Also at CAS-PKU Joint Beijing Astrophysical Centre, Beijing 100871, China; and Astronomical and Astrophysical Center of East China, Nanjing University, Nanjing 210093, China.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

Humboldt research fellow.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... follows

The luminosity and especially the photon index are taken from Baganoff et al. (2001b), which are slightly different from those in Baganoff et al. (2001a) where the spectral models used did not account for dust scattering; see Baganoff et al. (2001b) for details.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... rate

In Yuan (2000), the Chandra flux is produced with a higher accretion rate, $1.5 \times 10^{-4}~\dot{M}_{\rm Edd}$ . The discrepancy in accretion rate is because in Yuan (2000) the accretion is Bondi-like, while in the present paper it is disk-like. For the same accretion rate, the density in a Bondi-like accretion flow is much lower than in a disk-like flow. Since bremsstrahlung emission is proportional to the square of the density the X-ray emission can be largely different between the two cases.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... shocked

We note in this context that Falcke (1996b) and Beckert & Duschl (1997) demanded that $T \approx 10^{11}$ K based on the SgrA^* spectrum alone.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.