1
$\begingroup$

Given the following Legendre series, I wanted to test its convergence at $x^2=1$ when $l$ is not an integer \begin{align} S = \sum_{j=0}^{\infty} u_j && u_{j+2} = \dfrac{(j+1)(j+2)-l(l+1)}{(j+2)(j+3)} x^2 u_{j} \end{align} I tried using Gauss' test for the following Legendre Series \begin{align} \lim\limits_{j \rightarrow \infty} \left| \dfrac{u_{j}}{u_{j+2}} \right|_{x^2 = 1} = \lim\limits_{j \rightarrow \infty} \dfrac{(j+2)(j+3)}{ (j+1)(j+2)-l(l+1)} = \\ \lim\limits_{j \rightarrow \infty} \dfrac{ j^2 + 5j+6 }{ j^2+3j+2-l(l+1) } = \lim\limits_{j \rightarrow \infty} \dfrac{ [j^2+3j+2-l(l+1)] +[2j+4+l(l+1)] }{ j^2+3j+2-l(l+1) } \\= \lim\limits_{j \rightarrow \infty} 1+ \dfrac{ 2j+4+l(l+1) }{ j^2+3j+2-l(l+1) } = 1 + \dfrac{2}{j} + \mathcal{O}(j^{-2}) = 1 + \dfrac{h}{j} + \mathcal{O}(j^{-2}) \end{align} I identify the value of Gauss' coefficient $h$ from the power series expansion of the of the ratio of the two adjacent series coefficients as illustrated above. I keep getting a factor of 2 for Gauss' coefficient $h$ which states that the series convergence, however my friends tell me that the series diverges. Is it possible that power series expansion cannot be used to evaluate Gauss' coefficient?

$\endgroup$
  • $\begingroup$ Where is the series? $\endgroup$ – Did Sep 2 '14 at 22:13
  • $\begingroup$ @Did I provided more information on the series. Let me know if more information is required. I am so puzzled by this. $\endgroup$ – linuxfreebird Sep 2 '14 at 22:19
2
$\begingroup$

Your mistake is to apply a recipe valid when the ratios $$\left|\frac{u_j}{u_{j+1}}\right|$$ are considered, to a setting where you estimate the ratios $$\left|\frac{u_j}{u_{j+2}}\right|.$$ Renaming $u_{2k}=v_k$, your computations yield $$\left|\frac{v_k}{v_{k+1}}\right|=1+\frac2{2k}+O\left(\frac1{k^2}\right)=1+\frac{\color{red}{1}}{k}+O\left(\frac1{k^2}\right),$$ that is, the case $c=\color{red}{1}$ of the recipe you are trying to apply. (Hence you need to get one more term in the expansion to be able to conclude but at least now, the prediction of your friends is a possible outcome.)

$\endgroup$

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.