We report an interesting integral that connects the uniform mixture of bivari- ate normal densities to Khintchine's theorem on mixing random variables through multiplication. This integral also has an Bayesian interpretation that if we have a uniform prior over the correlation ; the resulting marginal density depends only on the maximal magnitude of the two variables. We report our study about this distribution and its generalizations in higher dimensions, as well as its applications.