The Gaming Den

Okay, I have to say I am totally fucking stumped here. I've designed a circuit that turns on/off a laser (standard pen laser pointer from radio shack) based on whether I turn a certain pin on my computer's parallel port on or off. At least, that's the way it's supposed to work. But of course things are never that easy. (And before you ask, yes, I made sure that the parallel port's voltage actually switched).

So here's what I created:

(The box represents the laser pointer.)

I was going by memory when I made the drawing; the 2.6V should actually be 3.3 and the 1.6 should be 2.1. Anyway...

The top circuit is what it's supposed to look like. It doesn't work. What I really don't get here is that the second circuit actually works (the LED lights up). The third circuit, which is just the batteries connected to the laser, also works. But when I put the positive voltage and the laser on the collector (and connect the emitter to ground), everything stops working. And I'm really fucking confused as to why.

It was suggested to me that it wasn't working because the voltage on the collector is lower than the voltage on the base, which gives a reverse bias. But if that's the case, I seriously don't get why the LED would stop lighting up; it lights up just fine when there's no +2.1 volts on the collector side, and +2.1 is obviously a higher value than +0 (so there's still a reverse bias).

So basically I'm really confused, and if someone here knows about this shit I'd really appreciate you pointing me in the right direction.

For reference, the transistor is a 2N3904 small-signal transistor. The measured resistance of the laser pointer (while it was off) was 24 kOhms.