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

The Crab flux in the 20-40 keV energy band was obtained for 3 pointings of revolution 300, namely pointings 39, 40 and 41, with OSA 6.0 software. A mean Crab flux of 117 count s^-1 was used to convert the XTE J1739-302 flux from count s^-1 to Crab units.

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

... extraction

The OSA software reports the ISGRI detections in a significance scale in which a source is considered as detected if its fitted with a detection level of 7. This detection level of 7 is equivalent to the classical 3 $\sigma$ level above the measured noise.

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

... observations

Craig Markwardt, private communication.

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

... value

If KP2 was carried out during an unusually inactive period of this source while KP1 occurred during an unusually high activity period, it would make sense to consider an average outburst frequency of $\nu_{\rm outb}\approx 1.5$ d^-1. Then, maintaining XTE J1739-302 outside the smooth wind zone of Fig. 7 (left panel) would require $L_{0}\approx 0.25$ . The values of r and $P_{\rm orb}$ do not change too much and the terms of the discussion are nearly the same.

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

... model

And also with the outburst frequency found by Walter & Zurita Heras (2007). These authors were detecting only the strong outbursts.

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