The authors have found that mineral oil forms an Fe 3 0 4 film on steel and the surfaces becomes smoother during runin. Their work supports earlier research and emphasizes the role of oxygen in boundary lubrication. I will comment on some of their techniques, point out some old papers that came to the same conclusion and discuss the important role of dissolved oxygen. Techniques Describing a mineral oil by viscosity only is inadequate characterization in boundary lubrication studies. The paraffinic, naphthenic, sulfur, oxygen and nitrogen content would be helpful. It was a mistake to compare a mineral oil and a synthetic engine oil probably containing zinc dialkyldithiophosphate. Anti-wear compounds should have been added to the mineral oil, or a mineral oil based engine oil used for a better comparison. The identification of Fe 3 0 4 described reference [10] was, in my opinion, weak. The use of X-ray or electron diffraction would have been positive. The cycling of the friction traces is probably an apparatus problem. At one time the Chevron Research Company pin-ondisk tribometer developed similar friction cycling. We found that a journal bearing was rubbing the disk support shaft on one side due to misalignment and vibrating the disk once per revolution [16]. Vibration would also increase contact resistance by momentary separation of surfaces. The reduction of contact resistance with continued sliding shown in Fig. 3 was surprising. One would expect a run-in surface to induce thicker elasto hydrodynamic lubrication films and thus increase contact resistance.