Invariant-mass distribution of $B_c^+\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \pi^+$ candidates (black data points). The maximum likelihood fit of the $B_c^+$ signal is superimposed (blue solid line). Individual fit components are also shown: (dashed blue line) the signal, (red long-dashed line) the background and (green dotted line) $B_c^+\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} K^+$ feeddown.

Distribution of $m_{ { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu}$ for $B_c^+\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu^+\nu_\mu$ candidates selected in simulated event samples of (blue filled points) the signal, (green filled points) the $B_c^+$ feeddown and (red filled squares) the $B_{u,d,s}$ backgrounds. Relative normalization is derived from the fit to the data described later in the text. The part of the spectrum included in the fit is indicated with a vertical dashed black line. The $B_c^+$ feeddown distribution is also shown after magnifying its normalization by a factor of ten (green dashed histogram).

Invariant-mass distribution of $ { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu^+$ pairs from $B_c^+\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu^+\nu_\mu$ candidates (black data points) for (top left) the data, (bottom left) $B_c^+\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu^+\nu_\mu$ signal simulation, (top right) $B_{u,d,s}\rightarrow { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} X$ background simulation and (bottom right) $B_c^+$ feeddown simulation. The unbinned maximum likelihood fit of the $B_c^+$ signal is superimposed (blue solid line). Individual fit components are also shown: (blue short-dashed line) the signal, (red long-dashed line) the background and (green dotted line) $B_c^+$ feeddown.

The measured value of $\cal R$ (horizontal solid line) and its $\pm1\sigma$ uncertainty band (dashed lines) compared to the predictions (diamonds). A nonrelativistic reduction of the Bethe-Salpeter equation is used in the predictions of Chang {\slshape et al.\/} [17], El-Hady \eta [20], and Colangelo \eta [21], while the latter also utilizes heavy quark symmetry. A light-front constituent quark model is used by Anisimov \eta [18] and Ke \eta [25]. QCD sum rules are used by Kiselev \eta [22], a relativistic quasipotential Schr\"odinger model is used by Ebert \eta [23], and a relativistic constituent quark model in used by Ivanov \eta [24].

Animated gif made out of all figures.

Values of the parameters affecting the estimate of the feeddown fraction in the fit to the $m_{ { J \mskip -3mu/\mskip -2mu\psi \mskip 2mu} \mu}$ distribution. For $B_c^+\rightarrow \chi_{cJ}\mu^+\nu_\mu$, $\sum_{J=0,1,2} R_{{\it f J}} \cal B_{\rm casc {\it f J}}$ is listed.

Summary of systematic uncertainties. The total systematic errors are obtained by adding in quadrature the individual contributions.