Hello everyone, first post

I'm planning on making a H bridge for reverse motor control, so to get started, I bought a few transistors to play with. For reference, the transistor has BC546 written on its flat part, so shoving that into Google brings up this datasheet:
http://www.datasheetcatalog.org/datasheet/philips/BC546_547_3.pdf
My first plan (just to get the hang of working with transistors) was to have the positive lead from my 9v battery connected to the collector, the base connected to pin 6 from the arduino, and the emitter connected to one end of the motor. The remaining ends of the motor and the battery both go to ground, which goes to the GND slot in my arduino. I anticipated that switching the battery on without powering the arduino whatsoever would have no effect on the motor, since the lack of current flowing through the base would not allow any current at all to flow through from the collector to the emitter, and then to the motor. Conversely, sending a signal down from the arduino to the base (with the battery on) should have turned the motor on due to the current flowing into the base.
However, I noticed that even when the base was NOT connected to pin 6 and both grounds from the motor and the battery were also not connected to the arduino, (in other words, the arduino is completed separated from the circuit) the motor begins to turn at some frequency when the battery is in the 'on' position. That is, instead of continuous power, the motor turns on, then off, then on, and so forth. I approximate the time between consecutive ons to be between about 0.4 of a second.
This makes absolutely no sense to me, because at school I always thought that no current would flow through from the collector to the emitter if no current flowed through the base. So if anyone could shed some light on this issue, it would be greatly appreciated.
Thanks