One of the CDAW 9C substorms is investigated in this paper using the database reported by Hones et al. and supplemented with magnetogram inversion technique (MIT) 2 data. These latter have provided information about the dynamics of the open tail magnetic flux, current systems in the ionosphere, and the size and dynamics of the current wedge. We have identified the growth, expansion, and recovery phases of this substorm, with characteristics expected from a generally accepted scenario. However, specific signatures were observed in the interval (0919--0935) UT. i.e., between the growth and expansion phases, indicating the concurrent development of the substorm onset and corresponding instabilities in the innermost current sheet, and small-scale cross-tail current disruptions without the open tail reconnection. In addition to signatures of small-scale dipolarization, an increase of the open tail magnetic flux, and a current system of the type close to DP 2 were observed at (0919--0935) UT, which is more likely to suggest predominance of the tail-stretching process than magnetic collapse. This fact was interpreted in terms of a relevant simple model as a signature of the growth of the energy input from the solar wind which ensures the observable disturbance power. Hence the disturbance at (0919--0935) UT was more likely a driven one than an unloading one. The aforementioned signatures make it possible to identify the interval (0919--0935) UT as the phase of multiple onsets or (equivalently) the first active phase. which was previously defined by Mishin [1991, and references therein> as one of the four standard phases of a typical substorm (in addition to the expansion phase). Thus the case study supports the substorm scenario with two active phases and, accordingly, with two different kinds of physics. This case study illustrates also the informativity of MIT 2 data and their ability to effectively complement the database traditionally used in substorm studies.¿ 1997 American Geophysical Union