... listed

Observations G and H are listed for completeness, but are of scarce utility due to their short integration time and presence of contaminating sources in the PDS instrument; they will not be discussed further.

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

... analysis

The spectral results for the bursts detected in observation E are reported in Feroci et al. (2004).

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

... range

See the Cookbook for BeppoSAX NFI Spectral Analysis, http://www.asdc.asi.it/bepposax/software/cookbook/

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

... area

We assume for SGR 1900+14a distance of 15 kpc (Vrba et al. 2000).

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

... XTE J1908+94

The source went in outburst in February 2002 and reached its flux peak about two months later (in't Zand et al. 2002).

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

... larger

The uncertainty in the relative calibration of the two satellite in the energy band considered here is of $\approx$10% (Kirsch et al. 2005).

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