[split] Taccola spillover thread

[Jorge_Stolfi]

 If so, it must have been just a "theoretical invention".  Such a siphon must be primed by filling it with water.  That would be tricky for a siphon of that size...

You bung both ends of the pipe, fill up the pipe by pouring in water at the apex, seal over the hole in the apex, then remove the bungs.

Yes, that is the trick.   But it is tricky for a siphon of that size, because the water pressure at the openings would have pushed out any bungs.  You need some contraption at each end that can close the pipe and withstand the pressure while the siphon is being filled.  And you also need an opening at the top that can be hermetically sealed, and a platform where people could stand while filling it (because the operation would have to be repeated periodically).   

Also the pipe must be substantially airtight and able to withstand the negative pressure that would be trying to crumple it.  And the siphon would not work if the height of the bridge exceeded 10 meters -- which would be the case if we took the the scale of the other details (ship, river width, town) literally.

I say that it is a "theoretical" invention because the drawing does not even hint at those necessary details.

Then let pipes be installed on the bridge. Let the intake of the rising pipe be smaller in length than the discharge pipe. Make the discharge pipe greater in length by an extra one-fourth.

This bit too is a bit suspicious.  

The key parameters are 

• (A) the level of the water on the pool at the entrance side -- not the level of its opening there, which can be at any depth below the water; and 
  
• (B) the level of the water on the pool at the exit side, if the exit opening is submerged,  or the level of that opening, if it is above water.  
  

Then B should be lower than A by some difference L.  The flow of water (volume per time) would be determined by L and the diameter of the pipe D.  

The explanation above seems to say that it is only the relative lengths of the two pipes that matter. (Could that be just bad translation?) 

Anyway the necessary level difference L is rather small and not related to the height of the bridge; and the explanation does not mention the diameter.   

Also note that the drawing labeled A by Koen shows the opening of the ascending pipe above the water level in the pool at left.  That would definitely not work; the opening must be submerged.  

As it stands, that explanation does not seem to be based on actual experience with such siphons.

All the best, --stolfi