Would this work?

It depends on the temp. of the smoke after it travels through the water, which would depend on numerous things such as the shape of the bong you have, how much water you have, the temperature of the water before you're hitting it, how long the hit is, how hard you hit it, and so on.

Technically, speeding up the smoke after it exits the water would be the only way to further cool the smoke, without using a coolant, such as ice, right? If you find the highest velocity you can hit it while still keeping the bong at max. cooling effeciency, you may be able to engineer a system to speed up the smoke to cool it to a maximum at the given velocity. I don't think a pinch in the tube will work though. I think it would do well to have maybe 4 cones and suck it through the cones, backward.
<| left to right of this funnel.
Maybe have 4 in a square. Forcing it through a small area and then undergoing rapid expansion is a fairly simple way to cool the air while maintaining decent air flow.

Any other ideas? This is an interesting thought.

 

I could see some smoke getting trapped on the backside of it as you pulled air through, maybe some very tiny holes would help that though? Also wouldn't using this idea (the expansion idea not the cone idea) require some minimum velocity? I would guess that you would need to pull at least a little hard to get the air to move fast enough for the expansion to matter.

 

I would think that someone hitting the bong provides the starting force? I mean, no point in creating this bong if we won't hit it, right? Plus, the cone idea IS the expansion idea. Imagine 4-5 cones with air being sucked in through the small hole and out the big one. Forced through a small opening, with an exponential expansion after said compression. The smoke will be mostly cool, anyway, as it has already gone through the water.

 

would all the smoke go through the cones? been trying to figure out if the vacuum caused by inhaling would be enough, because if you have 4 cones arranged as such (top view)
OO
OO

wouldnt smoke build-up around and above the small holes in the cone? or would the vacuum draw it all the way through? if you fail to understand this i'll draw do some mspaint skills and draw a propper pic.

 

But is that starting force fast enough that when the air/smoke runs over the cone it will be traveling fast enough to get a cooling effect from the backside of the cone. If it's traveling too slow there won't be any effect (like a plane needs a particular speed before the wings provide lift). That's also what I meant by expansion idea.

I only have a general knowledge of this sort of stuff so enjoy this exciting video describing what the Venturi effect is. (I looked it up because I didn't know for sure and she bored me so much I thought I'd share it with you guys)

[ame=http://www.youtube.com/watch?v=5zlfZD2cjXE]YouTube - Venturi Effect[/ame]

I wonder if when you speed it up going through the holes if you could send it back through the water again, I think the water does a good job of cooling it but several passes through water would be better?

 

Then fill it in with glass, and make 5 holes because a bong should be round, anyway. The top would be less filling. Not too hard.

 

In that case you're better off with a double perculator. You can't really cool the smoke much more than that if you want a double chamber, no coolant. The point is to speed and expand the smoke rapidly to cool is MORE than the water, not the same temp. Thus, water then expansion.