Note on the Method of Least Squares and the Thermoelectric Power of Rhodium