Calculating The Length of a Path
June 20, 2006 – 2:46 amWhen I learned calculus in high-school, integrals were introduced as a way of measuring the area below a function’s graph. I was quite struck by this concept, and I tried to find other geometric properties of functions I could calculate. The first thing that came to mind was measuring the length of the function’s path.
I remember thinking about how to go about this until I almost fell asleep. Then the solution came to me, and I immediately got up and wrote it down. Here it is: I first looked at a very small interval dx. In a sufficiently small interval, I assumed, the function would be almost linear, so it would look like this:

The length (dl) of this path element is then:



And the path’s length is the sum of all the path elements:

I verified this result by using it to calculate the circumference of a circle by setting
and
. I didn’t know how to do many integrals back then so I used an integrals table, and I was thrilled to get the correct result of
for the half-circle.
In retrospect it seems quite trivial, but back then I was very proud of this accomplishment. I started thinking about whether this might be a new result, and whether I should try to publish it. So I picked up a calculus book I had on my shelf to see if, by any chance, it mentioned this problem. And sure enough it was there: A section on measuring the length of a path. The proof was a bit involved (much longer than mine) and I couldn’t follow it, but the same result was obtained. And so after the thrill of discovery, I felt the pain of rediscovery…
Today I’m familiar with other ways of getting this result. The quickest (and least rigorous, I guess) is this lovely one-liner:

Subscribe by email