Characteristic Variation of Sco X -1

Rao, Rao and Jayanthi 1 have published results of their autocorrelation analysis of the optical photometry of Sco X -1, which they claim show a clear periodicity of about 3.0 h. They treated individual nights separately, finding respective periods of 4.43, 3.07, 2.04, 3.25, 3.00 and 0.96 h on six of the longer runs available. It seems that one should be able to say little about periodicities of the order of three hours when continuous coverage of at most eight hours is at hand, when the correlation coefficients are not very significant, and when the periods for only three of the six nights are close to three hours. Regardless, however, of opinions as to whether any significance should be attached to their analysis, we wish to point out that a much more profitable approach to studying the photometry of Sco X -1 is afforded by a simple inspection of the time variation. Suppose, for the moment, that the brightness changes of Sco X -1 were nearly random but that at irregular intervals a characteristic variation were to appear for, say, a few hours. Such a variation would not be periodic and should not lead to significant periodic terms in a power spectral analysis of a long continuous sample, but it could be the signature of a physical process which, although happening at unpredictable times, behaves in a characteristic way each time that it does occur. Furthermore, in a restricted data sample, such as we have at our disposal from the published photometry, such a characteristic variation could mimic a weak periodicity without, of course, having any obligation to continue being periodic during the gaps in the observations.