The paper is based upon the results of tests made by the Purdue Engineering Experiment Station to study the effect upon engine performance of varying the proportions of fuel to air in the mixture, and its object is to determine the variation in the mixture requirements of an engine at different rates of flow of air through the carbureter. The method of conducting the tests is described. The results are plotted in the charts shown and are discussed in some detail, special discussion regarding the effect of speed and load being presented, and the facts brought out by the tests are summarized. In the general discussion that follows, four definite conclusions regarding the richness of the fuel mixture in its relation to the maximum power are stated, and a like number of definite conclusions concerning the richness of the mixture in relation to maximum efficiency are also given. The further conclusions are that it is impossible for a carbureter to supply an engine with the mixture required for the highest efficiency at all speeds and loads, and that the Purdue tests show that the maximum power is obtained when using a wet mixture, the coldest that can be carbureted satisfactorily.