Skip to content
Tutorials Center
My learning

Topics

Lesson 10 of 37 · Sequences and Limits

The Definition of a Limit of a Sequence

Master the epsilon-N definition of convergence, prove limits like 1/n and (3n+1)/(n+2), and learn to prove that a sequence diverges.

A sequence is an infinite list of real numbers a_1, a_2, a_3, ..., formally a function from ℕ to ℝ. We write (a_n). The central question is: does the list settle down toward a single number? The answer needs a definition precise enough to prove things with. This is the famous epsilon-N definition, and it is the single most important definition in the course.

The definition

A sequence (a_n) converges to L ∈ ℝ, written a_n → L or lim a_n = L, if

for every ε > 0 there exists N ∈ ℕ such that for all n ≥ N, |a_n - L| < ε.

Read it as a game. An opponent picks any tolerance ε. You must respond with a cutoff N such that from N onward, every term is within ε of L. If you can always win, the limit is L. If the opponent can choose some ε that you cannot answer, L is not the limit.

A sequence that converges to some limit is convergent; otherwise it is divergent. The limit, if it exists, is unique (two different limits L, L' would force |L - L'| < ε for every ε, so L = L').

A standard proof template

To prove a_n → L:

  1. Let ε > 0 be arbitrary.
  2. Work out (on scratch paper) how large n must be so that |a_n - L| < ε. Choose N accordingly.
  3. Write the proof forward: for n ≥ N, show |a_n - L| < ε.

Example 1. 1/n → 0. Let ε > 0. By the Archimedean property pick N with N > 1/ε. For n ≥ N: |1/n - 0| = 1/n ≤ 1/N < ε. ∎

Example 2. (3n + 1)/(n + 2) → 3. Scratch work: |(3n+1)/(n+2) - 3| = |(3n + 1 - 3n - 6)/(n + 2)| = 5/(n + 2) < 5/n. We want 5/n < ε, i.e. n > 5/ε. So given ε, choose N > 5/ε. For n ≥ N, |a_n - 3| = 5/(n+2) < 5/n ≤ 5/N < ε. ∎

Python Compute N for a given epsilon

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Geometric sequences

For |r| < 1, rⁿ → 0. Proof for 0 < r < 1: write r = 1/(1 + h) with h > 0. By Bernoulli's inequality, (1 + h)ⁿ ≥ 1 + nh > nh, so rⁿ < 1/(nh). Pick N > 1/(εh) and the result follows. For negative r use |rⁿ| = |r|ⁿ.

Python How fast do geometric sequences shrink?

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The closer |r| is to 1, the larger N must be, but an N always exists.

Proving divergence

To show a_n does not converge to L, negate the definition (see Logic and Proof Techniques): there is some ε > 0 such that for every N there is n ≥ N with |a_n - L| ≥ ε.

Example. a_n = (-1)ⁿ does not converge. Suppose it converged to L. Take ε = 1. Then from some N on, all terms lie in (L - 1, L + 1), an interval of length 2 that cannot contain both 1 and -1 (they are exactly 2 apart, and the interval is open). But the terms alternate between them forever. Contradiction. ∎

Another way a sequence can diverge: a_n = n is unbounded. Fact: every convergent sequence is bounded. Proof: with ε = 1, all terms from some N on lie within 1 of L; only finitely many terms remain, so take the maximum of |a_1|, ..., |a_{N-1}|, |L| + 1. Therefore an unbounded sequence cannot converge.

The converse is false: (-1)ⁿ is bounded but diverges. Bounded does not imply convergent.

flowchart LR
  A["Opponent picks ε"] --> B["You choose N"]
  B --> C["Check: n ≥ N gives |a_n − L| < ε"]
  C --> D["Win for all ε means a_n → L"]

The epsilon-N definition as a game: your N may depend on ε.

Check your understanding

Quick check · 4 questions

of 4 answered You scored / 4 · perfect!

  1. 1. In the definition of a_n → L, how does N depend on ε?

    Smaller tolerances typically need a later cutoff.
  2. 2. For a_n = (3n+1)/(n+2) and L = 3, what is |a_n − 3|?

    (3n+1 − 3(n+2))/(n+2) = −5/(n+2), so the absolute value is 5/(n+2).
  3. 3. Which sequence diverges?

    (−1)^n alternates between 1 and −1 and never settles.
  4. 4. Which statement is true?

    Convergence forces boundedness; the converse fails, as (−1)^n shows. Limits are unique.

Summary

  • a_n → L means: for every ε > 0 there is N with |a_n − L| < ε for all n ≥ N.
  • Proofs: fix ε, find N by scratch work, then verify forwards.
  • Limits are unique; convergent sequences are bounded.
  • To disprove convergence, negate the definition: some ε is never answered.
  • rⁿ → 0 when |r| < 1.

Next: Limit Laws and the Squeeze Theorem