Positive manifold exists in all cognitive test batteries regardless of design intent. Anti-g tests like Woodcock-Johnson still yield strong g loadings. Thurstone's and Guilford's independent abilities theories failed due to positive correlations. British Ability Scales and CAS battery confirm prominent g despite diverse content. Sternberg's triarchic intelligence tests intercorrelated highly with g. MISTRA's 42 diverse tests show all positive correlations. Piagetian tasks load strongly on same g as IQ tests. g accounts for most common variance and nearly all predictive validity. Higher g-loading means higher heritability and genetic correlation near 1.0. g predicts job performance, academics, income, crime, health better than specifics. Specific abilities add no validity over g in most jobs. Training boosts specifics but not g. Brain size, white/grey matter, conduction velocity correlate with g. Sampling model fails to explain high g-loadings in dissimilar tests or dissociations in impairments. g persists as unidimensional behavioral scale despite neural fractionation.