05.01.2013 Views

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

12.7. From power series to absolute convergence 247<br />

more general theorem in order to illustrate that the requirements <strong>of</strong> his own theorem<br />

were necessary. However, as F. MERTENS (1840–1927) was later to show, 49 if scru-<br />

tinized more carefully, the example illustrated that the multiplication theorem could<br />

fail if both factors were non-absolutely convergent. This led P. L. WANTZEL (1814–<br />

1848) to prove a version <strong>of</strong> the multiplication theorem which only assumed absolute<br />

convergence <strong>of</strong> one <strong>of</strong> the factors (the other factor just being assumed convergent).<br />

<strong>The</strong> real beginning <strong>of</strong> a concept <strong>of</strong> absolute convergence came with DIRICHLET’S<br />

paper on primes in arithmetic progression which was published in 1837. 50 In that<br />

paper, DIRICHLET introduced a separation <strong>of</strong> convergent series into two classes based<br />

on the convergence <strong>of</strong> the series which resulted when the terms were replaced by<br />

the absolute values: Either the series <strong>of</strong> absolute values remained bounded or it was<br />

unbounded. For series <strong>of</strong> the first class (absolutely convergent series), DIRICHLET<br />

stated that their convergence and sum remained unaffected if the order <strong>of</strong> terms was<br />

altered. In particular, in double (and multiple) sums, the order <strong>of</strong> summation wold<br />

not effect the result. Dirichlet observed that these properties — which were certainly<br />

nice and expected — could fail to hold for series <strong>of</strong> the second class, and he gave two<br />

examples <strong>of</strong> what could happen: a convergent series could either become divergent<br />

or alter its sum if its terms were rearranged. 51<br />

For his Habilitation in 1854, RIEMANN presented a paper on the representability <strong>of</strong><br />

functions by trigonometric series. 52 <strong>The</strong> paper is a milestone in the theory <strong>of</strong> trigono-<br />

metric series and the theory <strong>of</strong> integrals and also contains interesting remarks on the<br />

concept <strong>of</strong> absolute and non-absolute convergence. In the historical preface, RIEMANN<br />

outlined the previous developments in the field and claimed that DIRICHLET’S impor-<br />

tant 1829 paper on the convergence <strong>of</strong> trigonometric series was directly inspired by<br />

DIRICHLET’S discovery <strong>of</strong> the distinction between absolute and non-absolute conver-<br />

gence. 53 RIEMANN expressed his belief that the prevalence <strong>of</strong> power series in analysis<br />

was the reason why those concepts had not previously been separated. 54 <strong>The</strong>re are no<br />

obvious traces <strong>of</strong> the alleged inspiration visible in DIRICHLET’S paper <strong>of</strong> 1829 but —<br />

as mentioned — the distinction became very explicit in a paper with a different topic<br />

in 1837.<br />

RIEMANN advanced a step beyond DIRICHLET’S observation <strong>of</strong> the differences be-<br />

tween absolutely and non-absolutely (conditionally) convergent series when he de-<br />

scribed a very simple method by which the partial sums <strong>of</strong> a conditionally convergent<br />

series could be made to approach any given value by proper rearrangement <strong>of</strong> the<br />

terms <strong>of</strong> the series. Central to RIEMANN’S argument was the realization that if a se-<br />

ries ∑ an was conditionally convergent, the series <strong>of</strong> its positive and negative terms<br />

49 (Mertens, 1875).<br />

50 (G. L. Dirichlet, 1837), see also e.g. (I. Grattan-Guinness, 1970b, 94–95).<br />

51 See also (ibid., 94–95).<br />

52 (B. Riemann, 1854).<br />

53 (ibid., 235) DIRICHLET’S paper is (G. L. Dirichlet, 1829).<br />

54 (B. Riemann, 1854, 235). See below.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!