4 Results

 

4.1 $H{\scriptstyle I}$ optical, and near-infrared detections and identifications

  The 16 brightest VLA HI sources of our primary list were re-observed in HI with the Nançay radio telescope. We confirm the presence of the 21 cm line at the position and velocity of all 16 VLA objects. The reduced Nançay spectra are shown in Fig. 9, after smoothing in velocity to a resolution of 15.6 km s^-1. Global HI line parameters were measured using standard Nançay data analysis software. The resulting central velocities, integrated line fluxes, line widths at 20% and 50% of peak maximum and the rms noise at the abovementioned velocity resolution are listed in Table 2. No corrections for instrumental resolution were applied to these values. The large telescope beam, HPBW $3\hbox{$.\mkern-4mu^\prime$}6$$\times 22\hbox{$^\prime$}$ ($\alpha \times \delta$), could in principle cause confusion with line signals from nearby HI-rich objects within the 1200 km s^-1 bandwidth. In fact, a second emission line is seen in the spectra of 3 objects (H1031-2632, H1035-2502 and H1038-2733; see Fig. 9). However, in all cases the confusion problem could be solved and emission lines disentangled using the Nançay and original VLA imaging data in combination with optical radial velocities from the NED and LEDA databases (see Sect. 5). In general, the VLA and Nançay data agree quite well as far as the integrated line flux and the central profile velocity is concerned; a direct comparison of the line widths given in Table 2 is difficult, given the low velocity resolution of the VLA data and the fact that the widths quoted by McMahon (1993) are indicative only (see Sect. 2).

  

Figure 9: 21 cm HI line spectra of the 12 HI-selected dwarf galaxies in the Hydra cluster re-observed at Nançay. The x-axis is the heliocentric velocity in km s-1, in the optical convention, and the y-axis is the flux density in mJy. The velocity resolution of the spectra is 15.6 km s-1

  

Figure 10: Nançay HIspectra at various positions around H1032-2819, a VLA-HI source for which we found no clear optical counterpart. In the finding chart of the field (Fig. 3) we identified other sources in the vicinity that may in principle contribute to the emission line seen towards H1032-2819: ESO 436-G046, 4$\hbox{$.\mkern-4mu^\prime$}3$ to the West and ESO 436-IG042 7$\hbox{$.\mkern-4mu^\prime$}2$ to the West of H1032-2819. The optical centers of these two ESO galaxies were used as pointing centers for the Nançay HIspectra shown here; the spectra labeled 2E and 2W were taken with pointing centers 2$^\prime$East and West, respectively, of the VLA HIsource. Clearly the reality of the HIdetection of H1032-2819 suffers no ambiguity

  

Figure 11: Nançay HI spectra of 4 other galaxies that missed our dwarf galaxy selection criteria; see Fig. 9 for details

  

Figure 12: B-band images of the 15 HI-selected dwarf galaxies in the Hydra I cluster displayed with the same spatial and intensity scales. The galaxies are ordered according to their integrated B-I color index, with the bluest objects displayed at the top left and the reddest at the bottom right. The lowest contour is 24.5 ${\rm mag\,\,arcsec^{-2 }}$and the interval is 0.5 mag. Note the diversity of the morphologies, ranging from compact objects to large low-surface brightness galaxies

  

Figure 13: K'- band images of the 12 dwarf galaxies observed at 2.2 $\mu$m wavelength. The images have been smoothed with a circular Gaussian function and calibrated. The surface brightness levels of the overlayed contours in ${\rm mag\,\,arcsec^{-2 }}$ are given at the bottom of each image. The arrows indicate the position of the dwarf galaxies in ambiguous cases. The size of each image is 1$^\prime$ $\times$ 1$^\prime$. North is up and East left

  

Figure 14: H1031-2818. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness level of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profiles in B (filled circles), I (open circles) and K' (triangles). The vertical error bars correspond to 2$\sigma$uncertainties. The FWHM of point sources is indicated by vertical lines of decreasing thickness for B, I and K', respectively. The B-band SBP is decomposed into an LSB component (thick curve) plus the luminosity excess above the LSB profile (small circles). The luminosity of the LSB component can be described by a modified exponential distribution with (b,q)=(1.8,0.65). (bottom-right) Radially averaged B-I profile

  

Figure 15: H1031-2632. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 16: H1032-2638. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. The crosses indicate the position of two spectroscopically identified foreground stars that were removed when computing the SBPs. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

We found an optical counterpart to all but one of our 20 primary targets - at the location of the HI source H1032-2819 no extended stellar object was found, to an I-band surface brightness limit of 25 ${\rm mag\,\,arcsec^{-2 }}$(see Fig. 3). The closest optical object is a faint galaxy seen in the I-band 7$\hbox{$^{\prime\prime}$}$ East of the central position of H1032-2819 and undetected in the B-band; it is most likely a red background galaxy. H1032-2819 is 4$\hbox{$.\mkern-4mu^\prime$}3$ and 134 km s^-1 in radial velocity away from an HI-rich spiral galaxy, ESO436-G046. In the VLA data, this source is only a $\sim 4\sigma$ detection in a single channel map, but our Nançay observations pointed towards H1032-2819, ESO 436-G046 and two positions 2$\hbox{$^\prime$}$ East and West of H1032-2819 (see Figs. 3 and 10) confirmed the reality of this HI source: it cannot be a sidelobe detection as its velocity does not match with that of any of the closeby sources. These observations will be discussed in a future Paper in these series.

Three candidate dwarf galaxies of the primary list appear on the CCD images as small face-on spirals: H1031-2734 (= ESO 436-G038; Fig. 2), H1033-2506 (= the barred spiral ESO 501- G032; Fig. 5) and H1034-2553 (included in the catalog of Richter et al. (1982), where it was misidentified as ESO 501- G040, however; see Fig. 5). Another galaxy with a more irregular morphology, H1033-2726 (ESO 501- G022; Fig. 2), turns out to have an HI width (FWHM 168 km s^-1) considerably larger than the 130 km s^-1 limit of our initial selection. The latter four objects have therefore all been excluded from further analysis. One galaxy, H1038-2730, with an HI line width (FWHM 155 km s^-1) exceeding slightly our selection criterion has nevertheless been kept in the sample as its absolute magnitude ($\mbox{$M_{\tiny B}$}=-15.8$) lies clearly within the range of dwarf galaxies.

Among the 15 remaining dwarf galaxies, only 3 were already listed as catalogued elsewhere in the NED and LEDA databases (see Table 1): H1031-2818 as [R87] 103156.2-281833, PGC 31270 (Richter & Huchtmeier 1987; Paturel et al. 1989), H1034-2558 as CTS 1015, M02.13 (Maza et al. 1991; Peña et al. 1991) and H1034-2758 as [R82] 103458.9-275843, PGC 31541 (Richter et al. 1982; Paturel et al. 1989). The calibrated images of all objects in our final list are shown in Figs. 14 to 28, and a thumbnail picture representation of all objects is displayed in Fig. 12. In the near-infrared K' band, 5 galaxies out of the 12 observed were detected clearly, 4 only marginally and 3 not at all (see Fig. 13). Limiting factors for detection were the central surface brightness of the galaxies as well as the stability of the sky background and, in general, weather conditions.

4.2 Surface brightness profiles

Surface brightness profiles (SBPs) were computed for the B and I-bands following Papaderos et al. (1996ab), and displayed in Figs. 14 to 28 for the 15 dwarfs of our final list. The SBPs were corrected for foreground Galactic absorption. The extrapolated central surface brightness $\mu_0$and exponential scale length $\alpha$ of the low surface brightness (LSB) disk component were derived from linear fits to the outer parts of each profile, weighted by the photometric uncertainties of each point. For three galaxies with sufficiently high surface brightness, an SBP could be derived in the K' band with the same method as used in the optical. It should be noted that the errors indicated for the K' profiles are purely statistical and do not take into account the systematic errors indicated above. Relevant photometric properties of the observed galaxy sample, corrected for foreground Galactic absorption, are listed in Table 4. Column 3 contains the extrapolated central surface brightness of the LSB disk component in units of ${\rm mag\,\,arcsec^{-2 }}$ and Col. 4 the exponential disk scale length in pc. Columns 5 and 6 list the apparent luminosity and dimension of each dwarf determined at the surface brightness level of 25 mag arcsec^-2. Column 7 contains the total apparent luminosity of each dwarf, as determined from a combination of numerical integration of each SBP out to the outermost point and the flux measurements within polygonal apertures, with contaminating foreground objects screened out. For K', the total magnitude, which could be determined for 5 objects only, has been measured within a polygonal aperture. Column 8 lists the effective radius for each profile in kpc, and Col. 9 the concentration index CI, defined, after James (1991), as $\log(r_{80}/r_{20})$, where r₈₀ and r₂₀ denote the radii within which 80% and 20% of the total luminosity are contained, respectively. In Col.10 we list the seeing, i.e. the FWHM (in arcsec) of stellar sources in the field of each dwarf. Column 11 contains the adopted Galactic extinction in the B-band and the range in radius used for the linear disk fit to each profile.

  

Table 4: Photometric properties of the HI selected Hydra dwarfs

$\parbox{17cm}{ {\bf Notes}\\ [1pt] $\mu_0$: \emph{Extrapolated} centr... ...extinction in $B$.\\ Fit Range: Range in radius in which fits were performed. }$

  

Figure 17: H1032-2722. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. The B-band SBP is decomposed into an LSB component (thick curve) plus the luminosity excess above the LSB profile (small circles). The luminosity of the LSB component can be described by a modified exponential distribution with (b,q)=(1.8,0.7). (bottom-right) Radially averaged B-I profile

  

Figure 18: H1033-2707. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles), I (open circles) and K' (triangles). An offset of 4 mag has been added to the K' SBP. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM in each band is indicated at the bottom left. (bottom-right) Radially averaged B-I profile

  

Figure 19: H1033-2642. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 20: H1033-2722. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 21: H1034-2558. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. A decomposition into a pure exponential LSB component and a starburst component in excess of the LSB component is shown by the solid line and the small circles, respectively. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 22: H1034-2758. (top-left) I-band image. The thick contour shows the extent of each galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. The HI-dwarf is indicated by the vertical thick lines. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile, showing a nearly constant color

  

Figure 23: H1035-2756. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 24: H1035-2605. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. The HI-dwarf is indicated by the horizontal thick lines. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles), I (open circles) and K' (triangles). An offset of 5 mag has been added to the K' SBP. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources in each band is indicated by the vertical lines. (bottom-right) Radially averaged B-I profile

  

Figure 25: H1035-2502. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 26: H1035-2740. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 27: H1038-2733. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure 28: H1038-2730. (top-left) I-band image. The thick contour shows the extent of the galaxy at the Galactic extinction-corrected I surface brightness of 23 mag arcsec-2. Contour levels are separated by 0.5 mag. (top-right) : B-I map of the galaxy. The overlayed contours correspond to the B-band image. (bottom-left) Surface brightness profile in B (filled circles) and I (open circles), corrected for Galactic extinction. The vertical error bars correspond to 2$\sigma$ uncertainties. The FWHM of point sources is indicated by the vertical thick and thin lines for B and I, respectively. (bottom-right) Radially averaged B-I profile

  

Figure: a) Absolute B luminosity vs. extrapolated central surface brightness of the LSB component for the HI-detected Hydra dwarfs. For comparison, dwarf irregulars and the LSB component of BCDs from the samples of Patterson & Thuan (1996) and Papaderos et al. (1996b), respectively, have been added. Note that roughly 1/3 of the surface brightness profiles of the HI-detected d warfs belong to the type-V class; e.g. they exhibit an outer exponential slope and a flattening towards the center. In these cases, the extrapolated central surface brightness $\mu_0$ may differ significantly from the value actually observed at the center of the galaxy. b) Absolute B luminosity vs. exponential scale length. All quantities shown for the Hydra dwarfs are corrected for foreground Galactic extinction

The LSB component of the SBPs can be divided in three main categories:

i. Pure exponential profiles
The surface brightness distribution of the LSB component of these dwarfs can be approximated, in at least one band, by a normal exponential law of the form:

$$I(R^*) = I_0\,\exp\left(-\frac{R^*}{\alpha}\right)$$ (1)

where R^* is the radius, I₀ the central intensity and $\alpha$ the exponential disk scale length. About half the dwarfs in our sample belong to this category and show, within the 1$\sigma$ uncertainties, a profile shape compatible with Eq. (1). These objects are H1032-2638, H1033-2707, H1033-2722, H1034-2558, H1034-2758, H1035-2740, H1038-2733 and H1038-2730.

ii. Exponential profiles flattening near the center
These profiles, designated as type-V in the nomenclature of Binggeli & Cameron (1991), exhibit an intensity depression with respect to a pure exponential distribution in their inner parts, typically for radii .In these cases the extrapolated central surface brightness of the LSB component, $\mu_0$, may differ significantly from the actually observed central surface brightness value.

The LSB component of these galaxies can be described by a modified exponential distribution (Papaderos et al. 1996b) of the form

$$I(R^*) = I_0\,\,\exp\left(-\frac{R^*}{\alpha}\right)\,\,\{1-q\exp(-P_3(R^*))\}$$ (2)

where P₃(R^*) is defined as

$$P_3(R^*) = \left(\frac{R^*}{b\alpha}\right)^3+\left(\frac{R^*}{\alpha}\,\,\frac{1-q}{q}\right).$$ (3)

The parameters $q=\Delta I/I_0$ and b denote, respectively, the depression of the flattened exponential distribution with respect to a pure exponential distribution and the radial extent within which the flattening occurs.

Dwarfs with this type of profile constitute $\sim\!\!\! 1/3$ of our sample. They are H1031-2818, H1031-2632, H1032-2722, H1035-2605 and H1035-2502.

For two of these, we have illustrated the decomposition of the B-band profile using Eq. (2), as well as the residual light remaining after fitting. As shown in Figs. 14 and 17, the q-parameter can be as high as 0.7 in the profile fits, implying an observed central surface brightness $\sim 1.3$ mag fainter than the extrapolated central value of the outer exponential profile. The physical origin of this type of profile is unclear; it has been conjectured that it may be due to early extensive mass loss via galactic winds (Fukunaga-Nakamura & Tosa 1989).

iii. Peculiar profiles
These are observed in two cases: H1033-2642 and H1035-2756. The former shows an exponential intensity distribution for intensities brighter than $\mu_B\sim 24.5$ ${\rm mag\,\,arcsec^{-2 }}$ which is truncated in the outer regions. This truncation may be due to a tidal interaction with a nearby S0 galaxy. In H1035-2756 the LSB component appears filamentary and patchy, and it dominates the light for intensities fainter than $\mu_B\sim 25$ ${\rm mag\,\,arcsec^{-2 }}$.

4.3 Color distribution

The B-I color maps are displayed at the upper right-hand panels of Figs. 14-28. The overlaying contours correspond to the B-band images. Note, that the color distribution of the dwarfs is not always spatially correlated with the B-band luminosity pattern; e.g. the intensity maximum does not necessarily coincide with a region exhibiting extreme colors with respect to the mean color of the galaxy. In most cases the bluest colors are found off-center, at the location of spectroscopically confirmed HII regions (see Paper II). Examples of dwarfs with off-center color minima are H1031-2818, H1033-2722, H1034-2758, H1035-2605, H1035-2502 and H1038-2733.

Radial color profiles (Figs. 14-28, lower-right panels) were computed by direct subtraction of the I-band SBP from the B-band SBP, after matching both profiles to the same resolution. In general, no strong color gradients are observed in the sample. The only exception is H1034-2558, a blue compact dwarf galaxy showing a inward bluing by 1.8 mag kpc^-1 within its starburst nucleus. The LSB components show a variety of radially averaged color distributions, ranging between weak outward bluing of the order of mag kpc^-1 and zero color gradient.

4.4 Classification of the Hydra dwarfs

Table 5 summarizes some integrated properties of the primary list objects. The absolute blue magnitude (Col. 2) and the B-I color index (Col. 5) were calculated using the total magnitudes listed in Table 4. The HI masses (Col. 3) were derived from the Nançay data, when available; otherwise, the original VLA line fluxes were used (those values are listed in parentheses in Table 5). All galaxies in our final list of HI-selected dwarf galaxies are fainter than absolute blue magnitude -17.5 and have HI masses ranging from $0.7 \, 10^8 \mbox{$M_{\odot}$}$ to $7.3 \, 10^8 \mbox{$M_{\odot}$}$. The mean HI-mass-to-blue luminosity ratios of the 15 Hydra dwarf galaxies in our final list is about 1 $M_{\odot}/L_{\odot,\,B}$, a rather typical value for gas-rich dwarf irregulars (Thuan 1985; Skillman 1996; Hoffman et al. 1996), though for individual objects it ranges from a low 0.1 (for H1031-2818 and H1035-2605), a value rather befitting quiescent lenticulars, to values exceeding 2 (for H1035-2756 and H1038-2733) which are not unexceptional for dwarf irregulars, however.

From the morphological point of view, the dwarfs of the HI-selected sample show a striking variety of properties. This diversity is clearly illustrated in Fig. 12, where surface brightness images of all 15 dwarfs are displayed together with the same spatial and intensity scales. In Figs. 29a,b we compare, respectively, the M_B vs. $\mu_0$ and the M_B vs. $\log(\alpha)$ correlation of our sample galaxies with those of other dwarf irregulars (dIrrs) and blue compact dwarfs (BCDs) taken from various other samples (see Patterson & Thuan 1996; Papaderos et al. 1996a, and references therein). In both diagrams the Hydra dwarfs span a wide range in the $\mu_0$ and $\log(\alpha)$ parameter space, from BCD-like objects to LSB dIrrs. Objects BCD H1034-2558 and LSB H1033-2642 are the two most extreme cases.

  

Table 5: Integrated properties