Astron. Astrophys. 364, L97-L100 (2000)

4. Conclusions

It is shown that any selection criteria imposed on the datasets may cause severe changes of the nature of the expected isotropic curve. We found that the isotropic distribution curve remains cosine for the polar angle and a straight line for the azimuthal angle (as previously adopted) only when there is no selection on B, L, P and i. The analytical method of Flin & Godl owski 1986 can be used only, when there are no selections on i. However, in all available databases both the positional and inclination effect are present. So simulations should be performed. In the following we summarize how selection effects will change the shape of the isotropic distributions:

1. The isotropic polar angle distribution is independent of L but changes with B when making selections on i.

2. The isotropic distribution for the polar angle is independent of B and L only when the range of i is full (i.e. no selections made); this distribution is independent of i, if there are no selections on positions.

3. The azimuthal angle distribution is independent of the range of B, provided the range of i and L is full. It changes with L even when the range of i and B is full.

4. The isotropic polar angle distribution is different for edge-on and face-on galaxies when making selections on B. However, the isotropic azimuthal angle distribution remains unchanged for face-on and edge-on galaxies when the range of L is full.

© European Southern Observatory (ESO) 2000

Online publication: January 29, 2001
helpdesk.link@springer.de