next previous
Up: High-velocity interstellar clouds towards


4 Discussion

Our primary motivation in obtaining these spectra was to study the HVC towards M 15 on a small angular scale, and identify suitable stars for further studies at high spectral resolution. In Fig. 4 we have plotted the HVC equivalent widths and velocities versus the galactic latitude b for a sample of eight stars in a $\sim 25\hbox{$^{\prime\prime}$}$strip (i.e. the stars between the dotted lines in Fig. 1). For the relative velocities $v_{\rm rel}$, we have a taken an uncertainty of $\pm 5 \,{\rm km\,s^{-1}}$. Since all the measured velocities do not have uncertainties greater than $5 \,{\rm km\,s^{-1}}$, and this value corresponds to half a pixel, it seems quite reasonable to assign this uncertainty to $v_{\rm rel}$. On this small scale, the equivalent widths vary randomly from 0.05 to 0.20 Å, and the velocities from 65 to $80 \,{\rm km\,s^{-1}}$, with mean values of $ \sim 0.12$ Å  and $70 \,{\rm km\,s^{-1}}$. An inspection of Fig. 4 reveals that any variations in these quantities over small scales are probably on the order of or less than the measurement errors. Higher quality data are needed to quantify the magnitude of these variations. In particular, closely spaced pairings of sight lines such as the K-064, K-084, K-143 combination (identified in Fig. 4) provide the possibility of measuring HVC absorption over scales of $50\hbox{$^{\prime\prime}$}$.These data could then be combined with high spatial resolution Hi maps to provide reliable estimates of Hi column density at similar locations. In this work, we have shown that absorption is detectable across the face of the cluster and that many of these sight lines are suitable for such studies.

Such observations will also allow us to investigate the fractal structure of the HVC on scale-sizes of a few arcseconds. According to Vogelaar & Wakker (1994) all structural details will disappear if the resolution is not sufficient to resolve individual cloudlets.

  
\begin{figure}
\vspace*{3mm}
\includegraphics[height=11cm,width=7.2cm]{7947.f4.eps}\end{figure} Figure 4: Caii equivalent widths and velocities of the HVC as a function of galactic latitude b for the subset of stars indicated in Fig. 1
Also the cloudlets within HVCs have been suggested as sites for star formation in the halo via cloudlet-cloudlet collisions (Dyson & Hartquist 1983). This mechanism has been invoked to explain the existence of several early-type stars far from the plane of the Galaxy, whose kinematics and z-distances are incompatible with disc ejection (Keenan 1992). However these computations are extremely model-dependent (see, for example, Christodoulou et al. 1997) and require as input detailed information on the sizes, number density and physical conditions of HVC cloudlets, which are currently very uncertain (Wakker & van Woerden 1997), and that our future observations should provide.
next previous
Up: High-velocity interstellar clouds towards

Copyright The European Southern Observatory (ESO)