I totally agree with Ahmadi Murjani. However, a good manual is the Introduction to Time Series Analysis by Brockwell and Davis, which can be found online for free in a version from 2002. Suggesting anything without knowing the type of data and the research objective is quite pointless.
Before performing complicated analysis, I would look for possible correlations between your variables. For example, if you have a time series about earthquakes you could start by looking if there is any correlation between the magnitude of the earthquake and the time until the next, or something like that.