All Figures

$\begin{figure} \par\includegraphics[angle=90,width=8.8cm]{IRAS.ps} \end{figure}$ Figure 1: IRAS colour-colour plot of the target objects as filled asterisk with errorbars, except the [12]-[60] error of IRAS 17175-3544, for which the 60 $\mu$ m IRAS measurement is of bad quality. Also indicated are the colours of 1. UC H II regions compiled by Wood & Churchwell (1989), 2. Planetary Nebulae (Hughes & MacLeod 1989), 3. T Tauri stars (Morgan & Bally 1991). Figure adapted from Osterloh et al. (1997).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[angle=90,width=8.8cm]{NIR_CC.ps} \end{figure}$ Figure 2: 2MASS near infrared colour-colour diagram. Target object are indicated again by asterisks which are larger than the errorbars ( $\pm$ 0.06 mag or smaller for 2MASS colours, $\pm$ 0.1 for Tapia et al. 1996). The short curve shows colours for main sequence stars (Koornneef 1983). The long straight arrow corresponds to displacements of an early type object suffering an extinction equal to $A_{\rm V}=30$ . Displacement to the right of this line indicates a near-infrared excess. The small open circles are 2MASS detected objects, within 1 $^\prime$ radius from the targets. Note the strong extinction towards our targets.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[bb=96 186 517 605,width=5cm,clip]{h17175.eps... ...s}\par\includegraphics[bb=94 188 515 606,width=5cm,clip]{h18079.eps}\end{figure}$ Figure 3: 2MASS H-band image of IRAS 17175-3544 (top), IRAS 17441-2910 (centre) and IRAS 18079-1756 (bottom). North is up, East to the left. The field size is 2.5 $^\prime$ square.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=5cm,width=5cm]{IRAS17175_4m.eps}\par\... ..._4m.eps}\par\includegraphics[height=5cm,width=5cm]{IRAS18079_4m.eps}\end{figure}$ Figure 4: 2MASS K-band image of IRAS 17175-3544 (top), IRAS 17441-2910 (centre) and IRAS 18079-1756 (bottom). North is up, East to the left. The field size is 4 $^\prime$ square. In the top panel (IRAS 17175-3544), the North-Eastern part of the image is NGC 6334 I; in the South-Western part is the star forming region NGC 6334 II (see text). Note the strong extinction, obvious from the small number of stars near the target objects, which are indicated.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{spec.eps}}\end{figure}$ Figure 5: H-band spectra of the target objects, covering the complete wavelength range 16715-17490 Å, and corrected for telluric absorptions (except near 1.70 $\mu$ m: see text). Upper large panel: IRAS 17175-3544. Central: IRAS 17441-2910. Lower: IRAS 18079-1756. Note that the flux scale (counts) is different for IRAS 17441-2910, in order to show the strong emission lines. A normalised template telluric spectrum used is shown in the bottom smaller panel.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{model.eps}}\end{figure}$ Figure 6: Normalised spectra of telluric standards for IRAS 17175-3544 (top) and IRAS 18079-1756 (bottom) with best fit model spectra (dashed lines). See text for the model parameters.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{fig7.eps}}\end{figure}$ Figure 7: Left panel: comparison of the region of the He I + $\oplus$ line in the telluric standard for IRAS 17175-3544 with the best-fit synthetic spectrum. (dashed line, see Fig. 6 for the complete synthesis). Right panel: the He I line in IRAS 17175-3544 (dashed line) together with a pure telluric line derived by division of the two spectra in the left panel (dotted line). The heavy line shows the results of the division, clearly showing the narrow He I emission in the IRAS 17175-3544 spectrum.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{5cm}{!}{\includegraphics[clip]{fig8.eps}}\end{figure}$ Figure 8: Detection of the He II line at $\sim$ 1.693 $\mu$ m in absorption in IRAS 17175-3544. The telluric spectrum in the same region (offset for clarity) is also shown (dashed line) and suggests the He II detection is not a result of imperfect $\oplus$ cancellation.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{fig9.eps}}\end{figure}$ Figure 9: Normalised spectrum of IRAS 18079-1756 overlain with the H-band spectrum (dashed line) of the fast rotator HR 5191 (spectral type B3V - see text). Good agreement in the region of the He I absorption can be seen. Mismatches in the strength of the H lines results from filling in of the line profile by emission.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{fig10.eps}}\end{figure}$ Figure 10: Left panel: the blended He I + $\oplus$ profile in the telluric standard spectrum for IRAS 18079-1756 (dashed line) together with the adopted model He I line (dotted line; see text). The solid line shows the pure telluric line derived by division. Right panel: this pure telluric line (dotted) overlain with the uncorrected data for IRAS 18079-1756 (dashed line). The heavy line shows the corrected He I profile for this object.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[bb=96 186 517 605,width=5cm,clip]{h17175.eps... ...s}\par\includegraphics[bb=94 188 515 606,width=5cm,clip]{h18079.eps}\end{figure}$	Figure 3: 2MASS H-band image of IRAS 17175-3544 (top), IRAS 17441-2910 (centre) and IRAS 18079-1756 (bottom). North is up, East to the left. The field size is 2.5 $^\prime$ square.
Open with DEXTER

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{model.eps}}\end{figure}$	Figure 6: Normalised spectra of telluric standards for IRAS 17175-3544 (top) and IRAS 18079-1756 (bottom) with best fit model spectra (dashed lines). See text for the model parameters.
Open with DEXTER

$\begin{figure} \par\resizebox{5cm}{!}{\includegraphics[clip]{fig8.eps}}\end{figure}$	Figure 8: Detection of the He II line at $\sim$ 1.693 $\mu$ m in absorption in IRAS 17175-3544. The telluric spectrum in the same region (offset for clarity) is also shown (dashed line) and suggests the He II detection is not a result of imperfect $\oplus$ cancellation.
Open with DEXTER

$\begin{figure} \par\resizebox{8.8cm}{!}{\includegraphics[clip]{fig9.eps}}\end{figure}$	Figure 9: Normalised spectrum of IRAS 18079-1756 overlain with the H-band spectrum (dashed line) of the fast rotator HR 5191 (spectral type B3V - see text). Good agreement in the region of the He I absorption can be seen. Mismatches in the strength of the H lines results from filling in of the line profile by emission.
Open with DEXTER