THE purpose of this paper is to call attention to the need for fundamental wear investigations and to show that wear does not correlate with oiliness.A testing machine suitable for measuring both friction and wear is described. The machine, which is a modification of one previously reported by the author, uses two sets of frictional surfaces - one in the form of a track having two concentric rails and the other consisting of three small buttons with recessed centers and flat tracks on the outer edges. An important feature of the machine is that the rubbing surfaces are maintained automatically at an almost uniform degree of surface smoothness by the lapping action produced by the combined rotating and sliding motion of the buttons.The results presented lead to the following conclusions within the thin-film range investigated: (1) Both wear and friction vary directly with load. (2) Total wear reaches a maximum, in some cases, at one particular speed, whereas wear rate (metal removed per unit of linear surface rubbed) decreases generally as speed is increased. (3) No direct relation exists between wear and friction. A new term, “coefficient of wear,” relating wear rate and load is proposed.No attempt is made to correlate the data obtained with other types of operation or metal combinations as data on these problems are incomplete at the present time.