Calculus, Infinite Series, and the Bernoullis

Philosopher, diplomat, and co-inventor of calculus — and the infinite series that grew out of it.

Louis XIV, King of France (Hyacinthe Rigaud, 1701).

The Stepped Reckoner (Staffelwalze), Leibniz’s mechanical calculator, designed to perform addition, subtraction, multiplication, and division automatically — part of why he’s sometimes called a forerunner of computer science (see “A Universal Genius”).
The triangular numbers count dots arranged in a triangle: \(1, 3, 6, 10, 15, \dots\) Huygens asked Leibniz to find:
\[1 + \frac{1}{3} + \frac{1}{6} + \frac{1}{10} + \frac{1}{15} + \dots = ?\]
\[1 + \frac{1}{3} + \frac{1}{6} + \frac{1}{10} + \dots = 2\]
The telescoping argument is persuasive, but Leibniz also gave a version that holds up to modern standards, using a finite partial sum and a limit.
A note on caution: not every “obvious” infinite sum behaves this well — see the next slide.
Jacob Bernoulli proved that the harmonic series — the sum of every reciprocal, \(1 + \frac{1}{2} + \frac{1}{3} + \frac{1}{4} + \dots\) — grows without bound, unlike Leibniz’s series.
The proof starts from half of Leibniz’s series and regroups it in a way that relates directly to the harmonic series.
The harmonic series diverges — but very slowly.
It takes an enormous number of terms just to pass modest totals like 20 or 30. The sum is unbounded, but only just — it grows roughly as slowly as the logarithm of the number of terms.

