Bekki Yahagi (2006) was limited to a rudimentary GC formation prescription, formal uncertainties on these total cluster GC masses are taken to be."641. Cluster halo masses are taken from the following sources: Virgo and Hydra clusters (Girardi et al., Cluster halo masses are taken from the following sources: Virgo and Hydra clusters (Girardi et al.642" 1998), NGC 1407 (Brough et al."," 1998), NGC 1407 (Brough et al."643" 2006), Antlia (Nakazawa et al."," 2006), Antlia (Nakazawa et al."644" 2000), and Fornax (Drinkwater, Gregg Colless 2001)."," 2000), and Fornax (Drinkwater, Gregg Colless 2001)."645 In Fig., In Fig.646" ὃ we show that GC system masses (Macs) are directly proportional to the total halo mass of its host galaxy, with a scatter comparable to the observational uncertainties."," \ref{fig3} we show that GC system masses $M_{GCS}$ ) are directly proportional to the total halo mass of its host galaxy, with a scatter comparable to the observational uncertainties."647 The form of the line in Fig., The form of the line in Fig.648 3 is logMnato=logMacs+ 4.15., \ref{fig3} is $log~M_{halo} = log~M_{GCS}+4.15$ .649" This can be related to the initial total baryon mass of a galaxy, by assuming the universal baryon fraction (i.e. Miaryon/Mnato£2Ώυ/Ώπι&0.17; Komatsu et al."," This can be related to the initial total baryon mass of a galaxy, by assuming the universal baryon fraction (i.e. $M_{baryon}/M_{halo}\approx\Omega_{b}/\Omega_{m}\approx0.17$; Komatsu et al."650 2008) applies on all galactic scales in the early Universe., 2008) applies on all galactic scales in the early Universe.651" For galaxies with Mnato>5x10!!Mo, the relationship in Fig."," For galaxies with $M_{halo}>5\times10^{11} M_{\odot}$, the relationship in Fig."652 3 appears to be invariant to the local environment and to the morphological type of the galaxy (see discussion in Spitler et al., \ref{fig3} appears to be invariant to the local environment and to the morphological type of the galaxy (see discussion in Spitler et al.653 2008)., 2008).654" In contrast, the statistical relationships between stellar and halo mass depend on whether the galaxy is a spiral or an elliptical type (see Fig. 1))."," In contrast, the statistical relationships between stellar and halo mass depend on whether the galaxy is a spiral or an elliptical type (see Fig. \ref{fig1}) )."655" Furthermore, the statistical relationship between galaxy stellar mass and the halo mass is strongly non-linear."," Furthermore, the statistical relationship between galaxy stellar mass and the halo mass is strongly non-linear."656 'This means that for very massive galaxies their halo masses derived from the stellar mass relation are poorly constrained., This means that for very massive galaxies their halo masses derived from the stellar mass relation are poorly constrained.657 'The direct proportionalitybetween GCsystem masses and their host halo is consistent with our current understanding of GC formation., The direct proportionalitybetween GCsystem masses and their host halo is consistent with our current understanding of GC formation.658 While stars in a galaxy, While stars in a galaxy659the integral of f between the maximum aud minimum value of μις,the integral of $f$ between the maximum and minimum value of $N_{HI}$.660" We normalize the theoretical distribution such as to give a number of detectious with Nyy,>1.6κ104 ? equal to the observed one.", We normalize the theoretical distribution such as to give a number of detections with $N_{HI}\ge 1.6\times 10^{17}$ $^{-2}$ equal to the observed one.661 Two maxima for the Likelihood are fouud: ΤΙe >68.356. > . aud >99% coulidence levels in the X—a plane are shown in Figure | where the filled dots indicates the location of the maxima as in eq. (15))," Two maxima for the Likelihood are found: The $>68.3\%$, $>95.5\%$ , and $>99\%$ confidence levels in the $X-\alpha$ plane are shown in Figure 4 where the filled dots indicates the location of the maxima as in eq. \ref{bestfit1}) )"662 and (16))., and \ref{bestfit2}) ).663 The self gravitaing gas solution (4j=0. a= 1.63. X= 3.33) lies wel outside the >99% conticence leve (we woild need the 99.999% confidence level to include it) aixd therefore it is not consistent witl the data.," The self gravitating gas solution $\eta=0$, $\alpha=4.63$ , $X=3.33$ ) lies well outside the $>99\%$ confidence level (we would need the $\%$ confidence level to include it) and therefore it is not consistent with the data."664 We have also checked that a similar concusion hoks if we use |ve exact self gravitating eas soltou. as giveu by eq. (7)).," We have also checked that a similar conclusion holds if we use the exact self gravitating gas solution, as given by eq. \ref{sigma_sg}) ),"665 in deriving the JVμιNu !elation., in deriving the $N_{HI\perp}-N_{H\perp}$ relation.666 For this case both N au the best fit a value are withi 1% of the values obalned usie eq. (5)), For this case both $X$ and the best fit $\alpha$ value are within $1\%$ of the values obtained using eq. \ref{density}) )667 aicy =0 for the vertica gas stratification., and $\eta=0$ for the vertical gas stratification.668 In Figure 5 we compare te observed valie of te cunmulative funetion with the theoretical ones derived from the integral of te projected HI coluuii deusity: we show resuls for the two best fit inodels (the two maxima in FEigure [) ard fo ‘two models cor'esponcding t«) the highest and lowest X values on the >95.5% conidence level of Figure {., In Figure 5 we compare the observed value of the cumulative function with the theoretical ones derived from the integral of the projected HI column density; we show results for the two best fit models (the two maxima in Figure 4) and for two models corresponding to the highest and lowest X values on the $>95.5\%$ confidence level of Figure 4.669 For points which lave Lo defined μι. Le. which have large errors. we tleni coiupiute treir best clistribttiou [or a give vf N spreadiug the data in the alowed range of Ij accoring o f weighted wih he redslit yatha.," For points which have no defined $N_{HI}$ , i.e. which have large errors, we then compute their best distribution for a given $f$ by spreading the data in the allowed range of $N_{HI}$ according to $f$ weighted with the redshift path."670 Between all the possible periiiatious of poins with undetermined Nyy; we t choose hose wlich satisfies best the UÜ-tes on the deviatious between tlie observed aud the ex)ecLed eundative functions over the error interval. /2. and over Nyy. C (see Paper ΠΠ lor more detals and for the use of a numerical siiulatio1 to proof the validity of this approac1).," Between all the possible permutations of points with undetermined $N_{HI}$ we then choose those which satisfies best the $U$ -test on the deviations between the observed and the expected cumulative functions over the error interval, $R$, and over $N_{HI}$, $C$ (see Paper II for more details and for the use of a numerical simulation to proof the validity of this approach)."671 For X—« viide the 99% confidence level of Figure 1 he Ix-8 tests on f aud C are satisfied o the 99.9% level., For $X-\alpha$ inside the $99\%$ confidence level of Figure 4 the K-S tests on $R$ and $C$ are satisfied to the $\%$ level.672 We |ave proved that tle gas clistribution beween the LLS axl the DLS region follows a single power law with index a>2 if the ionization level is such that less than 1% of the total gas is neutral when Nyy=1.6x10! > 7., We have proved that the gas distribution between the LLS and the DLS region follows a single power law with index $\alpha > 2$ if the ionization level is such that less than $1\%$ of the total gas is neutral when $N_{HI}=1.6 \times 10^{17}$ $^{-2}$ .673 There is 1o need of a dist{bution more complicated than a power law once one takes iito account lonizatiol effects., There is no need of a distribution more complicated than a power law once one takes into account ionization effects.674 Our results on a and X still hold even if we do not iuclude in the daa set the clampect lires with 5€W<104A oor if we exclude a certain percentage of these liies due to possible bleudiug withsinaller lines., Our results on $\alpha$ and $X$ still hold even if we do not include in the data set the damped lines with $5\le W\le 10$ or if we exclude a certain percentage of these lines due to possible blending withsmaller lines.675 We shall discussbere some results relative to the best-fit values ofWY and a as given in eq. (15)), We shall discusshere some results relative to the best-fit values of$X$ and $\alpha$ as given in eq. \ref{bestfit1}) )676 aud in eq. (16)).," and in eq. \ref{bestfit2}) ),"677 and to the twomost extreme values of XN ou the >95.5% confidence level.," and to the twomost extreme values of $X$ on the $>95.5\%$ confidence level,"678is constrained form the observed level of [Ti/Fe] among halo stars (seeKobayashietal.2006).,is constrained form the observed level of [Ti/Fe] among halo stars \citep[see][]{Kobayashi_06}.679". For [O/Fe] and the observed properties, i.e., a large scatter at a [Mg/Fe],low-metallicity and an overall decreasing trend, are well reproduced."," For [O/Fe] and [Mg/Fe], the observed properties, i.e., a large scatter at a low-metallicity and an overall decreasing trend, are well reproduced."680" For [Ti/Fe], the observed data around [Ti/Fe] ~+0.3 is lacking."," For [Ti/Fe], the observed data around [Ti/Fe] $\sim$ +0.3 is lacking."681" However, medium-resolution observations suggest the presence of stars exhibiting such a high ratio (Kirbyetal.2010)."," However, medium-resolution observations suggest the presence of stars exhibiting such a high ratio \citep{Kirby_10}."682". Note that an unusual upward Ti/Fe feature for [Fe/H] -1 is reproduced by the combination of low and high Ti/Fe yields in SNe II and SNe Ia, respectively."," Note that an unusual upward Ti/Fe feature for [Fe/H] -1 is reproduced by the combination of low and high Ti/Fe yields in SNe II and SNe Ia, respectively."683" Since Ca is also synthesized highly in SNe Ia, our model predicts a similar upward feature for Ca/Fe."," Since Ca is also synthesized highly in SNe Ia, our model predicts a similar upward feature for Ca/Fe."684 This is indeed observed in the Fnx dSph (Letarteetal.2010)., This is indeed observed in the Fnx dSph \citep{Letarte_10}.685". Our model anticipates that the observed dispersion in both and at a low-metallicity is an end result of [Ba/Fe]the IMF [Mg/Fe]variation which yields a high [Ba/Fe] and a low [Mg/Fe] by m,=25 and a reverse correlation by m,=50 Mo..", Our model anticipates that the observed dispersion in both [Ba/Fe] and [Mg/Fe] at a low-metallicity is an end result of the IMF variation which yields a high [Ba/Fe] and a low [Mg/Fe] by $m_u$ =25 and a reverse correlation by $m_u$ =50.686" Figure 6 demonstrates that the predicted correlation of [Ba/Fe] and [Mg/Fe] is broadly compatible with the observed one given by the data of the Fnx GC, though this analysis should be validated by much more data."," Figure 6 demonstrates that the predicted correlation of [Ba/Fe] and [Mg/Fe] is broadly compatible with the observed one given by the data of the Fnx GC, though this analysis should be validated by much more data."687" Observed unusual Ba enhancement relative to Fe, a- elements, and Eu in some dSphs as well as in the LMC at their late evolution is investigated by modeling the Fornax dSph galaxy case."," Observed unusual Ba enhancement relative to Fe, $\alpha$ -elements, and Eu in some dSphs as well as in the LMC at their late evolution is investigated by modeling the Fornax dSph galaxy case."688" Our claim is that its effect occurs because the IMF's truncate a high mass end at around 25Mo,, that causes the reduction of a-elements and Fe but no (little) influence on r- or s-process elements in their ejection, associated with the death of stars covering a wide range of masses."," Our claim is that its effect occurs because the IMFs truncate a high mass end at around 25, that causes the reduction of $\alpha$ -elements and Fe but no (little) influence on $r$ - or $s$ -process elements in their ejection, associated with the death of stars covering a wide range of masses."689 Such a truncated IMF is assured by the theoretical agument given by the IGIMF scheme in which the number of massive stars depends on the SFR in galaxies., Such a truncated IMF is assured by the theoretical agument given by the IGIMF scheme in which the number of massive stars depends on the SFR in galaxies.690" In addition, the star formation history of the Fnx dSph revealed by recent surveys, together with a large dispersion in elemental ratios such as [a/Fe] and suggests a rather complex chemical history not [Ba/Fe],represented by a single model but composed of a few evolutionary paths resulting from different speed of star formation as well as from a differenta form of the IMF."," In addition, the star formation history of the Fnx dSph revealed by recent surveys, together with a large dispersion in elemental ratios such as $\alpha$ /Fe] and [Ba/Fe], suggests a rather complex chemical history not represented by a single model but composed of a few evolutionary paths resulting from a different speed of star formation as well as from a different form of the IMF."691" Previous work on the chemical evolution of dSphs interpret the lower [a/Fe] ratios in these galaxies as being due to a combination of the time-delay model and a low SFR (e.g.,Carigietal.2002;Ikuta&ArimotoLanfranchi&Matteucci2003;Kirbyetal. 2011)."," Previous work on the chemical evolution of dSphs interpret the lower $\alpha$ /Fe] ratios in these galaxies as being due to a combination of the time-delay model and a low SFR \citep[e.g.,][]{Carigi_02, Ikuta_02, Lanfranchi_03, Kirby_11}."692". This interpretation can explain the observed [Ba/Eu] ratios in dSphs with the inclusion of the galactic wind effect (Lanfranchietal.2008);; however, it does not favor the same increasing [Ba/Fe] trend as [Ba/Eu] (see,however,tionongalacticwindfortheSculptor dSph).."," This interpretation can explain the observed [Ba/Eu] ratios in dSphs with the inclusion of the galactic wind effect \citep{Lanfranchi_08}; however, it does not favor the same increasing [Ba/Fe] trend as [Ba/Eu] \citep[see, however,][as the model with a different assumption on galactic wind for the Sculptor dSph]{Fenner_06}."693" Further, it confronts the fact that a massive dwarf galaxy, i.e., the LMC, exhibits the same level of increase in [Ba/Fe] as"," Further, it confronts the fact that a massive dwarf galaxy, i.e., the LMC, exhibits the same level of increase in [Ba/Fe] as"69411330).,1334).695 Finally. the GIS30AT aud CAGOAL spectra were co-aligned aud. co-added using lines of the same species .11260 vs .11526).," Finally, the G130M and G160M spectra were co-aligned and co-added using lines of the same species 1260 vs 1526)."696 Co-addition was performed by a snuple stim of counts. after aliguiment.," Co-addition was performed by a simple sum of counts, after alignment."697 For cach pixel. we tracked the wavelength. umber of counts upper aud lower error estimates due to Poisson statistics. and the effective exposure time.," For each pixel, we tracked the wavelength, number of counts, upper and lower error estimates due to Poisson statistics, and the effective exposure time."698 We applied a flat-field correction by takine the STScI COS team 1d flats (D. Alassa. priv.," We applied a flat-field correction by taking the STScI COS team 1d flats (D. Massa, priv."699 col).," comm.),"700 processed with a low-pass filter. to remove the high-frequency noise.," processed with a low-pass filter, to remove the high-frequency noise."701 These flats account for the erid-wire pattern in the COS FUV detectors. aud are applied to the effective exposure tine for cach pixel (note that they are not applied to the counts directly. since we use the counts to caleulated the Poisson errors).," These flats account for the grid-wire pattern in the COS FUV detectors, and are applied to the effective exposure time for each pixel (note that they are not applied to the counts directly, since we use the counts to calculated the Poisson errors)."702 Since we acciunulate counts. the spectra contain sliarp discontinuities in regions where the waveleneth οποιο resulted imm a larger effective exposure time. or iu regions affected by the eridewire pattern.," Since we accumulate counts, the spectra contain sharp discontinuities in regions where the wavelength dithering resulted in a larger effective exposure time, or in regions affected by the grid-wire pattern."703 Thus our counts spectra do not have a sinooth coutimmun but the countrate spectrin does., Thus our counts spectra do not have a smooth continuum but the count spectrum does.704 Finally. we normalized the count-rate spectrum by piecewise fitting of low-order Chebyshev polvuomials to chunks of spectrum ~50LOOA lone.," Finally, we normalized the count-rate spectrum by piecewise fitting of low-order Chebyshev polynomials to chunks of spectrum $\sim 50-100\A$ long."705 Our procedure vields a normalized spectrum with both upper aud lower error estimates., Our procedure yields a normalized spectrum with both upper and lower error estimates.706 In regions of hieh counts (2 30). these estimates couveree to the usual eaussian approximation of VN. with the upper aud lower estimates the same (1.6. converging to the usual lo errors).," In regions of high counts $\simgt 30$ ), these estimates converge to the usual gaussian approximation of $\sqrt{N}$, with the upper and lower estimates the same (i.e. converging to the usual $\sigma$ errors)."707 To be couservative. when quoting errors on nieasured quantities; we quote the larger of these two error estimates.," To be conservative, when quoting errors on measured quantities, we quote the larger of these two error estimates."708 The S/N of the resulting spectra varies from ~| per resolution at the shortest waveleugths (~ A). peaking at ~8 per rescl at ~LLOOA. and decliniug to ~203 per resel at the red euc of the spectrmm (1500 A).," The S/N of the resulting spectra varies from $\sim 4$ per resolution at the shortest wavelengths $\sim 1135\A$ ), peaking at $\sim 8$ per resel at $\sim 1400\A$, and declining to $\sim 2-3$ per resel at the red end of the spectrum $\sim 1800\A$ )."709 By choosing multiple appropriate central waveleneth settings for cach erating. we ensured continuous waveleneth coverage (ie. no gaps) at the expense of lower S/N in the waveleneth reeious with only one erating setting.," By choosing multiple appropriate central wavelength settings for each grating, we ensured continuous wavelength coverage (i.e. no gaps), at the expense of lower S/N in the wavelength regions with only one grating setting."710" We obtained a high resolution optical spectrumof oon 26 March 2010 using the Tieh Resolution. Echelle Spectrometer (IIIRES:?). on Weck 1. We employed the UV cross-disperser aud the C1 decker (slit width 0.86"")). resultiug in a resolution of ~L800 or —Glins|."," We obtained a high resolution optical spectrumof on 26 March 2010 using the High Resolution Echelle Spectrometer \citep[HIRES;][]{vogt-etal-94-HIRES} on Keck I. We employed the UV cross-disperser and the C1 decker (slit width ), resulting in a resolution of $\sim71148000$ or $\sim 6\kms$."712 The erating angles were set to give coverage down to 3050À. in order to detect low-: aabsorptiou.," The grating angles were set to give coverage down to $3050\A$, in order to detect $z$ absorption."713 The data were reduced. with the ITIBRedux pipeline included with the NIDLpackage., The data were reduced with the HIRedux pipeline included with the XIDL.714.. Tudividual orders were normalized after extraction with a series of Chebyshev polvnouuals to remove the echelle blaze function. and then combined iuto a final 1D spectrin.," Individual orders were normalized after extraction with a series of Chebyshev polynomials to remove the echelle blaze function, and then combined into a final 1D spectrum."715 After the (ος obscrvations were conducted we obtained spectra of several galaxies in the field close to the QSO with the Low Resolution Imaging Spectrometer (LRIS:?) ou the Keck I telescope., After the COS observations were conducted we obtained spectra of several galaxies in the field close to the QSO with the Low Resolution Imaging Spectrometer \citep[LRIS;][]{oke-etal-95-LRIS} on the Keck I telescope.716 We obtained long-slit spectra of 1 galaxies on 25 March 2010., We obtained long-slit spectra of 4 galaxies on 25 March 2010.717 We eiiploved the delichroic.A with the 600/7500 erating on the red side. aud the 600/L000 eria on the blue side.," We employed the dichroic, with the 600/7500 grating on the red side, and the 600/4000 grism on the blue side."718 The slit size was1.. yielding a FWIIM resolution of L7A (~200]aus.+) over a wavelength range of 5600200A on the red side: the blue side covered the range 00605500A at a resolution of 3.9LIA FWOAL (~300k1misly," The slit size was, yielding a FWHM resolution of $4.7\A$ $\sim 200\kms$ ) over a wavelength range of $5600 - 8200\A$ on the red side; the blue side covered the range $3000 - 5500\A$ at a resolution of $3.9-4.1\A$ FWHM $\sim 300\kms$ )."719 The data were reduced using the LRIS pipeline iu the NIDL package., The data were reduced using the LRIS pipeline in the XIDL package.720 Fig 1l shows an nuage of the field taken from the SDSS. centered ou the QSO position.," Fig \ref{fig: keck_field} shows an image of the field taken from the SDSS, centered on the QSO position."721 Objects classified as galaxies iu SDSS are labeled in a polar co-ordinate svstena. with a position anuele (degrees east of north) and an aneular distance from the QSO (iu aresec).," Objects classified as galaxies in SDSS are labeled in a polar co-ordinate system, with a position angle (degrees east of north) and an angular distance from the QSO (in arcsec)."722 Galaxies for which we obtained spectroscopic redshifts are euclosed in red boxes. aud the measured redshifts iucluded. in the label," Galaxies for which we obtained spectroscopic redshifts are enclosed in red boxes, and the measured redshifts included in the label."723 Objects classified as galaxies in SDSS but which lack spectroscopic redshifts are enclosed in erecu diunonds., Objects classified as galaxies in SDSS but which lack spectroscopic redshifts are enclosed in green diamonds.724 The maxim range of the SDSS photometric redshift estimates are listed., The maximum range of the SDSS photometric redshift estimates are listed.725 We have also examined the photometric redshift probability distributions for these objects produced by. 7.., We have also examined the photometric redshift probability distributions for these objects produced by \citet{cunha-etal-09-SDSS-photoz}.726 None of these objects are cousistent with being at the redshift of the absorbers. although we note that contamination of objects very close to the QSO by the QSO light may be an issue. and it would be ideal to obtain further spectroscopy follow-up for these galaxies.," None of these objects are consistent with being at the redshift of the absorbers, although we note that contamination of objects very close to the QSO by the QSO light may be an issue, and it would be ideal to obtain further spectroscopy follow-up for these galaxies."727 Uulabele objects in Figure 1— are classified as stars., Unlabeled objects in Figure \ref{fig: keck_field} are classified as stars.728 Higher resolution imaeimug with ce. TST. is au ongoing conrponeut of our multiiustrüment project. aud woul be helpful in the future for quantitative micasures of ealaxy morphology.," Higher resolution imaging with e.g. HST, is an ongoing component of our multi-instrument project, and would be helpful in the future for quantitative measures of galaxy morphology."729 Iu Figure 2. we show the absorption lines affiliated with the ealaxy227_19., In Figure \ref{fig: stack-plot} we show the absorption lines affiliated with the galaxy.730". We plot the contimmin jorinalized spectra iu a velocity space. with respect to he spectroscopic redshitt of the galaxy. z,4]cea=0.3529 (all velocities iu this paper are with respect to this "," We plot the continuum normalized spectra in a velocity space, with respect to the spectroscopic redshift of the galaxy, $\zgal = 0.3529$ (all velocities in this paper are with respect to this zero-point)."731Ate=|365kms| we detect strong saturated aabsorptiou., At $v = +365\kms$ we detect strong saturated absorption.732 The first 5 lines are saturated. and the svsteni breaks iuto two hain conponeuts in the higher order μαι lines. at velocities of e=|365.Ebims.!.," The first 5 lines are saturated, and the system breaks into two main components in the higher order Lyman lines, at velocities of $v733= +365, +445\kms$."734 Both coniponeuts lave associated weak aabsorptiou., Both components have associated weak absorption.735 The spectriun also shows a hint of possible aabsorptiou. but these features are only significant at the 20 level. aud we consider them non-detectious.," The spectrum also shows a hint of possible absorption, but these features are only significant at the $2\sigma$ level, and we consider them non-detections."736 Higher quality data would be required to confirm these features., Higher quality data would be required to confirm these features.737" Significantly. we detect no aabsorption in the TIRES data to very sensitive upper Πατ»,"," Significantly, we detect no absorption in the HIRES data to very sensitive upper limits."738 Due to its lack of strong metal liue. absorption. we describe this cloud as “metal-poor. a name we will justify in Section [.1..," Due to its lack of strong metal line absorption, we describe this cloud as “metal-poor”, a name we will justify in Section \ref{subsec: metal-poor}. ."739 This svstem also contains a acl, This system also contains a much740The differential equation possesses the right properties both at small and at large distances.,The differential equation possesses the right properties both at small and at large distances.741 At a large distance from the disk. the modulus in the complete elliptic integrals tends to zero.," At a large distance from the disk, the modulus in the complete elliptic integrals tends to zero."742 We have then . Since the total mass of the disk is given by the expression the aboveVain ODE can then be rearranged into As this equation must be satisfied for any s». we must have which is the expected behavior (the disk is no longer distinguishable from a point mass).," We have and then Since the total mass of the disk is given by the expression the above ODE can then be rearranged into As this equation must be satisfied for any $s$, we must have which is the expected behavior (the disk is no longer distinguishable from a point mass)."743 This also implies that rn perform ⋀↾⋅≏⊔∣⋯∣↑⋔⋋↾∐∏∁⊖⋅≏∐⋪⋂∐⋯↿⋔⊖⋂⊓∶↔↾⋯⋖∣↜⊜⊽∣⋅≪∠∣⋯⋟⋅∖∖⇁⊜∁∐∏ a second order expansion of the S-term by expanding the elliptic integral accordingly (?)..," This also implies that At a short distance around the origin (i.e. $r \ll \ain$ ), we can perform a second order expansion of the $S$ -term by expanding the elliptic integral accordingly \citep{gradryz65}."744 We find and then the ODE becomes whose solution is At second order. the potential in the inner domain (7< diy) IS quadratic with the cylindrical radius R (while. the gravitational acceleration is linear).," We find and then the ODE becomes whose solution is At second order, the potential in the inner domain $r \ll \ain$ ) is quadratic with the cylindrical radius $R$ (while the gravitational acceleration is linear)."745 As in ?.. we consider a vertical stratification of the form for |<]€f (and 0 elsewhere). where po is the density at the disk midplane. / the local semi-thiekness (both a function of the radius « ?..," As in \cite{hp09}, we consider a vertical stratification of the form for $|z| \le h$ (and $0$ elsewhere), where $\rho_0$ is the density at the disk midplane, $h$ the local semi-thickness (both a function of the radius $a$ \cite{hp09},"746objects randomly. distributed within their allowed: volume.,objects randomly distributed within their allowed volume.747 We now calculate. cach galaxys isophotal magnitude. a corrected magnitude and their total magnitude and sum the final number distribution according to absolute magnitude.," We now calculate each galaxy's isophotal magnitude, a corrected magnitude and their total magnitude and sum the final number distribution according to absolute magnitude."748 This is plotted in Fig 4 for total. corrected and isophotal absolute magnitucles.," This is plotted in Fig \ref{fig:N_M} for total, corrected and isophotal absolute magnitudes."749 We then reconstruct. the luminosity function using a ων prescription (as our simulations contain no clustering this should be an optimal estimator)., We then reconstruct the luminosity function using a $1/V_{Max}$ prescription (as our simulations contain no clustering this should be an optimal estimator).750 Fig., Fig.751 5 shows the recovered luminosity. functions., \ref{fig:lfs} shows the recovered luminosity functions.752 The LEs of 5, The LFs of Fig.753 demonstrate. the impact of surface. brightness selection as they are all drawn from the same DDE: the only cilferenee is the limiting isophote and the choice of magnitude measurement., \ref{fig:lfs} demonstrate the impact of surface brightness selection as they are all drawn from the same BBF; the only difference is the limiting isophote and the choice of magnitude measurement.754 The range of published values is shown as the shaded. area (excluding the LORS Lin ct al., The range of published values is shown as the shaded area (excluding the LCRS Lin et al.755 1996)., 1996).756 Also shown is the limit solution. for our model BBE.," Also shown is the limit solution, for our model BBF."757 The left panel assumes isophotal magnitudes are measured. the central panel assumes. Caussian corrected: magnitudes were used and the right panel assumes some procedure has been implemented. to recover the total magnitudes.," The left panel assumes isophotal magnitudes are measured, the central panel assumes Gaussian corrected magnitudes were used and the right panel assumes some procedure has been implemented to recover the total magnitudes."758 The results are also tabulatec in Table 2.., The results are also tabulated in Table \ref{table2}.759 If isophotal magnitudes. are adopted. and the surface brightness limit is bright. the luminosities of galaxies are severely uncerestimated.," If isophotal magnitudes are adopted and the surface brightness limit is bright, the luminosities of galaxies are severely underestimated."760 Thus both the number density and the A value are severely unclerestimatec (see Table. 2))., Thus both the number density and the $M^*$ value are severely underestimated (see Table \ref{table2}) ).761 The variation in O° is upto and in M? upto 1.0 mags., The variation in $\phi^*$ is upto and in $M^*$ upto 1.0 mags.762" This tallies well with the range of Schechter values recovered (see $2) over the range tested. (230κqua,<26).", This tallies well with the range of Schechter values recovered (see 2) over the range tested $23<\mu_{lim}<26$ ).763 To some extent is it surprising that @° is not more drastically ellected: this is because the observed distribution of galaxies is skewed towards the faint end. see Fig. 4..," To some extent is it surprising that $\phi^*$ is not more drastically effected; this is because the observed distribution of galaxies is skewed towards the faint end, see Fig. \ref{fig:N_M}."764 As à simple l/V4í; correction or maximum likelihood estimator based on the isophotal magnitudes alone does not take into account surface brightness issues. especially Dight loss. a smaller volume is calculated than for total magnitudes. leacling to an overestimate of the number density. see Fig 3...," As a simple $1/V_{max}$ correction or maximum likelihood estimator based on the isophotal magnitudes alone does not take into account surface brightness issues, especially light loss, a smaller volume is calculated than for total magnitudes, leading to an overestimate of the number density, see Fig \ref{fig:V_M}."765 This is tempered by a lower number density at. brighter absolute magnituces., This is tempered by a lower number density at brighter absolute magnitudes.766 Perhaps most surprising is the robustness of the faint end slope whose value is recovered correctly regardless of the isophote., Perhaps most surprising is the robustness of the faint end slope whose value is recovered correctly regardless of the isophote.767 Alost surveys attempt to correct their isophotal magnitudes to total magnitudes., Most surveys attempt to correct their isophotal magnitudes to total magnitudes.768 We used a Gaussian. correction as described. above., We used a Gaussian correction as described above.769 Fig., Fig.770 5— and Table 2— demonstrate that corrected. magnitudes recover of the Iuminosity density at 24. mag 2 compared to. the that isophotal magnitudes recover and the that total magnitudes. recover., \ref{fig:lfs} and Table \ref{table2} demonstrate that corrected magnitudes recover of the luminosity density at 24 mag $^{-2}$ compared to the that isophotal magnitudes recover and the that total magnitudes recover.771 As with isophotal magnitudes. corrected. magnitudes give a luminosity function. biased. at all values of AZ. although the bias has been significantly. reclucecL.," As with isophotal magnitudes, corrected magnitudes give a luminosity function biased at all values of $M$, although the bias has been significantly reduced."772 If some method is emploved to correct the galaxies to total magnitudes (e.g. Kron magnitudes or Petrosian magnitucdes) we find that the parameters are very. robust for. fts=24 mag arcsec, If some method is employed to correct the galaxies to total magnitudes (e.g. Kron magnitudes or Petrosian magnitudes) we find that the parameters are very robust for $\mu_{lim} \geq 24$ mag $^{-2}$.773" llowever. at nj,=28 mag arcsec2 the number density is. uncderestimatecl throughout the distribution."," However, at $\mu_{lim}=23$ mag $^{-2}$ the number density is underestimated throughout the distribution."774 Fig 3.illustrates why this occurs., Fig \ref{fig:V_M} illustrates why this occurs.775" The volume has almost no surface brightness dependeney for the thresholds 24<fii,«26 provided Alκ 14. but it has significant surface brightness dependeney for 22<M«<ld at fas= 23."," The volume has almost no surface brightness dependency for the thresholds $24<\mu_{lim}<26$ provided $M<-14$ , but it has significant surface brightness dependency for $-22<M<-14$ at $\mu_{lim}=23$ ."776 The bright, The bright777Iu this Appendix we display analytic forms for the convolution of the clitmensiouless prolile shape discussed in 83.,In this Appendix we display analytic forms for the convolution of the dimensionless profile shape discussed in 3.778 These analytic expressions are useful for computiug the nonlinear two-poiut correlation fuuctioun € of the mass deusity field. which is dominated by the 1-halo term £45 in and is related to A by For the type-I profile «t; of(2).. the angular integration in equation (A1)) is analytic. auc A is reduced to a simple integral For the special case p—1. this integral can be further reduced to the analytical form For«;; of(2).. we are able to simplify A to where the function.Cr.y) represents the angular partof the integration in equation (À1)) aud The integral in £j cau be reduced to analytic formisfor special values of p.," These analytic expressions are useful for computing the nonlinear two-point correlation function $\xi$ of the mass density field, which is dominated by the 1-halo term $\xi_{1h}$ in and is related to $\lambda$ by For the type-I profile $u_I$ of, the angular integration in equation \ref{lamb2}) ) is analytic, and $\lambda$ is reduced to a simple integral For the special case $p=1$, this integral can be further reduced to the analytical form For$u_{II}$ of, we are able to simplify $\lambda$ to where the function$F_p(x,y)$ represents the angular partof the integration in equation \ref{lamb2}) ) and The integral in $F_p$ can be reduced to analytic formsfor special values of $p$ ."779 Here we display the six cases p— 0. 1/2. 1. 3/2. 2. and 5/2:," Here we display the six cases $p=0$ , $1/2$ , 1, $3/2$ , 2, and$5/2$ :"7803.2). we find an upper init which is AZ(HTI)1.3(1.5)«10LM. with (οιt) photoionization incluled.,"3.2), we find an upper limit which is $\Ti44 \lsim 1.3 (1.5) \times 10^{-4} \Msun$ with (without) photoionization included."781 These lnasscs are for a box-shaped line profile. aux therefore provide conservative liuits.," These masses are for a box-shaped line profile, and therefore provide conservative limits."782 However. there is good reason to believe that the line profile should be simil:w to those for the lines observed bv Iaas et al. (," However, there is good reason to believe that the line profile should be similar to those for the lines observed by Haas et al. ("783199|) 100 davs after the explosion.,1990) $\sim 400$ days after the explosion.784 The srougest Lue olserved by Haas et al., The strongest line observed by Haas et al.785 was [Fo TI] 17.9an aid it had CFAVIDM(2900£80)kins |.," was [Fe II] $\mum$ and it had $v_{\rm FWHM} = (2\,900 \pm 80) \kms$ ."786 Although it scaled jus short off [1000kins and thus was uot svmunetric around the rest velocity of the super10Va. it rack t15 general appearance of the expected line profile tfxiucd by a filled sphere with coustanut cussion throughou the sphere.," Although it peaked just short off $+ 1\,000 \kms$, and thus was not symmetric around the rest velocity of the supernova, it had the general appearance of the expected line profile formed by a filled sphere with constant emission throughout the sphere."787 For a sphere extendius out to 20)X)ans logi6 peak ds 1.5 times higher than for the JON xofile usec Lin Table 2 (which is valid for a hollow sphere). aud CRWILN2S28kins +.," For a sphere extending out to $2\,000 \kms$, the peak is 1.5 times higher than for the box profile used in Table 2 (which is valid for a hollow sphere), and $v_{\rm FWHM} = 2\,828 \kms$ ."788 Using such a liic xofile im our τηoper linits on AJ(UTi) we dusCac Otal1 MUTE0.91.0by10.NI. with Gvithort) photoionization iucluded.," Using such a line profile in our upper limits on $\Ti44$, we instead obtain $\Ti44 \lsim 0.9 (1.0) \times 10^{-4} \Msun$ with (without) photoionization included."789 Wihin the framework of our iuochi1οB. a Conservative lait (ic.. the case when photoiozation is uninm]xytant) for a plausible line profile Is therefore ACHTHS1hy101 ALL.," Within the framework of our modeling, a conservative limit (i.e., the case when photoionization is unimportant) for a plausible line profile is therefore $\Ti44 \lsim 1.0 \times 10^{-4} \Msun$ ."790 In Fig., In Fig.791 1. we have inchided the ex]ected line Cluission for such à uxxdel with tjs limiting mass of !HTi.," 1, we have included the expected line emission for such a model with this limiting mass of $^{44}$ Ti."792 We will evaluate this init on M(Ti in Sect., We will evaluate this limit on $\Ti44$ in Sect.793 Ll., 4.1.794 There are several uncertainties involved in our modeling of the line fluxes., There are several uncertainties involved in our modeling of the line fluxes.795 We have already checked the effect of the lifetime of 14Ti (Sect., We have already checked the effect of the lifetime of $^{44}$ Ti (Sect.796 3.1)., 3.1).797 We have also studie Lthe effect of photoionization. and found that it introdices rather uuld uncertainties.," We have also studied the effect of photoionization, and found that it introduces rather mild uncertainties."798 A similar level of uncertaiuty is due to the distance to the supernova., A similar level of uncertainty is due to the distance to the supernova.799 This is still Inaccurate to the level of 5.10% (Lundqvist Souueοι 1999: Walker 1999). which means an wucertaiuty iut1ο line flux of ~1020%.," This is still inaccurate to the level of $5 - 10 \%$ (Lundqvist Sonneborn 1999; Walker 1999), which means an uncertainty in the line flux of $\sim 10 - 20 \%$."800" Atomic data of ion are notoriously «ifficult to calculate accurately,", Atomic data of iron are notoriously difficult to calculate accurately.801 This is therefore another source of error m our modeling. especially for individual lines.," This is therefore another source of error in our modeling, especially for individual lines."802" Most oeuportant for the 267221 line is the collision strength of that transition. QO»6,4,."," Most important for the $26 \mum$ line is the collision strength of that transition, $\Omega_{26 \mum}$."803 Oue normally assigns an uncertainty in the collision streugth for the strongest iron lines to 30% (cses Verner et al.," One normally assigns an uncertainty in the collision strength for the strongest iron lines to $\sim 30\%$ (e.g., Verner et al."804 1999)., 1999).805 Tn our models we have used OQ2041u=h.8 (Zhang Pradhan 1995)., In our models we have used $\Omega_{26 \mum} = 5.8$ (Zhang Pradhan 1995).806" To study the effect inc etail we have tested a model with AZ(HTi)10EAN, aud O.26411=2.9. We find that there is not a linear scaling between fob auld Qo6,44. as oue night naively believe."," To study the effect in detail we have tested a model with $\Ti44 = 10^{-4} \Msun$ and $\Omega_{26 \mum} = 2.9$ We find that there is not a linear scaling between $f_{26 \mum}$ and $\Omega_{26 \mum}$ , as one might naively believe."807 Tuscad. we find from) linear interpolation tha a 30% decrease in collision streugth gives a ~1πιοτά lower F26jiu- alc a correspondingv higher estimae of AM(CUTI).," Instead, we find from linear interpolation that a $30\%$ decrease in collision strength gives a $\sim 17\%$ lower $f_{26 \mum}$, and a correspondingly higher estimate of $\Ti44$."808 The reason for this is that the gas is slightly hotter iu t1e niodol with red1ced collision strength. boosting the exponcutial teri in tιο collisional rate so that it somewhat counteracts the reducd collision streugth.," The reason for this is that the gas is slightly hotter in the model with reduced collision strength, boosting the exponential term in the collisional rate so that it somewhat counteracts the reduced collision strength."809" We have receutly found out (A. Praciur private communication) that the preferred valie for Όρυμιι at fie low teruperatires jn SN 198TÀ is lost ise5v closer to ~7.0. Ίνοι, higher than we lave used."," We have recently found out (A. Pradhan, private communication) that the preferred value for $\Omega_{26 \mum}$ at the low temperatures in SN 1987A is most likely closer to $\sim 7.0$, i.e., higher than we have used."810 Des js. we have generously assigned an uncertainty of 1! pwiuds) iu MtLUTO) due to atomic data.," Despite this, we have generously assigned an uncertainty of $15\%$ (upwards) in $\Ti44$ due to atomic data."811 Anotjer source of uncertainty could be the explosion mode sec., Another source of uncertainty could be the explosion model used.812 Iu these calculations we have used t abuances from the ΤΟΠ explosion model., In these calculations we have used the abundances from the 10H explosion model.813 A comparix between the two models LOM (Woosev Weaver 1986: Woosev 1988) and LIEL (Shigevama et al., A comparison between the two models 10H (Woosley Weaver 1986; Woosley 1988) and 11E1 (Shigeyama et al.814 1988) was do1ο in Kozma Fransson (1998y, 1988) was done in Kozma Fransson (1998b).815 There i was fond that the iron lines are uot sensitive to the exdosion model used. because the ion core mass is the saue 1 both models.," There it was found that the iron lines are not sensitive to the explosion model used, because the iron core mass is the same in both models."816 The iron core mass is set xw the amount of Ni which is accurately determined. fi1u the booletric lieht curve., The iron core mass is set by the amount of $^{56}$ Ni which is accurately determined from the bolometric light curve.817 The choice of explosion model thus «oes not ποσα to be a lnajor source of uucertaiity when inodeliug these iron les., The choice of explosion model thus does not seem to be a major source of uncertainty when modeling these iron lines.818 Tn our calculations we assume a loca deposition of he positrous originating frou the radioactive decays of ARCHTi)., In our calculations we assume a local deposition of the positrons originating from the radioactive decays of $\Ti44$.819 We believe this is a good approximation since optical aud near-IR liebt curves of Fe Τα Fe II lines show hat trapping must occur (Chiugai ct al., We believe this is a good approximation since optical and near-IR light curves of Fe I and Fe II lines show that trapping must occur (Chugai et al.820 1997: IKoziua Frausso1 L998b)., 1997; Kozma Fransson 1998b).821 Actualv. there is no obvious sign of a cakaee of positrons. ucither from broad-band lighteurves (IKXE99). nor from the opical Fe I lines at 6300 uuutil tie last data poi at 3597 davs in Wozma (1999).," Actually, there is no obvious sign of a leakage of positrons, neither from broad-band lightcurves (KF99), nor from the optical Fe I lines at $6\,300$ until the last data point at $3\,597$ days in Kozma (1999)."822 Although the trapping may well be fully complete. we have assigned an error to this assumption by," Although the trapping may well be fully complete, we have assigned an error to this assumption by."823 Another approximation in ο models is the assuniptiou of a honiogeneous densitv in cach Fe-rich shell of the model core., Another approximation in our models is the assumption of a homogeneous density in each Fe-rich shell of the model core.824 To test the seusitivity to this asstuuption. we have run a model similar to M2. with photoionization included (see Table 2). but where we have divided the mass in the Fe-vich ejecta iuto two COMPONCs of equal mass but with different deusitics.," To test the sensitivity to this assumption, we have run a model similar to M2, with photoionization included (see Table 2), but where we have divided the mass in the Fe-rich ejecta into two components of equal mass but with different densities."825 The denser componucut is set to be nearly five times more deuse thai the other., The denser component is set to be nearly five times more dense than the other.826"Despite the siguificantlydifferent deusity distribution in this model compared to that in M2. the differences in fogj4, all foa, between the models","Despite the significantlydifferent density distribution in this model compared to that in M2, the differences in $f_{26 \mum}$ and $f_{24 \mum}$ between the models"827is à galaxy which is not member of any geometric pair.,is a galaxy which is not member of any geometric pair.828 The single galaxies are field galaxies in the environment. of geometric pairs., The single galaxies are field galaxies in the environment of geometric pairs.829 Every single galaxy has the own neighbours. single galaxies ancl geometric pair members can be among them.," Every single galaxy has the own neighbours, single galaxies and geometric pair members can be among them."830 According to thesecond-order Voronoi tessellation the larger is the degree of galaxy isolation. the e&reater is the number of neighbours. (see Fie.," According to thesecond-order Voronoi tessellation the larger is the degree of galaxy isolation, the greater is the number of neighbours (see Fig."831 Ibb in comparison with Fig., \ref{fig1}b b in comparison with Fig.832 2bb). but these neighbours locate farther.," \ref{fig2}b b), but these neighbours locate farther."833 The best parameter which cleseribes the isolation degree of the single galaxy is the mean value. of all parameters p; of this galaxy: where A is the number of neighbours., The best parameter which describes the isolation degree of the single galaxy is the mean value of all parameters $p_{j}$ of this galaxy: where $k$ is the number of neighbours.834 Therefore. the smaller is 5 value. the more isolated is the single galaxy.," Therefore the smaller is $s$ value, the more isolated is the single galaxy."835 Methocl of the third-order Voronoi tessellation can be introduced the same as the second-order approach (Fig., Method of the third-order Voronoi tessellation can be introduced the same as the second-order approach (Fig.836 Lec)., \ref{fig1}c c).837 All points of the common triplet's cell are the most closer to galaxies of this triplet than to other galaxies., All points of the common triplet's cell are the most closer to galaxies of this triplet than to other galaxies.838" Similarly to the parameter Dij [or pairs we set up the parameter Figs for triplets: where D - space dimension. Vij. - the area (for 2D) or volume (for 3D) of coll. m;j. mj,4. Πατ distances between galaxies in the triplet."," Similarly to the parameter $p_{i, j}$ for pairs we set up the parameter $t_{i, j, u}$ for triplets: where $D$ - space dimension, $V_{i, j, u}$ - the area (for 2D) or volume (for 3D) of cell, $m_{i,j}$, $m_{i,u}$, $m_{j,u}$ - distances between galaxies in the triplet."839 in the third-order Voronoi tessellation contains three ealaxies that have the common cell and the same maximal parameters ρα) f., in the third-order Voronoi tessellation contains three galaxies that have the common cell and the same maximal parameters $t_{max}(1)=t_{max}(2)=t_{max}(3)=t$ .840" Phe parameter / characterizes aisolation.. We defined parameter of [as the mean value of parameters (;(1). £;(2) and £,(3). except { from three sets: here in case of third-order Voronoi tessellation &. n and q denote the number of which contain ealaxies 1. 2 and 3.2 respectively."," The parameter $t$ characterizes a. We defined parameter of $t_{e}$ as the mean value of parameters $t_{i}(1)$, $t_{j}(2)$ and $t_{u}(3)$, except $t$ from three sets: here in case of third-order Voronoi tessellation $k$, $n$ and $q$ denote the number of which contain galaxies 1, 2 and 3, respectively."841 Pherefore &|on|g 3i is number of neighbouring triplets for certain triplet. which contain at least one galaxy from this triplet.," Therefore $k+n+q-$ 3 is number of neighbouring triplets for certain triplet, which contain at least one galaxy from this triplet."842 lt can be seen from Lie., It can be seen from Fig.843 Pee and (4). (5) that for the riplet with highest degree of standing oul against a background. the isolation parameter / has the highest value.," \ref{fig2}c c and (4), (5) that for the triplet with highest degree of standing out against a background, the isolation parameter $t$ has the highest value."844" At the same time. if the triplet neighbours locate far from it. parameter /,. has the small value."," At the same time, if the triplet neighbours locate far from it, parameter $t_{e}$ has the small value."845 'arameters p. s. fare the basic ones and. define the isolation degree of galaxy pair. single galaxy or triplet. in comparison with background. respectively.," Parameters $p$, $s$, $t$ are the basic ones and define the isolation degree of galaxy pair, single galaxy or triplet in comparison with background, respectively."846" Parameters. p, and ἐν are additional ones and contain information about the distribution of the neighbouring galaxies (environment).", Parameters $p_{e}$ and $t_{e}$ are additional ones and contain information about the distribution of the neighbouring galaxies (environment).847 Similarly to the seconc- ancl third-order Voronoi tessellation it is possible to apply more high-order Voronoi tessellation for the identification of galaxy quartets. quintets and so on.," Similarly to the second- and third-order Voronoi tessellation it is possible to apply more high-order Voronoi tessellation for the identification of galaxy quartets, quintets and so on."848 For our investigation we used. Northern part. of the SDSS DRS spectroscopic survey., For our investigation we used Northern part of the SDSS DR5 spectroscopic survey.849 Our sample is volume-IHmited and consist of objects that classified. as galaxies., Our sample is volume-limited and consist of objects that classified as galaxies.850 The primary sample hacl contained. approximately 11000. galaxies with radial velocities from 2500 kim s.| to 10000 kim sf. Lo = 75 km Alpe|l.," The primary sample had contained approximately 11000 galaxies with radial velocities from 2500 km $^{-1}$ to 10000 km $^{-1}$, $H_{0}$ = 75 km $^{-1}$ $^{-1}$."851 lt is known that compliteness of SDSS is poor for the bright galaxics because of spectroscopic selection criteria and the οασεν of obtaining correc photometry for object with large angular size., It is known that compliteness of SDSS is poor for the bright galaxies because of spectroscopic selection criteria and the diffuculty of obtaining correct photometry for object with large angular size.852 We trice to decrease of this cllect’s inlluence. duc to limiting of our sample 1j2 2500 km 1. Lie. in the way not το take into account nearest objects with the large angular diameter.," We tried to decrease of this effect's influence due to limiting of our sample $V_{h} > $ 2500 km $^{-1}$, i.e. in the way not to take into account nearest objects with the large angular diameter."853 Such a volume limiting also helps us to avo influence of Virgo cluster where strong peculiar motion exists., Such a volume limiting also helps us to avoid influence of Virgo cluster where strong peculiar motion exists.854 We checked: additionally all pairs of galaxies with a small angular resolution and. excluded. identical. objects (parts of galaxies). which are presented twice and more in SDSS survey.," We checked additionally all pairs of galaxies with a small angular resolution and excluded identical objects (parts of galaxies), which are presented twice and more in SDSS survey."855" All galaxy velocities Vj, were corrected. for the Local Group centroidVc; accordingly to Ixarachentsev Makarov (1996).", All galaxy velocities $V_{h}$ were corrected for the Local Group centroid$V_{LG}$ accordingly to Karachentsev Makarov (1996).856 When we πας applied the high-order Voronoi tessellation method to SDSS catalogue we limited our sample 3000 kim Vier: 9500 kni +., When we had applied the high-order Voronoi tessellation method to SDSS catalogue we limited our sample 3000 km $^{-1}$ $\leq V_{LG} \leq$ 9500 km $^{-1}$ .857 We did not, We did not858"probability is μπα,",probability is small.859 A very promising possibility is a damped absorber iu (2206-199. at 2=2.559. with a low metallicity aud very narrow lines (Pettini et al.," A very promising possibility is a damped absorber in Q2206-199, at $z=2.559$, with a low metallicity and very narrow lines (Pettini et al."860 1991)., 1994).861 The iuterloper problem is πιο simaller here. both because of the very ligh column aud the low redshift.," The interloper problem is much smaller here, both because of the very high column and the low redshift."862 The low redshift requires UST. but with STIS the ecutive Lyman series can be seen at once so this is a practical program. currently approved for cycle 7.," The low redshift requires HST, but with STIS the entire Lyman series can be seen at once so this is a practical program, currently approved for cycle 7."863 There is certainly HIT absorption between the identified lines of the a forest., There is certainly HI absorption between the identified lines of the $\alpha$ forest.864 As spectra of higher signal-to-noise ratio are obtained they reveal absorption of progressively lower optical depth., As spectra of higher signal-to-noise ratio are obtained they reveal absorption of progressively lower optical depth.865 In HII however. even at the highest S/N so far available (about 100 at high resolution). the IIT absorption does not vet fill redshift space.," In HI however, even at the highest S/N so far available (about 100 at high resolution), the HI absorption does not yet fill redshift space."866 Limits ou III coutiunous optical depth (or “Camu-Peterson effect”) provide useful coustraiunts on diffuse eas density. subject," Limits on HI continuous optical depth (or “Gunn-Peterson effect”) provide useful constraints on diffuse gas density, subject"867brown-dwarf companion and a well-determined astrometric orbit.,brown-dwarf companion and a well-determined astrometric orbit.868" After determining the orbital inclination from astrometry, we compared it to the orientation of the stellar spin axis."," After determining the orbital inclination from astrometry, we compared it to the orientation of the stellar spin axis."869" The inclination i;4 of the stellar spin axis is defined with respect to the line of sight (i4,=0? or 180? for a pole-on view) and can be derived from the spectroscopic estimate of 4.2 km s! (?)..", The inclination $i_\mathrm{rot}$ of the stellar spin axis is defined with respect to the line of sight $i_\mathrm{rot}=0\degr$ or $180\degr$ for a pole-on view) and can be derived from the spectroscopic estimate of $ \upsilon \sin i_\mathrm{rot} = 4.2$ km $^{-1}$ \citep{Santos:2010fk2}.870" The authors did not give an error bar for this measurement and we assumed a conservative uncertainty of 1 km s!. On the basis of the activity indicator logRink determined from the HARPS spectra, we used the calibration of ? to derive the stellar rotation period Py."," The authors did not give an error bar for this measurement and we assumed a conservative uncertainty of 1 km $^{-1}$ On the basis of the activity indicator $\log R'_{H,K}$ determined from the HARPS spectra, we used the calibration of \cite{Mamajek:2008fk} to derive the stellar rotation period $P_\mathrm{rot}$ ."871" Assuming an effective temperature of Teg=6297+32 (?),, the apparent visual magnitude my=6.839+0.001 (?),, and the parallax @ given in Table 2,, we derived the star’s absolute magnitude, luminosity(L=4.5 0.919). and radius (R=1.8+0.1 Re), using standard formulae and Monte Carlo resampling."," Assuming an effective temperature of $T_\mathrm{eff} = 6297 \pm 32$ \citep{Santos:2010fk2}, the apparent visual magnitude $m_V = 6.839 \pm 0.001$ \citep{:2007kx}, , and the parallax $\varpi$ given in Table \ref{tab:2}, we derived the star's absolute magnitude, luminosity$L = 4.5 \pm 0.3\, L_{\sun}$ ), and radius $R = 1.8 \pm 0.1 \,R_{\sun}$ ), using standard formulae and Monte Carlo resampling."872 Bolometric corrections were computed using the ? parameters given by ?.., Bolometric corrections were computed using the \cite{Flower:1996qy} parameters given by \cite{Torres:2010uq}.873" Using the stellar radius and rotation period, we derived the equatorial rotation velocity to be v=10+3 ss!, and then the inclination of the spin axis by calculating ijo;=arcsin(vsinijo: /v)."," Using the stellar radius and rotation period, we derived the equatorial rotation velocity to be $\upsilon = 10 \pm 3$ $^{-1}$, and then the inclination of the spin axis by calculating $i_\mathrm{rot} = \arcsin( \upsilon \sin i_\mathrm{rot} / \upsilon)$ ."874" Because we cannot determine the star's sense of rotation, the angle ig, hasa 180° ambiguity and we first considered the value that falls into the same quadrant as ioi, ie. a prograde configuration."," Because we cannot determine the star's sense of rotation, the angle $i_\mathrm{rot}$ hasa $180^\circ$ ambiguity and we first considered the value that falls into the same quadrant as $i_\mathrm{orbit}$, i.e. a prograde configuration."875" The astrometric analysis yielded the distribution of init, obtained from 1000000 Monte Carlo simulations."," The astrometric analysis yielded the distribution of $i_\mathrm{orbit}$, obtained from 000 Monte Carlo simulations."876" To obtain the i;4,-distribution, we performed 106 Monte Carlo simulations."," To obtain the $i_\mathrm{rot}$ -distribution, we performed $10^6$ Monte Carlo simulations."877 We finally compared these two distributions by drawing 2:107 pairs of values [iomit. trot] and obtained the distribution of the orbit obliquity or angle y by computing y=ioi—dot.," We finally compared these two distributions by drawing $2\cdot10^7$ pairs of values $[i_\mathrm{orbit}$ , $i_\mathrm{rot}]$ and obtained the distribution of the orbit obliquity or angle $\psi$ by computing $\psi = i_\mathrm{orbit} -i_\mathrm{rot}$."878" Figure 3. shows the probability density functions (PDF) of ioi, ioi, and y."," Figure \ref{fig:inclinations} shows the probability density functions (PDF) of $i_\mathrm{orbit}$ , $i_\mathrm{rot}$, and $\psi$."879" The io,pix-distribution is very narrow and peaks around 178.3°, whereas the ij4-distribution is broad with its maximum at 155? and the two distributions show a very small overlap."," The $i_\mathrm{orbit}$ -distribution is very narrow and peaks around $ 178.3^\circ$ , whereas the $i_\mathrm{rot}$ -distribution is broad with its maximum at $ \sim\!155^\circ$ and the two distributions show a very small overlap."880" The median values and confidence intervals of τοι, igi, and the obliquity y are given in Table 3.."," The median values and confidence intervals of $i_\mathrm{rot}$, $i_\mathrm{orbit}$, and the obliquity $\psi$ are given in Table \ref{tab:spin}."881" We found that wy is larger than 4.4? and 19.3? with a probability of and 68.3%,, respectively, and its nominal value with 1c confidence intervals is y.=29105 in prograde configuration, which indicates that the orientations of stellar spin and orbital axis differ substantially."," We found that $\psi$ is larger than $4.4\degr$ and $19.3\degr$ with a probability of and 68.3, respectively, and its nominal value with $1\,\sigma$ confidence intervals is $\psi= 23_{- 8}^{+10\,\circ}$ in prograde configuration, which indicates that the orientations of stellar spin and orbital axis differ substantially."882" In retrograde configuration, the obliquity would be y=153130."," In retrograde configuration, the obliquity would be $\psi'=153_{-10}^{+ 8\,\,\circ}$."883" We note that these values of y and w’ are lower limits, becausethe ascending node £X, of the spin axis is not constrained and we assumed that Q=€, in their derivation."," We note that these values of $\psi$ and $\psi'$ are lower limits, becausethe ascending node $\Omega_\mathrm{rot}$ of the spin axis is not constrained and we assumed that $\Omega = \Omega_\mathrm{rot}$ in their derivation."884" As illustrated in Fig. 4,,"," As illustrated in Fig. \ref{fig:inc}, ,"885" y increases if the two ascending nodes Ώχοι and Q do not coincide, creating an additional uncertainty in y of 2-(180?—iowit)=3.4? in the prograde configuration."," $\psi$ increases if the two ascending nodes $\Omega_\mathrm{rot}$ and $\Omega$ do not coincide, creating an additional uncertainty in $\psi$ of $2\cdot (180\degr-i_\mathrm{orbit}) = 3.4 \degr$ in the prograde configuration."886"For a retrograde orbit, thisuncertainty is larger at 51°.","For a retrograde orbit, thisuncertainty is larger at $51 \degr$ ."887" In summary, the obliquity is y=23*0E941 ?and ψ’=1535511? for prograde and retrogradeorbits, respectively, where theadditional uncertainties are denotedin square brackets."," In summary, the obliquity is $\psi= 23_{- 8}^{+10\,[+3.4]\,\circ}$ and $\psi'=153_{-10}^{+ 8\, [+51]\,\circ}$ for prograde and retrogradeorbits, respectively, where theadditional uncertainties are denotedin square brackets."888" We performed these calculations using the Pyo.-calibration by ?,, which yielded a value of 8.8+1.6 days and an obliquity ofyΞ22133, in agreementwith the result obtained with the"," We performed these calculations using the $P_\mathrm{rot}$ -calibration by \cite{Noyes:1984qy}, , which yielded a value of $8.8 \pm 1.6$ days and an obliquity of $\psi= 22_{- 7}^{+9\,\circ}$, in agreementwith the result obtained with the"889Permanent superhumps have been observed so far in about 20 cataclysmic variables (CVs) (Patterson 1999).,Permanent superhumps have been observed so far in about 20 cataclysmic variables (CVs) (Patterson 1999).890 These show superhumips (quasi-perioclicitics shifted by a few percent from their orbital periods) in their optical light curves during normal brightness state., These show superhumps (quasi-periodicities shifted by a few percent from their orbital periods) in their optical light curves during normal brightness state.891 In contrast. SU UMa systems (see Warner 1995 for a review of SU UMa systems and CVs in general) have superhumps only during their bright chwarl nova outbursts (superoutbursts)," In contrast, SU UMa systems (see Warner 1995 for a review of SU UMa systems and CVs in general) have superhumps only during their bright dwarf nova outbursts (superoutbursts)."892" Permanent superhumps can either be a few percent longer than the orbital periods and they are called ""positive superhumps'. or shorter ""negative superhumps'."," Permanent superhumps can either be a few percent longer than the orbital periods and they are called $\bf positive$ $\bf superhumps$ ', or shorter – $\bf negative$ $\bf superhumps$ '."893 The positive superhump is explained. as the beat between the binary motion and the precession of an accretion disc in, The positive superhump is explained as the beat between the binary motion and the precession of an accretion disc in894LEGOs observed in the field of 0322 fill out the sampled volume. with a mean redshift of 3.155 and a standard eviation of 0.019.,"LEGOs observed in the field of $-$ 0322 fill out the sampled volume, with a mean redshift of 3.155 and a standard deviation of 0.019."895 On the contrary. the redshifts of LEGOs —1 the field of 4427 have a mean of 2.858 and a tandard deviation of only 0.006. corresponding to a velocity Cvispersion of 470 km !.," On the contrary, the redshifts of LEGOs in the field of $-$ 4427 have a mean of 2.858 and a standard deviation of only 0.006, corresponding to a velocity dispersion of 470 km $^{-1}$."896" A significant part of this velocity vwpread must be caused by peculiar velocities or offsets between the Lya and systemic redshifts. and therefore. the Hubble flow ""Septh of the structure should be even less."," A significant part of this velocity spread must be caused by peculiar velocities or offsets between the $\alpha$ and systemic redshifts, and therefore, the Hubble flow depth of the structure should be even less."897 The mean observed redshift is close to the redshift of the DLA absorber toward 4427 (z=2.851)., The mean observed redshift is close to the redshift of the DLA absorber toward $-$ 4427 (z=2.851).898 This indicates the presence of a large-scale structure of galaxies. e.g. a pancake-like structure at the redshift of the DLA absorber. surrounded by voids.," This indicates the presence of a large-scale structure of galaxies, e.g. a pancake-like structure at the redshift of the DLA absorber, surrounded by voids."899 Independent evidence for this comes from the observation of strong metal absorption lines at the same redshift in two nearby QSOs (Francis Hewitt 1993: D'Odorico et al., Independent evidence for this comes from the observation of strong metal absorption lines at the same redshift in two nearby QSOs (Francis Hewitt 1993; D'Odorico et al.900 2002)., 2002).901 The redshift distribution of LEGOS in the field of 4427 is similar to that of LEGOs in the fields of radio galaxies (Pentericei et al., The redshift distribution of LEGOs in the field of $-$ 4427 is similar to that of LEGOs in the fields of radio galaxies (Pentericci et al.902 2000: Venemans et al., 2000; Venemans et al.903 2002)., 2002).904 In a subsequent paper (Ledoux et al..," In a subsequent paper (Ledoux et al.,"905 in prep.).," in prep.),"906 we will address the properties of the DLA absorber., we will address the properties of the DLA absorber.907 Most of the candidates that we did not confirm are faint in the narrow-band images and/or have low EWs (see Fig. 4))., Most of the candidates that we did not confirm are faint in the narrow-band images and/or have low EWs (see Fig. \ref{select}) ).908 These candidates could either have been missed by the slitlets. be too faint for the follow-up spectroscopy (due to the bad seeing we did not reach the planned detection limit). or simply not be LEGOs.," These candidates could either have been missed by the slitlets, be too faint for the follow-up spectroscopy (due to the bad seeing we did not reach the planned detection limit), or simply not be LEGOs."909 As seen in Fig. 3..," As seen in Fig. \ref{colcol},"910 the unconfirmed candidates tend to have redder colours R(AB) = 0.9-1.8) than the confirmed candidates (see Fig. 3))., the unconfirmed candidates tend to have redder colours $-$ R(AB) = 0.9–1.8) than the confirmed candidates (see Fig. \ref{colcol}) ).911 The confirmed emission- sources detected in the broad bands indeed have blue colours (typically R(AB)<0.8)., The confirmed emission-line sources detected in the broad bands indeed have blue colours (typically $-$ $<$ 0.8).912 However. the fact that one of the confirmed candidates. LEGO2138.331. has colours and flux very similar to the unconfirmed candidates makes it," However, the fact that one of the confirmed candidates, 31, has colours and flux very similar to the unconfirmed candidates makes it"913"Some simple properties of the Dirac-Milne universe are common to those of a purely linear cosmology studied in ?,, ??,, and ?..","Some simple properties of the Dirac-Milne universe are common to those of a purely linear cosmology studied in \citet{Simmering98}, , \citet{Kaplinghat99, Kaplinghat00}, , and \citet{Sethi05}."914 We briefly recall some of these properties., We briefly recall some of these properties.915 Using Eq. (5))," Using Eq. \ref{fried1}) ),"916" it is straightforward to obtain the relation between the age of the Universe and the redshift, hence the temperature where Tp is the present temperature of the Universe, as measured by CMB experiments, To=2.725+0.001K (?).."," it is straightforward to obtain the relation between the age of the Universe and the redshift, hence the temperature where $T_0$ is the present temperature of the Universe, as measured by CMB experiments, $T_0=2.725\pm 0.001\;\rm{K}$ \citep{Fixsen2002}."917 This relation between time and temperature is valid throughout the whole history of the universe and implies that the thermal history of the Dirac-Milne universe is drastically modified from the evolution in the standard ACDM cosmology., This relation between time and temperature is valid throughout the whole history of the universe and implies that the thermal history of the Dirac-Milne universe is drastically modified from the evolution in the standard $\Lambda$ CDM cosmology.918 Figure (1)) represents the age of the Universe as a function of the temperature for the Dirac-Milne and the ACDM models., Figure \ref{age}) ) represents the age of the Universe as a function of the temperature for the Dirac-Milne and the $\Lambda$ CDM models.919" It can be seen that, at high temperatures, the Dirac-Milne universe is much older than the corresponding ACDM cosmology."," It can be seen that, at high temperatures, the Dirac-Milne universe is much older than the corresponding $\Lambda$ CDM cosmology."920" For instance, the traditional 1 MeV-1 sec approximation for the standard model becomes 1 MeV - 3.3 years in the Dirac-Milne cosmology."," For instance, the traditional 1 $\sim$ 1 sec approximation for the standard model becomes 1 MeV $\sim$ 3.3 years in the Dirac-Milne cosmology."921" As noted in ?,, this difference has profound implications for big-bang nucleosynthesis calculations, and is discussed in section 4.."," As noted in \citet{Simmering98}, this difference has profound implications for big-bang nucleosynthesis calculations, and is discussed in section \ref{sec_bbn}."922 Another temperature of interest is the temperature of the quark-gluon-plasma (QGP) transition., Another temperature of interest is the temperature of the quark-gluon-plasma (QGP) transition.923" ? proposed that matter-antimatter separation occurred around that temperature, owing to a putative repulsive interaction between nucleons and antinucleons."," \citet{Omnes1972} proposed that matter-antimatter separation occurred around that temperature, owing to a putative repulsive interaction between nucleons and antinucleons."924 The maximum size of a domain of (anti)matter was controlled by the diffusion of neutrons., The maximum size of a domain of (anti)matter was controlled by the diffusion of neutrons.925" ? found a maximum size of 7x107cm at a temperature of T~330 MeV, which was at this epoch the estimated temperature of the QGP transition."," \citet{Aly_sep1974} found a maximum size of $7\times 10^{-4}\;\rm{cm}$ at a temperature of $T\sim330 $ MeV, which was at this epoch the estimated temperature of the QGP transition."926 This size was later found to differ from the minimum size a domain should have in order to ensure a production of primordial helium compatible with observations (??)..," This size was later found to differ from the minimum size a domain should have in order to ensure a production of primordial helium compatible with observations \citep{Combes75,Aly1978}."927 The situation is rather different in the Dirac-Milne universe as the timescale of the QGP transition is much longer., The situation is rather different in the Dirac-Milne universe as the timescale of the QGP transition is much longer.928" Since the temperature of the transition is estimated today to be around T~170 MeV (?),, it corresponds to an age of 6x10?sec in the Dirac-Milne universe, which is a factor ~10!° older than in the standard case."," Since the temperature of the transition is estimated today to be around $T \sim 170$ MeV \citep{Schwarz2003}, it corresponds to an age of $6\times 10^{5}\;\rm{sec}$ in the Dirac-Milne universe, which is a factor $\sim 10^{10}$ older than in the standard case."929 This implies that the maximum size to which a domain could possibly grow (assuming the existence of an efficient separation mechanism) is five orders of magnitude higher., This implies that the maximum size to which a domain could possibly grow (assuming the existence of an efficient separation mechanism) is five orders of magnitude higher.930" For this reason, the Dirac-Milne universe is far more weakly constrained by observations than the Omnéss cosmology."," For this reason, the Dirac-Milne universe is far more weakly constrained by observations than the Omnèss cosmology."931 A fundamental example of the modifications induced by a linear scale factor can be seen in the epoch of decoupling of the weak interactions., A fundamental example of the modifications induced by a linear scale factor can be seen in the epoch of decoupling of the weak interactions.932" This example was analyzed extensively in ?,, so we only provide here the main In the standard model, weak decoupling occurs at a temperature T~1MeV."," This example was analyzed extensively in \citet{Simmering98}, so we only provide here the main In the standard model, weak decoupling occurs at a temperature $T\sim 1\;\rm{MeV}$."933" In the Dirac-Milne universe, this decoupling happens at a lower temperature of around T~80keV, because of the slower variation and the lower value of the expansion rate."," In the Dirac-Milne universe, this decoupling happens at a lower temperature of around $T\sim 80 \;\rm{keV}$, because of the slower variation and the lower value of the expansion rate."934 This effect is illustrated in Fig. 2.., This effect is illustrated in Fig. \ref{wkrates}.935 Weak interactions control the n«€p equilibrium., Weak interactions control the $n \leftrightarrow p $ equilibrium.936" At low temperatures, this reaction is limited to the free neutron disintegration (green short-dashed line), but at temperatures sx than 80 keV, the equilibrium between proton and higheneutrons remains possible."," At low temperatures, this reaction is limited to the free neutron disintegration (green short-dashed line), but at temperatures higher than 80 keV, the equilibrium between proton and neutrons remains possible."937 The long-dashed line represent the proton conversion rate as a function of the temperature for the Dirac-Milne universe (red) and thestandard cosmology (blue)., The long-dashed line represent the proton conversion rate as a function of the temperature for the Dirac-Milne universe (red) and thestandard cosmology (blue).938 The analytical expressions for these reaction rates come from ? and ?.., The analytical expressions for these reaction rates come from \citet{Wagoner69} and \citet{Dicus82}.939 Weak interactions decouple when the expansion rate becomes higher than the p©n rate., Weak interactions decouple when the expansion rate becomes higher than the $p \leftrightarrow n$ rate.940 The small difference in the p€n rate between the two cosmologies is caused by a difference in the neutrino background temperature., The small difference in the $p \leftrightarrow n$ rate between the two cosmologies is caused by a difference in the neutrino background temperature.941" As the weak interactions decouple at a temperature of T~80 keV, neutrinos indeed also decouple from the photonbackgroundalso at this temperature, but only after the annihilation of most"," As the weak interactions decouple at a temperature of $T\sim 80\;\rm{keV}$ , neutrinos indeed also decouple from the photonbackgroundalso at this temperature, but only after the annihilation of most"942" AGNs in Section 4 (see, for example, Figure 5)) is a narrow-line/linelessphysical effect, robust beyond the choice of black hole mass estimator.","narrow-line/lineless AGNs in Section 4 (see, for example, Figure \ref{fig:acchistogram}) ) is a physical effect, robust beyond the choice of black hole mass estimator."943" We highlight the range and limitations of the AGN sample in Figure 4,, which shows bolometric luminosities and black hole masses for the broad-line, narrow-line, and lineless AGNs."," We highlight the range and limitations of the AGN sample in Figure \ref{fig:lbolmbh}, which shows bolometric luminosities and black hole masses for the broad-line, narrow-line, and lineless AGNs."944" Objects in the upper left have the highest specific accretion rates, while those in the lower right are weakly accreting AGNs."," Objects in the upper left have the highest specific accretion rates, while those in the lower right are weakly accreting AGNs."945" While the total sample spans 3 orders of magnitude in both luminosity and black hole mass, our narrow-line and lineless AGNs are generally less luminous and more massive than broad-line AGNs."," While the total sample spans 3 orders of magnitude in both luminosity and black hole mass, our narrow-line and lineless AGNs are generally less luminous and more massive than broad-line AGNs."946 The lack of low-mass narrow-line and lineless AGNs is due to the selection limits of the survey: such objects are too faint to be detected in COSMOS., The lack of low-mass narrow-line and lineless AGNs is due to the selection limits of the survey: such objects are too faint to be detected in COSMOS.947" It is suggestive that these higher mass narrow-line and lineless AGNs are at z«1 and are less luminous: this is consistent with "" downsizing,"" with more massive AGNs becoming less active at lower redshift"," It is suggestive that these higher mass narrow-line and lineless AGNs are at $z<1$ and are less luminous: this is consistent with ” downsizing,” with more massive AGNs becoming less active at lower redshift."9482007).. Figure 4 shows that at a given mass or luminosity there are generally all types of AGNs present in our sample., Figure \ref{fig:lbolmbh} shows that at a given mass or luminosity there are generally all types of AGNs present in our sample.949 For this reason we do not expect that the differences between broad-line and AGNs are biased by selected samples from different narrow-line/linelessmasses or luminosities., For this reason we do not expect that the differences between broad-line and narrow-line/lineless AGNs are biased by selected samples from different masses or luminosities.950" In addition, despite the different redshifts of most broad-line and AGNs, we do not expect their differences to be narrow-line/linelesscaused by redshift."," In addition, despite the different redshifts of most broad-line and narrow-line/lineless AGNs, we do not expect their differences to be caused by redshift."951" There is evidence that AGN obscuration properties depend on redshift2009a),, but these AGNs are unobscured."," There is evidence that AGN obscuration properties depend on redshift, but these AGNs are unobscured."952" The AGN central engine, meanwhile, does not change with redshift in terms of ionization parameters2004), spectral energy distributions(Vignali2008),, or metallicity 2010)."," The AGN central engine, meanwhile, does not change with redshift in terms of ionization parameters, spectral energy distributions, or metallicity ."953". Limiting the sample to z< 1,85«log(Mpg)9, or 44«log(Lint)45 does not significantly change the differences between the broad-line and AGN samples seen in Figures 5, 6, 7, narrow-line/linelessor 8."," Limiting the sample to $z<1$, $8.5<\log(M_{BH})<9$, or $44<\log(L_{int})<45$ does not significantly change the differences between the broad-line and narrow-line/lineless AGN samples seen in Figures 5, 6, 7, or 8."954" We estimate errors for each of our specific accretion rates, propagating the errors from both the intrinsic luminosity estimate and the black hole mass estimate."," We estimate errors for each of our specific accretion rates, propagating the errors from both the intrinsic luminosity estimate and the black hole mass estimate."955 Our intrinsic luminosityis subject to three major uncertainties:, Our intrinsic luminosityis subject to three major uncertainties:956values up to about 12LI... most notably the large hole.,"values up to about 12, most notably the large hole."957 The high column density region is fairly well demarcated by the 8 cecontour., The high column density region is fairly well demarcated by the 8 contour.958 Outside of the optical galaxy the velocity dispersion drops. and values in the extended eas are of order 5|.," Outside of the optical galaxy the velocity dispersion drops, and values in the extended gas are of order 5."959 Thus. the velocity dispersions in the in DDO 43 appear to be quite normal.," Thus, the velocity dispersions in the in DDO 43 appear to be quite normal."960 llowever. (he second moment map of the galaxy does not tell the whole story.," However, the second moment map of the galaxy does not tell the whole story."961 To better examine (he kinematics in the kknots and holes. using the ccube we plotted (he spectra of pixels (averaged across a beamwicltli) for several of the knots and the large hole.," To better examine the kinematics in the knots and holes, using the cube we plotted the spectra of pixels (averaged across a beamwidth) for several of the knots and the large hole."962 The spectra of the knots are complex., The spectra of the knots are complex.963 Many exhibit a central double peak. often with smaller peaks. sometimes on both sides of the bright peak. sometimes only ralsible as broad wings on one or both sides.," Many exhibit a central double peak, often with smaller peaks, sometimes on both sides of the bright peak, sometimes only visible as broad wings on one or both sides."964 We fil gaussians to the ILanning-smoothed beam-averaged spectra of the brightest knot and the laree hole: the spectra are shown in Figure 28.., We fit gaussians to the Hanning-smoothed beam-averaged spectra of the brightest knot and the large hole; the spectra are shown in Figure \ref{fig:spectra}.965 For the knot. we were able to successfully fit tree components to 19 of the 28 spectra and two components to six spectra.," For the knot, we were able to successfully fit three components to 19 of the 28 spectra and two components to six spectra."966 Three spectra had large uncertainties to the fits aud so were not included in our analysis., Three spectra had large uncertainties to the fits and so were not included in our analysis.967 For the 19 spectra with good fits to three components. the average dillerence in velocity between the central and (wo side components is ~£15kms...," For the 19 spectra with good fits to three components, the average difference in velocity between the central and two side components is $\sim \pm 15$."968 The average amplitucles are 2. 11. ancl 4 mJv/D going trom hieh to low velocity.," The average amplitudes are 2, 11, and 4 mJy/B going from high to low velocity."969 The central and low velocity components are broad. with an average width of ~17!.," The central and low velocity components are broad, with an average width of $\sim 17$."970. The hieh velocity component is narrower. wilh an average width of roughly half (hat.," The high velocity component is narrower, with an average width of roughly half that."971 The beam-averaged spectra of the laree hole are more complex (han those of the |vs10s. with most being fit by four components.," The beam-averaged spectra of the large hole are more complex than those of the knots, with most being fit by four components."972 Three of the 13 spectra were fit with three components. and one was fit wil two.," Three of the 18 spectra were fit with three components, and one was fit with two."973 Labelling the components 1.+ from high to low central velocities. Components 1 and 3 have average widths of approximately 10!|.. while components 2 and 4 have average widths of around 16|.," Labelling the components 1–4 from high to low central velocities, components 1 and 3 have average widths of approximately 10, while components 2 and 4 have average widths of around 16."974 The average amplitudes are low as expected for a depression in the gas. ranging Irom 2.2 (component 1) to 5.2 mJv/B (component 2).," The average amplitudes are low as expected for a depression in the gas, ranging from 2.2 (component 1) to 5.2 mJy/B (component 2)."975 The components are separated by 18. 1H. and 13 oon average.," The components are separated by 18, 14, and 13 on average."976 The high dispersions associated with this region are reflected in (lese spectra: the eas seems to be somewhat churned up., The high dispersions associated with this region are reflected in these spectra; the gas seems to be somewhat churned up.977 With these complex spectra. we do not believe that we detect expansion in the hole.," With these complex spectra, we do not believe that we detect expansion in the hole."978 The hole itself is not clearly idenüfiable in the channel maps. nor is (here any leature in the position-velocitv diagrams at the location and velocity (7 350-360 1) of the hole that would indicate expansion or even blowout Walter Brinks 1999).," The hole itself is not clearly identifiable in the channel maps, nor is there any feature in the position-velocity diagrams at the location and velocity $\sim$ 350-360 ) of the hole that would indicate expansion or even blowout Walter Brinks 1999)."979where dV. is the volume element ancl We define Aly=OEmy(lknisty8238.Ot. so that AM.M.r=AlonM(7bet)Eσι,"where $dV$ is the volume element and We define $\dot M_0 = \pi G^2 {\rm M}_\odot^2 m_p (1 {\rm km~s^{-1}})^{-3}980= 3.8 \times 10^{14}$, so that $\dot M(n,M,v) = 981\dot M_0 n M^2 (v^2+c_s^2)^{-3/2}$."982 pora elven mass and sound speed. we can define the minimum number density required. for accretion at a rate greater than A: /(ALjAL=).," For a given mass and sound speed, we can define the minimum number density required for accretion at a rate greater than $\dot M$ : $n > n_0 = \dot M c_s^3/(\dot M_0 M^2)$ ."983 Using equations (1)) ane (3)). wecan first carry out the b integration analytically. where rg=CMonM7/M)ος.," Using equations \ref{bondi}) ) and \ref{dndm}) ), wecan first carry out the $v$ integration analytically, where $v_0^2=(\dot M_0 n M^2/\dot M)^{2/3} -c_s^2$."984 The remaining two integrals we compute numerically., The remaining two integrals we compute numerically.985 2mm In Figure 1: we plot the function «EN/d4M for the various phases of the interstellar medium at the solar circle., 2mm In Figure 1 we plot the function $dN/d\dot M$ for the various phases of the interstellar medium at the solar circle.986 The densest eas (GMCs) dominates the highest accretion rates. while the hottest gas dominates the lowest accretion rates.," The densest gas (GMCs) dominates the highest accretion rates, while the hottest gas dominates the lowest accretion rates."987 For the hot LLL accretion is subsonic and p is assumed to be constant. so Mx07.," For the hot HII, accretion is subsonic and $n$ is assumed to be constant, so $\dot M \propto M^2$."988" hus. dN/dAMxM.3777, spanning a range in Al of (AdoΑΛ}. or two decades."," Thus, $dN/d\dot M \propto 989\dot M^{-(1+\gamma)/2}$, spanning a range in $\dot M$ of $(M_2/M_1)^2$, or two decades."990 This is consistent with the numerical spectrum shown in Figure 1., This is consistent with the numerical spectrum shown in Figure 1.991 The other phases have more complicated accretion-rate distributions since a.2 ὃς., The other phases have more complicated accretion-rate distributions since $\sigma_v > c_s$ .992 To compute the expected. total number of black holes in the Galaxy accreting at a rate M. we integrate the Luminosity functions (assumed to be constant) over the gas filling fraction times the number density of black holes as a function of position over the entire Galaxy. Figure2(a) shows N(.M)=fydal’(dNdal) for both black holes anc neutron stars.," To compute the expected total number of black holes in the Galaxy accreting at a rate $\dot M$, we integrate the luminosity functions (assumed to be constant) over the gas filling fraction times the number density of black holes as a function of position over the entire Galaxy, Figure 2(a) shows $N(>{\dot M})=\int_{\dot M}^\infty d\dot M' 993(dN/d\dot M')$ for both black holes and neutron stars."994" For neutron stars. we choose Al=14AL. δ0 per cent with a. =175kms tand14 per cent with a,=100 km t (CordesChernolI 1998). N=5.2.107 "," For neutron stars, we choose $M=1.4 {\rm M}_\odot$ , 86 per cent with $\sigma_v = 175$ km $^{-1}$ and14 per cent with $\sigma_v=700$ km $^{-1}$ (CordesChernoff 1998), $N_\odot = 5.2\times 10^5$ "995The lisht variation of SN 2001V was followed ou 18 nights. starting from f=S days (with respect to B. maxi). extending up tot=162 davs.,"The light variation of SN 2001V was followed on 18 nights, starting from $t = -8$ days (with respect to $B-$ maximum), extending up to $t = +62$ days."996 The applied telescopes aud detectors are listed in Table 1., The applied telescopes and detectors are listed in Table 1.997 All data were collected through standard Jolinsou-Cousis BVRE filters., All data were collected through standard Johnson-Cousins $BVRI$ filters.998" The CCD-frames have been reduced inLRAE"".", The CCD-frames have been reduced in.999. First. the instrumental magnitudes of the SN and the sclected conrparison stars (Fie...) were derived with aperture photometry using the taskdyiphot/apphot.," First, the instrumental magnitudes of the SN and the selected comparison stars (Fig.1) were derived with aperture photometry using the task."1000 The radius of the aperture was 6 pixels (about 2 x FWHAL see Table 1). while the sky level was determined in a 5 pixel- aunulus with iuuer radius of LO pixels.," The radius of the aperture was 6 pixels (about 2 x FWHM, see Table 1), while the sky level was determined in a 5 pixel-wide annulus with inner radius of 10 pixels."1001 Because the SN is located close to the “tip” of NGC 3987. most of the jxels in the annulus were uot affected significaitlv by the ight from tie host ealaxy.," Because the SN is located close to the “tip” of NGC 3987, most of the pixels in the annulus were not affected significantly by the light from the host galaxy."1002 The sky level was «etermined w calculating the modal average (3 X median 2x nean) of the intensities in the annulus., The sky level was determined by calculating the modal average (3 x median $-$ 2 x mean) of the intensities in the annulus.1003" The calculations were done interactively, aud the results were uotted on he screen and inspected visually in order to detect any obvious systematic errors."," The calculations were done interactively, and the results were plotted on the screen and inspected visually in order to detect any obvious systematic errors."1004 The average sky level arouud he SN was always very close to the background. around he comparison stars. no clear svstematic effect could be omc.," The average sky level around the SN was always very close to the background around the comparison stars, no clear systematic effect could be found."1005 Secoud. the whole dataset have been reaeuced Using PSF-photometry (diyiphot/duophot). as also advised by he referee of this paper.," Second, the whole dataset have been re-reduced using PSF-photometry ), as also advised by the referee of this paper."1006 PSF-photometry is a superior yhotometric method if the background strongly varies around the objects. and its removal is complicated.," PSF-photometry is a superior photometric method if the background strongly varies around the objects, and its removal is complicated."1007 Since SN 2001V is somewhat contaminated by its host. the use of PSF-photometry nay be useful to separate the light of the SN from that of he galaxy.," Since SN 2001V is somewhat contaminated by its host, the use of PSF-photometry may be useful to separate the light of the SN from that of the galaxy."1008 This method requires bright stars with hieh S/N to cousruct a good PSF., This method requires bright stars with high S/N to construct a good PSF.1009 Because there are only a few of such stars iu the fiedof NGC 3987. and also the field of view of nost of the telescopes used in this study was quite αμα]. a lot of framies contained ouly lL - 5 stars around the SN.," Because there are only a few of such stars in the field of NGC 3987, and also the field of view of most of the telescopes used in this study was quite small, a lot of frames contained only 4 - 5 stars around the SN."1010" Nevertjicless. the PSF of each frame was determines luteractively,"," Nevertheless, the PSF of each frame was determined interactively."1011 The analytic component of the PSF was approximated by the »üilt-in function iu.daoplhiot. but im most cases it was based on ouly 2-3 stars.," The analytic component of the PSF was approximated by the built-in function in, but in most cases it was based on only 2-3 stars."1012 The funes mace with the Schinidt-telescope (Table 1) contained unch more field objects. l1t the PSF ou these frames showed siguificaut positional dependence.," The frames made with the Schmidt-telescope (Table 1) contained much more field objects, but the PSF on these frames showed significant positional dependence."1013 Therefore. a secoud-order variable PSF model has been constructed for tlese pictures. while cousaut PSF was cletermined for all other frames.," Therefore, a second-order variable PSF model has been constructed for these pictures, while constant PSF was determined for all other frames."1014 Tudividua skv levels were calculated for a] objects. and i6 backeround was subtracted iteratively durius the fitting of 10 PSF indeophot/allstar.," Individual sky levels were calculated for all objects, and the background was subtracted iteratively during the fitting of the PSF in."1015 Then. the residuals were exandued visually on the subtracted frames.," Then, the residuals were examined visually on the subtracted frames."1016" The stars as well as the SN were adequately removed from most yanues, but slight residuals were preseut at the position of ie brightest stars on some frames."," The stars as well as the SN were adequately removed from most frames, but slight residuals were present at the position of the brightest stars on some frames."1017 This was probably caused by the muacertaimty of the PSF due to the small nuuber of PSF-stars., This was probably caused by the uncertainty of the PSF due to the small number of PSF-stars.1018 Differential magnitudes of ον 2001V have been colmputed using the comparison stars labeled in Fie.l (see also Table 2 below)., Differential magnitudes of SN 2001V have been computed using the comparison stars labeled in Fig.1 (see also Table 2 below).1019 After trausforming them to he staudard svsteii (Vinkóetal. 2001)). the brightuess of the SN was calculated from the calibratedΤΠ uaenitudes of cach comparison star (see below).," After transforming them to the standard system \cite{2ke}) ), the brightness of the SN was calculated from the calibrated magnitudes of each comparison star (see below)."1020 Then. he SN maguitudes belonging to the same frame were averaged.," Then, the SN magnitudes belonging to the same frame were averaged."1021 lu order to search for auv systematic effec caused w the reduction. procedure. we have compared the SN uaenitudes from the aperture- and PSE-phoolnetry.," In order to search for any systematic effect caused by the reduction procedure, we have compared the SN magnitudes from the aperture- and PSF-photometry."1022 Fig.2 preseuts the aperture minus PSFAanagnuitudes in all filters as a function of time., Fig.2 presents the aperture minus PSF-magnitudes in all filters as a function of time.1023 It is apparent that most of the differences are within 40.1 mae., It is apparent that most of the differences are within $\pm 0.1$ mag.1024 Naturally. the differences are higher at later phases. when the SN became faiuter. mt there is no visible systematic trend im the data.," Naturally, the differences are higher at later phases, when the SN became fainter, but there is no visible systematic trend in the data."1025 We couclude that boh the aperture- aud PSF-photometry of SN 2001V preseued here is affected by apxoxinatelv the παλιο amount of randoni errors at the ΕΕ. mag level. mainly due to the technical liuutations of the applied iustrumnenuts (lower S/N. smal field of view. very few PSE-stars).," We conclude that both the aperture- and PSF-photometry of SN 2001V presented here is affected by approximately the same amount of random errors at the $\pm 0.1$ mag level, mainly due to the technical limitations of the applied instruments (lower S/N, small field of view, very few PSF-stars)."1026" In order to reduce the ""uncertainties introduced "" the reduction iiethod. the SN magnitudes resulted from both aperture- aud PSF-phoolnetry were averaged. aud these magnitudes were accepted as the final result."," In order to reduce the uncertainties introduced by the reduction method, the SN magnitudes resulted from both aperture- and PSF-photometry were averaged, and these magnitudes were accepted as the final result."1027 At first. the magnitudes of the comparison stays were calibrated via Landolt standards. observed at Calar Alto Observatory with twe 1.2 in Casseerain. under photometric conditions ou Augll. 2001.," At first, the magnitudes of the comparison stars were calibrated via Landolt standards, observed at Calar Alto Observatory with the 1.2 m Cassegrain, under photometric conditions on Aug.11, 2001."1028 The rcliabilitw of this dataset have been checked by using the calibrated photometry of some of the field stars made by the CEA Supernova Croup at the F. L. Whipple Observatory., The reliability of this dataset have been checked by using the calibrated photometry of some of the field stars made by the CfA Supernova Group at the F. L. Whipple Observatory.1029 Because the comparison stars used iu this study. and," Because the comparison stars used in this study, and"1030"If the target photon field is characterized by a peak energy 5, (hen (he maximum injection rate in the blob occurs at energy llowever. we have to note that depending on the slope of the primary 5-ray spectrum. this value can change significantly.","If the target photon field is characterized by a peak energy $\varepsilon$, then the maximum injection rate in the blob occurs at energy However, we have to note that depending on the slope of the primary $\gamma$ -ray spectrum, this value can change significantly."1031 Since the svnchrotron cooling time of these electrons. is very short (compared to both the tvpical time scales for the svstem and the Compton cooling time of electrons). the entire absorbed energy will be immediately released by secondary electrons through the svuchrotvon channel.," Since the synchrotron cooling time of these electrons, is very short (compared to both the typical time scales for the system and the Compton cooling time of electrons), the entire absorbed energy will be immediately released by secondary electrons through the synchrotron channel."1032 In the case of large internal absorption or high bulk Lorentz Iactor. the secondary svuchrotron component has a broad distribution centerecl al The variability time-scale of (he svnchirotron radiation of secondary pairs is determined bv the change of the injection. ie. by the change of primary y-ray component.," In the case of large internal absorption or high bulk Lorentz factor, the secondary synchrotron component has a broad distribution centered at The variability time-scale of the synchrotron radiation of secondary pairs is determined by the change of the injection, i.e. by the change of primary $\gamma$ -ray component."1033 In (he case of small internal opacity ancl assuming (hat. protons are distributed over (he energy interval between 1GeV. and 10*TeV with E7 7-tvpe spectrum. the luminosity of the secondary svnchrotron radiation is estimated as," In the case of small internal opacity and assuming that protons are distributed over the energy interval between $1\,\rm GeV$ and $10^7\, \rm TeV$ with $E^{-2}$ -type spectrum, the luminosity of the secondary synchrotron radiation is estimated as"1034"where ko,(6000) is expressed in mag airmass!, and N(O3) is expressed in DU.","where $k_{\rm{O}_3}(6000)$ is expressed in mag $^{-1}$ , and $_3$ ) is expressed in DU."1035 From this we derive an average O3 column density of 258 DU (the RMS deviation is 14 DU)., From this we derive an average $_3$ column density of 258 DU (the RMS deviation is 14 DU).1036" To cross-check this result, we run an independent analysis of the ozone variability using the satellite data provided by the Ozone Monitor Instrument (OMI) on board of NASAAURA?."," To cross-check this result, we run an independent analysis of the ozone variability using the satellite data provided by the Ozone Monitor Instrument (OMI) on board of NASA."1037. OMI derives the daytime ozone column density by comparing the amount of back-scattered solar radiation in the UV and in the optical., OMI derives the daytime ozone column density by comparing the amount of back-scattered solar radiation in the UV and in the optical.1038" The OMI O; column density over Paranal collected between 2007 and 2009 clearly shows a seasonal trend (see Fig. 8)),"," The OMI $_3$ column density over Paranal collected between 2007 and 2009 clearly shows a seasonal trend (see Fig. \ref{fig:ozone}) ),"1039" with maxima attained around August, September and October (~300 DU), and minima reached in February, March and April (~240 DU)."," with maxima attained around August, September and October $\sim$ 300 DU), and minima reached in February, March and April $\sim$ 240 DU)."1040" The average value during the PARSEC campaign was about 260 DU, which is slightly larger than the value given by the LBLRTM model (~240 DU; see previous section)."," The average value during the PARSEC campaign was about 260 DU, which is slightly larger than the value given by the LBLRTM model $\sim$ 240 DU; see previous section)."1041 A closer inspection to the OMI data shows that the ozone content steadily decreased during the time covered by our observations., A closer inspection to the OMI data shows that the ozone content steadily decreased during the time covered by our observations.1042 The peak-to-peak variation is about25%., The peak-to-peak variation is about.1043". Thisturns into a maximum variation of ~0.01 mag airmass""! at 6000 during the time covered by our observations.", Thisturns into a maximum variation of $\sim$ 0.01 mag $^{-1}$ at $\sim$ 6000 during the time covered by our observations.1044" An inspection of the time evolution of ko,(6000) shows no traces of such a steady decrease, and the fluctuations appear to be dominated by short timescale variations, partially attributable to random errors."," An inspection of the time evolution of $k_{\rm{O}_3}(6000)$ shows no traces of such a steady decrease, and the fluctuations appear to be dominated by short timescale variations, partially attributable to random errors."1045" Molecular oxygen shows two main O» vibrational absorption bands centred at 6870 and7605A,, usually indicated as B and A bands, respectively (Fig. 9))."," Molecular oxygen shows two main $_2$ vibrational absorption bands centred at 6870 and, usually indicated as B and A bands, respectively (Fig. \ref{fig:h2o}) )."1046" Their typical equivalent widths (EW) are ~6A and ~28A,, which make them easily detectable in low resolution spectra."," Their typical equivalent widths (EW) are $\sim$ and $\sim$, which make them easily detectable in low resolution spectra."1047" To quantify the variability of O2 column density, and its effect on the extinction, we have measured the EW of the B band in the red setting spectra."," To quantify the variability of $_2$ column density, and its effect on the extinction, we have measured the EW of the B band in the red setting spectra."1048" The A band is severely affected by fringing, and so no very accurate measurements were possible."," The A band is severely affected by fringing, and so no very accurate measurements were possible."1049" However, the integrated strengths of the two bands are well correlated, as demonstrated by a series of LBLRTM simulations run for different values of the column density N(O2)."," However, the integrated strengths of the two bands are well correlated, as demonstrated by a series of LBLRTM simulations run for different values of the column density $_2$ )."1050" Additionally, they both follow a linear dependency on airmass, down to X—2.5."," Additionally, they both follow a linear dependency on airmass, down to $X$ =2.5."1051" The measurements clearly show the EW airmass dependency, which is well reproduced by the following best fit relation: where EWeg79 is expressed inA."," The measurements clearly show the EW airmass dependency, which is well reproduced by the following best fit relation: where $EW_{6870}$ is expressed in."1052". With the aid of this relationship one can correct the observed values to zenith, and derive the column density using a standard curve of growth procedure."," With the aid of this relationship one can correct the observed values to zenith, and derive the column density using a standard curve of growth procedure."1053 For this purpose we computed a number of LBLRTM models varying N(02) between 8.4x10? cm? and 6.7x10? cm?., For this purpose we computed a number of LBLRTM models varying $_2$ ) between$\times$ $^{23}$ $^{-2}$ and $\times$ $^{24}$ $^{-2}$ .1054" Subsequently we measured the EW of the B band on the output spectra, after convolving them with"," Subsequently we measured the EW of the B band on the output spectra, after convolving them with"1055severe for only one extra parameter.,severe for only one extra parameter.1056 It follows from our decision rule that the Type error rate is quite low., It follows from our decision rule that the Type I error rate is quite low.1057" Indeed, for the standard decisive ratio of e?I—148, the Type error rates are exceedingly small — the decision rule is very conservativeI at these noise ratios."," Indeed, for the standard decisive ratio of $e^5=148$, the Type I error rates are exceedingly small – the decision rule is very conservative at these signal-to-noise ratios."1058" On the other hand, clearly if Ηι is true we will often find values below the critical value and so the power is not large."," On the other hand, clearly if $H_1$ is true we will often find values below the critical value and so the power is not large."1059 Ultimately we can trace this to the width of the prior on the satellite line height; we are too vague about what we are looking for to have high power., Ultimately we can trace this to the width of the prior on the satellite line height; we are too vague about what we are looking for to have high power.1060 This point arises again in the next example., This point arises again in the next example.1061" The utility of the proposed decision rule is summarized in Fig. 7,,"," The utility of the proposed decision rule is summarized in Fig. \ref{figure6},"1062 which shows the power and Type I error rate as a function of decision threshold (the chosen critical evidence ratio) and signal-to-noise ratio., which shows the power and Type I error rate as a function of decision threshold (the chosen critical evidence ratio) and signal-to-noise ratio.1063" This diagram is specific to the problem at hand, but interesting points emerge."," This diagram is specific to the problem at hand, but interesting points emerge."1064" Evidently, the combination of the critical evidence and the signal-to-noise ratio determines where the decision rule places one in the diagram."," Evidently, the combination of the critical evidence and the signal-to-noise ratio determines where the decision rule places one in the diagram."1065 Standard decisive thresholds like e? result in low power and a very small Type I error rate — less than 1/500 with our number of repetitions of the Monte Carlo simulation., Standard decisive thresholds like $e^5$ result in low power and a very small Type I error rate – less than 1/500 with our number of repetitions of the Monte Carlo simulation.1066 This may not be what is needed., This may not be what is needed.1067" For comparison, we also apply a Bayesian Information Criterion (BIC; see e.g. Liddle 2007)."," For comparison, we also apply a Bayesian Information Criterion (BIC; see e.g. Liddle 2007)."1068 In our case this means we pick the model with the smallest value of the normalized sum of squares plus the penalty term In(number of data points) x (number of model parameters)., In our case this means we pick the model with the smallest value of the normalized sum of squares plus the penalty term $\ln$ (number of data points) $\times$ (number of model parameters).1069 The number of data points is the number of spectral channels — evidently this number is somewhat vague as not all channels are equally informative., The number of data points is the number of spectral channels – evidently this number is somewhat vague as not all channels are equally informative.1070" The BIC rule, while offering no choices, sits in a useful place in the diagram for this relatively simple problem and is no worse in power than the evidence ratio."," The BIC rule, while offering no choices, sits in a useful place in the diagram for this relatively simple problem and is no worse in power than the evidence ratio."1071" Finally, we note that different decision rules (for example, accepting Ho if the evidence for it is bigger than the evidence for H1) result in a different diagram."," Finally, we note that different decision rules (for example, accepting $H_0$ if the evidence for it is bigger than the evidence for $H_1$ ) result in a different diagram."1072" For a second example, we consider trying to decide if a line profile is Gaussian (Ho) or Lorentzian (H1)."," For a second example, we consider trying to decide if a line profile is Gaussian $H_0$ ) or Lorentzian $H_1$ )."1073" Here we have The simulation proceeds very much as in the first case, except that we assume the priors are the same for the two models; this"," Here we have The simulation proceeds very much as in the first case, except that we assume the priors are the same for the two models; this"1074from that of the MP.,from that of the MP.1075 Note that the position angle of the AIP raciation is determined by the magnetic Geld direction in the emission region. whereas the position angle of the LEC radiation should reflect the magnetic field orientation in the scattering region.," Note that the position angle of the MP radiation is determined by the magnetic field direction in the emission region, whereas the position angle of the LFC radiation should reflect the magnetic field orientation in the scattering region."1076 Because of the magnetosphere rotation. the ray emitted alone the magnetic Geld makes he angle r/2r; with the local magnetic field. direction (seePetrova2008a.andSect.4above)..," Because of the magnetosphere rotation, the ray emitted along the magnetic field makes the angle $\sim r/2r_L$ with the local magnetic field direction \citep[see][and Sect. 4 above]{p07a}."1077 Thus. the magnetic ield orientations in the scattering and emission regions ciller woσε. the dillerence between the position angles of he original8 and scattered. radiation. being& approximatelyap he same.," Thus, the magnetic field orientations in the scattering and emission regions differ by $\sim r/2r_L$, the difference between the position angles of the original and scattered radiation being approximately the same."1078 For the scattering taking place close to the ligh cvlinder this dillerence is 30°., For the scattering taking place close to the light cylinder this difference is $\sim 30^\circ$.1079 Ht is worthy to. poin out that in our consideration the position angle shift. of he scattered. component roughly equals. its. longitudina separation from the MP., It is worthy to point out that in our consideration the position angle shift of the scattered component roughly equals its longitudinal separation from the MP.1080 Ες is indeed the case for the LEC ol the Crab pulsar (Molfett&LIankins1999) and is believe o be a distinctive feature of the scattered. components in other pulsars., This is indeed the case for the LFC of the Crab pulsar \citep{mh98} and is believed to be a distinctive feature of the scattered components in other pulsars.1081 Lligh percentage of linear polarization is another characteristic feature of the scattered. components., High percentage of linear polarization is another characteristic feature of the scattered components.1082 In case of the scattering in a strong magnetic field. the scatterec radiation is dominated by the waves of ordinary polarization. whose electric vector lies in the plane of the wavevector anc he external magnetic field.," In case of the scattering in a strong magnetic field, the scattered radiation is dominated by the waves of ordinary polarization, whose electric vector lies in the plane of the wavevector and the external magnetic field."1083 In the original radio beam. only he ordinary waves are subject to the scattering below the resonance. whereas both types of waves. the ordinary. an extraordinary ones. undergo equally etlicient. scattering a he harmonies of the gvrofrequencey.," In the original radio beam, only the ordinary waves are subject to the scattering below the resonance, whereas both types of waves, the ordinary and extraordinary ones, undergo equally efficient scattering at the harmonics of the gyrofrequency."1084 In the Crab. as well as in other pulsars. the PR component is known to have almos complete linear polarization.," In the Crab, as well as in other pulsars, the PR component is known to have almost complete linear polarization."1085 Although the percentage. of inear polarization of the LEC is lower. ~40%. it stil exceeds that of the ALD.," Although the percentage of linear polarization of the LFC is lower, $\sim 40\%$, it still exceeds that of the MP."1086 One can expect that the LEC radiation sullers depolarization., One can expect that the LFC radiation suffers depolarization.1087 This can be understood. as follows., This can be understood as follows.1088 A noticeable sweep of the position angle across the LEC (Mollett&Llankins1999) implies that this component is formed. by the radiation coming from somewhat cilferent altitudes in the magnetosphere. which results. from. the scattering of somewhat cdilferent frequencies.," A noticeable sweep of the position angle across the LFC \citep{mh98} implies that this component is formed by the radiation coming from somewhat different altitudes in the magnetosphere, which results from the scattering of somewhat different frequencies."1089 Lone take into account the finite width of the MP. at a fixed pulse longitude within the LEC there should be radiation from cillerent altitucles and. therefore with cillerent position angles.," If one take into account the finite width of the MP, at a fixed pulse longitude within the LFC there should be radiation from different altitudes and therefore with different position angles."1090 The superposition of the waves with dillerent position angles may actually lead to a substantial depolarization of the resultant raciation., The superposition of the waves with different position angles may actually lead to a substantial depolarization of the resultant radiation.1091 The LEC and PR of the Crab. pulsar are. known to exhibit. pronounced. frequeney. evolution (Molfett&Lank-ins 1996)., The LFC and PR of the Crab pulsar are known to exhibit pronounced frequency evolution \citep{mh96}.1092. Phe PR component is significant at the lowest frequencies. the LEC becomes strong at. frequencies ~1 CGlIIz. and at higher frequencies both components vanish.," The PR component is significant at the lowest frequencies, the LFC becomes strong at frequencies $\sim 1$ GHz, and at higher frequencies both components vanish."1093" To analvze the spectral behaviour of the scattered components in our model let us turn to equation (28) and consider the ratio of the scattering elliciencies given that the frequencies of the scattered: radiation are equal. £j,νι."," To analyze the spectral behaviour of the scattered components in our model let us turn to equation (28) and consider the ratio of the scattering efficiencies given that the frequencies of the scattered radiation are equal, $\nu_{1_0}=\nu_{1_s}$."1094 Phen we have νο~GonybengfaePe+t ie. the role of the scattering at the harmonies of the evrofrequeney increases with frequeney. Py/Pyxvf3.," Then we have $\Gamma_s/\Gamma_0\sim(\nu_s\eta/\nu_{1_s}\eta_1)^{1-\alpha}1095(\gamma_0/\theta\gamma)^{2s-4}$, i.e. the role of the scattering at the harmonics of the gyrofrequency increases with frequency, $\Gamma_s/\Gamma_0\propto\nu_{1_s}^{\alpha-1}$."1096" πας, the LEC is expected to dominate at somewhat higher frequencies. which is in accordance with the observed. trend."," Thus, the LFC is expected to dominate at somewhat higher frequencies, which is in accordance with the observed trend."1097 On the way in the magnetosphere. both scattered components. the PR and. LEC. may be further subject to scattering.," On the way in the magnetosphere, both scattered components, the PR and LFC, may be further subject to scattering."1098 Because of magnetosphere rotation their inclination to the ambient magnetic field rapidly. increases with distance. while the magnetic field. strength rapidly decreases.," Because of magnetosphere rotation their inclination to the ambient magnetic field rapidly increases with distance, while the magnetic field strength rapidly decreases."1099 Similarly to the. MI the components may. be involved in the induced transverse scattering in a moderately strong magnetic field. in which case the radiation is scattered backwards (forthegeneraltheoryofthisprocessseePetrova2008b).," Similarly to the MP, the components may be involved in the induced transverse scattering in a moderately strong magnetic field, in which case the radiation is scattered backwards \citep[for the general theory of this process1100see][]{p07b}."1101. The consequences of the backward: scattering of the components will be studied. in detail in à separate paper., The consequences of the backward scattering of the components will be studied in detail in a separate paper.1102 Lt will be shown that this process may give rise to the high-frequency. components in the profile of the Crab pulsar: the backscattering of the PR can account for the LP’. whereas the scattering of the LEC can explain the LECT and 1Ο2.," It will be shown that this process may give rise to the high-frequency components in the profile of the Crab pulsar: the backscattering of the PR can account for the IP', whereas the scattering of the LFC can explain the HFC1 and HFC2."1103 Lt will also be demonstrated that the elieiency of the induced. transverse scattering increases with frequency. and. corresponcdinelv. at high enough frequencies the PLR and LEC vanish. their intensities being almost completely transferred to the backward components.," It will also be demonstrated that the efficiency of the induced transverse scattering increases with frequency and, correspondingly, at high enough frequencies the PR and LFC vanish, their intensities being almost completely transferred to the backward components."1104 The scattering elliciencies given by equations (26) and (27) depend on the intensity of the incident. radio beam and on the number density and the characteristic Lorentz-[actor of the scattering plasma particles., The scattering efficiencies given by equations (26) and (27) depend on the intensity of the incident radio beam and on the number density and the characteristic Lorentz-factor of the scattering plasma particles.1105 All these quantities may show marked: pulse-to-pulse Γιοπατος. so that. the scattering elliciencies may vary as well.," All these quantities may show marked pulse-to-pulse fluctuations, so that the scattering efficiencies may vary as well."1106 I£ the scattering is so strong that the component growth is at the stage of saturation. .r »1. the luctuations of E do not alfect the intensity of the scattered: component significantly.," If the scattering is so strong that the component growth is at the stage of saturation, $x\gg 1$ or $y\gg11071$, the fluctuations of $\Gamma$ do not affect the intensity of the scattered component significantly."1108 On condition that pr~Lory 1. however. even small Ductuations of the scattering ellieiency may [ead o drastic variations of the scattered component.," On condition that $x\sim 1$ or $y\sim 1$, however, even small fluctuations of the scattering efficiency may lead to drastic variations of the scattered component."1109 According o equations. (26)-(27). this condition may be. satisfied in a number of pulsars. which have large enough racio uminosities. strong magnetic fields and short periods.," According to equations (26)-(27), this condition may be satisfied in a number of pulsars, which have large enough radio luminosities, strong magnetic fields and short periods."1110 Such oulsars are believed to exhibit. occasional activity at. the απο longitudes. preceding the MP., Such pulsars are believed to exhibit occasional activity at the pulse longitudes preceding the MP.1111 Namely. these pulsars are expected to show the transient. components with the spectral and polarization properties similar to those known or the PR and LEC of the Crab pulsar.," Namely, these pulsars are expected to show the transient components with the spectral and polarization properties similar to those known for the PR and LFC of the Crab pulsar."1112 Furthermore. the ransient components resulting from the higher-harmonic scattering can also be present in pulsar profiles. in particular. in the Crab pulsar.," Furthermore, the transient components resulting from the higher-harmonic scattering can also be present in pulsar profiles, in particular, in the Crab pulsar."1113 The components of pulsar. profiles outside of the MP. are known to exhibit a number of peculiar. properties. and at the same time the pulse-to-pulse Iuctuations generally testify to a physical relation of these components to the ALD.," The components of pulsar profiles outside of the MP are known to exhibit a number of peculiar properties, and at the same time the pulse-to-pulse fluctuations generally testify to a physical relation of these components to the MP."1114 We believe that the components outside of the MP originate as a result. of induced: scattering of the pulsar radio beam into the background. with different types ofthe components corresponding to cilferent scattering regimes.," We believe that the components outside of the MP originate as a result of induced scattering of the pulsar radio beam into the background, with different types of the components corresponding to different scattering regimes."1115 In the present. paper. we have considered. the magnetized induced scattering olf the spiraling particles. which may be present in the outer magnetosphere of a pulsar.," In the present paper, we have considered the magnetized induced scattering off the spiraling particles, which may be present in the outer magnetosphere of a pulsar."1116 In this case the scattering at the harmonics of the particle &vrofrequeney may be eflicient., In this case the scattering at the harmonics of the particle gyrofrequency may be efficient.1117 Our investigation is aimed at explaining. at least partially. the extremely complex radio emission pattern of the Crab. pulsar.," Our investigation is aimed at explaining, at least partially, the extremely complex radio emission pattern of the Crab pulsar."1118 Lt has been demonstrated. that. the scattering from the first harmonic of the evrofrequency into the state below the resonance can account for the formation, It has been demonstrated that the scattering from the first harmonic of the gyrofrequency into the state below the resonance can account for the formation1119The global. simultaneous fit gives a reduced y7=1.055.,"The global, simultaneous fit gives a reduced $\chi^2_{\nu}$ =1.055."1120 Overall there are 25 free parameters common to all time intervals that. therefore. are constrained by the full 2 10° photon statistics.," Overall there are 25 free parameters common to all time intervals that, therefore, are constrained by the full 2 $^5$ photon statistics."1121 Each time interval has its own set of 8 free parameters. three for the continuum (Ny>. CF and powerlaw normalization) and five for the absorption Fe lines (velocity and individual depth).," Each time interval has its own set of 8 free parameters, three for the continuum $\rm N_{H,2}$, $\rm CF$ and powerlaw normalization) and five for the absorption Fe lines (velocity and individual depth)."1122 Fig., Fig.1123 3. shows the spectra and the resulting fit in two representative time intervals (T8 and T12. as marked in Fig. 2)).," \ref{spec_ind} shows the spectra and the resulting fit in two representative time intervals (T8 and T12, as marked in Fig. \ref{eclipses}) ),"1124 clearly displaying a strong variation of absorption between the two time intervals., clearly displaying a strong variation of absorption between the two time intervals.1125 In the following we will mostly focus on the variations of the column density Ny» and of the Covering Factor (CF) of the partial absorber. whose resulting best-fit values in the various time intervals are shown in Fig.," In the following we will mostly focus on the variations of the column density $\rm N_{H,2}$ and of the Covering Factor $\rm CF$ ) of the partial absorber, whose resulting best-fit values in the various time intervals are shown in Fig."1126 2bb and c. One possible concern is whether these two quantities are degenerate. given the reduced statistics available in each time interval.," \ref{eclipses}b b and c. One possible concern is whether these two quantities are degenerate, given the reduced statistics available in each time interval."1127 We found this not to be the case in any of the time intervals (except for the last interval. which does not require a second absorber).," We found this not to be the case in any of the time intervals (except for the last interval, which does not require a second absorber)."1128 As an example. Fig.," As an example, Fig."1129 4. shows the confidence levels in the Ny> versus CF plane. for the time interval T8: although there is some correlation between the two parameters," \ref{contour} shows the confidence levels in the $\rm N_{H,2}$ versus $\rm CF$ plane, for the time interval T8: although there is some correlation between the two parameters"1130regions of the Galaxy.,regions of the Galaxy.1131 Although the model has some points of similarity with the present study. there are also large differences.," Although the model has some points of similarity with the present study, there are also large differences."1132 The disk in Chandran’s model is assumed to have a uniform rotation and (he gravitational potential is assumed (o correspond (ο a constant background mass densitv (so that the gravitational acceleration increases linearly with radius)., The disk in Chandran's model is assumed to have a uniform rotation and the gravitational potential is assumed to correspond to a constant background mass density (so that the gravitational acceleration increases linearly with radius).1133 In contrast. we assume that the disk is differentiallv rotating and (hat the gravitational potential corresponds to that of a compact mass al the center.," In contrast, we assume that the disk is differentially rotating and that the gravitational potential corresponds to that of a compact mass at the center."1134 ILowever. Chandran considers a compressible eas whereas we simplify our problem by taking (he gas to be incompressible.," However, Chandran considers a compressible gas whereas we simplify our problem by taking the gas to be incompressible."1135 has derived a formula (his eqs. |, has derived a formula (his eqs. [113691] and [92]) for the oscillation frequency when there is no magnetic field and the density contrast parameter j(=+1.,91] and [92]) for the oscillation frequency when there is no magnetic field and the density contrast parameter $\mu =\pm 1$.1137 His results are consistent with our analvtical results for a disk with constant angular velocity (see our eq. 1960)., His results are consistent with our analytical results for a disk with constant angular velocity (see our eq. ]).1138 In particular. the results confirm that the disk vorticity has the effect of stabilizing the modes.," In particular, the results confirm that the disk vorticity has the effect of stabilizing the modes."1139 The authors thank the anonvmous referee for several useful comments., \acknowledgements The authors thank the anonymous referee for several useful comments.1140 LXL's research was supported by NASA through Chandra Postdoctoral Fellowship grant number PFE1-20018 awarded by the Chandra X-ray Center. which is operated by the Smithsonian Astrophysical Observatory for NASA under contract NASS-39073.," LXL's research was supported by NASA through Chandra Postdoctoral Fellowship grant number PF1-20018 awarded by the Chandra X-ray Center, which is operated by the Smithsonian Astrophysical Observatory for NASA under contract NAS8-39073."1141 RN was supported in part by NASA erant. NACG5-10780 and NSF grant. AST 0307433., RN was supported in part by NASA grant NAG5-10780 and NSF grant AST 0307433.1142The broad component of the Bahuer lines iu this τν=0.1511 quasar are bluc-shüfted (~3100 En with respect to NLs.,"The broad component of the Balmer lines in this $z_{\rm NL}=0.1514$ quasar are blue-shifted $\sim 3\,100$ ) with respect to NLs."1143 The line profile is boxy. with no sjeuificaut asvunuetiv.," The line profile is boxy, with no significant asymmetry."1144 The fux ratio is —5. constant over the velocity rauge.," The flux ratio is $\sim 5$, constant over the velocity range."1145 This object was not iucluded inthe analysis by Shenetal.(2010).., This object was not included inthe analysis by \citet{shen10a}.1146 Stratevaetal.(2003). and Bianetal.(2007) listed this source as a DPE., \citet{strateva03} and \citet{bian07} listed this source as a DPE.1147 The bulk of the BLs of this quasars is redshifted (~1200 +)) with respect to NLs.," The bulk of the BLs of this quasars is redshifted $\sim 1\,200$ ) with respect to NLs."1148 The red wine is brighter., The red wing is brighter.1149 The flux ratio is [in the blue wing and around 3 iu the red wine., The flux ratio is $\sim4$ in the blue wing and around 3 in the red wing.1150 This object was labeled asa DPE caudidate by Sheuetal.(2010).., This object was labeled as a DPE candidate by \citet{shen10a}.1151. The line of this source peaks at ~23700 bluewrds of the NLs. aud shows an extended red wing.," The line of this source peaks at $\sim 3\,700$ blue-wards of the NLs, and shows an extended red wing."1152 The hue profile is similar., The line profile is similar.1153 The properties of this quasar are half the wav between the objects with asvuunetric line profiles (0.9... JLIS1)0131) aud the typical DPEs. though this source has not been included in any previous compilation of DPEs.," The properties of this quasar are half the way between the objects with asymmetric line profiles (e.g., J1154+0134) and the typical DPEs, though this source has not been included in any previous compilation of DPEs."1154 The peculiar properties of this object were first reported bx Borosou&Lauer(2009)., The peculiar properties of this object were first reported by \citet{boroson09}.1155. The broad dines show two peaks. one consistent with the rest-frame of the ealaxy as set by NLs. the other significantly bluc-slüfted (~3100m 13).," The broad lines show two peaks, one consistent with the rest-frame of the galaxy as set by NLs, the other significantly blue-shifted $\sim3\,400$ )."1156 Boroson&Lauer(2009) proposed the BUB interpretation for this source.," \citet{boroson09}1157 proposed the BHB interpretation for this source."1158 However. following observations covering the red wing of revealed the presence of a bump in the line wing (Choruock 2010).. a feature conunouly observed in DPEs.," However, following observations covering the red wing of revealed the presence of a bump in the line wing \citep{chornock10}, a feature commonly observed in DPEs."1159emitter or From the spectroscopic point of view. the properties of this source are simular to those of J0927|29[3.," or From the spectroscopic point of view, the properties of this source are similar to those of J0927+2943."1160 The Spectruni presents three sets of lines at two different redshifts: Broad bahuer dines (driving the redshift estimate by the SDSS. pipeline) aud faint narrow lines are detected at τι=0.1993., The spectrum presents three sets of lines at two different redshifts: Broad balmer lines (driving the redshift estimate by the SDSS pipeline) and faint narrow lines are detected at $z_1=0.1993$.1161 Another set of (brigliter) narrow lues is observed at το=0.2263., Another set of (brighter) narrow lines is observed at $z_2=0.2263$.1162 The corresponding velocity shift is ~6600.," The corresponding velocity shift is $\sim11636\,600$."1164 A careful inspection of the SDSS image of this source reveals an extended stellar wing South-wards of the quasar., A careful inspection of the SDSS image of this source reveals an extended stellar wing South-wards of the quasar.1165 If this belongs to the quasar host galaxy. then it would reveal a strongly perturbed morphology.," If this belongs to the quasar host galaxy, then it would reveal a strongly perturbed morphology."1166 Ou the other hand. it could be that this is a superposed galaxy.," On the other hand, it could be that this is a superposed galaxy."1167 Iu this case. since τμ]<<te. the gsalaxw would be iu the backerouud of the quasar.," In this case, since $z_{\rm BL}<z_2$, the galaxy would be in the background of the quasar."1168 This scenario is usually extremely uulikely. eiven the high velocity differences (Dotti&Ruszkowski 2010).," This scenario is usually extremely unlikely, given the high velocity differences \citep{dotti10}."1169. However. the SDSS image reveals the presence 6 τα dich galaxy cluster South-West of the quasar. which may chhance the galaxy deusity on the skv plain by few orders of magnitudes.," However, the SDSS image reveals the presence of a rich galaxy cluster South-West of the quasar, which may enhance the galaxy density on the sky plain by few orders of magnitudes."1170 Follow-up observations aimed at directly measting the redshiftof the stellar wing are needed to fully understand the nature of this source., Follow-up observations aimed at directly measuring the redshiftof the stellar wing are needed to fully understand the nature of this source.1171candidate orothers., or.1172 The and. broad lines of this :xj=0.5929 quasar show a small blue-shüft (~500 Ly with respect to NLs., The and broad lines of this $z_{\rm NL}=0.5929$ quasar show a small blue-shift $\sim 500$ ) with respect to NLs.1173 The line profiles are similar aud do not show anv significant asvuumetry., The line profiles are similar and do not show any significant asymmetry.1174 Our shift estimates are consistent with those reported by Shenetal.(2010).., Our shift estimates are consistent with those reported by \citet{shen10a}.1175 νο the small velocity difference. this source 15 probably a normal quasar (Bounine.Shieldsder2007).," Given the small velocity difference, this source is probably a normal quasar \citep{bonning07}."1176. The Balmer lues of this source show a clear red-shift (~1300kuns1.consistentwiththevaluesreportedinShenetal.2010)..," The Balmer lines of this source show a clear red-shift \citep[$\sim 1\,300$ \kms, consistent with the values reported in][]{shen10a}."1177. The line profiles are svnuuectric., The line profiles are symmetric.1178 The flux ratio is around 3., The flux ratio is around 3.1179 or We presented. the outcome of our automatic and systematic search for massive DIIDs., or We presented the outcome of our automatic and systematic search for massive BHBs.1180 We have found 9 DIIB candidates in the SDSS DR7., We have found 9 BHB candidates in the SDSS DR7.1181 Of these. 5 have already been extensively discussecl in literature.," Of these, 5 have already been extensively discussed in literature."1182 The 1 new candidates are J1012|2613. 1151)0131. J1539|3333 aud JI7L013327.," The 4 new candidates are J1012+2613, J1154+0134, J1539+3333 and J1714+3327."1183 For each one of them a BOB is not the oulv possible explanation: The peculiar spectrum of J1012]|2613 can be explained also as au extreme case of double peaked euütter: J1151L10131 has a too noisy spectrmm to exclude other explanations: J1539|3333 πας be a rare superposition of a quasar and a galaxv: The small shift between broad aud narrow lines im J17111332 (z1300 km 5) does not necessarily imply the presence of a BIB.," For each one of them a BHB is not the only possible explanation: The peculiar spectrum of J1012+2613 can be explained also as an extreme case of double peaked emitter; J1154+0134 has a too noisy spectrum to exclude other explanations; J1539+3333 may be a rare superposition of a quasar and a galaxy; The small shift between broad and narrow lines in J1714+332 $\approx 1\,300$ km $^{-1}$ ) does not necessarily imply the presence of a BHB."1184 A more detailed understanding of the expected spectral features of BUBs and observational follow-ups are needed to confrii or dixuduss the DIID hypothesis for all the 9 candidates presented here., A more detailed understanding of the expected spectral features of BHBs and observational follow-ups are needed to confirm or dismiss the BHB hypothesis for all the 9 candidates presented here.1185 Qur imethod also automatically detected: a umber of other interesting objects with peculiar spectral features:i- | objects show strong asviunetries iu the line profiles. with a peak offset Z2000 (either red- or bluc-wauds) aud a longer wing im the opposite velocity range with no secondary peak.," Our method also automatically detected a number of other interesting objects with peculiar spectral features: 4 objects show strong asymmetries in the line profiles, with a peak offset $\gsim 2\,000$ (either red- or blue-wards) and a longer wing in the opposite velocity range with no secondary peak."1186H- 3 objects have BL properties analogous to what typically observed in DPEs. even if the secondary peak is not prominent.," 3 objects have BL properties analogous to what typically observed in DPEs, even if the secondary peak is not prominent."1187 None of them appeared in the compilation by Stratevaetal.(2003)., None of them appeared in the compilation by \citet{strateva03}.1188Hi- We provide strong evidence of a new class of extreme double-peaked oenmütters. with very broad (FEWIIN-100000 +3) and rather faint emission lines.," We provide strong evidence of a new class of extreme double-peaked emitters, with very broad $>$ 000 ) and rather faint emission lines."1189 The main peak of these lues show huge velocity shifts (>5000 1)) with respect to the NLs.," The main peak of these lines show huge velocity shifts $>5\,000$ ) with respect to the NLs."1190" For a conparison. oulv 5 objects out of 138 in Stratevaetal.(2003). have a shift of the brighter peak of Luger than 50000 ον, and none of them exceed τος Ἐ"," For a comparison, only 5 objects out of 138 in \citet{strateva03} have a shift of the brighter peak of larger than 000 , and none of them exceed 000 ."1191"ν, Note that the vextreme double-peaked enütter"" explanation is possible also for oue of the BITB candidates already discussed m literature 2010).."," Note that the “extreme double-peaked emitter” explanation is possible also for one of the BHB candidates already discussed in literature \citep[J1000+2233,][]{decarli_4c2225}. ."1192As discussed above. the sizes of the known populaloli of both AIBCs and disrupted asteroids. from which these generalizatious are drawn. are extremely sniall. and so the typical caveats associated with siiall-uuuber statistics certainly apply.,"As discussed above, the sizes of the known populations of both MBCs and disrupted asteroids, from which these generalizations are drawn, are extremely small, and so the typical caveats associated with small-number statistics certainly apply."1193 However. for the above reasons. we sugecstfelon] that repeated activity is the least ambiguous and most reliably obtainable indicator available a the current time that comet-like activity is sublimation-driven for a particular object.," However, for the above reasons, we suggest that repeated activity is the least ambiguous and most reliably obtainable indicator available at the current time that comet-like activity is sublimation-driven for a particular object."1194" While ouly two o| the seven currently known comet-like main-belt objects(he, both MIBCs aud disrupted asteroids) have been obxcyved to exhibit recurrent activity to date. we note tha the remaining five objects were discovered ax come-like bodies recently enough that they have not actually vet conipleted. full orbits since their respective disco‘TIES,"," While only two of the seven currently known comet-like main-belt objects, both MBCs and disrupted asteroids) have been observed to exhibit recurrent activity to date, we note that the remaining five objects were discovered as comet-like bodies recently enough that they have not actually yet completed full orbits since their respective discoveries."1195 As such. continued monitoring of all of these objects to search for recurrent activity will be importaut for validating their identi&cation as AIBCs or disrupted asteroids.," As such, continued monitoring of all of these objects to search for recurrent activity will be important for validating their identification as MBCs or disrupted asteroids."1196 We lave conducted a photometric. spectroscopic. ancl ανασα. study of comet-like main-belt asteroid (5096) Scheila. and report the following findings:," We have conducted a photometric, spectroscopic, and dynamical study of comet-like main-belt asteroid (596) Scheila, and report the following findings:"1197with y and help to exploit fully the ability of the XLF data to constrain γ.,with $\gamma$ and help to exploit fully the ability of the XLF data to constrain $\gamma$.1198" To demonstrate the impacts of these datasets, Figure 5 shows the constraints in the €,"" (left panel) and cs,"" (right panel) planes for various subsets of data."," To demonstrate the impacts of these datasets, Figure \ref{fig:datasets} shows the constraints in the $\Omega_{\rm m}, \gamma$ (left panel) and $\sigma_8, \gamma$ (right panel) planes for various subsets of data."1199 The red contours show the constraints from the fzas data (i.e. only cluster data); the green contours show the results from adding SNIa data (i.e. fzas--SNIa); the blue contours from adding BAO to the XLF--rest; and the gold (smallest) contours from adding CMBdata??., The red contours show the constraints from the $f_{\rm gas}$ data (i.e. only cluster data); the green contours show the results from adding SNIa data (i.e. $f_{\rm gas}$ +SNIa); the blue contours from adding BAO to the rest; and the gold (smallest) contours from adding CMB.1200". The left panel of Figure 5 demonstrates again the absence of any strong correlation between (34, and ¥ (see also Section 6.1)).", The left panel of Figure \ref{fig:datasets} demonstrates again the absence of any strong correlation between $\Omega_{\rm m}$ and $\gamma$ (see also Section \ref{sec:growth}) ).

Showing the first 1,200 of 10271 lines. Download the file for the rest.