Cornelis Velsen - the experiment that he may have carried out


The experimental write up below is not based on any historical research carried out. It is purely an hypotheses of how Cornelis Velson might have carried out any experiment of this type.


To determine the relationship between the slope of a stream bed and the velocity of water in the stream. To define a formulae that will provide a close approximation with the actual stream velocity.


  1. A fairly straight length of river or stream

  2. Two assistants prepared to get their feet wet

  3. A telescope on a tripod

  4. A long pole graduated in units of length, such as metres or perhaps decimetres

  5. Surveyors or Gunter’s chain

  6. A pocket watch

  7. An orange. The orange is very important.

Cornelis Velsen Experiment
Image by The Enviro Engineer - Sketch showing how Cornelis Velsen may have carried out the stream channel survey,

Description of stream survey

Cornelis would need to find the stream reach to use for the experiment. He would then need to measure the water course distance of 50.0 metres using a surveyor’s chain. The first assistant stands at the top of the measured distance in the middle of the stream channel. He has the tripod and telescope. The second assistant then stands in the stream 50.0 metres downstream of the first assistant with the graduated pole held vertically. The first assistant looks at the vertical pole through the telescope while making sure the telescope is held horizontally. He then records the graduation mark he sees in the telescope. Once the height of the graduation mark on the pole above the stream bed is known the height from the stream bed to the telescope lens is deducted to obtain the vertical drop along the channel, let’s assume it measures 2.0 metres.

Cornelis now needs the horizontal distance between the two assistants, since the 50.0 metres is the distance along the slope of the stream. In order to obtain the horizontal length, Cornelis would most likely have used Pythagoras’s theorem to find the horizontal length. This gives a horizontal distance of 49.96 metres, see calculation sheet equation (2). Now, the slope of the stream channel can be calculated by dividing the vertical drop by the horizontal length. Cornelis would have found that the stream bed slope was 0.04, see calculation sheet equation (4).

Pythagoras’ Theorem:

Pythagoras’ theorem

Equation (2):

Equation (4): 

Stream Velocity

Next, he would need to find out what the actual flow velocity was. That’s when he would use the orange. With his assistants still standing in the stream channel, he would hand the orange to the assistant at the upstream location. The assistant would place the orange in the water and Cornelis would note the time on his pocket watch. As soon as the orange floated downstream and passed the second assistant, he would shout out to Cornelis, who would check his pocket watch again. Cornelis now had a distance of 50 metres that the orange had travelled downstream and how long it had taken. If we assume it took 4.0 minutes for the orange to flow along the 50 metres of the stream, then Cornelis could work out that the actual velocity of the stream was 0.208 metres per second, see calculation sheet equation (3).


He will now be able to work out an equation using the slope of 0.4 to get a result approximately the same as the actual velocity in the river. The square root of 0.04 is 0.200, see calculation sheet equation (5) which is reasonably close to the actual velocity 0.208. In actual fact, it is probable that Cornelis repeated the experiment multiple times on different rivers or different reaches of the same river before publishing his book in 1749.

Equation (5):


While I have used values that show a close approximation between the experimental results and Cornelis Velsen’s equation, it was a fairly crude first attempt. Finally, I have an Empirical Formulae calculation sheet that shows the equations he may have used. You can change the slope of the river and time it took for the orange to flow downstream, and you will see that his equation was a very crude approximation.