Harmonic series (mathematics)
Part of a series of articles about  
Calculus  





Specialized


In mathematics, the harmonic series is the divergent infinite series:
Its name derives from the concept of overtones, or harmonics in music: the wavelengths of the overtones of a vibrating string are 1/2, 1/3, 1/4, etc., of the string's fundamental wavelength. Every term of the series after the first is the harmonic mean of the neighboring terms; the phrase harmonic mean likewise derives from music.
Contents
History
The fact that the harmonic series diverges was first proven in the 14th century by Nicole Oresme,^{[1]} but this achievement fell into obscurity. Proofs were given in the 17th century by Pietro Mengoli,^{[2]} Johann Bernoulli,^{[3]} and Jacob Bernoulli.^{[4]}^{[5]}
Historically, harmonic sequences have had a certain popularity with architects. This was so particularly in the Baroque period, when architects used them to establish the proportions of floor plans, of elevations, and to establish harmonic relationships between both interior and exterior architectural details of churches and palaces.^{[6]}
Applications
The harmonic series can be counterintuitive to students first encountering it, because it is a divergent series even though the limit of the nth term as n goes to infinity is zero. The divergence of the harmonic series is also the source of some apparent paradoxes. One example of these is the "worm on the rubber band".^{[7]} Suppose that a worm crawls along an infinitelyelastic onemeter rubber band at the same time as the rubber band is uniformly stretched. If the worm travels 1 centimeter per minute and the band stretches 1 meter per minute, will the worm ever reach the end of the rubber band? The answer, counterintuitively, is "yes", for after n minutes, the ratio of the distance travelled by the worm to the total length of the rubber band is
(In fact the actual ratio is a little less than this sum as the band expands continuously.)
Because the series gets arbitrarily large as n becomes larger, eventually this ratio must exceed 1, which implies that the worm reaches the end of the rubber band. However, the value of n at which this occurs must be extremely large: approximately e^{100}, a number exceeding 10^{43} minutes (10^{37} years). Although the harmonic series does diverge, it does so very slowly.
Another problem involving the harmonic series is the Jeep problem, which (in one form) asks how much total fuel is required for a jeep with a limited fuelcarrying capacity to cross a desert, possibly leaving fuel drops along the route. The distance that can be traversed with a given amount of fuel is related to the partial sums of the harmonic series, which grow logarithmically. And so the fuel required increases exponentially with the desired distance.
Another example is the blockstacking problem: given a collection of identical dominoes, it is clearly possible to stack them at the edge of a table so that they hang over the edge of the table without falling. The counterintuitive result is that one can stack them in such a way as to make the overhang arbitrarily large, provided there are enough dominoes.^{[7]}^{[8]}
A simpler example, on the other hand, is the swimmer that keeps adding more speed when touching the walls of the pool. The swimmer starts crossing a 10meter pool at a speed of 2 m/s, and with every cross, another 2 m/s is added to the speed. In theory, the swimmer's speed is unlimited, but the number of pool crosses needed to get to that speed becomes very large; for instance, to get to the speed of light (ignoring special relativity), the swimmer needs to cross the pool 150 million times. Contrary to this large number, the time required to reach a given speed depends on the sum of the series at any given number of pool crosses (iterations):
Calculating the sum (iteratively) shows that to get to the speed of light the time required is only 97 seconds. By continuing beyond this point (exceeding the speed of light, again ignoring special relativity), the time taken to cross the pool will in fact approach zero as the number of iterations becomes very large, and although the time required to cross the pool appears to tend to zero (at an infinite number of iterations), the sum of iterations (time taken for total pool crosses) will still diverge at a very slow rate.
Divergence
There are several wellknown proofs of the divergence of the harmonic series. A few of them are given below.
Comparison test
One way to prove divergence is to compare the harmonic series with another divergent series, where each denominator is replaced with the nextlargest power of two:
Each term of the harmonic series is greater than or equal to the corresponding term of the second series, and therefore the sum of the harmonic series must be greater than the sum of the second series. However, the sum of the second series is infinite:
It follows (by the comparison test) that the sum of the harmonic series must be infinite as well. More precisely, the comparison above proves that
This proof, proposed by Nicole Oresme in around 1350, is considered by many in the mathematical community^{[by whom?]} to be a high point of medieval mathematics. It is still a standard proof taught in mathematics classes today. Cauchy's condensation test is a generalization of this argument.
Integral test
It is possible to prove that the harmonic series diverges by comparing its sum with an improper integral. Specifically, consider the arrangement of rectangles shown in the figure to the right. Each rectangle is 1 unit wide and 1/n units high, so the total area of the infinite number of rectangles is the sum of the harmonic series:
Additionally, the total area under the curve y = 1/x from 1 to infinity is given by a divergent improper integral:
Since this area is entirely contained within the rectangles, the total area of the rectangles must be infinite as well. More precisely, this proves that
The generalization of this argument is known as the integral test.
Rate of divergence
The harmonic series diverges very slowly. For example, the sum of the first 10^{43} terms is less than 100.^{[9]} This is because the partial sums of the series have logarithmic growth. In particular,
where γ is the Euler–Mascheroni constant and ε_{k} ~ 1/2k which approaches 0 as k goes to infinity. Leonhard Euler proved both this and also the more striking fact that the sum which includes only the reciprocals of primes also diverges, i.e.
Partial sums
n  Partial sum of the harmonic series, H_{n}  

expressed as a fraction  decimal  relative size  
1  1  1 


2  3  /2  1.5 

3  11  /6  ~1.83333 

4  25  /12  ~2.08333 

5  137  /60  ~2.28333 

6  49  /20  2.45 

7  363  /140  ~2.59286 

8  761  /280  ~2.71786 

9  129 7  /520 2  ~2.82897 

10  381 7  /520 2  ~2.92897 

11  711 83  /720 27  ~3.01988 

12  021 86  /720 27  ~3.10321 

13  145993 1  /360 360  ~3.18013 

14  171733 1  /360 360  ~3.25156 

15  195757 1  /360 360  ~3.31823 

16  436559 2  /720 720  ~3.38073 

17  142223 42  /252240 12  ~3.43955 

18  274301 14  /084080 4  ~3.49511 

19  295799 275  /597520 77  ~3.54774 

20  835135 55  /519504 15  ~3.59774 

21  858053 18  /173168 5  ~3.64536 

22  093197 19  /173168 5  ~3.69081 

23  316699 444  /982864 118  ~3.73429 

24  347822955 1  /948592 356  ~3.77596 

25  052522467 34  /923714800 8  ~3.81596 

26  395742267 34  /923714800 8  ~3.85442 

27  536252003 312  /313433200 80  ~3.89146 

28  404588903 315  /313433200 80  ~3.92717 

29  227046511387 9  /329089562800 2  ~3.96165 

30  304682830147 9  /329089562800 2  ~3.99499 

The finite partial sums of the diverging harmonic series,
are called harmonic numbers.
The difference between H_{n} and ln n converges to the Euler–Mascheroni constant. The difference between any two harmonic numbers is never an integer. No harmonic numbers are integers, except for H_{1} = 1.^{[10]}^{:p. 24}^{[11]}^{:Thm. 1}
Related series
Alternating harmonic series
The series
is known as the alternating harmonic series. This series converges by the alternating series test. In particular, the sum is equal to the natural logarithm of 2:
The alternating harmonic series, while conditionally convergent, is not absolutely convergent: if the terms in the series are systematically rearranged, in general the sum becomes different and, dependent on the rearrangement, possibly even infinite.
The alternating harmonic series formula is a special case of the Mercator series, the Taylor series for the natural logarithm.
A related series can be derived from the Taylor series for the arctangent:
This is known as the Leibniz series.
General harmonic series
The general harmonic series is of the form
where a ≠ 0 and b are real numbers, and b/a is not zero or a negative integer.
By the limit comparison test with the harmonic series, all general harmonic series also diverge.
pseries
A generalization of the harmonic series is the pseries (or hyperharmonic series), defined as
for any real number p. When p = 1, the pseries is the harmonic series, which diverges. Either the integral test or the Cauchy condensation test shows that the pseries converges for all p > 1 (in which case it is called the overharmonic series) and diverges for all p ≤ 1. If p > 1 then the sum of the pseries is ζ(p), i.e., the Riemann zeta function evaluated at p.
The problem of finding the sum for p = 2 is called the Basel problem; Leonhard Euler showed it is π^{2}/6. The value of the sum for p = 3 is called Apéry's constant.
lnseries
Related to the pseries is the lnseries, defined as
for any positive real number p. This can be shown by the integral test to diverge for p ≤ 1 but converge for all p > 1.
φseries
For any convex, realvalued function φ such that
the series is convergent.^{[citation needed]}
Random harmonic series
The random harmonic series
where the s_{n} are independent, identically distributed random variables taking the values +1 and −1 with equal probability 1/2, is a wellknown example in probability theory for a series of random variables that converges with probability 1. The fact of this convergence is an easy consequence of either the Kolmogorov threeseries theorem or of the closely related Kolmogorov maximal inequality. Byron Schmuland of the University of Alberta further examined^{[12]} the properties of the random harmonic series, and showed that the convergent series is a random variable with some interesting properties. In particular, the probability density function of this random variable evaluated at +2 or at −2 takes on the value 999999999999999999999999999999999999999764…, differing from 0.1241/8 by less than 10^{−42}. Schmuland's paper explains why this probability is so close to, but not exactly, 1/8. The exact value of this probability is given by the infinite cosine product integral C_{2}^{[13]} divided by π.
Depleted harmonic series
The depleted harmonic series where all of the terms in which the digit 9 appears anywhere in the denominator are removed can be shown to converge and its value is less than 80.^{[14]} In fact, when all the terms containing any particular string of digits (in any base) are removed the series converges.
See also
Wikimedia Commons has media related to Harmonic series. 
References
 ^ Oresme, Nicole (c. 1360). Quaestiones super Geometriam Euclidis [Questions concerning Euclid's Geometry].

^ Mengoli, Pietro (1650). "Praefatio [Preface]". Novae quadraturae arithmeticae, seu De additione fractionum [New arithmetic quadrature (i.e., integration), or On the addition of fractions]. Bologna: Giacomo Monti.
Mengoli's proof is by contradiction: Let S denote the sum of the series. Group the terms of the series in triplets: S = 1 + (1/2 + 1/3 + 1/4) + (1/5 + 1/6 + 1/7) + (1/8 + 1/9 + 1/10) + … Since for x > 1, 1/x − 1 + 1/x + 1/x + 1 > 3/x, then S > 1 + 3/3 + 3/6 + 3/9 + … = 1 + 1 + 1/2 + 1/3 + … = 1 + S, which is false for any finite S. Therefore, the series diverges.
 ^ Bernoulli, Johann (1742). "Corollary III of De seriebus varia". Opera Omnia. Lausanne & Basel: MarcMichel Bousquet & Co. vol. 4, p. 8.
 ^ Bernoulli, Jacob (1689). Propositiones arithmeticae de seriebus infinitis earumque summa finita [Arithmetical propositions about infinite series and their finite sums]. Basel: J. Conrad.

^ Bernoulli, Jacob (1713). Ars conjectandi, opus posthumum. Accedit Tractatus de seriebus infinitis [Theory of inference, posthumous work. With the Treatise on infinite series…]. Basel: Thurneysen. pp. 250–251.
From p. 250, prop. 16: "XVI. Summa serei infinita harmonicè progressionalium, 1/1 + 1/2 + 1/3 + 1/4 + 1/5 &c. est infinita. Id primus deprehendit Frater:…"
 [16. The sum of an infinite series of harmonic progression, 1/1 + 1/2 + 1/3 + 1/4 + 1/5 + …, is infinite. My brother first discovered this…]
 ^ Hersey, George L. Architecture and Geometry in the Age of the Baroque. pp. 11–12, 37–51.
 ^ ^{a} ^{b} Graham, Ronald; Knuth, Donald E.; Patashnik, Oren (1989), Concrete Mathematics (2nd ed.), AddisonWesley, pp. 258–264, ISBN 9780201558029
 ^ Sharp, R. T. (1954). "Problem 52: Overhanging dominoes" (PDF). Pi Mu Epsilon Journal. 1 (10): 411–412.
 ^ Sloane, N. J. A. (ed.). "Sequence A082912 (Sum of a(n) terms of harmonic series is > 10n)". The OnLine Encyclopedia of Integer Sequences. OEIS Foundation.
 ^ Julian Havil, Gamma: Exploring Euler’s Constant, Princeton University Press, 2009.
 ^ Thomas J. Osler, “Partial sums of series that cannot be an integer”, The Mathematical Gazette 96, November 2012, 515–519. https://www.jstor.org/stable/24496876?seq=1#page_scan_tab_contents
 ^ Schmuland, Byron (May 2003). "Random Harmonic Series" (PDF). American Mathematical Monthly. 110: 407–416. doi:10.2307/3647827.
 ^ Weisstein, Eric W. "Infinite Cosine Product Integral". MathWorld. Retrieved 20101114.
 ^ "Nick's Mathematical Puzzles: Solution 72".
External links
 Hazewinkel, Michiel, ed. (2001) [1994], "Harmonic series", Encyclopedia of Mathematics, Springer Science+Business Media B.V. / Kluwer Academic Publishers, ISBN 9781556080104
 "The Harmonic Series Diverges Again and Again" (PDF). The AMATYC Review. 27: 31–43. 2006.
 Weisstein, Eric W. "Harmonic Series". MathWorld.
 Weisstein, Eric W. "Book Stacking Problem". MathWorld.
 Hudelson, Matt (1 October 2010). "Proof Without Words: The Alternating Harmonic Series Sums to ln 2" (PDF). Mathematics Magazine. 83 (4): 294. doi:10.4169/002557010X521831.