Is this a new post or revival of a very old one? Either way, this approach is not very good on many many levels. BrianT is dead-on.

Here are some questions to ask to start with:

No where is it mentioned how big this motor is. This will directly effect how big the transistor is. So there is the first question; how much current do you need and what is the in-rush current going to be? Ignore in-rush to your own peril (I've seen TO-3 cans vaporized). Once you answer this you can then decide whether an NPN will do it, or do you need something more like a Darlington device? You'll want to keep loses to a minimum and deal with the in-rush so at the very least we know we can use PWM. This can control in-rush as well as give speed-control (if you want).

LDR's are notoriously loose in their tolerance. Do you actually need an LDR or would a photo-diode be a better choice? In either case you have to consider hysteresis. If you do not then you can very easily get into a situation where the motor will stutter if a cloud or a person passes by. The simplest thing would a comparator (useful but) use an ADC and perform hysteresis in software. This is all going to be very low speed stuff so you have tons of time to measure the "photo-detector".

I see one of the 8 pin PIC's with pins left over to drive a led or two for whatever, measure battery voltage for an under-voltage lock-out, measure current, serial output...

Mike Tripoli