The motor will have a very low resistance, so point A will be pretty much the same voltage as the cap and solar cell positive, until the lower transistor turns hard on and runs it.
The "trigger" point is supposed to be when the capacitor voltage reaches about 0.6V more than the zener voltage, so current passes through the base-emitter junction of the left transistor and starts that turning on.
I can't get a simple circuit such as that to work, in simulation at least.. There are many far better circuits that will not give such problems, as Nigel says.