In this article, we re-examine the published experimental values of the Soret coefficients of a few water-alcohol systems at a mean temperature of 37.5 degrees C, because we feel that these published values (including microgravity values) are doubtful. The main technique applied is the use of thermogravitational columns to determine the thermodiffusion coefficient. The obtained values did not agree with the published values; worse, sometimes the sign was different.