*** Two posts went in while I was writing this, but I think it's still worth posting it. ***
If you want your reaction to be under nitrogen only, in the first place you need to remove air or any gas that you have in the apparatus.
There are two main ways to achieve that: one is based on doing vacuum/nitrogen cycles (which has a specific name, but I don't remember it), the other is based on diffusion.
In small scale reactions, people just seal the apparatus with a special rubber stopper that can be pierced by needles. Vacuum is applied, and then a rubber balloon filled with nitrogen, connected to a needle, is inserted. Nitrogen substitutes the vacuum, and you have little or no air left. You can repeat the procedure 2-3 times to be sure.
You can do the same in larger scale apparati, you just need to use a special piece of glassware (a gas inlet with stopcock). In this case, in order to keep the pressure at about 1 atm, you use the system described in previous posts. The bubbler is just a Drechsel bottle filled with (little!) silicone oil or similar stuff, and you need the oil only for detecting the flow. Basically you have your nitrogen source connected to a T joint. One arm is connected to the bubbler, the other to your reaction vessel, and that's it. Hardly rocket science!
When it's not practical (or not possible) to apply vacuum, the usual procedure consists in having a flow of nitrogen going through the apparatus (i.e. you need an inlet and an outlet). Based on the principle of diffusion, you can assume that after some time you have almost no air left, because nitrogen carried it all away.
The major drawback of this method (especially for small scale reaction) is that nitrogen carries away your solvent too, so you may end up with a different solvent composition, different concentration, or even with a solid (I say that from personal experience...).
What you can do then is to maintain the flow at the beginning, to remove air, and then switch to the static nitrogen system as above.
To answer beheada, yes, you do need a pressure gauge, if you connected directly the big cylinder to the line you would just blow everything up; reflux doesn't make a huge difference, you just need to use a good condenser and connect the nitrogen line to the top of the condenser rather than to a sidearm of the flask (or you will see the solvent condensing in the N2 line!).