Thomas Simpson's "On the Advantage of Taking the Mean of a Number of Observations, in Practical Astronomy