Astron. Astrophys. 319, 655-663 (1997)

4. Conclusions

We have presented new radio and X-ray observations of the SNR G 18.95-1.1. The X-ray data are in agreement with a pure thermal spectrum of a temperature and an interstellar absorption of . The variation across the source is for and for . About of the emission is concentrated in a diffuse component. The remaining part is visible as clumpy structures.

The new radio data at 10.55 GHz show that about of the emission is concentrated in a diffuse component. The central bar and the diffuse component have the same spectral index, whereas the spectrum of the outer arc-like features is significantly steeper. The integrated polarization is . Peak values reach . The polarization B-vectors mainly follow the arc structures.

There is some but occasional agreement between the radio and X-ray small-scale components. The radial profiles of the diffuse components in both domains are very similar when normalized to the peak value.

This latter result suggests a physical link between and responsible for the radio synchrotron emission and the number of thermal electrons as derived from the X-ray emission. It was shown that a radial magnetic field distribution does not fit this result. The magnetic field vectors are likely to follow threads in a ball of wool. Such a configuration could be presented if the magnetic field originates from the centre of the SNR and is wound by differential rotation of a central object, which we propose to be related to the observed central radio bar.

The radial profile of the diffuse X-ray component is discussed in terms of the evolution of SNRs in an interstellar medium with evaporating clouds as suggested by White and Long (1991). It is shown that the best fit with this model yields a shock velocity of , an age of , an ambient density of , a swept up mass of , and an initial explosion energy of . The observed X-ray clumps may be considered as some larger clouds which were hit by the SNR shock wave.

Compared with earlier results by Fürst et al. (1989) the ambient density turns out to be much lower. Fürst et al. (1989) derived an ambient density of about from the variation of the HI column density across the sources. If G 18.95-1.1 is the result of a type II or type Ib supernova explosion the preexisting stellar wind may have excavated the neighborhood. The currently observed remnant may just hit the wind shell but evolved in a sphere, the density of which is similar to that of a typical stellar wind bubble. The situation supports the assumption that in case of filled-centre SNRs the radio emission is dominated by relativistic electrons created by a central pulsar. Some of the outer arcs, which show a more typical shell-type radio spectral index, may demonstrate that the interaction of the SNR with the stellar wind shell is just starting.

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk@link.springer.de