Abstract
The first experimental evidence for electron correlation in nuclear electron-capture ratios is reported. The measured ML capture ratio of Ar37, 0.073±0.007, is inconsistent with Hartree-Fock calculations, but can be explained by final-state correlation, if the disputed 3s3p6 assignment in Cl i is adopted. The discrepancy in the experimental KβKα x-ray intensity ratio of Cl is removed, and the designation of L23 x-ray and 3s photoemission satellites in Cl is revised.