A search for the rare decay ${K}^{+}\ensuremath{\rightarrow}{\ensuremath{\pi}}^{+}{e}^{+}{e}^{\ensuremath{-}}$ has been carried out at the Lawrence Berkeley Laboratory using a sparkostrictive wire-chamber spectrometer. Analysis programs identified events with three tracks or with two tracks of which one was an electron and the other a positron. Analysis of the three-track events yields a branching ratio $\frac{\ensuremath{\Gamma}({K}^{+}\ensuremath{\rightarrow}{\ensuremath{\pi}}^{+}{e}^{+}{e}^{\ensuremath{-}})}{\ensuremath{\Gamma}({K}^{+}\ensuremath{\rightarrow}\mathrm{all})}<1.7\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}6}$ (90% confidence). The two-track events, analyzed by searching for sparks on a possible pion track, yield a limit of $\frac{\ensuremath{\Gamma}({K}^{+}\ensuremath{\rightarrow}{\ensuremath{\pi}}^{+}{e}^{+}{e}^{\ensuremath{-}})}{\ensuremath{\Gamma}({K}^{+}\ensuremath{\rightarrow}\mathrm{all})}<2.7\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}7}$ (90% confidence). The latter value implies that the coupling constant for a vector neutral current must be about three orders of magnitude smaller than the coupling constant for the charged vector current.

The development of a sparkostrictive wire-chamber spectrometer is described. The spectrometer consists of 9 chambers ranging in active width-times-height from 91 cm × 81 cm to 152 cm × 91 cm. The chambers are positioned between the 152-cm-diameter pole faces of the Lawrence Berkeley Laboratory M5 magnet, which has a 120-cm gap and a 6-kG field in this application. The grid wires are spaced 1 mm apart and on most chambers are oriented with wires running vertically and at ±30° angles to the vertical to provide three redundant coordinates per track. The root-mean-square spatial resolution is typically 1.6 mm. The chambers, while capable of one-track efficiencies of virtually 100% and three-track efficiencies averaging better than 98% per track, are extremely delicate and require constant attention. The various steps taken to achieve high efficiencies are analyzed, and the performance during a recent Ke4+−Kπee+ experiment involving more than 1.6 × 106 triggers is evaluated.