No, I don’t mean this: www.extrememusic.com. But now that I’ve got your attention, I want to return to a topic I wrote about last month: the Earth’s gravity and its effect on mechanical metronomes (or, more abstractly, its effect on pendulums). That post boiled down this fact:

Pendulums near the poles will swing more quickly than pendulums of the same length near the equator.

Of course, that begs the question—*how much more quickly*? At the time of my previous post, I hadn’t researched the mathematics required to do this. But now I’ve crunched some numbers and have some results to share. Before that, though, let me set up some of the basic ideas behind the calculations.

**1.*** The Earth isn’t a sphere*. We know today that the Earth is an oblate spheroid, and is flatter at the poles than at the equator. The verification of this fact dates back to the 1730s, and is credited to Pierre Louis Maupertuis. It’s worth mentioning that the oblateness follows from Newton’s mechanics, of which Maupertuis was a notable advocate. [Maupertuis did many other things, and you can read about them in this book.]

**2. Gravity varies according to latitude.** This follows quickly from point (1) above, though the first attempt at a precise formula was made by Alexis Claude de Clairaut in his 1743 treatise, *Théorie de la figure de la terre*. W. W. R. Ball has a short summary of Clairaut’s work. Today, we use a more precise formula, as part of the World Geodetic System:

Here, *G*_{E} ≈ 9.78033* *is gravity at the equator, *e *≈ 0.0818192 is the Earth’s eccentricity (which measures the “sphereness” of the ellipsoid, ranging from 0 to 1), and *k* ≈ 0.00193185 is a parameter that depends on gravity at the poles. (Details can be found here on page 4-2).* *

**3. Pendulum formula requires calculus. **I’m going to skip the derivation, and point you to Wikipedia on this one (grain of salt and all that). Here’s the formula for one “swoop” of a pendulum:

with *l* being the length of the pendulum and *θ*_{0} being its initial angle.

*4. Gravity decreases with altitude. *The experts use something called the *free-air correction* to compensate for this. For example, Reynolds’ *An Introduction to Applied and Environmental Geophysics* states that the proper way to adjust the formula in part (2) above is by subtracting (3.086×10^{-7})*h*, where *h* is your elevation above sea level* *(I’ve converted the units from Reynolds’ version).

*Now we’re ready for some results!* For all calculations, I took *θ*_{0} = π/4, and *l* ≈ 0.229060. This way, a pendulum at sea level at the equator will beat exactly 120 times per minute. Then, I picked five extreme points on the surface of the Earth: the summits of Chimborazo and Everest, and the Antarctic stations McMurdo, Russkaya, and Amundsen-Scott. And here’s what we have:

**Location** |
**Elevation** |
**Latitude** |
**Beats per Minute** |

Reference Point |
0 m. |
0° |
120 |

Chimborazo |
6268 m. |
1.469° S |
119.881 |

Everest |
8848 m. |
27.988° N |
119.902 |

Russkaya |
0 m. |
74.766° S |
120.296 |

McMurdo |
0 m. |
77.850° S |
120.304 |

Amundsen-Scott |
2835 m. |
90° S |
120.264 |

I’ve ignored the fact that the mass of the terrain that you’re standing on will affect the gravity at its summit (the *Bouguer anomaly*)—so the values for Everest and Chimborazo are probably a bit off. But, taking the numbers at their word, our hypothetical metronome will move the most quickly at McMurdo Station—which has the double-benefit of being both close to the pole (and thus closer to the center of the Earth) and having a low elevation. (In fact, McMurdo has the world’s southernmost port; you can go see what they’re up to right now.)

To conclude, it appears that a piece of music set to 120 beats per minute, that takes 5 minutes to perform, will finish almost exactly 1 beat sooner at McMurdo station than it would at the summit of Mount Everest. Not enough to mess anything up musically, but enough to notice. Extreme, right? (Not extreme enough for you? Then go read this comic.)