Experimental study of<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:msub><mml:mi>χ</mml:mi><mml:mi>b</mml:mi></mml:msub><mml:mo stretchy="false">(</mml:mo><mml:mn>2</mml:mn><mml:mi mathvariant="normal">P</mml:mi><mml:mo stretchy="false">)</mml:mo><mml:mo>→</mml:mo><mml:mi>π</mml:mi><mml:mi>π</mml:mi><mml:msub><mml:mi>χ</mml:mi><mml:mi>b</mml:mi></mml:msub><mml:mo stretchy="false">(</mml:mo><mml:mn>1</mml:mn><mml:mi mathvariant="normal">P</mml:mi><mml:mo stretchy="false">)</mml:mo></mml:math>

We have searched for the di-pion transition ${\ensuremath{\chi}}_{b}(2\mathrm{P})\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}{\ensuremath{\chi}}_{b}(1\mathrm{P})$ in the CLEO III sample of $\ensuremath{\Upsilon}(3\mathrm{S})$ decays in the exclusive decay chain: $\ensuremath{\Upsilon}(3\mathrm{S})\ensuremath{\rightarrow}\ensuremath{\gamma}{\ensuremath{\chi}}_{b}(2\mathrm{P})$, ${\ensuremath{\chi}}_{b}(2\mathrm{P})\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}{\ensuremath{\chi}}_{b}(1\mathrm{P})$, ${\ensuremath{\chi}}_{b}(1\mathrm{P})\ensuremath{\rightarrow}\ensuremath{\gamma}\ensuremath{\Upsilon}(1\mathrm{S})$, $\ensuremath{\Upsilon}(1\mathrm{S})\ensuremath{\rightarrow}{\ensuremath{\ell}}^{+}{\ensuremath{\ell}}^{\ensuremath{-}}$. Our studies include both ${\ensuremath{\pi}}^{+}{\ensuremath{\pi}}^{\ensuremath{-}}$ and ${\ensuremath{\pi}}^{0}{\ensuremath{\pi}}^{0}$, each analyzed both in fully-reconstructed events and in events with one pion undetected. We show that the null hypothesis is not substantiated. Under reasonable assumptions, we find the partial decay width to be $\ensuremath{\Gamma}({\ensuremath{\chi}}_{b}(2\mathrm{P})\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}{\ensuremath{\chi}}_{b}(1\mathrm{P}))=(0.83\ifmmode\pm\else\textpm\fi{}0.22\ifmmode\pm\else\textpm\fi{}0.08\ifmmode\pm\else\textpm\fi{}0.19)\text{ }\text{ }\mathrm{keV}$, with the uncertainties being statistical, internal CLEO systematics, and common systematics from outside sources.