Chapter VIII · The convergence of infinite series and infinite integrals
Series of positive terms and the comparison test
Why a series of positive terms can only converge or diverge to infinity, how to compare one such series with another, and Cauchy's root test.
Why this matters
Given an infinite series, the first question is whether it converges at all. In Chapter IV Hardy answered it for the geometric series and a few relatives. §§171–174 start the systematic theory, with series whose terms are all positive. They are the easiest case, and the most important one: the question for a series with terms of both signs, or complex terms, is usually settled by looking at the positive series of their absolute values (§191). Everything in this lesson rests on one fact about increasing sequences and one idea, comparison.
The idea
Write for the partial sums. If every , each new term can only push up or leave it where it is. An increasing sequence has only two possible fates (§69): either it is bounded, and then it converges, or it is not, and then it tends to . Oscillation, the third possibility in Chapter IV, cannot happen.
Try this. Compare the partial sums of (blue) and (ochre). Switch to 300 and then 1000 terms.
- sₙ for
- sₙ for
- the sum s
What to notice
Both sequences of partial sums only climb; neither can turn back. The ochre one levels off below , so it is bounded and converges. The blue one keeps climbing, slowly (about ), but with no ceiling, so it diverges. The terms tend to in both cases, so that alone decides nothing.
So to show that a positive series converges, it is enough to find some ceiling for its partial sums. The cheapest way to find a ceiling is to sit the series under another one whose sum we know. If for every and , then every partial sum of is at most the corresponding partial sum of , hence at most . That is the comparison test. Its most useful yardstick is the geometric series with , which converges to .
How do we tell whether ? Take -th roots: is the same as . This gives Cauchy’s root test.
Try this. Start with under each test: compare it with , then look at . Then choose and under the root test, and .
- the limit
What to notice
For the roots settle near , comfortably below the red line at , so the series converges. For and the roots creep up towards from below: the test is silent, although one series converges and the other diverges. For the roots also approach , yet the terms tend to , so the series diverges for a much simpler reason.
Precise statements
From now on “positive” includes zero, as it does in Hardy, and means (or when that is more natural). Hardy also writes
The four basic facts, all proved in Chapter IV (§77), are restated in §173.
A series of positive terms either converges or diverges to . It cannot oscillate.
A series of positive terms converges if and only if there is a number such that for every .
Let and be series of positive terms and a constant.
- If converges and for every , then converges, and .
- If diverges and for every , with , then diverges.
The conclusions about convergence and divergence remain true if the inequality holds only for . The inequality between the sums need not.
If , and tends to a limit other than zero as , then and both converge or both diverge.
Comparing with the geometric series gives the first two tests (§174).
Let be a series of positive terms.
- If for all sufficiently large , where , the series converges.
- (Cauchy’s test.) If for all sufficiently large , where , the series converges.
- If for infinitely many values of , the series diverges.
Notice what (2) asks: not merely , but for a fixed below . The roots must stay away from . For every root is less than , but the roots approach , and the test says nothing.
Proofs
The partial sums of a positive series form an increasing sequence, and Chapter IV classified increasing sequences completely: bounded ones converge, unbounded ones tend to .
The partial sums increase
, so .
If they are bounded, they converge
If for all , then by §69 the increasing sequence tends to a limit, at most . So the series converges.
If they are not bounded, they tend to
If no bounds them, then for any some exceeds , and since the sequence increases, for every . That is exactly .
Conclude
These two cases exhaust the possibilities, which proves A. And convergence happens exactly in the first case, which proves B.
Find a ceiling for the partial sums of using the ceiling we already have for , then apply Theorem B. For divergence, run the same inequality the other way.
Partial sums inherit the inequality
If for every , adding up gives .
A ceiling for
Since , increases to its limit , so . Hence for every . By Theorem B, converges, and letting in gives .
Divergence
If instead with and diverges, then and , so .
Only large matter
If the inequality holds only for , apply the argument to the series that begin at . Removing the finitely many earlier terms changes every partial sum by the same fixed amount, so it changes neither convergence nor divergence (§77). It does change the sums, so can fail: the discarded terms of may be enormous.
A non-zero limit pins the ratio between two positive constants for large , which gives comparison inequalities in both directions.
Squeeze the ratio
Let . Since the terms are positive, , and by hypothesis . Taking in the definition of a limit, for all .
Two comparisons
So and for . If converges, the first inequality and Theorem C give convergence of . If diverges, the second gives divergence.
Why is needed
If only the first inequality survives: may be much larger than . For and the ratio tends to , one series converges and the other diverges.
Translate the condition on into a comparison with a geometric series.
Convergence (parts 1 and 2)
If for with , Theorem C applies with the convergent series . If , then raising both (positive) sides to the power gives , which is part 1 with .
Divergence (part 3)
If then . If this happens for infinitely many , the terms do not tend to . But the terms of a convergent series must tend to (§77), since . So the series diverges.
Worked examples and exercises
Show that converges for every integer , starting from the convergent telescoping series .
Hint
Hint
Solution
First, , so : the series converges.
For : , so by Theorem D converges too. (Even more simply, since .)
For : for , so Theorem C gives convergence. Hardy’s own route uses , which also telescopes, and . Divergence for was proved in §77; for the terms do not tend to .
Let be positive integers. Show that converges if and diverges if . For example, .
Hint
Solution
By induction : , and forces . If then , and converges, so by Theorem C so does . If every term is at least , so the terms do not tend to and the series diverges. The example is the case , shifted to start at with ; the first term does not affect convergence.
If is a convergent series of positive terms, show that and converge.
Hint
Solution
Since , there is an with for ; then . By Theorem C (for large ) converges. For the second series no such care is needed: , so for every .
If and converges, show that converges.
Hint
Solution
Since , we get . Both and converge, so their sum does, and by Theorem C (with ) so does . Note that we never needed to know how large is compared with : the inequality covers both cases at once.
Prove by contradiction that diverges: suppose it converges to and split it into its odd and even terms.
Hint
Hint
Solution
Suppose . The odd-term series and the even-term series converge by Theorem C, since their terms are terms of the original series and their partial sums are at most . Adding partial sums, . Also , so .
But , a series of strictly positive terms whose first term is already . So , a contradiction. Hence diverges.
Check yourself
A series of positive terms whose terms tend to must converge.
False. is necessary for convergence, never sufficient: and diverge although their terms tend to . What Theorem A guarantees is only that such a series cannot oscillate.
If for all and , then converges and .
False, although only half of it fails. does converge, since only large matter for convergence. But the first 99 terms of are not controlled at all; take and afterwards. The inequality between the sums needs for every .
Suppose and for every . What follows?
Cauchy’s test needs for a fixed . Both and have for every and , but the first diverges and the second converges. Their roots tend to , so no fixed works.
Does converge?
Theorem D needs a limit other than zero. Against the ratio is , so the series behaves like and converges. The second option uses a limit , from which nothing follows. The root test would fail here (the roots tend to ).
For , what is ? (So Cauchy’s test gives convergence.)
. Indeed for every , so we may take in Cauchy’s test.
Hardy’s language
The theory of the convergence of series is comparatively simple when all the terms of the series considered are positive.
Summary
- The partial sums of a positive series increase, so the series converges exactly when they are bounded; otherwise it diverges to . It never oscillates.
- Comparison: a positive series below a convergent one converges; one above a divergent one diverges. Only large matter for convergence, but not for the size of the sum.
- If tends to a limit other than zero, the two series behave alike. A limit zero proves nothing.
- Cauchy’s (root) test: for large gives convergence. If the roots tend to , the test is silent.
- Always check first whether the terms tend to .