In laboratory data, the distributions are skewed and only rarely are they normal, Gaussian or bell-shaped. The authors describe what the great mathematicians of the past did about skewness: Poisson, Quetelet, Carnot, and Laplace, each, treated the problem differently. In modern times, Gram, Charlier, Edgeworth and Pearson made attempts to systematize the treatment of skewed data. Recently, Box, Cox, Johnson, Moestellar and Tukey devised transformations of skewed data to bell-shaped curved. The authors identify distribution functions that are flexible and fit both normal and skewed data. In addition the authors present a totally a priori calculation of a normal range or normal value or reference interval that encompasses not only the biological variations but also the analytical variables.