On the Time‐Correlation Problem, with Especial Reference to the Variate‐Difference Correlation Method