All Figures

$\begin{figure} \par\includegraphics[width=12cm,clip]{1662fig1.eps} \end{figure}$ Figure 1: 2-18 keV light curves as observed by IXAE ( upper panel) and the mean photon spectral index $s_\phi$ ( lower panel) in 1st, 3rd and 5th orbits of June 22nd, 1997 (see, Table 1). GRS1915+105 was a) in the $\kappa$ class; b) in an unknown class; and c) went to the $\rho$ class on that day. Lower panels show how $s_\phi$ distinctly change. Specifically it is noisy during the transition.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=12cm,clip]{1662fig2.eps} \end{figure}$ Figure 2: Class transition of GRS 1915+105 as observed by IXAE on the 25th of June, 1997 in three successive orbits (see Table 1). It was a) in an unknown class similar to $\chi$ ; b) in transition to $\rho$ -type class with very low count rate; and c) in $\rho$ class after stabilization in that class. In the lower panels are $s_\phi$ showing noisy behaviour during transition before settling down in c).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.8cm,clip]{1662fig3.eps} \end{figure}$ Figure 3: Class transition as seen from IXAE and RXTE observations on the 8th of June, 1999 in two successive orbits (marked). RXTE photon count rates are divided by 50 and shifted upward by 200/s for comparison. In a), for a period of about 3000 s, there was no significant variation in light curve or spectral index. The object was in a class similar to $\chi$ but the count rates were a factor of 10-20 times higher. There is a gap of 44 min in the two RXTE data presented in a) and b). In b), the object is distinctly in the $\theta$ class. RXTE data is less noisy than the IXAE data because of higher counts. It was binned in 2-6 and 6-15 keV before computing $s_\phi$ so that comparison with IXAE could be made.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=6.8cm,clip]{1662fig4.eps} \end{figure}$ Figure 4: The power density spectrum of the RXTE light curve of Fig. 3b. The signature of 0.1 Hz break and a kink (weak and broad QPO) at around 5.4 Hz are typical of a $\theta$ class. These, along with softening of the spectrum at the dips (prominent in both the data) indicate that a class transition has indeed taken place.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.8cm,clip]{1662fig5.eps} \end{figure}$ Figure 5: Two successive orbital IXAE data showing a class transition on the 25th of June, 2000. In a), the class is similar to $\rho$ but the count rate is higher and the recurring time-scale between bursts is large ( $\sim$ 100 s). In a), the spectral index is less noisy but in the beginning of b) it becomes noisy though superficially it is still in " $\rho$ '' class. After transition it went to the so-called $\alpha$ class and $s_\phi$ becomes very noisy.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8cm,clip]{1662fig6.eps} \end{figure}$ Figure 6: Time dependence of the power density spectrum for the medium mode (0.1 s time resolution) IXAE observations presented in Fig. 5b. Along X-axis is the time after the observation started. The contours of constant power (in logarithmic scale) are plotted (marked on the contours). A strong peak at $\log~(\nu_{\rm QPO}) \sim -2$ in the first half signifies the Low Frequency QPO (LFQPO) in the $\rho$ state and another weak peak at $\sim$ 0.1 Hz (break frequency, see text) is also present. After the transition, the peak occurs at $\sim$ 2.5 Hz.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=6.8cm,clip]{1662fig4.eps} \end{figure}$	Figure 4: The power density spectrum of the RXTE light curve of Fig. 3b. The signature of 0.1 Hz break and a kink (weak and broad QPO) at around 5.4 Hz are typical of a $\theta$ class. These, along with softening of the spectrum at the dips (prominent in both the data) indicate that a class transition has indeed taken place.
Open with DEXTER