2001ApJ...549.1135C

We present the timing analysis results for our observations of the X-ray dip source X1916-053 conducted with RXTE between February and October of 1996. Our goal was to finally measure the binary period–as either the X-ray dip period or the ∼1% longer optical modulation period–thereby establishing whether the binary has a precessing disk (SU UMa model) or a third star (triple model). Combined with historical data (1979-1996), the X-ray dip period is measured to be 3000.6508±0.0009 s with a 2 σ upper limit |P{dot}|≤2.06x10^–11. From our quasi-simultaneous optical observations (1996 May 14-23) and historical data (1987-1996), we measure the optical modulation period to be 3027.5510±0.0052 s with a 2 σ upper limit |P{dot}|≤2.28x10^–10. The two periods are therefore each stable (over all recorded data) and require a 3.9087±0.0008 day beat period. This beat period, and several of its harmonics, is also observed as variations in the dip shape. Phase modulation of X-ray dips, observed in a 10 consecutive day observation, is highly correlated with the ∼3.9 day dip shape modulation. The 1987-1996 optical observations show that the optical phase fluctuations are a factor of 3 larger than those in the X-ray. We discuss SU UMa versus triple models to describe the X1916-053 light-curve behavior and conclude that the X-ray dip period, with smaller phase jitter, is probably the binary period, so that the required precession is most likely similar to that observed in SU UMa and X-ray nova systems. However, the ``precession'' period stability, and especially the fact that the times of X-ray bursts may partially cluster to occur just after X-ray dips, continues to suggest that this system may be a hierarchical triple.