1

Suppose that we have a convergent iterated series $$ \sum_{i=1}^\infty\sum_{j=1}^\infty x_{i,j}, $$ where $\{x_{i,j}:i\ge1,j\ge1\}$ is a double sequence of real numbers (not necessarily non-negative). This iterated series doesn't converge absolutely. Does the iterated series $$ \sum_{j=1}^\infty\sum_{i=1}^\infty x_{i,j} $$ converge?

I've seen a few examples where both of the iterated series converges, but the limits are different. So I'm looking for an example where the first iterated series converges, but the second iterated series diverges.

Any help is much appreciated!

Calvin Khor
  • 34,903
Cm7F7Bb
  • 17,364

1 Answers1

2

Consider the sequence given by the infinite matrix $(x_{ij})$,

$$x = \begin{pmatrix}1 \\ -1 & 2 \\ &-2 & 3 \\ &&-3 & 4\\ &&&\ddots \end{pmatrix}$$ where omitted entries are 0. Doing a vertical sum first gives $0+0+0+… = 0$. A horizontal sum first gives $1+1+1+1+… = ∞$.

If $\infty$ isn't what you call divergent,

$$x = \begin{pmatrix}1 \\ -1 & -2 \\ &2 & 3 \\ &&-3 & -4\\ &&&\ddots \end{pmatrix}$$

Vertical sum first still gives 0, but now horizontal sum gives you $1-3+5-7+…$ which is divergent (and I believe, $(C,2)$-summable! )

In fact if you want one where a horizontal sum leaves you with something that is just $(C,1)$-summable, $$x = \begin{pmatrix}1 \\ -1 & -1 \\ &1 & 1 \\ &&-1 & -1\\ &&&\ddots \end{pmatrix}$$ will do.


Just for the record:

If one of them converges, and it does so absolutely in the sense that (say) $∑_i∑_j |x_{i,j}| < ∞ $, then so will the other and to the same value. This is a special case of Fubini's theorem, using the counting measure.

Calvin Khor
  • 34,903
  • Whoops, ok. :) I'll give it a think. – Calvin Khor Sep 23 '15 at 13:29
  • @V.C. I have edited in a proper answer :) – Calvin Khor Sep 23 '15 at 13:34
  • @V.C. Added two other examples with a different kind of divergence. – Calvin Khor Sep 23 '15 at 13:46
  • Thank you very much for your answer (+1)! This is exactly what I was looking for. If we have an iterated series that is not absolutely convergent, then changing the order of summation can lead to anything. You might want to have a look at my another question (http://math.stackexchange.com/questions/1446459/iterated-series-of-a-particular-form) that is related to this question. – Cm7F7Bb Sep 23 '15 at 13:55