4. NGC 2194
4.1. Proper motion study
For stars brighter than , some of the plates of NGC 2194 showed a systematic shift of the computed positions with respect to the ACT values. We therefore excluded all those stars from our input catalogue. However, this effect - which is different from the one described before for NGC 1960, since it perceptibly affects the positions - does not influence our membership determination as the region of NGC 2194 does not cover any stars of this brightness (see also Table 3).
Proper motions of 2,233 stars could be computed from the plates of NGC 2194. This figure is significantly higher than for NGC 1960, since this time the second epoch plates are of much higher quality, so that we reach fainter stars over the entire field. After four iterations of our proper motion determination, the systematic difference between ACT and our results were for the positions and for the proper motions. The standard deviation of the proper motions were and . The vector point plot diagram (Fig. 13) of the stars in the region of NGC 2194 shows that the stars are not as highly concentrated in the centre of the distribution of the cluster stars as for NGC 1960. This also explains the less distinct peak for high membership probabilities in the histogram shown in Fig. 9.
Although one finds more stars on the whole plates, in the inner region the proper motions of fewer objects were detected. This is caused by the lower number of sufficiently bright stars in and around the cluster (see Figs. 2 and 4). On the other hand, the low part of the main sequence is much more densely populated. We will see in Sect. 4.2 that the total number of members detected is higher for NGC 2194 than for NGC 1960.
We classified 149 members and 81 non-members. For this cluster, the separation between members and non-members was even more difficult as can be seen from the membership probability histogram plotted in Fig. 9: Approximately 50 stars show intermediate membership probabilities between 0.2 and 0.8. The result for the absolute proper motion of the cluster is
and for the field
The measured standard deviation of the cluster proper motion distribution again is the same as was determined for one individual object.
Table 9 shows a list of all proper motions computed.
Table 9. List of all stellar proper motions determined from the photographic plates of NGC 2194. Besides our internal numbering system, the numbers of del Rio's (1980) study are given. For more information see Table 7. The complete table is available online at the CDS archive in Strasbourg
4.2. Colour magnitude diagram properties
According to the field star subtracted CMD, presented as Fig. 14, NGC 2194 shows a prominent main sequence with a turn-off point near and a sparsely populated red giant branch. For this cluster, the proper motion study is not of great value for the isochrone fitting process: As the main sequence turn-off is located around , it is clear that the bright blue stars either do not belong to the cluster or do not help in finding the best isochrone fit because of their non-standard evolution like blue stragglers (see, e.g., Stryker 1993). We assume that the presence of the blue bright stars is mainly caused by the coincidence of the field and cluster proper motion centres which causes a certain number of fields stars to be mis-identified as cluster members. As expected in Sect. 4.1, this effect is more dramatic here than in the case of NGC 1960. From the comparison of the isochrones we derived the parameters for NGC 2194 given in Table 8.
Star No. 38 of our sample was first mentioned by del Rio (1980, star 160 therein) who considers this object a field star as a consequence of its location in the CMD. For the same reason, but with the opposite conclusion, Ahumada & Lapasset (1995) mention this object as a cluster member in their "Catalogue of blue stragglers in open star clusters". We find for this star a proper motion of and leading to a membership probability of 0.34. Therefore, we agree with del Rio and assume star 38 to be a field star, too. So far, no further information is known about the bright blue stars in Fig. 14, so that we cannot give any definite statement about the nature of these objects.
4.3. Initial mass function
The age of NGC 2194 of 550 Myr - together with its distance of almost 3 kpc - implies that the range of the observable main sequence is very limited. We therefore could compute the IMF only over the mass interval from (corresponding to ) to (or ). The slope determined is . The comparably higher error is a consequence of the smaller mass interval.
Comparing the resulting IMF with a histogram (see Fig. 16), one finds good agreement for three of the four bins (The bin width is the same as for NGC 1960). As the completeness drops rapidly within the leftmost bin, it has - in total - a high uncertainty. However, only stars with masses of (corresponding to a completeness of higher than 60%) were taken into consideration for the IMF computation, which is indicated by the limits of the IMF line in Fig. 16.
Note that as a consequence of the age of NGC 2194, the cluster may have encountered dynamical evolution during its lifetime, which has to be kept in mind when using the term "initial mass function". However, we follow Scalo's (1986) nomenclature, who uses the expression for intermediate age clusters, to discriminate between a mass function based on the initial and the present day stellar masses.
© European Southern Observatory (ESO) 2000
Online publication: June 5, 2000