The electromagnetic equations are put in their most general form and the following criteria for a useful set of equations and sets of units are established: (1) there should be only one set of fundamental equations, (2) there should be two sets of units, one c .g. s. for the physicist and one practical for the engineer, (3) the eight internationally recognized units should be retained in the practical system, (4) the fundamental equations should be symmetrical in form. It is shown that these criteria can be fulfilled in an infinite variety of ways, of which the Heaviside‐Lorentz system of equations provides the simplest. The relations between c. g. s. and practical units employing this system of equations are tabulated.