1: \documentclass[12pt,preprint]{aastex}

2: %\documentclass[letterpaper]{emulateapj}

3: \usepackage{amsmath}

4: %\usepackage[varg]{txfonts}

5: %%\usepackage[mtbold]{mathtime}

6: %% avoid conflicts between aastex and hyperref defs of \url

7: \newcommand{\aastexurl}{}

8: \let\aastexurl\url

9: \let\url\relax

10: \usepackage[hypertex]{hyperref}

11: %\usepackage[all]{hypcap}

12: \makeatletter

13: \renewcommand{\fps@figure}{tp}

14: \makeatother

15: \setkeys{Gin}{width=\linewidth,height=0.9\textheight,keepaspectratio=true}

16: 

17: 

18: \newcommand\TA{\tablenotemark{a}}

19: \newcommand\TB{\tablenotemark{b}}

20: \newcommand\TC{\tablenotemark{c}}

21: \newcommand\TD{\tablenotemark{d}}

22: \newcommand\TE{\tablenotemark{e}}

23: \newcommand\TF{\tablenotemark{f}}

24: \newcommand\nd{\nodata}

25: \newcommand\mcnd{\multicolumn{2}{c}{\nodata}}

26: \newcommand{\thin}{\thinspace}

27: \newcommand{\elecd}{$n_{\rm e}$}

28: \newcommand{\elect}{$T_{\rm e}$}

29: \newcommand{\hb}{H$\beta$}

30: \newcommand{\ha}{H$\alpha$}

31: \newcommand{\fci}{[\ion{C}{1}]}

32: \newcommand{\foiii}{[\ion{O}{3}]}

33: \newcommand{\fniii}{[\ion{N}{3}]}

34: \newcommand{\foi}{[\ion{O}{1}]}

35: \newcommand{\foii}{[\ion{O}{2}]}

36: \newcommand{\fsii}{[\ion{S}{2}]}

37: \newcommand{\fsiii}{[\ion{S}{3}]}

38: \newcommand{\fni}{[\ion{N}{1}]}

39: \newcommand{\fnii}{[\ion{N}{2}]}

40: \newcommand{\fariv}{[\ion{Ar}{4}]}

41: \newcommand{\fcliii}{[\ion{Cl}{3}]}

42: \newcommand{\fcliv}{[\ion{Cl}{4}]}

43: \newcommand{\fneiii}{[\ion{Ne}{3}]}

44: \newcommand{\ffeii}{[\ion{Fe}{2}]}

45: \newcommand{\ffeiii}{[\ion{Fe}{3}]}

46: \newcommand{\ffeiv}{[\ion{Fe}{4}]}

47: \newcommand{\oiii}{\ion{O}{3}}

48: \newcommand{\nitroi}{\ion{N}{1}}

49: \newcommand{\nii}{\ion{N}{2}}

50: \newcommand{\niii}{\ion{N}{3}}

51: \newcommand{\sili}{\ion{Si}{1}}

52: \newcommand{\silii}{\ion{Si}{2}}

53: \newcommand{\oi}{\ion{O}{1}}

54: \newcommand{\oii}{\ion{O}{2}}

55: \newcommand{\ci}{\ion{C}{1}}

56: \newcommand{\cii}{\ion{C}{2}}

57: \newcommand{\nei}{\ion{Ne}{1}}

58: \newcommand{\neii}{\ion{Ne}{2}}

59: \newcommand{\neiii}{\ion{Ne}{3}}

60: \newcommand{\sii}{\ion{S}{2}}

61: \newcommand{\siii}{\ion{S}{3}}

62: \newcommand{\niqi}{\ion{Ni}{1}}

63: \newcommand{\niqii}{\ion{Ni}{2}}

64: \newcommand{\ciii}{\ion{C}{3}}

65: \newcommand{\fciii}{\ion{C}{3}]}

66: \newcommand{\cliii}{\ion{Cl}{3}}

67: \newcommand{\fariii}{[\ion{Ar}{3}]}

68: \newcommand{\fei}{\ion{Fe}{1}}

69: \newcommand{\feii}{\ion{Fe}{2}}

70: \newcommand{\feiii}{\ion{Fe}{3}}

71: \newcommand{\feiv}{\ion{Fe}{4}}

72: \newcommand{\ari}{\ion{Ar}{1}}

73: \newcommand{\ariii}{\ion{Ar}{3}}

74: \newcommand{\ariv}{\ion{Ar}{4}}

75: \newcommand{\hi}{\ion{H}{1}}

76: \newcommand{\hii}{\ion{H}{2}}

77: \newcommand{\di}{\ion{D}{1}}

78: \newcommand{\hei}{\ion{He}{1}}

79: \newcommand{\heii}{\ion{He}{2}}

80: \newcommand{\hp}{H$^+$}

81: \newcommand{\hep}{He$^+$}

82: \newcommand{\hepp}{He$^{2+}$}

83: \newcommand{\tmthree}{10$^{-3}$}

84: \newcommand{\tmfour}{10$^{-4}$}

85: \newcommand{\tmfive}{10$^{-5}$}

86: \newcommand{\ts}{\emph{$t^2$}}

87: \newcommand{\mc}{\multicolumn}

88: 

89: \shorttitle{On the abundance discrepancy problem in {\hii} regions}

90: \shortauthors{Garc\'{\i}a-Rojas \& Esteban}

91: 

92: \received{}

93: \begin{document}

94: 

95: \title{On the abundance discrepancy problem in {\hii} regions.\footnotemark{}}

96: \author{Jorge Garc\'{\i}a-Rojas} 

97: \affil{Instituto de Astrof{\'\i}sica de Canarias, E-38200, La Laguna, Tenerife, Spain}

98: \email{jogarcia@ll.iac.es}

99: 

100: \and

101: 

102: \author{C\'esar Esteban}

103: \affil{Instituto de Astrof{\'\i}sica de Canarias, E-38200, La Laguna, Tenerife, Spain}

104: \email{cel@ll.iac.es}

105: 

106: \begin{abstract} 

107: The origin of the abundance discrepancy --i.e. the fact that abundances derived from recombination lines 

108: are larger than those from collisionaly excited lines-- is one of the key problems in the physics of photoionized nebula. 

109: In this work, we analize and discuss data for a sample of

110: Galactic and extragalactic {\hii} regions where this abundance discrepancy has been determined. 

111: We find that the abundance discrepancy factor (ADF) is fairly constant  

112: and of the order of 2 in all the available sample of {\hii} regions. This is a rather different behaviour than that 

113: observed in planetary nebulae, where the ADF shows a much wider range of values.

114: We do not find correlations between the ADF and the O/H, O$^{++}$/H$^+$ ratios, the ionization degree, 

115: $T_e$(High), $T_e$(Low)/ $T_e$(High), FWHM,

116: and the effective temperature of the main ionizing stars within the observational uncertainties. 

117: These results indicate that whatever mechanism is producing  the abundance

118: discrepancy in {\hii} regions it does not substantially depend on those nebular parameters. On the contrary, 

119: the ADF seems to be slightly dependent on the

120: excitation energy, a fact that is consistent with the predictions of the classical temperature fluctuations paradigm. 

121: Finally, we obtain that $T_e$ values obtained from

122: {\oii} recombination lines in {\hii} regions are in agreement with those obtained from collisionally excited line 

123: ratios, a behaviour that is again different from that

124: observed in planetary nebulae. These similar temperature determinations are in contradiction with the predictions 

125: of the model based on the presence of chemically

126: inhomogeneous clumps but are consistent with the temperature fluctuations paradigm. 

127: We conclude that all the indications suggest that the physical mechanism responsible

128: of the abundance discrepancy in {\hii} regions and planetary nebulae are different.

129: \end{abstract}

130: 

131: \keywords{line:identification. ISM:abundances---H II regions. }

132: 

133: \footnotetext[1]{Based on observations collected at the European Southern

134: Observatory, Chile, proposals number ESO 68.C-0149(A) and ESO 70.C-0008(A).}

135: 

136: \section{Introduction

137: \label{intro}}

138: 

139: {\hii} regions are essential tools for the study of the chemical composition and star formation in the Universe, 

140: specially in the 

141: extragalactic domain. It is necessary to be confident that our traditional techniques for deriving chemical 

142: abundances in 

143: ionized nebulae (based on the analysis of intensity ratios of colisionally excited lines, CELs) provide the real 

144: values. 

145: 

146: Forty years ago, \citet{peimbert67} characterized the temperature structure of a gaseous nebula to a second order 

147: by 

148: two parameters: the average temperature, $T_0$ and the mean square temperature fluctuation, {\ts}. At least, two 

149: independent 

150: methods of determining the electron temperature {\elect} of the nebula --with different weights 

151: in the high and low temperature regions-- are necessary to derive $T_0$ and {\ts}. 

152: In their observations, \citet{peimbert67} found that the temperature derived from the Balmer discontinuity, using 

153: the intensity ratio BJ/$I$({\hb}) where BJ is the value of the Balmer jump, was systematically lower than that 

154: derived from the 

155: {\foiii} 4363/5007 CELs ratio. This result has been corroborated by several subsequent works \citep[see e.g.][and 

156: references 

157: therein]{garciarojasetal06} and we will refer to this problem as the temperature discrepancy.

158: 

159: Later, \citet{peimbertcostero69} claimed that, in the presence of temperature variations over the observed 

160: volume of the nebula, the gaseous abundances derived from the analysis of intensity ratios of CELs 

161: are underestimated if these temperature fluctuations are not considered. 

162: This is due to the strong dependence of the intensity of those lines on the assumed {\elect}. 

163: This possibility is annoying, considering that the analysis of CELs is the standard method for deriving ionic 

164: abundances in 

165: ionized nebulae. On the other hand, intensity ratios of optical recombination lines (ORLs) are almost independent 

166: of the temperature structure 

167: of the nebula, and could be more appropiate to derive the ``real'' abundances of the nebulae. 

168: In the last years, several measurements of ORLs on {\hii} regions and planetary nebulae (PNe) have been done, 

169: and it has been found that 

170: abundance determinations from ORLs are systematically higher than those obtained using CELs, independently of the 

171: ion considered \citep[see e.g.][and references therein]{liu06,garciarojasetal06}. This is usually known as the 

172: abundance discrepancy problem.

173: 

174: Thanks to the fast technological progress --related to the development of large telescopes and better 

175: detectors-- and to the significant improvement in the understanding of the recombination theory for multielectronic 

176: atomic systems, many works have been dedicated to understand the origin of the abundance and temperature 

177: discrepancies. 

178: During the last years, it has been confirmed that both effects are real, and that they are not related neither to systematic 

179: biases in the 

180: observations nor in the basic atomic physics. In addition, it has been found that the two dichotomies can be 

181: intimately related 

182: \citep[see][]{liuetal00}. 

183: 

184: The high surface brightness of the brightest PNe has permitted to detect many ORLs of heavy elements in these 

185: objects. 

186: The pioneering work of \citet{wyse42} in NGC~7009, was the first in calling the attention on the different 

187: O$^{++}$ abundance obtained from {\oii} ORLs and {\foiii} CELs. Later, \citet{allermenzel45} and 

188: \citet{torrespeimbertpeimbert77} also remarked the excess of carbon observed in some PNe when calculating the 

189: C$^{++}$ abundance from {\cii} ORLs. Since the 

190: availability of the $IUE$ satellite, carbon abundances from UV CELs could be determined, obtaining abundances 

191: systematically smaller than those derived from ORLs \citep[see e.g.][]{harringtonetal80, kaler86, rolastasinska94, 

192: peimbertetal95}. During the last decade many efforts have been done to understand the physical conditions under 

193: which the different types of spectral lines are formed in PNe. In particular, the use of medium-large aperture 

194: telescopes, 

195: has allowed that the catalogue of detected {\oii} and {\cii} ORLs in PNe has increased spectacularly in the 

196: last years. The works dedicated to the study of the abundance discrepancy and/or the temperature structure in PNe 

197: are 

198: fairly numerous 

199: \citep[e.g.][and references therein]{garnettdinerstein01, mathisetal98, liuetal01, liuetal06, 

200: yliuetal04b, peimbertetal04, robertsontessigarnett05, tsamisetal04, wessonetal05}. 

201: The data collected in these works, indicate that the {\ts} parameter shows 

202: an extraordinary range of variation in PNe, reaching very high values in some objects. \citet{liuetal06}, found 

203: values 

204: of {\ts} $\sim$ 0.14 and 0.18 in Hf 2-2, from the abundance and temperature discrepancies, respectively. On the 

205: other hand, 

206: \citet{rubinetal02}, from $HST$ image and spectroscopy of PN NGC~7009, found that the {\elect} map of this nebula 

207: was practically uniform and that {\ts} throughout the nebula on the plane of the sky was small ({\ts}$_A$ = 

208: 0.0035). In spite of the low {\ts} value found, it could not completely be discarded the existence of 

209: temperature fluctuations throughout the line of sight that could account for the observed abundance discrepancies in 

210: this object 

211: \citep[see][]{liuetal95}.

212: 

213: The abundance discrepancy is commonly quantified using the so-called abundance discrepancy factor (ADF), that 

214: is defined as:

215: \begin{equation}

216: \label{defadf}

217: {\rm ADF}({\rm X}^{+i})=({\rm X}^{+i}/{\rm H}^{+})_{\rm ORLs}/({\rm X}^{+i}/{\rm H}^{+})_{\rm CELs}. 

218: \end{equation}

219: 

220: \citet{liuetal01} found a correlation between the ADF and the difference between {\elect}({\foiii}) and 

221: {\elect}(Bac). Qualitatively, this correlation is consistent with the existence of temperature fluctuations, 

222: although these authors argue that abundance and temperature discrepancies are qualitatively better explained on the basis of 

223: a chemically inhomogenous or two-phase model.

224: \citet{liuetal00} analized the emission line spectrum of the PN NGC~6153, and concluded that the nebula contains a 

225: component of the 

226: ionized gas previously unknown --cold and very metal-rich--. Later works have supported 

227: this model, sustaining it mainly in the fact that the {\foiii} fine structure far infrared (FIR) lines at 

228: $\lambda\lambda$ 52, 

229: 88 $\mu$m --which have very low excitation energies, of the order of 1000 K, and are almost insensible to the 

230: variations of {\elect}-- 

231: give ionic abundances comparable to the ones obtained from UV and optical CELs. These results are not compatible 

232: with the temperature fluctuations paradigm \citep[see][and references therein]{liu06}, but can be explained by a 

233: two-phase model (see below). 

234: 

235: The observational features that allow us to determine the {\ts} parameter (ORLs, {\hi} recombination continua) are 

236: much more 

237: difficult to be accurately measured in {\hii} regions than in PNe. 

238: The low surface brightness of most of these objects has caused 

239: that, until a few years ago, the {\hii} regions sample with {\ts} determinations available has been limited to a 

240: few Galactic {\hii} regions: the Orion nebula \citep{torrespeimbertetal80, osterbrocketal92, peimbertetal93, estebanetal98, 

241: rubinetal03}, 

242: M8 \citep{peimbertetal93b, estebanetal99b} and M17 \citep{peimbertetal92, estebanetal99a, tsamisetal03}, and to a few 

243: extragalactic {\hii} regions \citep{gonzalezdelgadoetal94, estebanetal02}. 

244: In the last four years, this sample has been extended with the works 

245: of \citet{tsamisetal03} on an additional Galactic {\hii} region (NGC~3576) and 3 extragalactic ones 

246: (1 in the Large Magellanic Cloud, LMC and 2 in the Small Magellanic Cloud, SMC); \citet{apeimbert03} on 30 Doradus; 

247: \citet{apeimbertetal05} on 2 NGC~6822 {\hii} regions; and \citet{lopezsanchezetal06} on the starburst galaxy 

248: NGC~5253. 

249: Aditionally, \citet{odelletal03} made a direct estimation of {\ts} in the plane of the sky of the Orion nebula from 

250: deep WFC/HST images, 

251: and making some geometrical considerations found a value of {\ts}=0.028$\pm$0.006, consistent with previous results \citep[{\ts} between 

252: 0.020 and 0.030, see e.g.][and references therein]{estebanetal04}. 

253: Finally, our group has obtained and analyzed extremely deep 

254: echelle spectra of a sample of 8 Galactic {\hii} regions \citep{estebanetal04, garciarojasetal04, 

255: garciarojasetal05, garciarojasetal06, 

256: garciarojasetal06b}. In 4 of them (M16, M20, S~311 and NGC~3603), {\cii} and {\oii} ORLs were detected, for the 

257: first time; 

258: in other two regions (Orion nebula and NGC~3576) {\neii} ORLs were also detected for the first time. 

259: 

260: On the contrary, some puzzling results for the temperature fluctuations hypothesis have been recently found for some 

261: extragalactic 

262: {\hii} regions. On the one hand, \citet{hageleetal06} have measured {\elect}(Bac) in three compact {\hii} galaxies 

263: of the second data emission release of the SDSS\footnote{Sloan Digital Sky Survey} reobserved with 

264: the 4.2m William Herschel Telescope (WHT), finding {\ts}'s values close to zero for two of the objects, and a 

265: relatively large value ({\ts}$\sim$0.066$\pm$0.026) for the third.

266: On the other hand, \citet{gusevaetal06, gusevaetal07} have made a detailed study of a 

267: sample of 47 low metallicity {\hii} regions in blue compact dwarf (BCD) galaxies 

268: and and 69 emission-line galaxies from the data release 3 of the SDSS and M101. 

269: These authors have measured the Balmer and Paschen jumps with the purpose of 

270: determining {\elect}({\hi}). From Monte Carlo simulations, they fit the spectral energy distribution (SED) 

271: of the galaxies and find that the temperatures in the O$^{++}$ zone --derived from  

272: {\foiii} $\lambda\lambda$(4959 5007)/$\lambda$4363 ratio-- are, considering the dispersion of the data, equal to 

273: the temperature of 

274: the H$^+$ zone, obtained from fitting the Balmer jump and the SED. In any case, the 

275: dispersion of the data does not allow to discard small temperature fluctuations of about $\sim$ 3\% -- 5\%. 

276: 

277: \citet{tsamisetal03} proposed a chemically inhomogeneous 

278: scenario for {\hii} regions similar to that proposed for PNe, but only very recently a physically reasonable model 

279: has been drawn \citep{tsamispequignot05, stasinskaetal07}. 

280: This scenario implies the presence of temperature fluctuations in the nebulae, but its resulting abundance pattern is 

281: not identical to the one found under the standard temperature fluctuations paradigm \citep[see][]{stasinskaetal07}. 

282: Those authors postulate the presence 

283: of a gas component of high metallicity, and consequently low {\elect} embedded in a less dense ambient gas, 

284: with lower metallicity and therefore larger temperature. According to \citet{tsamispequignot05} and \citet{stasinskaetal07}, 

285: this denser component could come from  the resulting material of type II SNe that has not been mixed with the bulk of ISM and 

286: that is in pressure balance with the normal composition ambient gas \citep[an scenario originally proposed by][]{tenoriotagle96}.

287: According to this model, these inclusions would be responsible for most the ORLs emission and they would not emit in CELs due 

288: to their low temperature. 

289: 

290: In this paper, we present a global analysis of some results of our study of a sample of

291: {\hii} regions mainly focussed on discussing the properties and behavior of the abundance discrepancy problem in these objects.

292: 

293: The structure of this paper is as follows: in \S~\ref{datos} we briefly describe the data we have analized. 

294: In \S~\ref{adfvsparam} we explore the behavior of the abundance discrepancy factor (ADF) with respect to some 

295: nebular parameters. In \S~\ref{discusiont2} we present some observational arguments against 

296: the two phase model in {\hii} regions. Finally in \S~\ref{conclu} we briefly draw the conclusions of this work.

297: 

298: \section{The data

299: \label{datos}}

300: 

301: The observations were made in two runs: March 2002 and March 2003 with the Ultraviolet Visual Echelle Spectrograph 

302: (UVES) at the Very 

303: Large Telescope (VLT) Kueyen unit in Cerro Paranal Observatory (Chile). A detailed description of the instrument 

304: and telescope 

305: configuration, as well as of data reduction and analysis, can be found in previous papers 

306: \citep{estebanetal04, garciarojasetal04, garciarojasetal05, garciarojasetal06, garciarojasetal06b, 

307: lopezsanchezetal06}.

308: 

309: With the aim of presenting a homogeneous data set, we have recomputed the physical conditions and ionic and total 

310: abundances of some of our published results for some objects (NGC~3576, Orion nebula, 

311: NGC~3603) using the same atomic data and ionization correction factors (ICF) scheme than for the rest of the sample 

312: nebulae. 

313: Also, we have corrected some misprints ocurring in 

314: previous papers. In Table~\ref{plasma}, we present the physical conditions of the ionized gas 

315: computed for our sample; it can be noted that only 

316: small corrections in the adopted {\elect} and {\elecd} have been done with respect to the previously published 

317: ones. 

318: In Table~\ref{t2}, we present the estimated {\ts} values, which are essentially the same as the previously 

319: published ones. Also, in Table~\ref{rls} we present the compilation of the ionic abundances from ORLs.

320: In Tables~\ref{ionic} and~\ref{total}, we present the recomputed ionic and total abundances from CELs. We have re-calculated 

321: the electron 

322: density for the Orion nebula, NGC~3576 and NGC~3603. These new densities imply new values for the ionic abundances for the most density dependent 

323: ionic abundances i. e. O$^+$, S$^+$, Cl$^{++}$, Fe$^{3+}$ and Ar$^{+3}$. Changes in the 

324: adopted atomic data for O$^+$ and S$^{++}$ \citep[see][]{garciarojasetal05} have affected the determination of the 

325: ionic abundances and physical conditions of NGC~3576 and the Orion nebula. 

326: However, the combined effect of these corrections is small for NGC~3576 and NGC~3603 (less than 0.03 dex). 

327: For the Orion nebula the effect is somewhat more important, reaching a factor of about 0.1 dex for O$^+$.

328: 

329: Finally, we have re-computed He abundances for all the objects adopting a new set of ICFs: for the low ionization objects 

330: we have adopted the ICF by \citet{peimberttorrespeimbert77}, whereas for the high ionization ones we have adopted the ICF by 

331: \citet{peimbertetal92}.

332: 

333: \setcounter{table}{0}

334: \begin{deluxetable}{l@{\hspace{10pt}}l@{\hspace{10pt}}l@{\hspace{10pt}}l@{\hspace{10pt}}l@{\hspace{10pt}}l@{\hspace{10pt}}} 

335: \tabletypesize{\scriptsize}

336: \tablecaption{Plasma diagnostics.

337: \label{plasma}}

338: \tablewidth{0pt}

339: \tablehead{

340: \colhead{Diagnostic} & 

341: \colhead{Line}  & 

342: \colhead{M16}  & 

343: \colhead{M8}  & 

344: \colhead{M17}  & 

345: \colhead{M20}}  

346: \startdata 

347: {\elecd} (cm$^{-3}$)& [N\sc I] ($\lambda$5198)/($\lambda$5200)			& 1100$^{+750}_{-400}$ 	& 1600$^{+750}_{-470}$  & 1200$^{+1250}_{-500}$	     & 560$^{+340}_{-220}$     \\

348: & {\foii} ($\lambda$3726)/($\lambda$3729)					& 1050$\pm$250       	& 1800$\pm$800   	& 480$\pm$150	 	     & 240$\pm$70	       \\

349: & {\fsii} ($\lambda$6716)/($\lambda$6731)					& 1390$\pm$550	       	& 1600$\pm$450        	& 500$\pm$220	       	     & 320$\pm$130	       \\ 

350: & {\ffeiii} 									& 540$^{+>1000}_{-500}$	& 2600$\pm$1450       	& 430$^{+>1000}_{-400}$      & 560$\pm$390	       \\

351: & {\fcliii} ($\lambda$5518)/($\lambda$5538)					& 1370$\pm$1000	       	& 2100$\pm$700        	& 270$^{+630}_{-270}$        & 350$^{+780}_{-350}$     \\ 

352: & {\fariv} ($\lambda$4711)/($\lambda$4740) 					& \nodata	       	& 2450:$^{\rm d}$   	& $>$800		     & \nodata  	       \\

353: & {\elecd} (adopted) 								& 1120$\pm$220	       	& 1800$\pm$350		& 470$\pm$120	 	     & 270$\pm$60	       \\

354: & 										& 			& 			& 			     &  		       \\

355: $T_{\rm e}$ (K)& {\fnii} ($\lambda$6548+$\lambda$6583)/($\lambda$5755)$^{\rm a}$& 8450$\pm$270       	& 8470$\pm$180        	& 8950$\pm$380         	     & 8500$\pm$240	       \\

356: & {\fsii} ($\lambda$6716+$\lambda$6731)/($\lambda$4069+$\lambda$4076)		& 7520$\pm$430        	& 7220$\pm$300        	& 7100$\pm$750         	     & 6950 $\pm$350	       \\      

357: & {\foii} ($\lambda$3726+$\lambda$3729)/($\lambda$7320+$\lambda$7330)$^{\rm a}$ & 8260$\pm$400	       	& 8700$\pm$350        	& 8750$\pm$550         	     & 8275$\pm$400	       \\

358: & $T_{\rm e}$ (Low) 								& 8350$\pm$200	       	& 8500$\pm$150        	& 8870$\pm$300         	     & 8400$\pm$200	       \\

359: & {\foiii} ($\lambda$4959+$\lambda$5007)/($\lambda$4363) 			& 7650$\pm$250       	& 8090$\pm$140        	& 8020$\pm$170         	     & 7800$\pm$300	       \\

360: & {\fariii} ($\lambda$7136+$\lambda$7751)/($\lambda$5192)			& \nodata 	       	& 7550$\pm$420  	& 8380$\pm$570         	     & 8730$\pm$920$^{\rm b}$  \\      

361: & {\fsiii} ($\lambda$9069+$\lambda$9532)/($\lambda$6312) 			& 8430$\pm$450        	& 8600$\pm$300$^{\rm c}$& 8110$\pm$400         	     & 8300$\pm$400	       \\

362: & $T_{\rm e}$ (High) 								& 7850$\pm$220	       	& 8150$\pm$120        	& 8050$\pm$150         	     & 7980$\pm$250	       \\

363: & $T_{\rm e}$ (O~{\sc II+III}) 							& 8180$\pm$300	       	& 8570$\pm$200        	& 8200$\pm$200         	     & 8230$\pm$350	       \\

364: & {\heii} 									& 7300$\pm$350	       	& 7500$\pm$200        	& 7450$\pm$200         	     & 7650$\pm$300	       \\

365: & Balmer discontinuity								& 5450$\pm$820	       	& 7100$\pm$1100       	& \nodata 	       	     & 6000$\pm$1300	       \\ 

366: & Paschen discontinuity								& 7200$\pm$1300	       	& 7750$\pm$900        	& 6500$\pm$1000        	     & 5700$\pm$1300	       \\ \hline

367: \noalign{\smallskip}

368: & & \mc{1}{c}{NGC~3576} & \mc{1}{c}{ORION} & \mc{1}{c}{NGC~3603} & \mc{1}{c}{S~311} \\

369: \noalign{\smallskip}

370: \hline

371: \noalign{\smallskip}

372: {\elecd} (cm$^{-3}$)& {\fni} ($\lambda$5198)/($\lambda$5200)			& \nodata	       	& 1700$\pm$600  	 & 4000:$^{\rm d}$			& 590$^{+260}_{-200}$	\\

373: & {\foii} ($\lambda$3726)/($\lambda$3729)					& 1900$\pm$400         	& 6300$^{+2800}_{-1600}$ & 3000$\pm$1000	 	& 260$\pm$110         \\

374: & {\fsii} ($\lambda$6716)/($\lambda$6731)					& 1300$^{+500}_{-300}$ 	& 6500$^{+2000}_{-1200}$ & 4150$^{+3350}_{-1650}$	& 360$^{+140}_{-120}$   \\ 

375: & {\ffeiii} 									& 3100$\pm$ 1300       	& 9300$\pm$2700  	 & 1330:$^{\rm d}$			& 390$\pm$220		\\

376: & {\fcliii} ($\lambda$5518)/($\lambda$5538)					& 3500$^{+900}_{-700}$ 	& 9400$^{+1200}_{-700}$  & 5600$^{+3900}_{-2400}$	& 550$^{+650}_{-550}$   \\ 

377: & {\fariv} ($\lambda$4711)/($\lambda$4740) 					& 4500$^{+2600}_{-1800}$& 6800$^{+1100}_{-1000}$ & $\leq 4000$  		& \nodata		\\

378: & {\elecd} (adopted) 								& 2300$\pm$300         	& 7800$\pm$600		 & 3400$\pm$850 		& 310$\pm$80		\\

379: & 										&		       	&			 &				&			\\

380: $T_{\rm e}$ (K)&{\fnii} ($\lambda$6548+$\lambda$6583)/($\lambda$5755)$^{\rm a}$	& 8500$\pm$200         	& 10150$\pm$350 	 & 11050$\pm$800	 	& 9500$\pm$250        \\

381: & {\fsii} ($\lambda$6716+$\lambda$6731)/($\lambda$4069+$\lambda$4076)		& 8400$^{+350}_{-600}$ 	& 9050$\pm$800  	 & 11050$^{+3550}_{-2050}$	& 7200$^{+750}_{-600}$\\

382: & {\foii} ($\lambda$3726+$\lambda$3729)/($\lambda$7320+$\lambda$7330)$^{\rm a}$ & 8050$\pm$450         	& 8700$\pm$500	 	 & 12350$\pm$1250	 	& 9800$\pm$600        \\

383: & $T_{\rm e}$ (Low) 								& 8400$\pm$200         	& 9600$\pm$300		 & 11400$\pm$700	 	& 9550$\pm$250        \\

384: & {\foiii} ($\lambda$4959+$\lambda$5007)/($\lambda$4363) 			& 8500$\pm$50	       	& 8300$\pm$40		 & 9060$\pm$200 		& 9000$\pm$200        \\

385: & {\fariii} ($\lambda$7136+$\lambda$7751)/($\lambda$5192)			& 8600$^{+450}_{-350}$ 	& 8300$\pm$400  	 & \nodata			& 8800$^{+700}_{-850}$  \\

386: & {\fsiii} ($\lambda$9069+$\lambda$9532)/($\lambda$6312) 			& 8750$\pm$550         	& 9700$^{+800}_{-1200}$  & 8800 $\pm$500$^{\rm c}$	& 9300$\pm$350$^{\rm c}$\\

387: & $T_{\rm e}$ (High) 								& 8500$\pm$50	       	& 8320$\pm$40		 & 9030$\pm$200 		& 9050$\pm$200         \\

388: & $T_{\rm e}$ (O~{\sc II+III}) 							& 8500$\pm$50	       	& 8430$\pm$130		 & 9600$\pm$200 		& 9600$\pm$450         \\

389: & {\heii} 									& 6800$\pm$400         	& 7950$\pm$200		 & 8480$\pm$200 		& 8750$\pm$500         \\

390: & Balmer discontinuity								& 6650$\pm$750         	& 7900$\pm$600  	 & \nodata			& 9500$\pm$900         \\ 

391: & Paschen discontinuity								& 6700$\pm$900         	& 8100$\pm$1400 	 & 6900$\pm$1100	 	& 8700$\pm$1100	\\ 

392: \enddata

393: \tablenotetext{a}{The recombination contribution to the auroral lines has been taken into account.}

394: \tablenotetext{b}{The [Ar~{\sc iii}] $\lambda$7751 line is severely blended with a telluric line.}

395: \tablenotetext{c}{[S~{\sc iii}] $\lambda$9069 or $\lambda$9530 affected by atmospheric absorption bands.}

396: \tablenotetext{d}{Colons indicate very high uncertainties. These values has not been taken into account in the 

397: adopted average.}

398: \end{deluxetable}

399: 

400: 

401: \setcounter{table}{1}

402: \begin{deluxetable}{l@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}} 

403: \tabletypesize{\scriptsize}

404: \tablecaption{{\ts} parameter.

405: \label{t2}}

406: \tablewidth{0pt}

407: \tablehead{

408: \colhead{Method} & 

409: \colhead{M16}  & 

410: \colhead{M8}  & 

411: \colhead{M17}  & 

412: \colhead{M20}}  

413: \startdata 

414: O$^{\rm +}$ (R/C)	& \nodata	      	& 0.031$\pm$0.017       & 0.109:  	        & 0.032$\pm$0.020       \\

415: O$^{\rm ++}$ (R/C)	& 0.046$\pm$0.007	& 0.045$\pm$0.005       & 0.034$\pm$0.005       & 0.038$\pm$0.016       \\

416: C$^{\rm ++}$ (R/C)	& \nodata		& 0.035$\pm$0.005 	& \nodata 	        & \nodata	       \\

417: Ne$^{\rm ++}$ (R/C)	& \nodata		& \nodata	 	& \nodata 	        & \nodata	       \\

418: He$^{\rm +}$ 		& 0.017$\pm$0.013	& 0.046$\pm$0.009 	& 0.027$\pm$0.014       & 0.017$\pm$0.010       \\

419: Bac/Pac--LEC 		& 0.045$\pm$0.014	& 0.022$\pm$0.015       & 0.035$\pm$0.021       & 0.049$\pm$0.019       \\ 

420: Adopted 		& {\bf 0.039$\pm$0.006}	& {\bf 0.040$\pm$0.004} & {\bf 0.033$\pm$0.005} & {\bf 0.029$\pm$0.007} \\ 

421: \hline

422: \noalign{\smallskip}

423: \mc{1}{c}{Method} & \mc{1}{c}{NGC~3576} & \mc{1}{c}{ORION} & \mc{1}{c}{NGC~3603} & \mc{1}{c}{S~311} \\

424: \noalign{\smallskip}

425: \hline

426: \noalign{\smallskip}

427: O$^{\rm +}$ (R/C)	& \nodata		   & 0.063$\pm$0.029       & \nodata		   & \nodata		   \\

428: O$^{\rm ++}$ (R/C)	& 0.038$\pm$0.006	   & 0.020$\pm$0.002	   & 0.042$\pm$0.009	   & 0.040$\pm$0.008     

429: \\

430: C$^{\rm ++}$ (R/C)	& \nodata		   & 0.039$\pm$0.011	   & \nodata		   & \nodata		   \\

431: Ne$^{\rm ++}$ (R/C)	& 0.036$^{+0.014}_{-0.024}$& 0.034$\pm$0.014	   & \nodata		   & \nodata		   \\

432: He$^{\rm +}$ 		& 0.023$\pm$0.019   	   & 0.022$\pm$0.002	   & 0.032$\pm$0.014	   & 0.034$\pm$0.010     

433: \\

434: Bac/Pac--LEC 		& 0.037$\pm$0.012      	   & 0.016$\pm$0.014	   & 0.056$\pm$0.023	   & 0.009:		   \\ 

435: Adopted 		& {\bf 0.038$\pm$0.009}    & {\bf 0.022$\pm$0.002} & {\bf 0.040$\pm$0.008} & {\bf 0.038$\pm$0.007}\\ 

436: \enddata

437: \end{deluxetable}

438: 

439: \setcounter{table}{2}

440: \begin{deluxetable}{l@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}c@{\hspace{10pt}}} 

441: \tabletypesize{\scriptsize}

442: \tablecaption{Heavy element ionic abundances from ORLs.

443: \label{rls}}

444: \tablewidth{0pt}

445: \tablehead{

446:  & \mc{4}{c}{12 + log(X$^{+i}$/H$^+$)}  \\ 

447: \noalign{\smallskip}

448: \hline

449: \noalign{\smallskip}

450: \colhead{Ion} & 

451: \colhead{M16}  & 

452: \colhead{M8}  & 

453: \colhead{M17}  & 

454: \colhead{M20}}  

455: \startdata 

456: O$^{\rm +}$ 		& \nodata	      	& 8.53$\pm$0.06         & \nodata  	        & 8.62$\pm$0.07       \\

457: O$^{\rm ++}$ 		& 8.30$\pm$0.04		& 8.23$\pm$0.02       	& 8.68$\pm$0.02         & 8.00$\pm$0.18       \\

458: C$^{\rm ++}$ 		& 8.40$\pm$0.03		& 8.30$\pm$0.02 	& 8.68$\pm$0.03 	& 8.18$\pm$0.05	       \\

459: Ne$^{\rm ++}$ 		& \nodata		& \nodata	 	& \nodata 	        & \nodata	       \\

460: \hline

461: \noalign{\smallskip}

462: \mc{1}{c}{Ion} & \mc{1}{c}{NGC~3576} & \mc{1}{c}{ORION} & \mc{1}{c}{NGC~3603} & \mc{1}{c}{S~311} \\

463: \noalign{\smallskip}

464: \hline

465: \noalign{\smallskip}

466: O$^{\rm +}$ 		& \nodata		   & \nodata       	   & \nodata		   & \nodata		   \\

467: O$^{\rm ++}$ 		& 8.62$\pm$0.05	   	   & 8.57$\pm$0.01	   & 8.71$\pm$0.05	   & 8.08$\pm$0.03     \\

468: C$^{\rm ++}$ 		& 8.45$\pm$0.06	   	   & 8.34$\pm$0.02	   & 8.48$\pm$0.07   	   & 8.00$\pm$0.04	   \\

469: Ne$^{\rm ++}$ 		& 7.90$\pm$0.14	 	   & 7.95$\pm$0.09	   & \nodata		   & \nodata		   \\

470: \enddata

471: \end{deluxetable}

472: 

473: 

474: \setcounter{table}{3}

475: \begin{deluxetable}{lcccccccc} 

476: \tabletypesize{\scriptsize}

477: \tablecaption{Heavy element ionic abundances from CELs.

478: \label{ionic}}

479: \tablewidth{0pt}

480: \tablehead{

481:  & \mc{8}{c}{12 + log(X$^{+i}$/H$^+$)}  \\ 

482: \noalign{\smallskip}

483: \hline

484: \noalign{\smallskip}

485:  & \mc{2}{c}{M16}  & 

486: \mc{2}{c}{M8}  & 

487: \mc{2}{c}{M17}  & 

488: \mc{2}{c}{M20} \\

489: \colhead{Ion} & 

490: \colhead{{\ts} = 0.000} & 

491: \colhead{{\ts} = 0.039} & 

492: \colhead{{\ts} = 0.000} & 

493: \colhead{{\ts} = 0.040} & 

494: \colhead{{\ts} = 0.000} & 

495: \colhead{{\ts} = 0.033} & 

496: \colhead{{\ts} = 0.000} & 

497: \colhead{{\ts} = 0.029}}  

498: \startdata 

499: N$^{0}$	   & 6.15$\pm$0.06 	& 6.33$\pm$0.07 &  5.81$\pm$0.05  	& 5.99$\pm$0.06 & 5.57$\pm$0.07 & 5.70$\pm$0.07		     	

500: & 5.90$\pm$0.07 & 6.03$\pm$0.08  \\

501: N$^{+}$	   & 7.71$\pm$0.05 	& 7.88$\pm$0.06 &  7.50$\pm$0.03  	& 7.67$\pm$0.04 & 6.82$\pm$0.10 & 6.94$\pm$0.10		     	

502: & 7.55$\pm$0.04 & 7.67$\pm$0.05  \\

503: O$^{0}$	   & 7.23$\pm$0.05 	& 7.40$\pm$0.06 &  6.40$\pm$0.03  	& 6.57$\pm$0.04 & 6.87$\pm$0.07 & 7.00$\pm$0.07		     	

504: & 6.60$\pm$0.05 & 6.72$\pm$0.06  \\

505: O$^{+}$	   & 8.47$\pm$0.08 	& 8.66$\pm$0.09 &  8.39$\pm$0.06  	& 8.58$\pm$0.07 & 7.84$\pm$0.09 & 7.98$\pm$0.09		     	

506: & 8.46$\pm$0.07 & 8.59$\pm$0.08  \\

507: O$^{++}$   &  7.85$\pm$0.07	& 8.18$\pm$0.10 &  7.86$\pm$0.03  	& 8.18$\pm$0.07 & 8.41$\pm$0.04 & 8.67$\pm$0.06		     	

508: & 7.67$\pm$0.08 & 7.90$\pm$0.10  \\

509: Ne$^{++}$  &  7.01$\pm$0.07	& 7.38$\pm$0.10 &  6.95$\pm$0.05  	& 7.30$\pm$0.07 & 7.64$\pm$0.04 & 7.93$\pm$0.07		     	

510: & 6.55$\pm$0.09 & 6.80$\pm$0.11  \\

511: S$^{+}$	   & 6.32$\pm$0.05 	& 6.49$\pm$0.06 &  5.93$\pm$0.04  	& 6.10$\pm$0.07 & 5.44$\pm$0.05 & 5.56$\pm$0.06		     	

512: & 6.17$\pm$0.05 & 6.29$\pm$0.06  \\

513: S$^{++}$   &  6.84$\pm$0.06	& 7.22$\pm$0.10 &  6.89$\pm$0.03  	& 7.25$\pm$0.07 & 6.90$\pm$0.04 & 7.19$\pm$0.06		     	

514: & 6.79$\pm$0.06 & 7.09$\pm$0.10  \\

515: Cl$^{+}$   &  4.77$\pm$0.05	& 4.91$\pm$0.07 &  4.53$\pm$0.04  	& 4.66$\pm$0.06 & 3.95$^{+0.09}_{-0.12}$& 

516: 4.06$^{+0.09}_{-0.12}$& 4.75$\pm$0.05 & 4.85$\pm$0.07  \\

517: Cl$^{++}$  &  5.04$\pm$0.06	& 5.36$\pm$0.08 &  5.02$\pm$0.04  	& 5.32$\pm$0.06 & 5.04$\pm$0.04 & 5.29$\pm$0.06		     	

518: & 4.99$\pm$0.07 & 5.21$\pm$0.08  \\

519: Cl$^{3+}$  &  \nodata	    	& \nodata	   &  \nodata	     	& \nodata       & 3.15:	     	& 3.35:			     	

520: & \nodata       & \nodata        \\

521: Ar$^{++}$  &  6.25$\pm$0.05	& 6.53$\pm$0.08 &  6.21$\pm$0.03  	& 6.48$\pm$0.05 & 6.35$\pm$0.04 & 6.57$\pm$0.06		     	

522: & 6.17$\pm$0.06 & 6.36$\pm$0.08  \\

523: Ar$^{3+}$  &  3.89$\pm$0.22	& 4.23$\pm$0.23 &  3.69$\pm$0.09  	& 4.01$\pm$0.10 & 4.15$^{+0.12}_{-0.18}$& 

524: 4.42$^{+0.13}_{-0.18}$& 4.01$\pm$0.18 & 4.24$\pm$0.19  \\

525: Fe$^{+}$   &  4.62:	   	& 4.78:	   	&  4.61: 	     	& 4.77:	     	& 4.05:	     	& 4.17:			     	& 

526: 4.51:	     	& 4.62:	      	\\       

527: Fe$^{++}$  &  5.07$\pm$0.04	& 5.41$\pm$0.08 &  5.58$\pm$0.04  	& 5.91$\pm$0.06 & 5.24$\pm$0.06	& 5.51$\pm$0.08		     	

528: & 5.23$\pm$0.10 & 5.47$\pm$0.12  \\

529: Fe$^{3+}$  &  \nodata	   	& \nodata	&  \nodata	     	& \nodata       & \nodata       & \nodata  		     	

530: & \nodata       & \nodata        \\

531: \hline

532: \noalign{\smallskip} 

533: & \multicolumn{8}{c}{12 + log(X$^{+i}$/H$^+$)} \\

534: \noalign{\smallskip}

535: \hline

536: \noalign{\smallskip}

537: & \mc{2}{c}{NGC~3576} & \mc{2}{c}{ORION} & \mc{2}{c}{NGC~3603} & \mc{2}{c}{S~311} \\

538: \noalign{\smallskip}

539: Ion      & {\ts} = 0.000 	& {\ts} = 0.038 	& {\ts} = 0.000   & {\ts} = 0.022 		& {\ts} = 0.000 & {\ts} = 

540: 0.040 & {\ts} = 0.000 & {\ts} = 0.038 				\\

541: \noalign{\smallskip} 

542: \hline

543: \noalign{\smallskip}

544: N$^{0}$	   &  \nodata	     &  \nodata 		& 5.69$\pm$0.06      & 5.76$\pm$0.06	     & 5.55$\pm$0.11 & 

545: 5.65$\pm$0.11 & 5.74 $\pm$ 0.06       & 5.88 $\pm$ 0.07       \\

546: N$^{+}$	   & 7.09$\pm$0.06   & 7.25$\pm$0.07		& 6.97$\pm$0.09      & 7.03$\pm$0.09	     & 6.45$\pm$0.07 & 

547: 6.55$\pm$0.07 & 7.25 $\pm$ 0.05       & 7.38 $\pm$ 0.06       \\

548: O$^{0}$	   & 6.35$\pm$0.04   & 6.51$\pm$0.06		& 6.22$\pm$0.05      & 6.28$\pm$0.05	     & 6.32$\pm$0.09 & 

549: 6.42$\pm$0.09 & 6.74 $\pm$ 0.06       & 6.87 $\pm$ 0.06       \\

550: O$^{+}$	   & 8.21$\pm$0.07   & 8.38$\pm$0.08		& 7.93$\pm$0.15      & 8.00$\pm$0.15	     & 7.42$\pm$0.11 & 

551: 7.52$\pm$0.11 & 8.26 $\pm$ 0.07       & 8.40 $\pm$ 0.08       \\

552: O$^{++}$   & 8.35$\pm$0.03   & 8.63$\pm$0.08		& 8.42$\pm$0.01      & 8.58$\pm$0.03	     & 8.42$\pm$0.05 & 

553: 8.68$\pm$0.08 & 7.81 $\pm$ 0.04       & 8.05 $\pm$ 0.06       \\

554: Ne$^{++}$  & 7.61$\pm$0.09   & 7.91$\pm$0.12		& 7.67$\pm$0.07      & 7.84$\pm$0.07	     & 7.72$\pm$0.05 & 

555: 8.00$\pm$0.08 & 6.81 $\pm$ 0.05       & 7.07 $\pm$ 0.07       \\

556: S$^{+}$	   & 5.75$\pm$0.08   & 5.91$\pm$0.09		& 5.44$\pm$0.06      & 5.50$\pm$0.06	     & 5.08$\pm$0.10 & 

557: 5.17$\pm$0.10 & 6.03 $\pm$ 0.05       & 6.15 $\pm$ 0.06       \\

558: S$^{++}$   & 6.86$\pm$0.10   & 7.17$\pm$0.10		& 6.95$\pm$0.04      & 7.12$\pm$0.05	     & 6.83$\pm$0.04 & 

559: 7.11$\pm$0.09 & 6.68 $\pm$ 0.07       & 6.95 $\pm$ 0.09       \\

560: Cl$^{+}$   & 4.13$\pm$0.08   & 4.26$\pm$0.08		& 3.80$\pm$0.11$^{\rm a}$& 3.85$\pm$0.11$^{\rm a}$& 

561: 3.46$\pm$0.07 & 3.54$\pm$0.07 & 4.56 $\pm$ 0.05       & 4.67 $\pm$ 0.05       \\

562: Cl$^{++}$  & 4.95$\pm$0.06   & 5.21$^{+0.10}_{-0.08}$	& 5.13$\pm$0.02      & 5.28$\pm$0.02	     & 

563: 5.07$\pm$0.05 & 5.31$\pm$0.08 & 4.85 $\pm$ 0.05       & 5.08 $\pm$ 0.05       \\

564: Cl$^{3+}$  & 3.21$\pm$0.07   & 3.42$\pm$0.09		& 3.66$\pm$0.12      & 3.78$\pm$0.12	     & 3.86$\pm$0.04 & 

565: 4.06$\pm$0.07 & \nodata  	     & \nodata  	     \\ 

566: Ar$^{++}$  & 6.34$\pm$0.05   & 6.57$\pm$0.08		& 6.37$\pm$0.05      & 6.50$\pm$0.05	     & 6.35$\pm$0.04 & 

567: 6.56$\pm$0.07 & 6.08 $\pm$ 0.04       & 6.28 $\pm$ 0.06       \\

568: Ar$^{3+}$  & 4.20$\pm$0.07   & 4.48$\pm$0.10		& 4.59$\pm$0.03      & 4.75$\pm$0.04	     & 4.85$\pm$0.06 & 

569: 5.11$\pm$0.08 & 3.42$^{+0.18}_{-0.30}$& 3.66$^{+0.18}_{-0.30}$\\

570: Fe$^{+}$   & 4.61:	     & 4.76:  			& 4.54:	     	     & 4.60:  	     	     & 4.04:	     & 4.13:	     

571: & 4.25:  	     	     & 4.37:   		     \\ 

572: Fe$^{++}$  & 5.57$\pm$0.05   & 5.85$\pm$0.09		& 5.43$\pm$0.06      & 5.59$\pm$0.06	     & 5.24$\pm$0.06 & 

573: 5.50$\pm$0.09 & 5.05 $\pm$ 0.06       & 5.30 $\pm$ 0.08       \\

574: Fe$^{3+}$  & 5.71$^{+0.17}_{-0.29}$& 5.95$^{+0.17}_{-0.29}$& 5.65$^{+0.19}_{-0.30}$ & 5.78$^{+0.19}_{-0.30}$& 

575: \nodata& \nodata	     & \nodata         	     & \nodata	 	     \\ 

576: \enddata

577: \tablenotetext{a}{These values were misprinted in \citet{estebanetal04}.}

578: \end{deluxetable}

579: 

580: 

581: 

582: \setcounter{table}{4}

583: \begin{deluxetable}{lcccccccc} 

584: \tabletypesize{\scriptsize}

585: \tablecaption{Total abundances.

586: \label{total}}

587: \tablewidth{0pt}

588: \tablehead{

589:  & \mc{8}{c}{12 + log(X$^{+i}$/H$^+$)}  \\ 

590: \noalign{\smallskip}

591: \hline

592: \noalign{\smallskip}

593:  & \mc{2}{c}{M16}  & 

594: \mc{2}{c}{M8}  & 

595: \mc{2}{c}{M17}  & 

596: \mc{2}{c}{M20} \\

597: & 

598: \colhead{{\ts} = 0.000} & 

599: \colhead{{\ts} = 0.039} & 

600: \colhead{{\ts} = 0.000} & 

601: \colhead{{\ts} = 0.040} & 

602: \colhead{{\ts} = 0.000} & 

603: \colhead{{\ts} = 0.033} & 

604: \colhead{{\ts} = 0.000} & 

605: \colhead{{\ts} = 0.029}}  

606: \startdata 

607: He	   	& 11.01$\pm$0.02& 10.97$\pm$0.02&  10.87$\pm$0.01	& 10.85$\pm$0.01      	& 10.97$\pm$0.01        & 

608: 10.97$\pm$0.01       	& 10.95$\pm$0.06& 10.92$\pm$0.06\\

609: C$^{\rm a}$	& 8.85$\pm$0.10 & 8.85$\pm$0.10 &  8.61/8.69$\pm$0.09 	& 8.69/8.69$\pm$0.09    & 8.77$\pm$0.04         

610: & 8.77$\pm$0.04        	& 8.66$\pm$0.11 & 8.66$\pm$0.11 \\

611: N	   	& 7.84$\pm$0.06 & 8.07$\pm$0.12 &  7.72$\pm$0.03	& 7.96$\pm$0.06       	& 7.62$\pm$0.12	      	& 

612: 7.87$\pm$0.13        	& 7.67$\pm$0.05 & 7.83$\pm$0.07 \\

613: O	   	& 8.56$\pm$0.07 & 8.78$\pm$0.07 &  8.51$\pm$0.05	& 8.73$\pm$0.05       	& 8.51$\pm$0.04	      	& 

614: 8.75$\pm$0.05        	& 8.53$\pm$0.06 & 8.67$\pm$0.07 \\

615: O$^{\rm b}$	& 8.70$\pm$0.06 & 8.70$\pm$0.06 &  8.71$\pm$0.05	& 8.71$\pm$0.05       	& 8.89$\pm$0.04	      	& 

616: 8.89$\pm$0.04        	& 8.71$\pm$0.09 & 8.71$\pm$0.09 \\

617: O$^{\rm c}$	& 8.81$\pm$0.07 & 8.81$\pm$0.07 &  8.71$\pm$0.04	& 8.71$\pm$0.04       	& 8.76$\pm$0.04	      	& 

618: 8.76$\pm$0.04        	& 8.69$\pm$0.10 & 8.69$\pm$0.10 \\

619: Ne$^{\rm d}$   	& 7.86$\pm$0.15 & 8.08$\pm$0.17 &  7.81$\pm$0.12	& 8.03$\pm$0.13       	& 7.74$\pm$0.07	      	

620: & 8.01$\pm$0.09        	& 7.83$\pm$0.16 & 7.97$\pm$0.18 \\

621: S	   	& 6.96$\pm$0.05 & 7.29$\pm$0.08 &  6.94$\pm$0.03	& 7.28$\pm$0.06       	& 7.01$\pm$0.04	      	& 

622: 7.33$\pm$0.06        	& 6.88$\pm$0.05 & 7.12$\pm$0.09 \\

623: Cl$^{\rm e}$	& 5.23$\pm$0.04 & 5.49$\pm$0.07 &  5.14$\pm$0.04	& 5.41$\pm$0.06       	& 5.08$\pm$0.04	    	

624: & 5.32$\pm$0.06	       	& 5.19$\pm$0.05 & 5.37$\pm$0.06 \\

625: Ar	   	& 6.70$\pm$0.06 & 6.84$\pm$0.08 &  6.52$\pm$0.05	& 6.69$\pm$0.06       	& 6.39$\pm$0.06	      	& 

626: 6.59$\pm$0.07        	& 6.65$\pm$0.07 & 6.70$\pm$0.09 \\

627: Fe$^{\rm f}$	& 5.17$\pm$0.11 & 5.53$\pm$0.13 &  5.69$\pm$0.09	& 6.04$\pm$0.10       	& 5.87$\pm$0.13	      	

628: & 6.22$\pm$0.15        	& 5.31$\pm$0.13 & 5.56$\pm$0.15 \\

629: Fe$^{\rm g}$	& 5.20$\pm$0.06 & 5.51$\pm$0.07 &  5.62$\pm$0.04	& 5.94$\pm$0.06  	& 5.27$\pm$0.06		& 

630: 5.52$\pm$0.08	      	& 5.31$\pm$0.09 & 5.52$\pm$0.10 \\

631: \hline

632: \noalign{\smallskip} 

633: & \multicolumn{8}{c}{12 + log(X$^{+i}$/H$^+$)} \\

634: \noalign{\smallskip}

635: \hline

636: \noalign{\smallskip}

637: & \mc{2}{c}{NGC~3576} & \mc{2}{c}{ORION} & \mc{2}{c}{NGC~3603} & \mc{2}{c}{S~311} \\

638: \noalign{\smallskip}

639:  & {\ts} = 0.000 	& {\ts} = 0.038 	& {\ts} = 0.000   & {\ts} = 0.022 		& {\ts} = 0.000 & {\ts} = 0.040 & 

640: {\ts} = 0.000 & {\ts} = 0.038 				\\

641: \noalign{\smallskip} 

642: \hline

643: \noalign{\smallskip}

644: He	   	& 10.97$\pm$0.03& 10.96$\pm$0.04& 10.95$\pm$0.004 & 10.95$\pm$0.004	& 10.99$\pm$0.01       	& 

645: 10.99$\pm$0.01        & 10.99$\pm$0.02& 10.97$\pm$0.02\\

646: C$^{\rm a}$	& 8.61$\pm$0.08 & 8.61$\pm$0.08 & 8.43$\pm$0.04 & 8.43$\pm$0.04 	& 8.51$\pm$0.07        	& 

647: 8.51$\pm$0.07         & 8.37$\pm$0.10 & 8.37$\pm$0.10\\

648: N	   	& 7.57$\pm$0.06 & 7.81$\pm$0.07 & 7.73$\pm$0.09 & 7.87$\pm$0.09 	& 7.62$\pm$0.13        	& 7.89$\pm$0.14         

649: & 7.43$\pm$0.06 & 7.61$\pm$0.07 \\

650: O	   	& 8.59$\pm$0.03 & 8.82$\pm$0.07 & 8.54$\pm$0.03 & 8.68$\pm$0.04 	& 8.46$\pm$0.05        	& 8.71$\pm$0.07         

651: & 8.39$\pm$0.05 & 8.56$\pm$0.06 \\

652: O$^{\rm b}$	& 8.74$\pm$0.06 & 8.74$\pm$0.06 & 8.71$\pm$0.03 & 8.71$\pm$0.03 	& \nodata	       	&  \nodata	       	

653: & \nodata	& \nodata       \\

654: O$^{\rm c}$	& 8.82$\pm$0.06 & 8.82$\pm$0.06 & 8.67$\pm$0.03 & 8.67$\pm$0.03 	& 8.72$\pm$0.05        	& 

655: 8.72$\pm$0.05         & 8.57$\pm$0.05 & 8.57$\pm$0.05 \\

656: Ne$^{\rm d}$	& 7.85$\pm$0.10 & 8.11$\pm$0.12 & 7.79$\pm$0.07 & 7.94$\pm$0.07 	& 7.76$\pm$0.08        	& 

657: 8.03$\pm$0.11         & 7.79$\pm$0.13 & 7.98$\pm$0.14 \\

658: S	   	& 6.92$\pm$0.10 & 7.23$\pm$0.10 & 7.04$\pm$0.04 & 7.23$\pm$0.04 	& 7.04$\pm$0.05        	& 7.36$\pm$0.08         

659: & 6.77$\pm$0.06 & 7.02$\pm$0.08 \\

660: Cl$^{\rm e}$	& 5.02$\pm$0.05 & 5.27$\pm$0.07 & 5.14$\pm$0.04	& 5.29$\pm$0.04 	& 5.11$\pm$0.05        	& 

661: 5.34$\pm$0.07         & 5.03$\pm$0.06 & 5.22$\pm$0.07 \\

662: Ar	   	& 6.44$\pm$0.06 & 6.65$\pm$0.09 & 6.50$\pm$0.06 & 6.63$\pm$0.06 	& 6.37$\pm$0.06        	& 6.58$\pm$0.08         

663: & 6.43$\pm$0.06 & 6.56$\pm$0.07 \\

664: Fe$^{\rm f}$	& 5.89$\pm$0.21 & 6.23$\pm$0.10 & 5.96$\pm$0.16	& 6.18$\pm$0.14 	& 6.14$\pm$0.16        	& 

665: 6.53$\pm$0.19         & 5.17$\pm$0.11 & 5.44$\pm$0.13 \\

666: Fe$^{\rm g}$	& 5.97$^{+0.11}_{-0.15}$ & 6.22$\pm$0.12& 5.88$\pm$0.12	& 6.02$\pm$0.12 & 5.27$\pm$0.06        	& 

667: 5.52$\pm$0.09         & 5.11$\pm$0.06	& 5.35$\pm$0.08 \\

668: \enddata

669: \tablenotetext{a}{For M8 and Orion nebula, ICFs from C~{\sc ii}] $\lambda$2525 line/ICF from photoionization models 

670: by \citet{garnettetal99}.}

671: \tablenotetext{b}{O$^+$/H$^+$ and O$^{++}$/H$^+$ from O~{\sc i} and O~{\sc ii} ORLs.}

672: \tablenotetext{c}{O$^+$/H$^+$ from [O~{\sc ii}] CELs and {\ts}$>$0.00. O$^{++}$/H$^+$ from O~{\sc ii} ORLs.}

673: \tablenotetext{d}{ICF from \citet{peimbertcostero69} for M17, NGC~3576 and NGC~3603. ICF from 

674: \citet{torrespeimbertpeimbert77} for M16, M8, M20 and S~311.}

675: \tablenotetext{e}{Cl/H=Cl$^+$/H$^+$ + Cl$^{++}$/H$^+$ for M16, M8, M20, and S~311. Cl/H=Cl$^+$/H$^+$ + 

676: Cl$^{++}$/H$^{+}$ + Cl$^{+3}$/H$^+$ for M17, NGC~3576, Orion nebula and NGC~3603.}

677: \tablenotetext{f}{ICF from \citet{rodriguezrubin05} equation (2).}

678: \tablenotetext{g}{Fe/H=Fe$^+$/H$^+$ + Fe$^{++}$/H$^+$ for M16, M8, M17, M20, NGC~3603 and S~311. Fe/H=Fe$^+$/H$^+$ 

679: + Fe$^{++}$/H$^{+}$ + Fe$^{+3}$/H$^+$ for NGC~3576 and Orion nebula.}

680: \end{deluxetable}

681: 

682: 

683: 

684: \section{Behavior of the ADF with respect to some nebular parameters.

685: \label{adfvsparam}} 

686: 

687: The different hypothesis proposed to explain the abundance discrepancy predict the existence or 

688: absence of correlations between the ADF and some nebular properties. 

689: Although these correlations have been exhaustively explored for PNe, 

690: there has not been a systematic study for {\hii} regions. 

691: The aim of this section is to analyze possible correlations between the ADF and different nebular properties 

692: in {\hii} regions (from our data and others available in the literature), with the purpose of discarding possible 

693: systematic 

694: errors and verifying the consistency of some hypothesis proposed to explain the abundance discrepancy. 

695: 

696: Hereafter, the ADF found by 

697: \citet{tsamisetal03} for the SMC N11B {\hii} region has been omitted in most of the figures and the discussion, 

698: because the ADF value for this region is much larger than for the rest of the {\hii} regions and may be 

699: considerably overestimated. For this object, \citet{tsamisetal03} 

700: made an attempt to correct the intensity of the RLs of multiplet 1 of {\oii} for the presence of stellar absorption features, mainly 

701: caused 

702: by B stars --that have a strong absorption {\oii} spectra-- in the area covered by the slit.  

703: In any case, this effect, which can be important in extragalactic objects, can only be corrected in the appropriate 

704: form if the spectral features of the stars are resolved, or if synthetic spectra are available, and this is not the 

705: case 

706: in this particular object. 

707: In fact, it is not possible to distinguish between the nebular emission and stellar absorption features in the 

708: spectra of \citet{tsamisetal03} of N11B. It is important to 

709: emphasize that this effect must be investigated whenever is possible to perform a suitable analysis when computing 

710: abundances from ORLs in extragalactic {\hii} regions. 

711: 

712: For all the fits, we have computed their Spearman's correlation coefficiens. These coefficients are the most appropriate ones

713: for non gaussian distributions, 

714: as well as the two-sided significance of its deviation from zero; the significance is a value in the interval [0.0, 1.0], and a small 

715: value indicates a significant correlation.

716: 

717: One important result of our data compilation for {\hii} regions is that the ADF is fairly constant for these objects and 

718: of the order of 2. This is a crucial difference with the behavior shown by this factor in PNe, where it can vary from values 

719: of 1 up to 20 (see \S~\ref{te_rls} for a further discussion and references).

720: 

721: \begin{figure}

722: \begin{center}

723: \epsscale{1.0}

724: \plotone{f1.eps} 

725: \figcaption{ADF(O$^{++}$) vs. the degree of ionization (upper panel) and O$^{++}$/H$^+$ determined from CELs (lower 

726: panel). 

727: The Galactic H~{\sc ii} regions of our sample 

728: are represented by filled dots; the rest of symbols represents data from the literature for 

729: extragalactic H~{\sc ii} regions: filled triangles: \citet{lopezsanchezetal06}; filled squares: 

730: \citet{tsamisetal03}; empty dots: \citet{estebanetal02}; empty triangles: \citet{apeimbertpeimbert05}; 

731: empty squares: \citet{apeimbert03}.

732: \label{adfoxy2}} 

733: \end{center} 

734: \end{figure} 

735: 

736: In Figure~\ref{adfoxy2} to ~\ref{adftemp} we 

737: have represented the ADF with respect to different nebular parameters, and in Table~\ref{coefcorr} we include the 

738: slope of the fits, their corresponding Spearman's coefficients, $r$, and the significance of the correlation, $P$, 

739: for each of the aforementioned diagrams. As it can be seen in 

740: Table~\ref{coefcorr} the correlation coefficients for all the fits are very low and they have quite poor statistical 

741: significance.

742: As in the rest of the cases, Figure~\ref{adfoxy2} shows that the ADF seems to 

743: be not related to O$^{++}$/H$^+$ nor O$^{+}$/O$^{++}$, at least within the observational uncertainties.

744: The metallicity dependency of the {\ts} parameter was suggested by \citet{garnett92}, who found 

745: that photoionization models could produce temperature fluctuations similar to the observed ones for 

746: nebulae with {\elect} $\leq$ 9000 K, increasing for colder (more metallic) nebulae. 

747: However, the observed {\ts} does not seem to depend on metallicity ,both for {\hii} regions and PNe. 

748: For PNe, several authors have found that the large range of abundance discrepancies found 

749: might be mainly due to 

750: the fact that ADF increases monotonically with metallicity \citep[see e.g.][]{liuetal00, yliuetal04b}. In the 

751: case of {\hii} regions, as it can be observed in Figure~\ref{adf_Orl_cel}, 

752: the ADF seems to remain relatively constant with the O abundance, determined either from CELs or ORLs, and 

753: within the uncertainties.

754: We have also represented the ADF $vs.$ the ratio between high and low ionization temperatures, finding that 

755: there is, again, no correlation within the observational errors (see Figure~\ref{adftemp} and Table~\ref{coefcorr}); 

756: this indicates that large scale 

757: variations of {\elect} 

758: due to natural temperature gradients throughout a nebula, do not seem to be related to the 

759: abundance discrepancy. As it can also been seen in Figure~\ref{adftemp}, we have verified that the ADF does not seem 

760: to depend on the assumed temperature, which discards systematic effects in the determination of abundances from CELs, 

761: at least at the precision of our data.

762: 

763: \begin{figure}

764: \begin{center}

765: \epsscale{1.0}

766: \plotone{f2.eps} 

767: \figcaption{ADF(O$^{++}$) $vs.$ total O abundance derived from CELs (upper panel) 

768: and ORLs (lower panel). 

769: Symbols are the same that in Figure~\ref{adfoxy2}. 

770: \label{adf_Orl_cel}} 

771: \end{center} 

772: \end{figure} 

773: 

774: \begin{figure}

775: \begin{center}

776: \epsscale{1.0}

777: \plotone{f3.eps} 

778: \figcaption{ADF(O$^{++}$) $vs.$ {\elect}(low)/{\elect}(high) (upper panel) and $vs.$ {\elect}(high) (lower panel). 

779: Symbols are the same that in Figure~\ref{adfoxy2}. 

780: \label{adftemp}} 

781: \end{center} 

782: \end{figure} 

783: 

784: 

785: \setcounter{table}{5}

786: \begin{deluxetable}{cccc} 

787: \tablecaption{Parameters of the linear least-squares fits of the data represented 

788: in Figures~\ref{adfoxy2} to~\ref{adftemp}.

789: \label{coefcorr}}

790: \tablewidth{0pt}

791: \tablehead{

792: \colhead{ADF(O$^{++}$) vs. } & 

793: \colhead{Slope} & 

794: \colhead{$r$} &

795: \colhead{$P$}}

796: \startdata 

797: O$^+$/O$^{++}$		& 0.099		& 0.204  	& 0.411   \\

798: 12+log(O$^{++}$)/H$^+$	& $-$0.164	& $-$0.198 	& 0.420   \\

799: 12+log(O/H)$_{CEL}$	& $-$0.049	& 0.064  	& 0.768   \\

800: 12+log(O/H)$_{ORL}$	& 0.088		& 0.283  	& 0.283   \\

801: {\elect}(Low)/{\elect}(High)& $-$0.379	& 0.225 	& 0.378   \\

802: {\elect}(High)		& 2.21$\times$10$^{-6}$& $-$0.113&0.646    \\

803: \enddata

804: \end{deluxetable}

805: 

806: 

807: \subsection{Testing the effect of gas kinematics.} 

808: 

809: Several authors have proposed that mechanical energy deposited by shocks is a 

810: mechanism to increase the heating of nebulae and a possible source of temperature fluctuations 

811: \citep[e.g.][]{peimbert95, peimbertetal91}. 

812: One of the manifestations of the shocks consists of line broadening or line splitting due to the presence of 

813: different kinematical components. Hence, a way to investigate whether the 

814: ADF can be related to excitation by shocks is representing it with respect to the full width at half 

815: maximum (FWHM) of different emission lines. 

816: In Figure~\ref{adf_fwhm} we represent the ADF $vs.$ the FWHM of {\hb}, {\foiii} $\lambda$4959 and 

817: {\fnii} $\lambda$6548. In order to extend the FWHM baseline, we have introduced the points corresponding to 

818: 3 slit positions of NGC~5253 analyzed by \citet{lopezsanchezetal06}. 

819: We have corrected the FWHM by the instrumental width. It is possible to 

820: note that FWHM are much larger for the extragalactic objects due to large scale gas movements, that are 

821: not observed in the Galactic {\hii} regions. With the available data, we cannot find a clear relation 

822: between the ADF and the represented FWHMs. It is obvious that the paucity of points with large FWHM does not 

823: permit to conclude any strong statement about this possible relation. Therefore, it is necessary to increase 

824: the number of extragalactic objects with good determinations of both quantities for making a suitable exploration.

825: 

826: \begin{figure}

827: \begin{center}

828: \epsscale{1.0}

829: \plotone{f4.eps} 

830: \figcaption{ADF(O$^{++}$) $vs.$ FWHM of {\hb} 

831: (upper panel), [O~{\sc iii}] $\lambda$4959 (medium panel) and [N~{\sc ii}] $\lambda$6548 (lower panel). 

832: Symbols are the same that in Figure~\ref{adfoxy2}. 

833: \label{adf_fwhm}} 

834: \end{center} 

835: \end{figure} 

836: 

837: \subsection{Testing the effect of dust grain heating.} 

838: 

839: Another possible effect that can affect the ADF was proposed by \citet{stasinskaszczerba01} and consists of 

840: photoelectric heating of dust grains in the zones near the ionizing star of a PN that increase {\elect} in the 

841: inner zones producing large {\ts} values. This is the case if the nebula is inhomogeneous in density, because the 

842: fluctuation in the ionization parameter induced by the variation of density, produce {\ts} due to differential dust and hydrogen 

843: heating. In the case of an homogeneous density, dust produces a temperature gradient but not real temperature fluctuations 

844: \citep{stasinskaszczerba01}. 

845: A way to verify if the effect of dust grains is significant in the observed ADF is to look for a correlation between the effective 

846: temperature, $T_{eff}$, of the ionizing star of the nebula --that gives an indication of the hardness of the radiation field-- 

847: and the ADF found in the associated nebula. 

848: \citet{robertsontessigarnett05} obtained that for PNe, the ADF was not correlated with $T_{eff}$ of the 

849: ionizing star. For {\hii} regions this is a complicated task because they are 

850: generally ionized by an OB association. In Table~\ref{adf_tipo} we include the $T_{eff}$ of the main ionizing star 

851: (the hottest one) of each studied {\hii} region, and it is clear that such quantity seems to be not correlated with the ADF of the 

852: associated nebula.

853: 

854: \setcounter{table}{6}

855: \begin{deluxetable}{cccc} 

856: \tablecaption{Comparison of the ADF with the spectral type of the main ionizing sources of each of our sample 

857: {\hii} regions.

858: \label{adf_tipo}}

859: \tablewidth{0pt}

860: \tablehead{

861: \colhead{Object$^{\rm a}$} & 

862: \colhead{Star/Cluster} & 

863: \colhead{Spectral type$^{\rm b}$} & 

864: \colhead{ADF(O$^{++}$)}}  

865: \startdata 

866: NGC~3603& NGC~3603/HST-38/40/A2/16& O3V		& 1.9  \\

867: 	& NGC~3603/HST-42/A3	& O2-3III	&   	\\

868: M16     & HD168075 		& O4V		& 2.8  \\

869: M17	& Kleinmann's star	& O4V		& 2.1  \\

870: M20	& HD164492		& O6V		& 2.1  \\

871: Ori\'on & $\theta^1$Ori C	& O6peV		& 1.4  \\

872: S~311	& HD~64315		& O6eV		& 1.9	\\

873: M8   	& H36 			& O7.5V		& 2.0  \\

874: NGC~3576& Obscured cluster 	& OB association & 1.8  \\

875: \enddata

876: \tablenotetext{a}{Ordered from earlier to later spectral type.}

877: \tablenotetext{b}{From the Galactic O star catalog by \citet{maizapellanizetal04}.}

878: \end{deluxetable}

879: 

880: \subsection{The ADF and recombination continuum/CELs temperature difference.}

881: 

882: The standard temperature fluctuations hypothesis predicts that the ADF 

883: is related to the difference between {\elect}(CELs) and {\elect}({\hi}) \citep{peimbertcostero69, 

884: torrespeimbertetal80}. 

885: Indeed, \citet{liuetal00} found a strong correlation between these two discrepancies from data for PNe that cannot 

886: be reproduced by the temperature fluctuation paradigm. The results 

887: of our work seem to indicate that there is also a slight correlation for {\hii} regions, but with a flatter 

888: slope than 

889: for PNe (see Figure~\ref{tfl_adf}). 

890: In Figure~\ref{tfl_adf} we have also included the results obtained for 2 extragalactic {\hii} regions: 

891: 30 Doradus \citep{apeimbert03} and NGC~2363\footnote{For this object the ADF has been obtained from the analysis by 

892: \citet{estebanetal02}, and temperatures were obtained from \citet{gonzalezdelgadoetal94}. 

893: We have included these data because they correspond to the same slit position, covering a very 

894: similar volume 

895: of nebula.} \citep{gonzalezdelgadoetal94, estebanetal02}. 

896: In the case of {\hii} regions, contrary to what happens in PNe, the relation between ADF and the temperature 

897: differences seem to be consistent with the standard temperature fluctuations hypothesis, 

898: and they are translated into moderate and similar values of the {\ts} parameter, without reaching the extreme 

899: values found in PNe. 

900: The fit obtained for our data is: 

901: 

902: \begin{eqnarray} 

903: \log[{\rm ADF}({\rm O}^{++}) ] & = & (0.184\pm0.022)+(0.609\pm0.208) \\ 

904: & & \times10^{-4}[T_e({\rm [ O~III]})-T_e({\rm H~I}) ], 

905: \end{eqnarray} 

906: 

907: \noindent with a correlation coefficient of $r$= 0.575. 

908: 

909: \begin{figure}

910: \begin{center}

911: \epsscale{1.0}

912: \plotone{f5.eps} 

913: \figcaption{ADF(O$^{++}$) $vs.$ the difference between {\elect}(CELs) and {\elect}(H~{\sc i}). 

914: Solid line corresponds to the fit found by \citet{liuetal00} for PNe, whereas pointed line corresponds 

915: to the fit to our data of {\hii} regions. We have added the data of two extragalactic H~{\sc ii} regions: 30 

916: Doradus 

917: \citep{apeimbert03} and NGC~2363 \citep{gonzalezdelgadoetal94, estebanetal02}. 

918: \label{tfl_adf}} 

919: \end{center} 

920: \end{figure} 

921: 

922: Finally, as we previously pointed out \citep[see][]{garciarojasetal06b}, we do not find the correlation between 

923: the ADF and the electron density, {\elecd}, obtained by \citet{robertsontessigarnett05} for a PNe sample. 

924: 

925: \section{Objections to the two phase model.

926: \label{discusiont2}} 

927: 

928: The fact that {\hii} regions as well as PNe are photoionized nebulae does not imply that photoionization is

929: responsible for all the observed properties of these objects. A common argument against temperature fluctuations in 

930: {\hii} regions is based on the difficulties of photoionization models to reproduce {\ts} values of the order of those 

931: estimated for most objects. 

932: However, photoionization models might not be sufficiently realistic, and/or 

933: additional energy sources could be necessary to explain the discrepancies between observations and model 

934: predictions 

935: \citep[see][and references therein]{viegas02}. Several authors have proposed the existence of 

936: a plasma component --colder and metal-rich-- (the so-called two-phase model) to explain the dichotomy between 

937: abundance and temperature determinations 

938: in PNe, where the standard temperature and/or density fluctuations hypothesis do not seem to explain the available observations 

939: \citep{liuetal01, tsamisetal04}. 

940: As we have already commented in \S~\ref{intro}, \citet{tsamispequignot05} have proposed a similar explanation for 

941: the abundance discrepancy in {\hii} regions although there are not direct observational evidences of the colder 

942: and metal-rich inclusions they propose\footnotetext{It should be remarked that the hypothesis involved in two phase scenario 

943: are in the classical temperature fluctuations paradigm are not of the same order, since the two phase scenario actually gives a 

944: natural explanation for the origin of the temperature fluctuations.}. 

945: In the following, we are going to present some objections to this model, based on the results obtained in this 

946: work.

947: 

948: In Table~\ref{adfcomp} we present values of the ADF computed for different ions in Galactic and extragalactic 

949: {\hii} regions. 

950: The table includes all the data available up to date, including our own results and others from the literature. 

951: As we commented before, the ADF for O$^{++}$ is quite similar in all the objects --Galactic or 

952: extragalactic-- and 

953: also similar to the ADFs for the other ions: C$^{++}$, O$^+$, and Ne$^{++}$. Comparing the results for {\hii} 

954: regions of 

955: different galaxies, 

956: we find that the ADF seems to be independent of the morphological type, mass, metallicity or even the star 

957: formation history 

958: of the host galaxy \citep[see also][]{lopezsanchezetal06}. In the case of the Galactic objects, the similarity of 

959: the ADFs suggests 

960: that the process that produces this phenomenon is independent of the conditions and properties of the Galactic 

961: disk, 

962: at least along the range of Galactocentric distances covered by our sample.  \citet{lopezsanchezetal06} have 

963: discussed this fact 

964: in a paper that presents the ADF values for several zones of NGC 5253, a dwarf Wolf-Rayet galaxy. 

965: 

966: \setcounter{table}{7}

967: \begin{deluxetable}{cccccccc} 

968: \tabletypesize{\scriptsize}

969: \tablecaption{Observed ADFs in Galactic and extragalactic H~{\sc ii} regions.

970: \label{adfcomp}}

971: \tablewidth{0pt}

972: \tablehead{

973: \colhead{ID} & 

974: \colhead{{\elecd} (cm$^{-3}$)} & 

975: \colhead{{\elect}({\foiii}) (K)}  &

976: \colhead{ADF(O$^{++}$)}  &

977: \colhead{ADF(O$^{+}$)}  &

978: \colhead{ADF(C$^{++}$)$^{\rm a}$}  &

979: \colhead{ADF(Ne$^{++}$)}  &

980: \colhead{Ref.$^{\rm b}$}  \\ 

981: \noalign{\smallskip} 

982: \hline

983: \noalign{\smallskip} 

984: \mc{8}{c}{Galactic H~{\sc II} regions}}

985: \startdata 

986: M16     & 1120 			& 7650			& 2.8		& \nodata      	& \nodata      	& \nodata      	& 1  	\\

987: M8     	& 1800 			& 8150			& 2.0		& 1.4	       	& 1.8     	& \nodata    	& 1 	\\

988:      	& 1750 			& 8050			& 2.0		& \nodata	& 2.1		& \nodata	& 2	\\

989: M17    	& 470			& 8050			& 2.1		& 4.2:$^{\rm c}$& \nodata      	& \nodata     	& 1 	\\

990:     	& 860, 520		& 8120, 8210		& 1.8, 2.2	& \nodata	& \nodata	& \nodata	& 3	\\

991:    	& 600--1500		& 8200			& 2.1		& \nodata	& \nodata	& \nodata	& 4	\\

992: M20    	& 270			& 7980			& 2.1		& 1.5	       	& \nodata      	& \nodata      	& 1 	\\

993: NGC~3576& 2300			& 8500			& 1.8		& \nodata      	& \nodata      	& 2.0      	& 1 	\\

994: 	& 1300--2700		& 8850			& 1.8		& \nodata	& \nodata	& \nodata	& 4	\\

995: Orion	& 7800			& 8320			& 1.4		& 1.6$^{\rm c}$	& 1.9          	& 1.9	     	& 1 	\\

996: 	& 4000, 5700		& 8300, 8350		& 1.3, 1.5	& \nodata      	& 2.5, 2.2	& \nodata      	& 5	\\

997: NGC~3603& 3400			& 9030			& 1.9		& \nodata      	& \nodata      	& \nodata      	& 1 	\\

998: S~311	& 310			& 9050			& 1.9		& \nodata      	& \nodata      	& \nodata      	& 1 	\\

999: \hline

1000: \noalign{\smallskip} 

1001: \multicolumn{8}{c}{Extragalactic H~{\sc II} regions} \\ 

1002: \noalign{\smallskip} 

1003: \hline

1004: \noalign{\smallskip} 

1005: LMC 30 Dor& 370--1800		& 10100			& 2.0--2.7	& \nodata	& \nodata     & \nodata       & 4     	\\

1006: 	& 316			& 9950			& 1.6		& \nodata	& \nodata     & \nodata       & 6     	\\

1007: LMC N11B& 80--1700		& 9400			& 4.9--8.2	& \nodata	& \nodata     & \nodata       & 4     	\\

1008: SMC N66 & 50--3700		& 12400			& 2.3		& \nodata	& \nodata     & \nodata       & 4     	\\

1009: NGC~604	& $\le$ 100		& 8150			& 1.6		& \nodata	& \nodata     & \nodata       & 7     	\\

1010: NGC~2363& 360--1200		& 15700			& 2.2		& \nodata	& \nodata     & \nodata       & 7     	\\

1011: NGC~5461& 300			& 8600			& 1.9		& \nodata	& \nodata     & \nodata       & 7     	\\

1012: NGC~5471& 220--1150		& 14100			& 1.6		& \nodata	& \nodata     & \nodata       & 7     	\\

1013: NGC~6822 V& 175			& 11900			& 1.9		& \nodata	& \nodata     & \nodata       & 8     	\\

1014: NGC~5253 A& 580			& 12100			& 1.8		& \nodata	& 2.6	      & \nodata       & 9     	\\

1015: NGC~5253 B& 610			& 12030			& 1.5		& \nodata	& 2.1	      & \nodata       & 9     	\\

1016: NGC~5253 C& 370			& 10810			& 1.9		& \nodata	& 2.1	      & \nodata       & 9     	\\

1017: \enddata

1018: \tablenotetext{a}{CELs abundances for C$^{++}$ were obtained from $IUE$ data of C~{\sc iii}] 

1019: $\lambda\lambda$1906+1909 line. 

1020: \citep[][for M8, the Orion nebula and NGC~5253, respectively]{peimbertetal93b, walteretal92, kobulnickyetal97}.}

1021: \tablenotetext{b}{1) This work; 2) \citet{estebanetal99b}; 3) \citet{estebanetal99a}; 4) \citet{tsamisetal03}; 5) 

1022: \citet{estebanetal98}; 

1023: 6) \citet{apeimbert03}; 7) \citet{estebanetal02}; 8) \citet{apeimbertetal05}; 9) \citet{lopezsanchezetal06}.}

1024: \tablenotetext{c}{Dubious value.}

1025: \end{deluxetable}

1026: 

1027: One of the observational arguments against pure temperature fluctuations in PNe by several authors 

1028: is the agreement between total abundances derived from  

1029: optical and FIR CELs in PNe \citep[see e.g.][]{liuetal01,tsamisetal04}. Due to their low excitation energy, 

1030: E$_{ex}$ $\sim$ 10$^3$ K, the fine structure infrared line emissivities 

1031: have a very weak dependency with {\elect}, therefore they must be insensible to the uncertainties introduced by the 

1032: presence of temperature fluctuations and, in the case of the existence of such fluctuations, they must give values 

1033: of abundance 

1034: similar to that obtained from ORLs. In the case of PNe, this comparison is usually feasible due to the small 

1035: angular size of these objects, both in the optical range

1036: \citep[using scanning techniques to cover the whole nebula, see e.g.][]{liuetal00}, 

1037: and in FIR and UV \citep[see e.g.][]{liuetal00,liuetal01}. 

1038: However, we think that this comparison is not so feasible in the case of extended Galactic {\hii} regions. 

1039: These objects cover much larger sky areas than PNe, 

1040: so large uncertainties may arise due to ionization stratification when comparing ionic abundances derived from 

1041: narrow slit optical spectroscopy and FIR observations obtained with space-borne facilities, that use larger 

1042: apertures. 

1043: In a previous paper \citep{garciarojasetal06} we have showed the difficulties found when comparing our derived 

1044: total 

1045: abundances for three objects of our sample (M17, NGC~3603 and the Orion nebula) with those derived from FIR 

1046: observations 

1047: \citep{simpsonetal95}.

1048: 

1049: \subsection{Dependence of the ADF on the excitation energy and the critical density.}

1050: 

1051: By definition, in the temperature fluctuations formulation, the abundance discrepancy must be related to 

1052: the excitation energy, E$_{ex}$ of the upper level in which the line originates 

1053: \citep[see basic equations in][]{peimbert67}; whereas in the presence of 

1054: density fluctuations the 

1055: abundance discrepancy is maximized if the abundance has been calculated from a CEL coming from a level with low 

1056: critical density 

1057: $n_c$. This last case implies that the ADF and $n_c$ are inversely related. 

1058: \citet{liuetal00, liuetal01}, comparing different abundance determinations in PNe (UV, optical and FIR CELs), 

1059: showed that the 

1060: ADF was not related to E$_{ex}$ nor to $n_c$. We cannot make comparisons between different types of lines 

1061: (UV, optical or FIR) of the same ion, since we do not have observations of the same zones in other spectral range 

1062: than the optical one; however, 

1063: we can compare the ADFs obtained for Galactic and extragalactic {\hii} regions, with E$_{ex}$ 

1064: and $n_c$ of the higher level of the main CELs of each ion. 

1065: In the case of C$^{++}$ the comparison between UV and optical data could be problematic due to the different volumes of nebula covered by the slits. 

1066: For NGC~5253, we are sure that optical and UV observations cover almost the same zones, as was pointed 

1067: out by \citet{lopezsanchezetal06}. In the case of M8 and the Orion nebula, the observations do not cover the same zone, but 

1068: we have compared the ionization correction factor for C$^{++}$ obtained from computing C$^{+}$/H$^{+}$ ratio from $IUE$ observations 

1069: of the UV C~{\sc ii}] $\lambda$2326 line \citep[][for M8 and the Orion nebula, respectively]{peimbertetal93b, walteretal92}  

1070: and by using the ionization correction factor (ICF) of \citet{garnettetal99}. From these data we have obtained ICFs of 2.48 and 2.45 

1071: for M8, and 1.20 and 1.24 for the Orion nebula, suggesting that there are not significant changes in the ionization structure 

1072: of the observed zones in the optical and UV.

1073: In Figure~\ref{adf_exc_ncr} it can be noted that, 

1074: taking into account the uncertainties, a slight correlation ($r$=0.7, $P$=0.188) between the ADF and E$_{ex}$ is apparent. 

1075: There is also an apparent correlation ($r$=0.7, $P$=0.188) with $n_c$, the opposite to that predicted by density fluctuations, 

1076: which suggests that density fluctuations should play a much less important role than temperature fluctuations in 

1077: {\hii} regions. 

1078: These results provide an additional argument that the mechanism that produces the abundance discrepancy in 

1079: {\hii} regions should be different to that 

1080: proposed for PNe and, therefore, we can not make the same considerations when working with PNe and {\hii} regions. 

1081: In Table~\ref{adf_exc}, we show the values of ADF, E$_{ex}$ and $n_c$ adopted for each ion. The values we have adopted are the 

1082: weighted mean of the ADFs computed for each ion in different objects. 

1083: We distinguish between O$^+$ abundances from {\foii} $\lambda\lambda$3726+3729 and $\lambda\lambda$7320+30 because 

1084: they have very different E$_{ex}$ and $n_c$. 

1085: 

1086: 

1087: \begin{figure}

1088: \begin{center}

1089: \epsscale{1.0}

1090: \plotone{f6.eps} 

1091: \figcaption{ADF (obtained for different ions) $vs.$ excitation energy of the upper level from which each line 

1092: originates (down) and critical density (up). In upper panel, the represented ions are, from left to right: O$^+$ 

1093: ({\foii} $\lambda\lambda$ 3726+29), O$^{++}$,  O$^+$ ($\lambda\lambda$ 7319+20+29+30), Ne$^{++}$ and C$^{++}$; in lower 

1094: panel the represented ions are, from left to right: O$^+$ ($\lambda\lambda$ 7319+20+29+30), 

1095: O$^{++}$, Ne$^{++}$, O$^+$ ({\foii} $\lambda\lambda$ 3726+29) and C$^{++}$.

1096: \label{adf_exc_ncr}} 

1097: \end{center} 

1098: \end{figure} 

1099: 

1100: \setcounter{table}{8}

1101: \begin{deluxetable}{cccc} 

1102: \tablecaption{ADFs, excitation energy and critical density of the CELs used to compute the abundances.

1103: \label{adf_exc}}

1104: \tablewidth{0pt}

1105: \tablehead{

1106: \colhead{Ion} & 

1107: \colhead{log(ADF)$^{\rm a}$} & 

1108: \colhead{E$_{ex}$ (cm$^{-1}$)} & 

1109: \colhead{$n_c$ (cm$^{-3}$)}}  

1110: \startdata 

1111: C$^{++}$	& 0.35$\pm$0.07	& 52400	 & 1.2$\times$10$^9$  \\

1112: Ne$^{++}$	& 0.28$\pm$0.09	& 25520	 & 1.0$\times$10$^7$  \\

1113: O$^{++}$	& 0.25$\pm$0.02	& 20000	 & 6.4$\times$10$^4$  \\

1114: O$^{+}$$^{\rm b}$& 0.16$\pm$0.06& 26800	 & 4.8$\times$10$^3$  \\

1115: O$^{+}$$^{\rm c}$& 0.15$\pm$0.06& 13660	 & 5.8$\times$10$^6$  \\

1116: \enddata

1117: \tablenotetext{a}{Weighted mean for all Galactic and extragalactic H~{\sc ii} regions.}

1118: \tablenotetext{b}{From [O~{\sc ii}] $\lambda\lambda$ 3726+29.}

1119: \tablenotetext{c}{From [O~{\sc ii}] $\lambda\lambda$ 7319+20+29+30.}

1120: \end{deluxetable}

1121: 

1122: \subsection{ {\oii} ORLs temperature.

1123: \label{te_rls}} 

1124: 

1125: The two phase model predicts that {\elect}(ORLs) $\le$ {\elect}({\hei}) $\le$ {\elect}({\hi}) $\le$ {\elect}(CELs) 

1126: \citep{liu03}, 

1127: with the difference between any of these two temperatures being proportional to the ADF (see e.g. 

1128: Figure~\ref{tfl_adf}). 

1129: 

1130: \citet{wessonetal03} made use --for the first time-- of the temperature sensitive ratio 

1131: $I$($\lambda$4089.29)/$I$($\lambda$4649.14) to determine {\elect} of the ionized gas in which {\oii} lines arise in 

1132: PNe.  

1133: These authors found very low {\elect} in two H-deficient knots in the PN Abell 30. Later, several authors have 

1134: found 

1135: similar results in other PNe \citep[e.g.][]{tsamisetal04, yliuetal04b, wessonetal05}. The intensities of the ORLs 

1136: that 

1137: originate from states with different orbital angular momentum present different dependencies with {\elect}. 

1138: For example, comparing the intensity of one 3$d$--4$f$ transition with respect to a 3$s$--3$p$ transition, is 

1139: possible to compute 

1140: the electron temperature. However, this method has its difficulties: first, the dependency of the intensity ratio 

1141: with the 

1142: temperature is weak, so these lines must be measured very accurately; second, the relative 

1143: intensities of the {\oii} lines can be affected by departures of the local thermodynamic equilibrium (LTE) 

1144: in the fundamental recombination level of the O$^{++}$ ion, $^3$P. \citet{tsamisetal04} consider the 

1145: ratio of the 3$d$--4$f$ transition of {\oii} $\lambda$4089.29 and the 3$p$--3$s$ transition {\oii} $\lambda$4649.14 

1146: of multiplet 1 

1147: because these lines arise from states of high total angular momentum J, $^3$P$_2$ and, therefore they must be 

1148: affected in a similar way by this effect. 

1149: 

1150: Determination of electron temperature from {\oii} ORLs in {\hii} regions is a non trivial problem because in these 

1151: objects ORLs are usually much weaker than in PNe. In Figure~\ref{temp_rls} we have represented the change of 

1152: $I$($\lambda$4089.29)/$I$($\lambda$4649.14) ratio with {\elect} for {\elect}=10$^4$ cm $^{-3}$, and we have 

1153: compared 

1154: that function with 

1155: the ratios obtained in NGC~3576 and the Orion nebula, the only objects of our sample where {\oii} $\lambda$4089.29 

1156: line has been measured. Additionally, we have represented the ratio measured in the giant {\hii} region 30 Doradus 

1157: in the 

1158: Large Magellanic Cloud (LMC) \citep{apeimbert03} and in several PNe taken from the literature: M~1-42 and M~2-36 

1159: \citep{liuetal01}, \citet{tsamisetal04} sample, PNe with ADF $>$ 4 of \citet{wessonetal05} sample, 

1160: NGC~5307 \citep{ruizetal03} and NGC~5315 \citep{peimbertetal04}. This last PN is also included in the PNe 

1161: sample of \citet{tsamisetal04} and we have considered it with the purpose of comparing the effect that 

1162: high spectral resolution observations introduce in the determination of {\elect}({\oii}). 

1163: NGC~5315 is appropiate because NLTE effects are small due to their high {\elecd} ($\sim$10$^4$ cm$^{-3}$) and 

1164: also presents an ADF which is similar to that measured in {\hii} regions. 

1165: Within the sample of PNe included in panel a) of Figure~\ref{temp_rls} there are objects with very high ADF 

1166: (between 5.0 and 

1167: 22) and in panel b) of Figure~\ref{temp_rls} there are other PNe that show values similar to that obtained in 

1168: {\hii} regions. NGC~5307 and NGC~5315 show a moderate 

1169: ADF (1.9 and 1.7 

1170: respectively), and have been studied from high resolution spectra (R$\sim$8800), which avoids the overlapping of 

1171: {\oii} multiplet 1 lines (in the other PNe, {\oii} $\lambda$4649.14 has been measured deblending the line using a 

1172: multiple Gaussian fit) and of {\oii} $\lambda$4089.29 line with other spectral features\footnote{In some of the PNe 

1173: of the 

1174: Tsamis et al. sample, the contribution of Si~{\sc iv} $\lambda$4088.85 line to the intensity of O~{\sc ii} 

1175: $\lambda$4089.29 

1176: line was corrected using the measured intensity of Si~{\sc iv} $\lambda$4116.10 line and adopting the theoretical 

1177: Si~{\sc iv} $\lambda$4088.85/Si~{\sc iv} $\lambda$4116.10 flux ratio (2:1).}. 

1178: As we can see in panel a) of Figure~\ref{temp_rls}, PNe with high ADFs show, in general, very low 

1179: {\elect}({\oii}), 

1180: which is consistent with the two phase model predictions. On the other hand, PNe with moderate ADFs (panel b) 

1181: of Figure~\ref{temp_rls}) 

1182: present temperatures that, in general, are also below {\elect}(CELs) and {\elect}({\hi}), except in the cases 

1183: of NGC~5882 \citep{tsamisetal04} and NGC~5315 \citep{peimbertetal04}, where {\elect}({\oii}) is consistent with 

1184: the derived {\elect} 

1185: from CELs. On the other hand, {\elect}({\oii}) obtained for NGC~5307 is extremely low, which can be due to 

1186: uncertainties in the measurement of the intensity of {\oii} $\lambda$4089.29 line\footnote{The NGC~5307 spectrum 

1187: was obtained in the same observation run as those of NGC~3576 and the Orion nebula. In the analyzed spectra of these 

1188: observations, 

1189: O~{\sc ii} $\lambda$4089.29 line is closely blended with a spectral feature produced by charge transfer effects in 

1190: the CCD due to the presence of a very bright line in a different order; this feature is clearly separated from the 

1191: O~{\sc ii} 

1192: $\lambda$4089.29 line in the Orion nebula, but not in NGC~3576 nor in 30 Doradus 

1193: (see Figure~\ref{linea4089}). An alternative form to correct 

1194: it, would be to adopt the theoretical ratio with other line of the same multiplet but unfortunately we do not know 

1195: the effect of NLTE to the 

1196: relative intensity of the lines of the multiplet.} (see Figure~\ref{linea4089}). In the case of the Galactic 

1197: {\hii} regions, 

1198: {\elect}({\oii}) obtained for the Orion nebula is somewhat higher to that obtained from CELs; for NGC~3576, 

1199: {\elect}({\oii}) is smaller than that 

1200: obtained from CELs, but {\oii} $\lambda$4089.29 line is affected by charge transfer effects in the CCD (see 

1201: Figure~\ref{linea4089}). 

1202: Similarly, $I$(4089)/$I$(4649)=0.825 ratio found for 30 Doradus \citep{apeimbert03} is extremely high, 

1203: due to a significant contribution of charge transfer to the intensity of the line (see Figure~\ref{linea4089}). 

1204: For these two regions we have tried to separate the line from the parasite feature, but it was not easy because of 

1205: the low signal-to-noise of the data in the spectral range of interest. As a first step we have identified the line 

1206: responsible of the 

1207: charge transfer effect in the CCD and then we have measured the flux of all the features 

1208: produced by charge transfer in the CCD (hereinafter CCT) in all the orders; finally, we have fitted the flux 

1209: decrement through the 

1210: different orders in order to 

1211: estimate the contribution of the CCT to the {\oii} $\lambda$4089.29 line. 

1212: Unfortunately, uncertainties 

1213: in the observed fluxes were high, so we can only make a rough estimation to the corrected line flux. For NGC~3576, 

1214: we have 

1215: estimated that CCT is about a 16\% of the observed flux of the {\oii} $\lambda$4089.29 line. 

1216: In Figure~\ref{temp_rls}c we have represented the new corrected value for NGC~3576, adopting the same uncertainty 

1217: in the line 

1218: flux that before correction. 

1219: In the case of 30 Doradus, we have estimated that the contribution to the observed flux of 

1220: {\oii} $\lambda$4089.29 line is between 30\% and 60\%.  

1221: We have represented an intermediate case (correction of 45\%) in Figure~\ref{temp_rls}d, assuming that the 

1222: uncertainty in the 

1223: $I$(4089)/$I$(4649) ratio of 30 Dor is about 30\% or even greater. 

1224: As it can be seen in Figure~\ref{temp_rls}c, applying the estimated corrections, 

1225: the {\elect}({\oii}) obtained for NGC~3576 is now consistent with {\elect} derived from CELs. For 30 Doradus, 

1226: taking into 

1227: account that the {\elecd} derived by \citet{apeimbert03} was $\sim$300 cm$^{-3}$, we have represented 

1228: the theoretical $I$($\lambda$4089.29)/$I$($\lambda$4649.14) ratio for {\elecd}= 250 cm$^{-3}$ (P.J. Storey, private 

1229: communication) 

1230: which is quantitatively different from that derived for {\elecd}=10$^4$ cm$^{-3}$; unfortunately, in this case, 

1231: the results are not as clear as in the case of NGC~3576: the large uncertainties in the adopted flux of the 

1232: {\oii} $\lambda$4089.29 line makes the {\oii} $I$(4089)/$I$(4649) ratio unreliable to compute {\elect} because 

1233: uncertainties 

1234: are compatible with variations of {\elect} from 600 K to 25000 K (see Figure~\ref{temp_rls}d). 

1235: In order to have additional {\hii} region data, we have represented the value of the intensity ratio for 

1236: non-published UVES datasets for two slit positions of the Orion nebula: 

1237: the position labeled as ``Orion 1'' coincides with one of the two zones studied by \citet{estebanetal98}, 

1238: and one slit position located on the Orion bar (24$\arcsec$ N and 12$\arcsec$ W of $\theta^2$Ori A). 

1239: From the position of the error boxes for these two regions in Figure~\ref{temp_rls}c, it can be seen 

1240: that the values of {\elect} are similar to that obtained for the Orion 2 slit position. 

1241: Considering {\elect}({\foiii}) calculated for these two regions and the relatively high densities that have been 

1242: computed in both regions 

1243: --{\elecd}({\fcliii}) = 7900$\pm$1300 and 5300$\pm$800 cm$^{-3}$ for Orion 1 and Bar, respectively, which minimizes 

1244: NLTE effects in the intensity of the lines--, we can affirm that {\elect}({\oii}) measured in different positions 

1245: from the Orion nebula are similar or even somewhat larger than that measured from CELs. On the other hand, it is true that

1246: in the presence of pure temperature fluctuations, {\elect}({\oii}) should be somewhat lower than that measured from CELs but, 

1247: taking into account that measured {\ts}'s in {\hii} regions are moderate, differences as large as those predicted by the two-phase 

1248: model are not expected.

1249: 

1250: \begin{figure*}

1251: \begin{center}

1252: \epsscale{0.48}

1253: \plotone{f7a.eps}

1254: \plotone{f7b.eps} 

1255: \plotone{f7c.eps} 

1256: \plotone{f7d.eps} 

1257: \figcaption{O~{\sc ii} $\lambda$4089.29/$\lambda$4649.14 line ratio as a function of the 

1258: electron temperature. Solid line in a), b)  and c) panels is the theoretical ratio for {\elecd} = 10$^4$ cm$^{-3}$. 

1259: The data are represented as error boxes (see text for the references). 

1260: In panel a) we have represented a sample of PNe with ADF $>$ 5.0; in panel b), we have 

1261: represented a sample of PNe  with moderate ADFs, similar to those found in our sample of H~{\sc ii} regions; in the 

1262: panel c), we show the two H~{\sc ii} regions for which it has been possible to measure O~{\sc ii} $\lambda$4089.29 3d--4f 

1263: transition: NGC~3576 and the Orion Nebula and two additional slit positions in the Orion nebula (see text) and, 

1264: finally, in panel d), we show the data for 30 Doradus \citep{apeimbert03}. 

1265: In this last diagram we have also included the theoretical behavior of the line ratio 

1266: with respect to electron temperature for {\elecd} = 250 cm$^{-3}$ (P.J. Storey, private communication) in order to 

1267: perform a more appropiate comparison in the case of 30 Doradus, which shows {\elecd} $\sim$ 300 cm$^{-3}$.

1268: For NGC~3576 and 30 Doradus, solid line boxes indicate the values found after correction by the presence of 

1269: charge transfer features in the CCD (see text), the dashed box indicates the position of NGC~3576 before the 

1270: correction. The horizontal arrows indicate the value of the line ratio for some objects which corresponding 

1271: temperature is outside the scale, whereas the vertical arrows indicate that the line ratio is outside the 

1272: represented scale. 

1273: \label{temp_rls}} 

1274: \end{center} 

1275: \end{figure*} 

1276: 

1277: \begin{figure}

1278: \begin{center}

1279: \epsscale{1.0}

1280: \plotone{f8.eps} 

1281: \figcaption{ O~{\sc ii} $\lambda$4089.19 line in 5 spectra of H~{\sc ii} regions 

1282: and in PNe NGC~5307. It can be noticed that this line is severely blended with another feature produced by charge 

1283: transfer along the CCD (CT) in NGC~3576 and 30 Doradus, whereas in NGC~5307 the signal to noise is too low. 

1284: The Orion nebula is the only H~{\sc ii} region in which this line has been well measured. 

1285: The calibrated spectra of 30 Doradus and NGC~5307 have been kindly provided by Antonio Peimbert. 

1286: \label{linea4089}} 

1287: \end{center} 

1288: \end{figure} 

1289: 

1290: Unfortunately, the available data for {\hii} regions are not enough to draw definitive general conclusions; 

1291: however our analysis allows us to enumerate a series of problems that we have found and to reach some 

1292: partial conclusions; a) first, it is necessary to develop atomic models that consider departures from LS coupling 

1293: in NLTE, with the purpose of 

1294: studying how this effect affects each line; b) it is necessary to carry out very deep high resolution spectrophotometry 

1295: in {\hii} regions to extend the catalogue of this type of objects with measured {\elect}({\oii}); c) the few 

1296: available data for {\hii} regions behave differently from those of most PNe and do not follow the predictions of the 

1297: two phase model as given in \citet{tsamispequignot05} for 30 Doradus.

1298: 

1299: Finally, a similar diagram has been constructed using {\cii} RLs ratios: $I$($\lambda$9903)/$I$($\lambda$4267) and 

1300: $I$($\lambda$6462)/$I$($\lambda$4267) 

1301: --that are not affected by NLTE effects-- but these ratios are much less temperature sensitive than 

1302: {\oii} $\lambda$4089.29/$\lambda$4649.14, and variations between 10$^3$ to 

1303: 15$\times$10$^3$ K are of the order of the observed error, so these ratios are not reliable to determine accurately 

1304: electron temperatures. 

1305: 

1306: 

1307: \section{Conclusions.

1308: \label{conclu}}

1309: 

1310: We have derived the abundance discrepancy factor, ADF, defined as: O$^{++}$/H$^+$(ORLs)/O$^{++}$/H$^+$(CELs) in eight 

1311: Galactic {\hii} regions, finding similar values for all of them. Furthermore, these values are similar with the ADF 

1312: determinations in extragalactic {\hii} regions available in the literature. This is a fairly different behavior 

1313: from the one shown by the ADF in PNe.

1314: 

1315: We have found that --within the uncertainties-- the ADF is not related to O/H, O$^{++}$/H$^+$, nor to the ionization degree. 

1316: As well, it is not related either to the assumed 

1317: {\elect}(High), so systematic effects in the abundance determination from CELs can be discarded. Also, the ADF is 

1318: not related to the {\elect}(Low)/{\elect}(High) ratio, suggesting that {\elect} natural gradients in {\hii} regions do 

1319: not produce the abundance discrepancies. 

1320:  

1321: In the case of {\hii} regions, contrary to the case of PNe, the ADF seems to be metallicity independent in the studied 

1322: range of O abundances. Also, there is no 

1323: evidence that the ADF is related to the effective temperature of ionizing stars --as should be implied by 

1324: photoelectric heating due to  the presence of dust near the stars--. 

1325: We have explored the behavior of the ADF with respect to the FWHM of the spectral lines finding no correlation 

1326: although the result is not conclusive due to the paucity of objects with lines of large FWHM.

1327:  

1328: We have found that the ADF seems to be slightly related to the excitation energy, $E_{ex}$ and to the critical density, $n_c$ 

1329: (in contrast to that observed in PNe), 

1330: a behavior that is not predicted by the two phase model initially proposed for PNe and recently extended to {\hii} 

1331: regions \citep{tsamispequignot05}, but consistent --at least in the first aspect-- with the predictions of the standard 

1332: temperature fluctuations paradigm. 

1333: On the other hand, electron temperatures derived from ORLs are consistent 

1334: with those derived from CELs, within the uncertainties, also contradicting the large differences predicted by the two phase model. 

1335: All these results suggest that the physical mechanism that produces the abundance discrepancy in PNe and {\hii} 

1336: regions might be different. Our results do not permit to discern what is the mechanism or 

1337: natural phenomenon that underlies the abundance discrepancy problem, but seem to be more consistent with the predictions of the 

1338: pure temperature fluctuations paradigm than with those of the two phase model. Therefore, it seems necessary to explore further 

1339: other alternatives for the origin of the temperature fluctuations and the abundance discrepancy in {\hii} regions.

1340:  

1341: Obtaining deep spectra of extragalactic {\hii} regions is fundamental to check the different models proposed to 

1342: explain the abundance discrepancy, permitting to extend the parameters space (metallicity, ionization budget, 

1343: kinematics, morphological complexity) of the objects. 

1344: 

1345: Finally, we want to stress again that, although the presence and survival of temperature fluctuations in gaseous 

1346: nebulae is a controversial fact, it is clear that the behavior in {\hii} regions and PNe should be explored in 

1347: independent ways, due to 

1348: their different origin, evolution time scales and physical processes that could affect both type of objetcs.

1349: 

1350: 

1351: 

1352: 

1353: JGR and CE would like to thank M. Peimbert, A. Peimbert, M. Rodr\'{\i}guez, A. Mampaso, V. Luridiana, E. P\'erez 

1354: and P.J. Storey for fruitful discussions and comments. We want also to thank all the organizers and participants to the 

1355: workshop on Deep Spectroscopy and Modeling of Gaseous Nebulae (held at Xiang Shan, Beijing on Apr 16-18 2007) for fruitful 

1356: discussions and 

1357: advice related to this work. We would also thank the referee, G. Stasi\'nska, for her carefully reading of the paper and 

1358: for her comments, that have increased significantly the quality of the paper. 

1359: This work has been funded by the Spanish Ministerio de Ciencia y Tecnolog\'{\i}a (MCyT) under project 

1360: AYA2004-07466.  

1361: 

1362: 

1363: \begin{thebibliography}{65}

1364: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

1365: 

1366: \bibitem[{{Aller} \& {Menzel}(1945)}]{allermenzel45}

1367: {Aller}, L.~H. \& {Menzel}, D.~H. 1945, ApJ, 102, 239

1368: 

1369: \bibitem[{{Esteban} {et~al.}(2004){Esteban}, {Peimbert}, {Garc{\'{\i}}a-Rojas},

1370:   {Ruiz}, {Peimbert}, \& {Rodr{\'{\i}}guez}}]{estebanetal04}

1371: {Esteban}, C., {Peimbert}, M., {Garc{\'{\i}}a-Rojas}, J., {Ruiz}, M.~T.,

1372:   {Peimbert}, A., \& {Rodr{\'{\i}}guez}, M. 2004, MNRAS, 355, 229

1373: 

1374: \bibitem[{{Esteban} {et~al.}(1998){Esteban}, {Peimbert}, {Torres-Peimbert}, \&

1375:   {Escalante}}]{estebanetal98}

1376: {Esteban}, C., {Peimbert}, M., {Torres-Peimbert}, S., \& {Escalante}, V. 1998,

1377:   MNRAS, 295, 401

1378: 

1379: \bibitem[{{Esteban} {et~al.}(1999{\natexlab{a}}){Esteban}, {Peimbert},

1380:   {Torres-Peimbert}, \& {Garc\'{\i}a-Rojas}}]{estebanetal99a}

1381: {Esteban}, C., {Peimbert}, M., {Torres-Peimbert}, S., \& {Garc\'{\i}a-Rojas},

1382:   J. 1999{\natexlab{a}}, Rev. Mexicana Astron. Astrofis., 35, 65

1383: 

1384: \bibitem[{{Esteban} {et~al.}(1999{\natexlab{b}}){Esteban}, {Peimbert},

1385:   {Torres-Peimbert}, {Garc\'{\i}a-Rojas}, \& {Rodr\'{\i}guez}}]{estebanetal99b}

1386: {Esteban}, C., {Peimbert}, M., {Torres-Peimbert}, S., {Garc\'{\i}a-Rojas}, J.,

1387:   \& {Rodr\'{\i}guez}, M. 1999{\natexlab{b}}, ApJS, 120, 113

1388: 

1389: \bibitem[{{Esteban} {et~al.}(2002){Esteban}, {Peimbert}, {Torres-Peimbert}, \&

1390:   {Rodr{\'{\i}}guez}}]{estebanetal02}

1391: {Esteban}, C., {Peimbert}, M., {Torres-Peimbert}, S., \& {Rodr{\'{\i}}guez}, M.

1392:   2002, ApJ, 581, 241

1393: 

1394: \bibitem[{{Garc{\'{\i}}a-Rojas} {et~al.}(2005){Garc{\'{\i}}a-Rojas}, {Esteban},

1395:   {Peimbert}, {Peimbert}, {Rodr{\'{\i}}guez}, \& {Ruiz}}]{garciarojasetal05}

1396: {Garc{\'{\i}}a-Rojas}, J., {Esteban}, C., {Peimbert}, A., {Peimbert}, M.,

1397:   {Rodr{\'{\i}}guez}, M., \& {Ruiz}, M.~T. 2005, MNRAS, 362, 301

1398: 

1399: \bibitem[{{Garc{\'{\i}}a-Rojas} {et~al.}(2007){Garc{\'{\i}}a-Rojas}, {Esteban},

1400:   {Peimbert}, {Rodr{\'{\i}}guez}, {Peimbert}, \& {Ruiz}}]{garciarojasetal06b}

1401: {Garc{\'{\i}}a-Rojas}, J., {Esteban}, C., {Peimbert}, A., {Rodr{\'{\i}}guez},

1402:   M., {Peimbert}, M., \& {Ruiz}, M.~T. 2007, Rev. Mexicana Astron. Astrofis.,

1403:   43, 3

1404: 

1405: \bibitem[{{Garc{\'{\i}}a-Rojas} {et~al.}(2006){Garc{\'{\i}}a-Rojas}, {Esteban},

1406:   {Peimbert}, {Costado}, {Rodr{\'{\i}}guez}, {Peimbert}, \&

1407:   {Ruiz}}]{garciarojasetal06}

1408: {Garc{\'{\i}}a-Rojas}, J., {Esteban}, C., {Peimbert}, M., {Costado}, M.~T.,

1409:   {Rodr{\'{\i}}guez}, M., {Peimbert}, A., \& {Ruiz}, M.~T. 2006, MNRAS, 368,

1410:   253

1411: 

1412: \bibitem[{{Garc{\'{\i}}a-Rojas} {et~al.}(2004){Garc{\'{\i}}a-Rojas}, {Esteban},

1413:   {Peimbert}, {Rodr{\'{\i}}guez}, {Ruiz}, \& {Peimbert}}]{garciarojasetal04}

1414: {Garc{\'{\i}}a-Rojas}, J., {Esteban}, C., {Peimbert}, M., {Rodr{\'{\i}}guez},

1415:   M., {Ruiz}, M.~T., \& {Peimbert}, A. 2004, ApJS, 153, 501

1416: 

1417: \bibitem[{{Garnett}(1992)}]{garnett92}

1418: {Garnett}, D.~R. 1992, AJ, 103, 1330

1419: 

1420: \bibitem[{{Garnett} \& {Dinerstein}(2001)}]{garnettdinerstein01}

1421: {Garnett}, D.~R. \& {Dinerstein}, H.~L. 2001, ApJ, 558, 145

1422: 

1423: \bibitem[{{Garnett} {et~al.}(1999){Garnett}, {Shields}, {Peimbert},

1424:   {Torres-Peimbert}, {Skillman}, {Dufour}, {Terlevich}, \&

1425:   {Terlevich}}]{garnettetal99}

1426: {Garnett}, D.~R., {Shields}, G.~A., {Peimbert}, M., {Torres-Peimbert}, S.,

1427:   {Skillman}, E.~D., {Dufour}, R.~J., {Terlevich}, E., \& {Terlevich}, R.~J.

1428:   1999, ApJ, 513, 168

1429: 

1430: \bibitem[{{Gonz\'alez-Delgado} {et~al.}(1994){Gonz\'alez-Delgado}, {P\'erez},

1431:   {Tenorio-Tagle}, {V\'{\i}lchez}, {Terlevich}, {Terlevich}, {Telles},

1432:   {Rodr\'{\i}guez-Espinosa}, {Mas-Hesse}, {Garc\'{\i}a-Vargas}, {D\'{\i}az},

1433:   {Cepa}, \& {Casta\~neda}}]{gonzalezdelgadoetal94}

1434: {Gonz\'alez-Delgado}, R.~M., {P\'erez}, E., {Tenorio-Tagle}, G.,

1435:   {V\'{\i}lchez}, J.~M., {Terlevich}, E., {Terlevich}, R., {Telles}, E.,

1436:   {Rodr\'{\i}guez-Espinosa}, J.~M., {Mas-Hesse}, M., {Garc\'{\i}a-Vargas},

1437:   M.~L., {D\'{\i}az}, A.~I., {Cepa}, J., \& {Casta\~neda}, H. 1994, ApJ, 437,

1438:   239

1439: 

1440: \bibitem[{{Guseva} {et~al.}(2007){Guseva}, {Izotov}, {Papaderos}, \&

1441:   {Fricke}}]{gusevaetal07}

1442: {Guseva}, N.~G., {Izotov}, Y.~I., {Papaderos}, P., \& {Fricke}, K.~J. 2007,

1443:   A\&A, in press, astro-ph/0701032

1444: 

1445: \bibitem[{{Guseva} {et~al.}(2006){Guseva}, {Izotov}, \& {Thuan}}]{gusevaetal06}

1446: {Guseva}, N.~G., {Izotov}, Y.~I., \& {Thuan}, T.~X. 2006, ApJ, 644, 890

1447: 

1448: \bibitem[{{H\"agele} {et~al.}(2006){H\"agele}, {P\'erez-Montero}, {D\'{\i}az},

1449:   {Terlevich}, \& {Terlevich}}]{hageleetal06}

1450: {H\"agele}, G.~F., {P\'erez-Montero}, E., {D\'{\i}az}, A.~I., {Terlevich}, E.,

1451:   \& {Terlevich}, R. 2006, MNRAS, 372, 293

1452: 

1453: \bibitem[{{Harrington} {et~al.}(1980){Harrington}, {Lutz}, {Seaton}, \&

1454:   {Stickland}}]{harringtonetal80}

1455: {Harrington}, J.~P., {Lutz}, J.~H., {Seaton}, M.~J., \& {Stickland}, D.~J.

1456:   1980, MNRAS, 191, 13

1457: 

1458: \bibitem[{{Kaler}(1986)}]{kaler86}

1459: {Kaler}, J.~B. 1986, ApJ, 308, 337

1460: 

1461: \bibitem[{{Kobulnicky} {et~al.}(1997){Kobulnicky}, {Skillman}, {Roy}, {Walsh},

1462:   \& {Rosa}}]{kobulnickyetal97}

1463: {Kobulnicky}, H.~A., {Skillman}, E.~D., {Roy}, J.-R., {Walsh}, J.~R., \&

1464:   {Rosa}, M.~R. 1997, ApJ, 477, 679

1465: 

1466: \bibitem[{{Liu}(2003)}]{liu03}

1467: {Liu}, X.-W. 2003, in IAU Symp. 209, Planetary Nebulae: Their Evolution and

1468:   Role in the Universe (San Francisco: ASP), ed. S.~{Kwok}, M.~{Dopita}, \&

1469:   R.~{Sutherland}, 339

1470: 

1471: \bibitem[{{Liu}(2006)}]{liu06}

1472: {Liu}, X.-W. 2006, in IAU Symp. 234, Planetary Nebulae in our Galaxy and Beyond

1473:   (San Francisco: ASP), ed. M.~J. {Barlow} \& R.~H. {M\'endez}, 219

1474: 

1475: \bibitem[{{Liu} {et~al.}(2006){Liu}, {Barlow}, {Zhang}, {Bastin}, \&

1476:   {Storey}}]{liuetal06}

1477: {Liu}, X.-W., {Barlow}, M.~J., {Zhang}, Y., {Bastin}, R.~J., \& {Storey}, P.~J.

1478:   2006, MNRAS, 368, 1959

1479: 

1480: \bibitem[{{Liu} {et~al.}(2001){Liu}, {Luo}, {Barlow}, {Danziger}, \&

1481:   {Storey}}]{liuetal01}

1482: {Liu}, X.-W., {Luo}, S.-G., {Barlow}, M.~J., {Danziger}, I.~J., \& {Storey},

1483:   P.~J. 2001, MNRAS, 327, 141

1484: 

1485: \bibitem[{{Liu} {et~al.}(1995){Liu}, {Storey}, {Barlow}, \&

1486:   {Clegg}}]{liuetal95}

1487: {Liu}, X.-W., {Storey}, P.~J., {Barlow}, M.~J., \& {Clegg}, R.~E.~S. 1995,

1488:   MNRAS, 272, 369

1489: 

1490: \bibitem[{{Liu} {et~al.}(2000){Liu}, {Storey}, {Barlow}, {Danziger}, {Cohen},

1491:   \& {Bryce}}]{liuetal00}

1492: {Liu}, X.-W., {Storey}, P.~J., {Barlow}, M.~J., {Danziger}, I.~J., {Cohen}, M.,

1493:   \& {Bryce}, M. 2000, MNRAS, 312, 585

1494: 

1495: \bibitem[{{Liu} {et~al.}(2004){Liu}, {Liu}, {Barlow}, \& {Luo}}]{yliuetal04b}

1496: {Liu}, Y., {Liu}, X.-W., {Barlow}, M.~J., \& {Luo}, S.-G. 2004, MNRAS, 353,

1497:   1251

1498: 

1499: \bibitem[{{L\'opez-S\'anchez} {et~al.}(2007){L\'opez-S\'anchez}, {Esteban},

1500:   {Garc{\'{\i}}a-Rojas}, {Peimbert}, \&

1501:   {Rodr{\'{\i}}guez}}]{lopezsanchezetal06}

1502: {L\'opez-S\'anchez}, A.~R., {Esteban}, C., {Garc{\'{\i}}a-Rojas}, J.,

1503:   {Peimbert}, M., \& {Rodr{\'{\i}}guez}, M. 2007, ApJ, 656, 168

1504: 

1505: \bibitem[{{Ma{\'{\i}}z-Apell{\'a}niz}

1506:   {et~al.}(2004){Ma{\'{\i}}z-Apell{\'a}niz}, {Walborn}, {Galu{\'e}}, \&

1507:   {Wei}}]{maizapellanizetal04}

1508: {Ma{\'{\i}}z-Apell{\'a}niz}, J., {Walborn}, N.~R., {Galu{\'e}}, H.~{\'A}., \&

1509:   {Wei}, L.~H. 2004, ApJS, 151, 103

1510: 

1511: \bibitem[{{Mathis} {et~al.}(1998){Mathis}, {Torres-Peimbert}, \&

1512:   {Peimbert}}]{mathisetal98}

1513: {Mathis}, J.~S., {Torres-Peimbert}, S., \& {Peimbert}, M. 1998, ApJ, 495, 328

1514: 

1515: \bibitem[{{O'Dell} {et~al.}(2003){O'Dell}, {Peimbert}, \&

1516:   {Peimbert}}]{odelletal03}

1517: {O'Dell}, C.~R., {Peimbert}, M., \& {Peimbert}, A. 2003, AJ, 125, 2590

1518: 

1519: \bibitem[{{Osterbrock} {et~al.}(1992){Osterbrock}, {Tran}, \&

1520:   {Veilleux}}]{osterbrocketal92}

1521: {Osterbrock}, D.~E., {Tran}, H.~D., \& {Veilleux}, S. 1992, ApJ, 389, 305

1522: 

1523: \bibitem[{{Peim\-bert}(1995)}]{peimbert95}

1524: {Peim\-bert}, M. 1995, in The Analysis of Emission Lines, ed. R.~{Williams} \&

1525:   M.~{Livio}, 165

1526: 

1527: \bibitem[{{Peimbert}(2003)}]{apeimbert03}

1528: {Peimbert}, A. 2003, ApJ, 584, 735

1529: 

1530: \bibitem[{{Peimbert} \& {Peimbert}(2005)}]{apeimbertpeimbert05}

1531: {Peimbert}, A. \& {Peimbert}, M. 2005, in Rev. Mexicana Astron. Astrofis. Conf.

1532:   Ser., ed. D.~{Torres-Peimbert} \& G.~{MacAlpine}, Vol.~23, 9

1533: 

1534: \bibitem[{{Peimbert} {et~al.}(2005){Peimbert}, {Peimbert}, \&

1535:   {Ruiz}}]{apeimbertetal05}

1536: {Peimbert}, A., {Peimbert}, M., \& {Ruiz}, M.~T. 2005, ApJ, 634, 1056

1537: 

1538: \bibitem[{{Peimbert}(1967)}]{peimbert67}

1539: {Peimbert}, M. 1967, ApJ, 150, 825

1540: 

1541: \bibitem[{{Peimbert} \& {Costero}(1969)}]{peimbertcostero69}

1542: {Peimbert}, M. \& {Costero}, R. 1969, Boletin de los Observatorios Tonantzintla

1543:   y Tacubaya, 5, 3

1544: 

1545: \bibitem[{{Peimbert} {et~al.}(2004){Peimbert}, {Peimbert}, {Ruiz}, \&

1546:   {Esteban}}]{peimbertetal04}

1547: {Peimbert}, M., {Peimbert}, A., {Ruiz}, M.~T., \& {Esteban}, C. 2004, ApJS,

1548:   150, 431

1549: 

1550: \bibitem[{{Peimbert} {et~al.}(1991){Peimbert}, {Sarmiento}, \&

1551:   {Fierro}}]{peimbertetal91}

1552: {Peimbert}, M., {Sarmiento}, A., \& {Fierro}, J. 1991, PASP, 103, 815

1553: 

1554: \bibitem[{{Peimbert} {et~al.}(1993{\natexlab{a}}){Peimbert}, {Storey}, \&

1555:   {Torres-Peimbert}}]{peimbertetal93}

1556: {Peimbert}, M., {Storey}, P.~J., \& {Torres-Peimbert}, S. 1993{\natexlab{a}},

1557:   ApJ, 414, 626

1558: 

1559: \bibitem[{{Peimbert} \& {Torres-Peimbert}(1977)}]{peimberttorrespeimbert77}

1560: {Peimbert}, M. \& {Torres-Peimbert}, S. 1977, MNRAS, 179, 217

1561: 

1562: \bibitem[{{Peimbert} {et~al.}(1993{\natexlab{b}}){Peimbert}, {Torres-Peimbert},

1563:   \& {Dufour}}]{peimbertetal93b}

1564: {Peimbert}, M., {Torres-Peimbert}, S., \& {Dufour}, R.~J. 1993{\natexlab{b}},

1565:   ApJ, 418, 760

1566: 

1567: \bibitem[{{Peimbert} {et~al.}(1995){Peimbert}, {Torres-Peimbert}, \&

1568:   {Luridiana}}]{peimbertetal95}

1569: {Peimbert}, M., {Torres-Peimbert}, S., \& {Luridiana}, V. 1995, Rev. Mexicana

1570:   Astron. Astrofis., 31, 131

1571: 

1572: \bibitem[{{Peimbert} {et~al.}(1992){Peimbert}, {Torres-Peimbert}, \&

1573:   {Ruiz}}]{peimbertetal92}

1574: {Peimbert}, M., {Torres-Peimbert}, S., \& {Ruiz}, M.~T. 1992, Rev. Mexicana

1575:   Astron. Astrofis., 24, 155

1576: 

1577: \bibitem[{{Robertson-Tessi} \& {Garnett}(2005)}]{robertsontessigarnett05}

1578: {Robertson-Tessi}, M. \& {Garnett}, D.~R. 2005, ApJS, 157, 371

1579: 

1580: \bibitem[{{Rodr{\'{\i}}guez} \& {Rubin}(2005)}]{rodriguezrubin05}

1581: {Rodr{\'{\i}}guez}, M. \& {Rubin}, R.~H. 2005, ApJ, 626, 900

1582: 

1583: \bibitem[{{Rola} \& {Stasi{\'n}ska}(1994)}]{rolastasinska94}

1584: {Rola}, C. \& {Stasi{\'n}ska}, G. 1994, A\&A, 282, 199

1585: 

1586: \bibitem[{{Rubin} {et~al.}(2002){Rubin}, {Bhatt}, {Dufour}, {Buckalew},

1587:   {Barlow}, {Liu}, {Storey}, {Balick}, {Ferland}, {Harrington}, \&

1588:   {Martin}}]{rubinetal02}

1589: {Rubin}, R.~H., {Bhatt}, N.~J., {Dufour}, R.~J., {Buckalew}, B.~A., {Barlow},

1590:   M.~J., {Liu}, X.-W., {Storey}, P.~J., {Balick}, B., {Ferland}, G.~J.,

1591:   {Harrington}, J.~P., \& {Martin}, P.~G. 2002, MNRAS, 334, 777

1592: 

1593: \bibitem[{{Rubin} {et~al.}(2003){Rubin}, {Martin}, {Dufour}, {Ferland},

1594:   {Blagrave}, {Liu}, {Nguyen}, \& {Baldwin}}]{rubinetal03}

1595: {Rubin}, R.~H., {Martin}, P.~G., {Dufour}, R.~J., {Ferland}, G.~J., {Blagrave},

1596:   K.~P.~M., {Liu}, X.-W., {Nguyen}, J.~F., \& {Baldwin}, J.~A. 2003, MNRAS,

1597:   340, 362

1598: 

1599: \bibitem[{{Ruiz} {et~al.}(2003){Ruiz}, {Peimbert}, {Peimbert}, \&

1600:   {Esteban}}]{ruizetal03}

1601: {Ruiz}, M.~T., {Peimbert}, A., {Peimbert}, M., \& {Esteban}, C. 2003, ApJ, 595,

1602:   247

1603: 

1604: \bibitem[{{Simpson} {et~al.}(1995){Simpson}, {Colgan}, {Rubin}, {Erickson}, \&

1605:   {Haas}}]{simpsonetal95}

1606: {Simpson}, J.~P., {Colgan}, S.~W.~J., {Rubin}, R.~H., {Erickson}, E.~F., \&

1607:   {Haas}, M.~R. 1995, ApJ, 444, 721

1608: 

1609: \bibitem[{{Stasi{\' n}ska} {et~al.}(2007){Stasi{\' n}ska}, {Tenorio-Tagle},

1610:   {Rodr\'{\i}guez}, \& {Henney}}]{stasinskaetal07}

1611: {Stasi{\' n}ska}, G., {Tenorio-Tagle}, G., {Rodr\'{\i}guez}, M., \& {Henney},

1612:   W.~J. 2007, A\&A, arXiv:astro

1613: 

1614: \bibitem[{{Stasi{\'n}ska} \& {Szczerba}(2001)}]{stasinskaszczerba01}

1615: {Stasi{\'n}ska}, G. \& {Szczerba}, R. 2001, A\&A, 379, 1024

1616: 

1617: \bibitem[{{Tenorio-Tagle}(1996)}]{tenoriotagle96}

1618: {Tenorio-Tagle}, G. 1996, AJ, 111, 1641

1619: 

1620: \bibitem[{{Torres-Peimbert} \& {Peimbert}(1977)}]{torrespeimbertpeimbert77}

1621: {Torres-Peimbert}, S. \& {Peimbert}, M. 1977, Rev. Mexicana Astron. Astrofis.,

1622:   2, 181

1623: 

1624: \bibitem[{{Torres-Peimbert} {et~al.}(1980){Torres-Peimbert}, {Peimbert}, \&

1625:   {Daltabuit}}]{torrespeimbertetal80}

1626: {Torres-Peimbert}, S., {Peimbert}, M., \& {Daltabuit}, E. 1980, ApJ, 238, 133

1627: 

1628: \bibitem[{{Tsamis} {et~al.}(2003){Tsamis}, {Barlow}, {Liu}, {Danziger}, \&

1629:   {Storey}}]{tsamisetal03}

1630: {Tsamis}, Y.~G., {Barlow}, M.~J., {Liu}, X.-W., {Danziger}, I.~J., \& {Storey},

1631:   P.~J. 2003, MNRAS, 338, 687

1632: 

1633: \bibitem[{{Tsamis} {et~al.}(2004){Tsamis}, {Barlow}, {Liu}, {Storey}, \&

1634:   {Danziger}}]{tsamisetal04}

1635: {Tsamis}, Y.~G., {Barlow}, M.~J., {Liu}, X.-W., {Storey}, P.~J., \& {Danziger},

1636:   I.~J. 2004, MNRAS, 353, 953

1637: 

1638: \bibitem[{{Tsamis} \& {P{\'e}quignot}(2005)}]{tsamispequignot05}

1639: {Tsamis}, Y.~G. \& {P{\'e}quignot}, D. 2005, MNRAS, 364, 687

1640: 

1641: \bibitem[{{Viegas}(2002)}]{viegas02}

1642: {Viegas}, S.~M. 2002, Rev. Mexicana Astron. Astrofis. Conf. Ser., 12, 219

1643: 

1644: \bibitem[{{Walter} {et~al.}(1992){Walter}, {Dufour}, \&

1645:   {Hester}}]{walteretal92}

1646: {Walter}, D.~K., {Dufour}, R.~J., \& {Hester}, J.~J. 1992, ApJ, 397, 196

1647: 

1648: \bibitem[{{Wesson} {et~al.}(2003){Wesson}, {Liu}, \& {Barlow}}]{wessonetal03}

1649: {Wesson}, R., {Liu}, X.-W., \& {Barlow}, M.~J. 2003, MNRAS, 340, 253

1650: 

1651: \bibitem[{{Wesson} {et~al.}(2005){Wesson}, {Liu}, \& {Barlow}}]{wessonetal05}

1652: ---. 2005, MNRAS, 362, 424

1653: 

1654: \bibitem[{{Wyse}(1942)}]{wyse42}

1655: {Wyse}, A.~B. 1942, ApJ, 95, 356

1656: 

1657: \end{thebibliography}

1658: 

1659: %\bibliographystyle{apj}

1660: %\bibliography{apjmnemonic,apjbib}

1661: 

1662: \end{document}

1663: