Remark on the Mulliken Approximation for Two‐Center Electron Distributions