4 Results

4.1 Effective radii analysis

Our sample covers a wide range of effective radii, varying from less than 1 kpc for very compact objects to about 9 kpc for the more extended ones; typical values are between 4 and 6 kpc. The effective brightness is distributed from 18 to 22 mag arcsec^-2, being concentrated between 19.0 and 20.5 mag arcsec^-2.

Figure 3: left: comparison between $R_{\rm e}$(J) and $R_{\rm e}$(H,$K_{\rm s}$); right: comparison between $R_{\rm e}$(B) and $R_{\rm e}$(J, H, $K_{\rm s}$). Open circles are referred to J, solid boxes to H and open triangles to $K_{\rm s}$.

Figure 3 (left) presents the comparison between effective radii for J, H and $K_{\rm s}$. We note little variation of effective radius in the three bands. We have also obtained from de Vaucouleurs (1991), for these galaxies, the effective radii determined from B band photometry, which we compared with those obtained in this paper. The comparison is shown in Fig. 3 (right). Despite the considerable dispersion, there is a strong tendency of the effective radii in B to be higher, by a factor of nearly 2, than the corresponding infrared effective radii. This result is compatible with that found by Pahre et al. (1998), who compared effective radii in V and K bands, finding a tendency of effective radii to be higher for shorter wavelengths. Some objects do not follow this tendency; NGC 7562 and IC 5105 are dislocated in the sense of smaller radii in B, while NGC 1600 goes in the opposite sense. This effect could be interpreted as being due to differences in the stellar population distribution that contributes to each band. The fact that the B-band effective radii are higher than its near-IR counterparts shows that the stellar population that dominates near-IR emission is more concentrated towards the central regions than the bluer one.

4.2 Fundamental plane

Figure 4: Edge-on views of the Fundamental Plane for J, H and $K_{\rm s}$.

One of the most important constraints on early-type galaxy formation and evolution is the existence the three-parametric scaling relation called the Fundamental Plane (FP) relation (Djorgovski & Davis 1987), which is mostly defined for the optical bands. This relation covers a wide range of magnitudes and is, as far as we can see, independent of the environment. The very existence of this relation, which presents a very low scatter, must reflect an important regularity in the evolutive stories of early-type systems, be they dominated by dissipative collapse or not.

From 2.2 $\mu$m photometry of elliptical galaxies of the Coma cluster, Mobascher et al. (1999) found that it is possible to define for the same sample the FP in the near-infrared, with a similar dispersion as in the optical bands. Since our work presents a) the first near-infrared data available for several objects, b) observations in three photometric bands, and c) a sample not limited to a specific environment, we found it interesting to analyze the behavior of our sample in relation to the FP. One can express the FP, in terms of the structural parameters and the central velocity dispersion $\sigma$, as

$$\log~R_{\rm e}=a~\log~\sigma+b~\mu_{\rm e}+{\rm constant},$$ (4)

where a and b are constants to be determined from the fitting.

We have used the central velocity dispersion for our objects from the catalogue of Prugniel & Simien (1996). A least-square method was applied in determining the best-fit regression. We considered the uncertainties in $R_{\rm e}$ and $\mu_{\rm e}$ as independent quantities, because the correlation coefficient between these quantities is small and largely dependent upon only one point (NGC 1600). We obtained the following relations:

$$\log\;R_{\rm e}=(1.44\pm{0.12})\;\log\;\sigma+(0.46\pm{0.03})~\mu_{\rm e}, \quad {\rm for}~J$$ (5)

$$=(1.60\pm{0.11})\;\log\;\sigma+(0.38\pm{0.02})~\mu_{\rm e}, \quad {\rm for}~H$$ (6)

$$=(1.42\pm{0.09})\;\log\;\sigma+(0.43\pm{0.02})~\mu_{\rm e}, \quad {\rm for}~K_{\rm s},$$ (7)

where $R_{\rm e}$ is in parsecs and $\sigma$ in km s^-1.

The a values are in agreement with the average determinations of this constant (Mobasher et al. 1999; Milvang-Jensen 1997). The b values are stable, and near to 0.40. Scodeggio et al. (1998) analyzed the wavelength dependence of the FP parameters for BVrIHK bands, including a correction for completeness of the samples, and discovered that the a values are strongly wavelength-dependent, varying from 1.13 in B to 1.51 to K, while the b values are constant and near to 0.35. While our a values are compatible with those of Scodeggio, the b values are higher by nearly 20%.

Figure 4 shows edge-on views of the Fundamental Plane for J, H and $K_{\rm s}$. Although the relationship is built on a small number of point (12 for J and $K_{\rm s}$, 10 for H), that causes the rms scatter of these relations to be relatively high (0.20 dex for J and H, and 0.23 dex for $K_{\rm s}$), the FP is well defined in the three bands. The galaxy mostly off the correlation in the three bands (below the line of correlation) is NGC 1400, which is the most compact galaxy of our sample. We conclude therefore that for our sample of objects from different environments, in a 3-band near-infrared photometry, the Fundamental Plane is well defined.

4.3 Comparison with globular clusters

In order to analyze the integrated stellar population dominating the infrared luminosity of our galaxies, we compared our data with typical infrared colors of globular clusters. Infrared colors of globular clusters from the Magellanic Clouds and from M 31 are represented in Fig. 4. Data for Magellanic Clouds clusters are from Persson et al. (1983); Andromeda Galaxy clusters are from Frogel et al. (1980) (Galactic clusters are distributed in the same region delimited by Andromeda Galaxy cluster; see Aaronson et al. 1978). Magellanic Clouds clusters presented here are divided into two groups (Searle et al. 1980): types II-III are rich in red supergiant and AGB luminous stars, which dominate its infrared colors; in types IV-VI, the contribution of AGB stars is still stronger, especially that of carbon stars.

We can see that the integrated colors of the galaxies of our sample, with some exceptions, are compatible with populations like those found in the more metal-rich clusters of the Andromeda galaxy and of our own Galaxy. In all cases, populations like those found in extremely red clusters of the Magellanic Clouds, which are dominated by giant stars of the asymptotic branch, can be ruled out. For galaxies NGC 7192, NGC 7562 and NGC 7619, the infrared colors are dominated by less metal-rich populations, compatible with the average colors of globular clusters of the Galaxy.

Figure 5: Flux ratios from integrated isophotal colors compared to globular clusters. Beneath the dashed line, types II-III clusters are concentrated; types IV-VI are scattered by the whole diagram, but are more concentrated above the dashed line. The dashed circle indicates the average colors of the sample of E/S0 galaxies from Frogel et al. (1978).