The First Log-Log Plot

In Chapter 2 of Intermediate Physics for Medicine and Biology, Russ Hobbie and I discuss log-log plots. Have you ever wondered who made the first log-log plot? The honor goes to French mathematician and engineer Léon Lalanne (1811–1892), who was interested in using infographics to aid in computation. Let me take you through his idea.

Start with a sheet of log-log graph paper, one cycle in each direction. 

The lines in the bottom left are far apart, so let’s add a few more so it’s easier to make accurate estimates. 

Next, following Lalanne, add a bunch of diagonal lines connecting points of equal value on the vertical and horizontal axes. Label them, so they’re easy to read. 

What we’ve just invented is a log-log plot to do multiplication. For example, suppose we want to multiply 3.2 by 6.8. We find the value of 3.2 on the vertical axis, and draw a horizontal line (solid red). Then we find 6.8 on the horizontal axis and draw a vertical line (dashed red). Where the two lines intersect gives the product. We estimate it by seeing what are the closest diagonal lines. The intersection is between 20 and 22.5. I would guess it’s a little closer to 22.5 than 20, so I’ll estimate the product as 22.0. I’m pretty confident that I have the result correct to within ± 0.5. If I do the calculation on an electronic calculator, I get 21.76. My answer is off by 1.1%. Not bad.

You can do other sorts of calculations with this one sheet of log-log paper. For instance, below I plot a green line with a slope of one half, which lets me calculate square roots. Really, this is just a plot of y = x^1/2 on log-log paper. Because my log-log plot is only one cycle in each direction, the green line lets me calculate square roots of the numbers one through ten. To get the roots of ten through one hundred, I need to add a second, parallel line (green dashed). 

To calculate the square root of 77, I find 7.7 on the horizontal axis, go up to the dashed line, and then extrapolate over to the vertical axis. I estimate the result is about 8.8. When I use my electronic calculator, I get 8.775, so my estimate was accurate to about 0.3%. 

Of course, you could do all sorts of other calculations. Lalanne included many in his “universal calculator” that he had printed and posted in public places. Basically, the universal calculator is meant to compete with the slide rule (see my discussion of IPMB and the slide rule here). His charts never were as popular as the slide rule, perhaps because it’s more fun to slide the little rules than it is to look at a busy chart.

Léon Lalanne’s “Universal Calculator,”
or “Abacus” (1843).