Theorem 2. Let \( \left\langle p_{n}\right\rangle_{n=1}^{\infty} \) be a Cauchy sequence in a metric space \( (E, d) \). Assume that this sequence has a convergent subsequence \( \left\langle p_{n_{k}}\right\rangle_{k=1}^{\infty} \). Then the original sequence \( \left\langle p_{n}\right\rangle_{n=1}^{\infty} \) converges.
Put more succinctly: Cauchy plus convergent subsequence implies convergent.
Problem 3. Prove Theorem 2. Hint: The \( \left\langle p_{n_{k}}\right\rangle_{k=1}^{\infty} \) converges, so there is a \( p \in E \) such that
\[
\lim _{k \rightarrow \infty} a_{n_{k}}=p .
\]
The basic idea of the proof is to note that by the triangle inequality that
\[
d\left(p, p_{n}\right) \leq d\left(p, p_{n_{k}}\right)+d\left(p_{n_{k}}, p_{n}\right) .
\]
Because the subsequence converges, the first term on the right side of the inequality can be made small by making \( k \) large. Because the original sequence is Cauchy the second term on the right can be made small by making both \( n_{k} \) and \( n \) large. Here is an outline of making this precise. Let \( \varepsilon>0 \).
(a) Explain why there is a \( N \) such that
\[
m, n \geq N \quad \text { implies } \quad d\left(a_{m}, a_{n}\right)<\frac{\varepsilon}{2} \text {. }
\]
(b) Explain there is a \( n_{k} \geq N \)
\[
d\left(p, p_{n_{k}}\right)<\frac{\varepsilon}{2} .
\]
(c) Show
\[
n \geq N \quad \text { implies } \quad d\left(p, p_{n}\right)<\varepsilon
\]
which completes the proof.