Tower investigated the influence of lubrication on friction at a high sliding velocity. Like ohter researchers, he found that the friction coefficient strongly varied with the load and velocity, contrary to what Coulomb had formulated. Dependent on the rotational velocity, a very low friction coefficient of µ=0.001 to 0.01 was found.

In the final phase of his research, he decided to drill an oil feed hole in the bearing. During the experiment, the oil was found to rise upwards in the feed hole and leaking over the top of the bearing cap. A wooden plug used to block the hole was pushed out by the oil. He then installed a pressure gauge and found it to be inadequate for measuring the high pressure levels.

In addition of Towers discovery, two years later Osborne Reynolds (1886) published a differential equation describing the pressure build up in the oil film. It took still many years before this equation was solved for journal bearings.

Sassenfeld en Walther (1954) numerically calculated characteristics for bearings with final length. The so called "Design Chart of Sassenfeld and Walther" is, despite of later improved solution methods still in use. Alternatively, an analytic method based on interpolation between asymptotic solutions can be used...