Ordinary methods of determining surface tension require a certain amount of apparatus and time, and often disturb the surface, thereby altering the tension of a water surface covered by a film of oil. As it was desired to make a number of observations on the amount of invisible contamination on the sea, a method depending on the spreading power of drops of mixtures of different fatty substances has been worked out. An oil will just spread against an amount of contamination which lowers the tension by an amount equal to the spreading force, or “spreading coefficient”, of the oil. Pure long chain hydrocarbon oils do not spread on clean water; if small amounts of substances containing water-attracting groups in the molecule are dissolved in the hydrocarbon, spreading occurs, the spreading force depending on the amount of the second substance with the water-attracting group, in the solution. The theory has been treated particularly thoroughly by Langmuir (1933); this paper is, however, concerned only with the application of solutions of different concentrations of a substance such as dodecyl alcohol, in a pure, rather heavy, hydrocarbon oil, to the determination of the surface tension of either fresh or salt water. It is found that the behaviour of a single drop of each of three or four such solutions, observed for half a minute or so, indicates the value of the surface tension within an accuracy of 1 dyne/cm. Calibration against water surfaces with oily contamination depressing the tension by known amounts is of course required; this is however quickly done, once for all, and it is possible to find substances for the spreading solutions which behave similarly on waters differing widely in acidity and in salt content.