I'm trying to prove whether or not the following series converges:

$$\sum_{n=1}^{\infty} \left(\frac{\sqrt{n+1}}{n+3}-\frac{\sqrt{n}}{n+4}\right).$$

I'm having some difficulty. I've been trying to compare it with something, but I'm having trouble making progress.


$$\sum_{n=1}^{\infty} \left(\frac{\sqrt{n+1}}{n+3}-\frac{\sqrt{n}}{n+4}\right)= \sum_{n=1}^{\infty} \left(\frac{\sqrt{n+1}-\sqrt{n-1}}{n+3}+\frac{\sqrt{n-1}}{n+3}-\frac{\sqrt{n}}{n+4}\right)= \sum_{n=1}^{\infty} \left(\frac{\sqrt{n+1}-\sqrt{n-1}}{n+3}\right)+ \sum_{n=1}^{\infty} \left( \frac{\sqrt{n-1}}{n+3}-\frac{\sqrt{n}}{n+4} \right)= \sum_{n=1}^{\infty} \left(\frac{\sqrt{n+1}-\sqrt{n-1}}{n+3}\right)= \sum_{n=1}^{\infty} \left(\frac{2}{(n+3)(\sqrt{n+1}+\sqrt{n-1})}\right) < \infty $$

  • 1
    $\begingroup$ Nice use of telescoping. $\endgroup$ – AndrewG Jan 6 '13 at 0:46

Let $A=\dfrac{\sqrt{n+1}}{n+3}$ and $B=\dfrac{\sqrt{n}}{n+4}$. Multiply the $n$-th term by $\dfrac{A+B}{A+B}$. This is the usual "rationalizing the numerator" trick.

After some manipulation, we arrive at $$\frac{3n^2+15n+16}{(n+4)^2(n+3)\sqrt{n+1}+(n+3)^2(n+4)\sqrt{n}}.$$ The denominator is greater than $2n^{7/2}$, and after a while the numerator is less than $4n^2$.

So if $n$ is large enough, the $n$-th term is $\lt \dfrac{2}{n^{3/2}}$.


$\frac{\sqrt{n+1}}{n+3}-\frac{\sqrt{n}}{n+4} ={1 \over \sqrt n}{\sqrt{1+1/n} \over 1+3/n} -{1 \over \sqrt n}{1 \over 1+4/n} ={1 \over \sqrt n}{(1+1/(2n)+O(1/n^2))(1-3/n+O(1/n^2)} -{1 \over \sqrt n}(1-4/n+O(1/n^2)) ={1 \over \sqrt n}(1-5/(2n)+O(1/n^2)) -{1 \over \sqrt n}(1-4/n+O(1/n^2)) ={1 \over \sqrt n}(-5/(2n)+4/n+O(1/n^2)) ={1 \over \sqrt n}(3/(2n)+O(1/n^2)) ={1 \over n^{3/2}}(3/2+O(1/n)) $.

Since each term is $O(1/n^{3/2})$, the sum converges.

The important thing is that the two terms in the summation are each ${1 \over \sqrt n}(1+O(1/n))$, so the ${1 \over \sqrt n}$ terms cancel, leaving only terms which are $O(1/n^{3/2})$, and the sum of these converges. As far as I am concerned, this would be a valid proof of convergence.

  • $\begingroup$ Indeed it would, it is the right way to approach things. Surprised at the downvote, hadn't head of "big O" allergies. $\endgroup$ – André Nicolas Jan 6 '13 at 1:54

$$\frac{\sqrt{n+1}}{n+3}-\frac{\sqrt n}{n+4}=\frac{(n+4)\sqrt{n+1}-(n+3)\sqrt n}{(n+3)(n+4)}$$ $$=\frac{(n+3)(\sqrt{n+1}-\sqrt n)}{(n+3)(n+4)}+\frac{\sqrt{n+1}}{(n+3)(n+4)}$$

$$=\frac{n+1-n}{(n+4)(\sqrt{n+1}+\sqrt n)}+\frac{\sqrt{n+1}}{(n+3)(n+4)}$$

Let $T_n=\frac 1{(n+4)(\sqrt{n+1}+\sqrt n)}$ and $S_n=\frac{\sqrt{n+1}}{(n+3)(n+4)}$

$\frac 1{(n+4)(\sqrt{n+1}+\sqrt n)}<\frac1{n(2\sqrt n)}=\frac1{2n^{\frac32}}$ or $\lim_{n\to \infty}\frac{\frac 1{(n+4)(\sqrt{n+1}+\sqrt n)}}{\frac1{n\sqrt n}}=2$

Using Comparison Test, with p-series (here $p=\frac32$), we can conclude $\sum_{n\to\infty}T_n $ is convergent.

Similarly, $\sum_{n\to\infty}S_n $ is convergent.

Applying $4$th property of this, $\sum_{n\to\infty}(S_n+T_n)$ is convergent.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.