Generate a set of pseudorandom numbers that are

distributed:
Compare the density histogram of the sample with the PDF of the estimated distribution:
In the limit the

distribution becomes symmetric:
In the limit the

distribution has the same kurtosis as
NormalDistribution: