5.1. Which direction?
The models of the SDG orbit indicate (see Fig. 11 Ibata et al. 1997 and references cited therein), together with the preliminary proper motion reported by Irwin et al. (1996), that the SDG is moving towards the galactic mid-plane. On the other hand, the ALRW91 C-stars appear to form the evidence that the SDG already crossed the galactic mid-plane. Furthermore, the Ibata et al. orbit at low galactic latitudes appears to be inconsistent with the position obtained from RR Lyrae stars, see Fig. 4 and Sect. 5.2. A study of the Galactic globular cluster Palomar 5 (Scholz et al. 1998) indicates further that the cluster is moving in the opposite direction 5 with respect to the Ibata et al. orbital motion of the SDG. Instead of looking for alternative explanations for the apparent contradictions, one should consider an independent determination of the proper motion with a zeropoint tied to one or more distant galaxies.
Recognizing that the present contradiction about the direction of motion of the SDG will not be solved until definite proper motions are available, and considering that reasonable grounds are given to use a lower weight for the preliminary value of the proper motion reported for the SDG, I assume for the remaining part of the paper that THE SDG IS NOT MOVING TOWARDS THE GALACTIC MIDPLANE, BUT CROSSED IT RECENTLY.
5.2. Where did the last crossing occurred?
The impact position at the galactic mid-plane is obtained from an unweighted least-squares fit through the distances determined for the SDG from RR Lyrae stars (Alcock et al. 1997, Alard 1996, NS97, and Mateo et al. 1995&1996) 6 as a function of the galactic latitude.
Fig. 3 displays the distances from the RR Lyrae stars as a function of galactic latitude. An unweighted least-squares fit gives D (kpc) = . One thus obtains 22.8 kpc for the distance towards the disc impact position of the SDG. Adopting a Solar galacto-centric distance of R0 = 8.0 kpc (Paczyski & Stanek 1998, Wesselink 1987) implies that THE CROSSING OF THE SDG THROUGH THE GALACTIC PLANE OCCURRED BEHIND THE GALACTIC BULGE AT 14.8 KPC FROM THE GALACTIC CENTRE .
5.3. Are the carbon stars related to the Galactic disc?
A consistency check is made for the suggestion that the SDG already crossed the Galactic plane and triggered a star formation event in the Galactic disc (Sect. 4.2.3). Such an event would imply that the disc metallicity at impact position has to be comparable with the photometric determination of the metallicity of the ALRW91 C-stars. To determine the disc metallicity at impact position one has to take the radial dependence of [Fe/H] into account. Carraro et al. (1998b) determined from open clusters the radial metallicity gradient in different age ranges.
For the clusters younger than 2 Gyr a present day gradient of - 0.063 0.013 dex kpc-1 was obtained. For all clusters in the sample the average gradient is - 0.085 0.008 dex kpc-1. The normalization at the Solar position is respectively - 0.18 0.12 dex and - 0.15 0.08 dex . The disc metallicity at the crossing position thus obtained is [Fe/H] = - 0.61 0.20 (Z 0.0050) or [Fe/H] = - 0.73 0.10 (Z 0.0035) with respectively the present day and the average radial metallicity gradient.
Irrespective of the present day or average metallicity gradient it is estimated that THE DISC METALLICITY AT THE CROSSING POSITION IS Z 0.0045 0.0010. WITHIN THE UNCERTAINTIES THE GALACTIC DISC METALLICITY AT THE IMPACT POSITION IS THE SAME AS METALLICITY OF THE ALRW91 C-STARS (Sect. 3.2).
5.4. Are the carbon stars near Sagittarius dwarf galaxy?
Another point to examine is how far the ALRW91 C-stars have moved from the impact since the crossing through the Galactic plane of the SDG. The C-stars were accelerated out of the Galactic midplane to their present average radial velocity of - 44 km s-1. In 0.1 Gyr they could have traveled about 4.4 kpc towards us. The actual distance traveled is considerably less, say 2.2 - 3.1 kpc, because the C-stars had a considerable smaller velocity in the past. The traveled distance is within the 10% - 15% uncertainty of the 22.8 kpc distance to the impact position.
The velocity dispersion of the C-stars provides further an indication about the average separation. Taking into account that the dispersion is smaller, due to a combination of turbulence and a lower velocities in the past, the separation is estimated to 5.6 - 8.0 kpc.
This can be compared with the distance between 2 of the 4 globular clusters associated with SDG is about 10 kpc, i.e. Terzan 8 at 21.1 kpc and Arp 2 at 31.0 kpc (Da Costa & Armandroff 1995).
The distance between the two carbon stars with well determined periods 7 is on the other hand about 5 kpc, i.e. 21.9 kpc for a carbon semiregular variable (NS97, Schultheis et al. 1998 and 26.7 kpc for a carbon Mira (Whitelock 1998). The separation between the ALRW91 C-stars is from the above consideration within acceptable limits.
THE ALRW91 C-STARS ARE STILL AT A DISTANCE RELATED WITH THE SDG .
5.5. ALRW91 C-stars = SDG C-stars?
Ibata et al. (1995) identified four new carbon stars in a field skimming over the Galactic Bulge. The radial velocities of the stars confirmed their membership to the SDG. Their average (J - K) colour is . The average (J - K) colour for the ALRW91 C-stars is . The colour difference indicates that there should be marked differences between the ALRW91 and the SDG C-stars 8, because comparable colours are expected if they originated from the same star formation event.
The difference between the two groups is an indication for differences in age and metallicity. The SDG C-stars should belong to a stellar population with an age comparable or younger than 4 Gyr. Only the youngest stellar population identified thus far for the SDG matches this constraint: the population has an age of 4 Gyr and a metallicity around Z 0.008 (Mighell et al. 1998). This is significantly different from the metallicity and age obtained for the ALRW91 C-stars: Z = 0.004 and 0.1 Gyr (see Sects. 3.2& 3.3).
THE ALRW91 C-STARS THE SDG C-STARS: THE SDG C-STARS ARE OLDER AND METAL-RICHER .
© European Southern Observatory (ESO) 1998
Online publication: September 14, 1998